Advances in Experimental Social Psychology, Volume 38 (Advances in Experimental Social Psychology)

EXPLORING THE LATENT STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

Penny S. Visser George Y. Bizer Jon A. Krosnick

Some attitudes are durable and impactful, whereas others are weak and inconsequential. Over the last few decades, researchers have identified roughly a dozen attributes of attitudes that diVerentiate the strong from the weak. However, considerable controversy remains regarding the relations among these attributes. Some scholars have suggested that the various strength‐related attributes reflect a small number of latent constructs, whereas others have suggested that each is a distinct construct in its own right. We review this ongoing controversy, and we then review a diverse set of recent studies that provide new evidence in support of the latter perspective. We consider the implications of our findings for the conceptualization of attitude strength and for the methods by which it is studied. Attitudes determine for each individual what he [or she] will see and hear, what he [or she] will think and what he [or she] will do .... They draw lines about, and segregate, an otherwise chaotic environment; they are our methods for finding our way about in an ambiguous universe. —Gordon W. Allport, 1935, p. 806

I. Introduction It has been nearly 70 years since Gordon Allport famously declared the attitude ‘‘the most distinctive and indispensable concept in contemporary 1 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 38 DOI: 10.1016/S0065-2601(06)38001-X

Copyright 2006, Elsevier Inc. All rights reserved. 0065-2601/06 $35.00

2

PENNY S. VISSER et al.

American social psychology’’ (Allport, 1935, p. 798). In the decades since this bold claim, a large literature has accumulated generally reinforcing the notion that attitudes often do, as Allport (1935) suggested, profoundly influence perception, cognition, and behavior. Equally clear from this literature, however, is that attitudes do not always do so. That is, although some attitudes exert a powerful impact on thinking and on behavior, others are largely inconsequential. Similarly, whereas some attitudes are tremendously durable, resisting change in the face of a persuasive appeal and remaining stable over long spans of time, others are highly malleable and fluctuate greatly over time. The term ‘‘attitude strength’’ is often used to capture this distinction, and researchers have identified roughly a dozen attributes of attitudes that are associated with their strength (Petty & Krosnick, 1995). These strength‐related attitude attributes include attitude importance, knowledge, elaboration, certainty, ambivalence, accessibility, intensity, extremity, structural consistency, and others. A large literature now exists documenting the relations of these attitude attributes with the four defining features of strong attitudes (i.e., resistance to change, stability over time, and a powerful impact on thought and behavior; Krosnick & Petty, 1995). For example, attitudes to which people attach more personal importance are better predictors of behavior (Budd, 1986; Parker, Perry, & Gillespie, 1974; Rokeach & Kliejunas, 1972), more resistant to change (Fine, 1957; Gorn, 1975; Zuwerink & Devine, 1996), and more powerfully influence liking of other people (Byrne, London, & GriYtt, 1968; Clore & Baldridge, 1968; Granberg & Holmberg, 1986; Krosnick, 1988b; McGraw, Lodge, & Stroh, 1990), inferences about others’ personality traits (Judd & Johnson, 1981), and many other cognitive processes. Similar sorts of relations have been documented between other strength‐related attitude attributes and the four defining features of strong attitudes (Krosnick & Petty, 1995). But considerably less is known about the relations among the strength‐related attitude attributes themselves. In fact, a controversy has emerged in the attitude literature regarding the underlying structure of these attributes. In this chapter, we review this controversy and the empirical evidence and conceptual assumptions that have fueled it. We then draw on a diverse set of recent studies to shed new light on the conceptual and practical utility of the competing perspectives. Finally, we consider the implications of our findings for the conceptualization of attitude strength and for the methods by which it is studied. We begin by defining each strength‐related attitude attribute and describing how it is typically measured.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

3

II. Defining Strength‐Related Attitude Attributes A. IMPORTANCE Attitude importance refers to the amount of psychological significance a person ascribes to an attitude (Boninger, Krosnick, Berent, & Fabrigar, 1995b). Because this construct is, by definition, a perception of an attitude, it has typically been measured by asking a person to indicate how personally important an object is to him or her or the extent to which he or she personally cares about the object. Reports of the importance of or concern about the object have been found to be extremely strongly correlated with reports of the importance of and concern about the attitude (Boninger et al., 1995b). B. KNOWLEDGE Knowledge refers to the information about an attitude object that is stored in memory, ranging in volume from very large to none at all (Wood, 1982; Wood, Rhodes, & Biek, 1995). Knowledge has often been measured by asking people to rate their subjective sense of the amount of knowledge they have about an object. Knowledge has also been measured by asking people to list everything they know about an object and directly assessing the quantity of information generated. And knowledge has sometimes been measured via performance on a factual quiz, on the assumption that people who have more correct information about an object also have more total volume of information about the object stored in memory. C. ACCESSIBILITY The accessibility of an attitude refers to how easily or quickly it can be retrieved from memory (Fazio, 1995). The speed of retrieval is presumed to indicate the strength of the link in memory between the representation of the object and the evaluation of it. Accessibility has most often been measured by assessing the length of time it takes a person to report his or her attitude. And accessibility has sometimes been measured by asking people to subjectively rate how quickly their attitudes come to mind when they think of the object. D. CERTAINTY Attitude certainty refers to the amount of confidence a person attaches to an attitude. It has usually been measured by asking people how certain

4

PENNY S. VISSER et al.

or how confident they are about their attitudes, or how sure they are that their attitudes are valid, accurate, or correct (Gross, Holtz, & Miller, 1995).

E. AMBIVALENCE Ambivalence refers to the degree to which a person has both favorable and unfavorable reactions to an object. For example, a person may recognize both admirable and despicable qualities in a particular individual, or may see both pros and cons of a proposed policy. Ambivalence, then, is the degree of evaluative conflict in a person’s responses to an object, with maximum ambivalence occurring when favorable and unfavorable responses are both maximally strong (Thompson, Zanna, & GriYn, 1995). Ambivalence has often been measured by asking people to report the degree to which they subjectively experience feelings of internal conflict. Ambivalence has also been measured by asking people to separately rate the extent of their positive and negative evaluations of an object, which can then be used to calculate ambivalence (Priester & Petty, 1996; Thompson et al., 1995).

F. STRUCTURAL CONSISTENCY Closely related to ambivalence are three aspects of structural consistency involving a person’s overall evaluation of an object, the evaluative implications of his or her beliefs about the object’s qualities, and the evaluative valence of his or her emotional reactions to the object. Evaluative‐aVective, evaluative‐cognitive, and aVective‐cognitive consistency are the terms used to refer to the three possible manifestations of consistency (Chaiken, Pomerantz, & Giner‐Sorolla, 1995). Usually, the three elements have been measured separately and then integrated mathematically to yield quantitative assessments of each type of consistency.

G. EXTREMITY Attitudes are typically conceptualized as lying on a continuum from very positive through neutral to very negative. Attitudes that lie toward either end of this continuum are considered to be extreme (Abelson, 1995). Extremity has usually been derived from reports of attitudes on rating scales.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

5

H. ELABORATION Some attitudes are formed as a result of in‐depth, highly elaborative thought processes. Others are formed through more superficial, cue‐driven processes that require relatively little thought (Petty, Haugtvedt, & Smith, 1995). The extent of prior elaboration about an attitude object has been gauged by asking people how much they have thought about the object previously.

I. INTENSITY Attitude intensity is the strength of the emotional reaction provoked by the attitude object in an individual. It has typically been measured using self‐ reports of the intensity of feelings a person says he or she has about the object (Cantril, 1944, 1946; StouVer et al., 1950).

III. Underlying Constructs? When presented in this fashion, these various attributes seem obviously distinct from one another, with fundamentally diVerent psychological natures and at least somewhat distinct origins and consequences. Some of these attributes are inherently subjective perceptions of the attitude‐holder. For example, attitude importance is a personal judgment of significance—for an individual to say that an attitude is extremely important is to say that he or she cares deeply about it and is presumably motivated to protect it, express it, and be faithful to it in action. Certainty is another subjective judgment, this time about the subjective sense of the justification for holding a particular attitude. Other dimensions reflect the content and structure of representations stored in long‐term memory. For example, knowledge volume is the stockpile of information about an object in memory, which may confer certain abilities to interpret, store, retrieve, and use information to particular ends. Accessibility represents the character of the relation between an object’s representation and its evaluation stored in memory, which regulates the speed and ease with which the attitude springs to mind upon encountering the object. Ambivalence and structural consistency refer to the degree of evaluative harmony among the various components of an attitude that are stored in memory.

6

PENNY S. VISSER et al.

Extremity is a feature of the attitude itself: where along a bipolar evaluative continuum an attitude falls. And elaboration refers to the nature of the cognitive work a person has previously devoted to an attitude object. Thus, it is a description of psychological processes that have evolved in the past. When described in these ways, it seems obvious that understanding attitude strength requires an in‐depth understanding of the workings of each of these various strength‐related attributes. On the other hand, the fact that all of these attributes relate in similar ways to the defining features of strong attitudes (e.g., resistance to change, persistence over time) may suggest that they are surface manifestations of a smaller number of underlying psychological constructs. If this is the case, understanding attitude strength simply requires identifying and investigating these more basic constructs. This issue has sharply divided attitude strength researchers.

A. COMMON‐FACTOR MODELS Many researchers have assumed, either implicitly or explicitly, that sets of strength‐related attitude attributes reflect a common underlying latent construct. For example, many early researchers treated measures of the various attributes as interchangeable. When setting out to gauge attitude ‘‘intensity,’’ some researchers measured certainty (Brim, 1955; Guttman & Suchman, 1947; Katz, 1944; McDill, 1959; Suchman, 1950), whereas others measured extremity (McDill, 1959; Tannenbaum, 1956). To measure ‘‘involvement,’’ some researchers assessed importance (Apsler & Sears, 1968; Borgida & Howard‐Pitney, 1983; Gorn, 1975; Howard‐Pitney, Borgida, & Omoto, 1986), others have assessed knowledge (Stember & Hyman, 1949–1950), and still others measured elaboration (Bishop, 1990; Petty & Cacioppo, 1979). And whereas attitude ‘‘salience’’ has sometimes been measured by indices of importance (Hoelter, 1985; Jackson & Marcus, 1975; Lemon, 1968; Powell, 1977; Tedin, 1980), it has also been gauged by measuring elaboration (Brown, 1974). Treating measures of diVerent strength‐related attitude attributes as interchangeable is reasonable if one assumes that they each reflect the same underlying construct. The notion of conceptual overlap between attributes has sometimes been advocated more explicitly. Roese and Olson (1994), for example, argued that people’s internal cues regarding subjective perceptions of their attitudes (e.g., how important the attitude is to them) may often be weak and ambiguous. As a result, when people are asked about the personal importance of an attitude, they may be forced to engage in self‐perception processes along the lines of those described by Bem (1972). Roese and Olson

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

7

(1994) suggested that one useful cue in such situations may be the speed and ease with which one’s attitude comes to mind. If an attitude comes to mind quickly, people may infer that it must be important to them, whereas if an attitude comes to mind slowly, people may infer that it must not be very important to them. Haddock and his colleagues (Haddock, Rothman, & Schwarz, 1996; Haddock, Rothman, Reber, & Schwarz, 1999) and Bassili (1996a) have made similar arguments, suggesting that reports of many strength‐related attitude attributes (e.g., importance, certainty, intensity) are all derived from a common set of cues, such as the experienced ease with which people are able to retrieve attitude‐supportive information from memory. 1. Evidence The most pervasive type of empirical evidence presented to advance the notion of overlap among strength‐related attributes has been factor analysis. In a number of studies, exploratory factor analyses or principal components analyses have been conducted in order to identify the factor structure underlying various strength‐related attitude attributes (Abelson, 1988; Bass & Rosen, 1969; Bassili, 1996a; Erber, Hodges, & Wilson, 1995; Kokkinaki, 1998; Lastovicka & Gardner, 1979; Pomerantz, Chaiken, & Tordesillas, 1995; Prislin, 1996). Although these studies (listed down the left side of Table I) each included somewhat diVerent sets of strength‐related attributes, most included measures of attitude importance, knowledge, accessibility, certainty, extremity, and elaboration. And in each of these studies, two or three latent factors emerged. Table I indicates upon which latent factor (labeled 1, 2, or 3) each attribute was found to load. Although attitudes toward disparate objects were examined in a variety of diVerent participant populations and at diVerent times spanning nearly three decades, some striking consistencies emerged: some pairs of attributes consistently loaded on the same factor, whereas other pairs tended to load on diVerent factors. For example, in every one of the 18 analyses that included importance and elaboration, these attributes loaded on the same latent factor (the first factor). And in 74% of the 19 analyses involving both importance and knowledge, these two attributes loaded on the same factor (again, the first factor). Likewise, knowledge and elaboration loaded on the same factor in nearly 70% of the analyses that included them both. Just as strikingly, importance and accessibility did not load on the same factor in a single 1 of the 11 instances in which they were included in the same analysis. Similarly, knowledge and extremity loaded on the same factor in only 2 of 17 analyses. Finally, importance and extremity loaded together in only 2 of 21 analyses. Thus, the cluster of importance, elaboration, and

COMPARISON

OF

RESULTS

TABLE I EXPLORATORY FACTOR ANALYSES

FROM

Strength‐related attitude attributes Study Prislin (1996)

Prislin (1996)

Prislin (1996)

Issue AYrmative action

Abortion

Euthanasia

Bassili (1996a)

Hiring quotas

Bassili (1996a)

Pornography

Bassili (1996a)

Hate speech

Erber et al. (1995) Pomerantz et al. (1995)

Ronald Reagan Capital punishment

Researchers’ name for factor Generalized attitude strength Extremity Ease of attitude expression Generalized attitude strength Extremity Ease of attitude expression Generalized attitude strength Extremity Ease of attitude expression Meta‐attitudinal Operative Meta‐attitudinal Operative Meta‐attitudinal Operative Conviction Factor 2 Embeddedness Commitment

Importance

Knowledge

Elaboration

Certainty

1

1

1

1

Extremity

Accessibility

2 3 1

1

1

1

2

2 3

1

1

1

1 2 3

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1 2 1

1

1 2

2

2

2

2

2

2

2

2

2

2

Pomerantz et al. (1995) Pomerantz et al. (1995) Pomerantz et al. (1995) Pomerantz et al. (1995) Kokkinaki (1998) Krosnick et al. (1993) Krosnick et al. (1993) Krosnick et al. (1993) Krosnick et al. (1993) Abelson (1988)

Abortion Environment Gay men Election forecasts European Monetary Union Abortion (Study 1) Death penalty Defense spending

Abortion (Study 3) South Africa, nuclear power, God, abortion, welfare, strategic defense initiative, Nicaragua, AIDS

Embeddedness Commitment Embeddedness Commitment Embeddedness Commitment Embeddedness Commitment Embeddedness Conviction Internal consistency Factor 1 Factor 2 Factor 1 Factor 2 Factor 1 Factor 2 Factor 3 Factor 1 Factor 2 Factor 3 Ego preoccupation Cognitive elaboration Emotional commitment

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

2

2 3

2

2

2

2

2

2

1 2

1

1

1

1

1

1

1

1 3

1 2

3 1 2

2

3 1

1 2 3

Continues

TABLE I Continued Strength‐related attitude attributes Study Lastovicka & Gardner (1979) Bass & Rosen (1969) Bass & Rosen (1969) Bass & Rosen (1969) Bass & Rosen (1969) Bass & Rosen (1969)

Issue

Researchers’ name for factor

Importance

Knowledge

Elaboration

Consumer products

Importance Familiarity

1

Study 1—George Romney Study 2— Academics Study 2— Quarter system Study 2—Farm subsidies Study 2—George Romney

Involvement Confidence Involvement Confidence Involvement Confidence Involvement/ confidence Involvement Confidence

1

1

1

1

1

1

1

1

1

1

Certainty

Extremity

2

2

2

2

2 1

2 1

2

2

Accessibility

2

Note: Cell entries of ‘‘1’’ indicate that an attitude attribute loaded on the first factor, ‘‘2’’ indicates that an attitude attribute loaded on the second factor, and ‘‘3’’ indicates that the attribute loaded on the third factor.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

11

knowledge seems distinct from accessibility and extremity when viewed through this lens.1 Based upon these sorts of factor analyses, many investigators have averaged together measures of multiple strength‐related attributes to create composite indices of attitude strength, which have then been used to investigate attitude properties and processes (Bassili, 1996a; Bassili & Roy, 1998; Eagly et al., 2000; Haddock et al., 1999; Lavine, HuV, Wagner, & Sweeney, 1998; Pomerantz et al., 1995; Prislin, 1996). For example, composites called ‘‘attitude strength’’ have been created by averaging measures of importance, certainty, and intensity (Haddock et al., 1996, 1999), measures of importance, certainty, knowledge (and other measures; Eagly et al., 2000; Theodorakis, 1994), measures of extremity, certainty, and accessibility (Bassili & Roy, 1998), measures of importance and certainty (and other measures; Holland, Verplanken, & van Knippenberg, 2002), or measures of importance, knowledge, certainty, and elaboration (and other measures; Bassili, 1996a; Prislin, 1996). Pomerantz et al. (1995) averaged importance and knowledge (and other measures) into a composite called ‘‘embeddedness,’’ and they averaged extremity and certainty (and other measures) to create a composite called ‘‘commitment.’’ Hodson, Maio, and Esses (2001) created an index that they also labeled ‘‘commitment’’ by averaging measures of importance, certainty, and personal relevance. Kokkinaki (1998) averaged measures of importance and elaboration (and other measures) to create an index of ‘‘embeddedness,’’ and she averaged measures of certainty, knowledge, ambivalence, and extremity to create an index of ‘‘conviction.’’ Abelson (1988) gauged ‘‘ego‐ preoccupation’’ by averaging measures of importance, elaboration, and other variables. And several scholars have created indices that they labeled ‘‘involvement’’ by averaging measures of importance, certainty, and elaboration (and other measures; Miller, 1965), measures of importance, knowledge, and elaboration (and other measures; Verplanken, 1989, 1991), or measures of importance, interest in information, and attitude‐expressive behaviors (and other measures; Thompson & Zanna, 1995). In each case, these investigators have then explored the cognitive and behavioral consequences of the composites they created. 1 Some inconsistencies across studies also emerged. For example, attitude importance and certainty loaded on the same factor in some studies (Erber et al., 1995; Prislin, 1996) but on diVerent factors in other studies (Abelson, 1988; Pomerantz et al., 1995; Visser, 1998). And even within a single investigation, inconsistent results sometimes appeared. For example, Bassili (1996a) found importance and certainty to load on a common factor for some attitude objects but to load on diVerent factors for other attitude objects. Similarly, Bass and Rosen (1969) found importance and certainty to load on a single factor for one attitude object and to load on diVerent factors for a second object.

12

PENNY S. VISSER et al.

B. A DIFFERENT PERSPECTIVE However, this is not the only approach present in the attitude strength literature. In fact, more common are studies that have examined one strength‐related attribute at a time, without combining attributes together into a composite index (Krosnick & Abelson, 1992; Raden, 1985). This approach can be justified by the obvious conceptual diVerences between the various strength‐related attitude attributes. For example, as our earlier review implied, attaching personal importance to an attitude is quite diVerent from simply possessing a large store of knowledge about the attitude object. To attach great importance to an attitude is to care tremendously about it and to be deeply concerned about it. And this deep concern is consequential—perceiving an attitude to be personally important leads people to protect it against attack and use it in processing information, making decisions, and choosing a course of action (Krosnick & Abelson, 1992). Thus, attitude importance seems to be primarily a motivator: of attitude protection and attitude use. In contrast, knowledge is simply a cache of information stored in memory. As such, knowledge per se seems less likely to be motivational in and of itself. Rather, its eVects seem more likely to be ability based in character. DiVerences of this sort in the psychological nature of the various strength‐related attitude attributes have led some scholars to question the wisdom of combining them into composite indices of attitude strength. 1. Evidence Supporting this view is evidence from a number of confirmatory factor analyses, which have avoided some of the pitfalls of exploratory factor analyses (Krosnick, Boninger, Chuang, Berent, & Carnot, 1993, Krosnick, Jarvis, Strathman, & Petty, 1994; Lavine et al., 1998; Visser, 1998). In particular, exploratory factor analysis is subject to distortion due to systematic measurement error, because people diVer from one another in how they interpret the meanings of the points on rating scales. When the same scale is used to measure diVerent constructs, as has often been the case in exploratory factor analytic investigations of attitude strength, these individual diVerences in scale point interpretation will produce an artifactual positive correlation between the measures of the constructs across participants (Bentler, 1969; Brady, 1985; Costner, 1969; Green, 1988; Green & Citrin, 1994). This would cause constructs that are perfectly orthogonal to appear to be positively correlated if they are measured using the same rating scale coded in the same direction (as was the case in most past studies). Exploratory factor analysis and related techniques will (incorrectly) presume that

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

13

this covariation is substantively meaningful; increasing the likelihood that variables measured with a common rating scale will appear to load on the same factor. To overcome this problem, Krosnick et al. (1993) advocated the use of confirmatory factor analyses that model and correct for systematic measurement error. Following the logic of the multitrait–multimethod approach (Campbell & Fiske, 1959), they proposed that multiple strength‐related attitude attributes should be assessed using several diVerent types of rating scales. The attributes can then be represented as latent variables (gauged by multiple indicators), and method factors can be included in the model to account for covariation between the measures that is due to a common method of measurement (Alwin & Krosnick, 1985; Andrews, 1984; Green & Citrin, 1994; Green, Goldman, & Salovey, 1993; Jarvis & Petty, 1996). The use of multiple indicators also permits disattenuation of correlations between latent constructs to correct for the impact of random measurement error. Krosnick et al. (1993) used this approach in three studies to estimate associations among various strength‐related attributes of people’s attitudes toward several social and political issues. Although a few pairs of attributes were quite strongly correlated, most were weakly or not at all correlated. Krosnick et al. (1993) explicitly tested the possibility that a common underlying construct could account for covariation among sets of strength‐related attributes. Specifically, they tested the goodness‐of‐fit of various structural equation models positing that pairs, trios, or larger sets of attributes load on a single underlying factor, each possibility derived from existing theories and empirical findings. In each case, Krosnick et al. (1993) compared the fit of the common‐factor model to that of a model representing the relevant strength‐related attributes as separate (albeit correlated) constructs. Across three studies, Krosnick et al. (1993) found virtually no evidence that a group of attributes reflected a common underlying factor. In fact, only two common‐factor models received consistent support across all tests of them: self‐reported knowledge and objective measures of people’s actual stores of knowledge about an object consistently loaded on a single factor, as did attitude extremity and attitude accessibility. But in most tests (30 of 42 tests, or 71%), the common‐factor models entailed significant compromises in goodness‐of‐fit of the model to the observed data. Lavine et al. (1998) conducted similar analyses and obtained similar results. Lavine et al.’s confirmatory factor analyses (1998) examined the underlying structure of six strength‐related attributes: importance, certainty, intensity, frequency of thought, extremity, and ambivalence. These researchers found that a model in which the six attributes were treated as separate constructs fit the data significantly better than did any model in which

14

PENNY S. VISSER et al.

subsets of attributes were treated as manifestations of a common latent factor. Empirical findings of this sort have reinforced the notion among some attitude strength researchers that the various strength‐related attitude attributes are distinct constructs in their own right rather than reflections of a smaller number of more general underlying factors. This perspective emphasizes the multidimensionality of attitude strength, and it suggests that eVorts to elucidate the origins and consequences of strength should focus on developing a fuller understanding of the origins and consequences of each individual strength‐related attribute.

C. RECONCILING THESE DIVERGENT PERSPECTIVES As the earlier discussion implies, the attitude strength literature seems to have come to an impasse. Two bodies of empirical evidence support two contradictory conceptualizations of attitude strength. Exploratory factor analyses suggest that there are two or three basic dimensions of attitude strength and that all of the various strength‐related attitude attributes discussed in the literature can be reduced to these basic dimensions. According to this perspective, distinguishing each and every strength‐related attribute is a trivial exercise in splitting hairs. In contrast, confirmatory factor analyses suggest that attitude strength is multifaceted and that the strength‐related attributes cannot be reduced to a smaller set of more general underlying factors. According to this perspective, combining strength‐related attributes into composite indices glosses over meaningful conceptual distinctions among them. Clearly, the debate over the underlying structure of the strength‐related attitude attributes is not likely to be resolved through additional eVorts to factor analyze the correlations among them. Such correlations can be used to support either of these two perspectives. In fact, even with a single data set, the decision to conduct exploratory versus confirmatory factor analyses can yield evidence that appears to unambiguously favor one perspective or the other (Krosnick et al., 1993; Visser, 1998). Instead, we propose that the solution may lie in a reformulation of the basic question that attitude strength researchers have set out to address. Up to this point, the debate over the relations among strength‐related attributes has been cast, at least implicitly, in absolute terms—either each strength‐related attribute is a distinct construct with unique antecedents and consequences, or sets of attributes are largely redundant, essentially interchangeable reflections of the same underlying construct. But the truth almost certainly lies somewhere between these extremes. Most pairs of

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

15

strength‐related attitude attributes are likely to be at least partially distinct— arising from at least some unique antecedents and setting into motion at least some distinct cognitive and behavioral consequences. But many pairs of strength‐related attributes may share some common variance as well, arising from some of the same antecedents and perhaps exerting some of the same eVects on thought and behavior. The question, then, is whether there is enough unique variance to justify distinguishing among the various strength‐related attitude attributes when building theories of the origins and consequences of attitude strength. If very little of the variance in each strength‐related attribute is unique and most is shared with other attributes in a set, those attributes must have largely redundant antecedents and consequences. In the interest of parsimony, measures of these attributes could be combined together into an index to more eYciently explore their workings in relation to other psychological constructs. But if the amount of unique variance in each attribute is substantial, this would indicate that the causes of the various attributes are quite diVerent, and it would raise the possibility that the attributes may also exert diVerent sorts of cognitive and behavioral eVects.2 And the more diVerent their origins and consequences, the more misleading the results of an investigation will be if measures of diVerent attributes are combined into a composite index of attitude strength. 1. A New Approach This logic makes clear that an eYcient alternative to factor analysis is direct, simultaneous exploration of the antecedents and consequences of various strength‐related attributes. If two attributes appear to be similarly aVected by many predictor variables and appear to exert similar kinds of eVects on thinking and action, there is little to be gained by maintaining sharp distinctions between them in empirical investigations of attitude strength—even if the attributes are far from being perfectly correlated. But if two attributes have diVerent causes and distinct eVects on thought and behavior, there is indeed utility in maintaining the distinction between them in theory‐building—even if the attributes are quite strongly correlated. More specifically, evidence that one strength‐related attitude attribute is related to a particular cognitive or behavior outcome whereas another

2 Of course, evidence that two attributes arise from diVerent antecedents would not necessarily imply that they also exert diVerent consequences. Distinct constructs can have overlapping sets of eVects. But two attributes that reflect a common underlying construct must arise from a common set of antecedents and exert a common set of consequences.

16

PENNY S. VISSER et al.

attribute is not related to that outcome would clearly challenge the practice of combining the two attributes into an omnibus index of attitude strength. Doing so would yield misleading evidence regarding the distinct functioning of these attributes. Similarly, if one strength‐related attribute is positively associated with a cognitive or behavioral outcome and another attribute is negatively associated with that outcome, researchers would be ill‐advised to combine the two attributes into a composite index. In this case, doing so is likely to mask entirely these countervailing relations, obscuring the functioning of these independent attributes. Finally, evidence that two strength‐ related attributes interact to produce a particular cognitive or behavioral outcome would also challenge the practice of combining them into an index. Such an index would yield an incomplete portrait of the relation between the attributes and the outcome. In our view, this approach—focusing not strictly on the correlations among attributes but on the degree of overlap in their antecedents and consequences—provides a better conceptual match to the basic questions regarding the structure and function of attitude strength that interest attitude researchers. Identifying sets of attributes that arise from a common set of antecedents and that set into motion a common set of cognitive and behavioral consequences would provide an empirically justified conceptual framework for consolidating disparate lines of research on these individual attributes. And findings of this sort would facilitate swift progress in the eYcient investigation of the workings of these clusters of attributes. Evidence that the various strength‐related attributes instead arise from distinct causal antecedents and exert diVerent cognitive and behavior eVects would also clarify the conceptualization of attitude strength and would facilitate progress in empirical investigations of it. Findings of this sort would suggest that not all strong attitudes are alike, and that careful attention to the bases of attitude strength will have useful payoVs for psychological theory building. It may be, for example, that some attitudes are strong because people attach a great deal of importance to them, which has a particular set of consequences for thinking and action. Other attitudes may be strong because they are based on a substantial volume of attitude‐relevant knowledge, which may set into motion a somewhat diVerent set of cognitive and behavioral consequences. And some attitudes may manifest strength because of the copresence of two or more strength‐related attributes, with unique consequences for thinking and action. This multidimensional conceptualization of attitude strength would suggest that composite indices of attitude strength, comprised of sets of strength‐related attributes, will often yield misleading evidence and inaccurate characterizations of strength‐ related processes, impeding the development of refined theory in this domain.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

17

D. OVERVIEW OF THIS CHAPTER In the remainder of this chapter, we review a set of studies that have directly assessed the degree of overlap in the antecedents and consequences of strength‐related attributes, providing a broad set of evidence about the conceptual and practical utility of maintaining distinctions among them. To build this chapter, we conducted a thorough search of the literature to identify all studies that have directly compared the causes or eVects of two or more strength‐related attitude features. We describe all such studies in the sections that follow. As will become evident, much of the existing work has compared attitude importance to other attitude features, so our chapter necessarily tilts in this direction. But even with this tilt, the studies we review seem to provide a broad and solid basis for drawing conclusions about the structure and function of attitude strength. 1. Importance and Knowledge We begin by describing a program of research exploring the workings of attitude importance and attitude‐relevant knowledge. Consistent with the fact that importance and knowledge loaded on the same factor in most exploratory factor analyses, many investigators have averaged together measures of importance and knowledge to yield an index when investigating attitude properties and processes (Bassili, 1996a; Eagly et al., 2000; Pomerantz et al., 1995; Prislin, 1996; Theodorakis, 1994; Verplanken, 1989, 1991). We will review evidence regarding the degree to which importance and knowledge arise from the same causal antecedents and exert the same kinds of cognitive and behavioral consequences, as the common‐factor model assumes. 2. Importance and Certainty We then review evidence of convergences and divergences in the cognitive and behavioral consequences of attitude importance and certainty. In seven past exploratory factor analyses, these attitude attributes loaded on the same factor (Bass & Rosen, 1969; Bassili, 1996a; Prislin, 1996). And many investigators have averaged importance and certainty to create composite indices (Bassili, 1996a; Eagly et al., 2000; Haddock et al., 1996, 1999; Hodson et al., 2001; Holland et al., 2002; Miller, 1965; Prislin, 1996; Theodorakis, 1994). The evidence we will review assesses the wisdom of this practice. 3. Importance and Accessibility Next, we review a program of research examining attitude importance and attitude accessibility, in light of Roese and Olson’s (1994) suggestion that

18

PENNY S. VISSER et al.

people may infer the importance of their attitudes by noting how quickly and easily they come to mind and Krosnick’s (1989) suggestion that importance may be a cause of accessibility. We review the array of available evidence regarding the antecedents of these attributes, and we review evidence exploring the causal relations between them to gauge their conceptual and procedural independence. 4. Certainty and Accessibility We then consider similar evidence regarding attitude certainty and attitude accessibility. Specifically, we explore the degree to which these two strength‐ related attributes arise from a common set of antecedents, and we assess the implications of each attribute for attitude‐congruent behavior. 5. Attitude Strength Composites Versus Their Constituents We then describe a series of studies that explicitly pitted individual strength‐ related attitude attributes against composites of them. These studies explored the extent to which these attributes (alone or in combination): (1) predicted two defining features of strong attitudes (resistance to change and stability), (2) responded similarly to ease of retrieval manipulations, and (3) regulated responsiveness to question wording and question order eVects in attitude measurement. In our review of these studies, we will pay particular attention to the degree to which composite indices may obscure the workings of individual strength‐related attributes.

IV. Attitude Importance and Knowledge A. HYPOTHESES 1. Common Antecedents and Consequences? Until recently, no existing evidence disputed the notion that importance and knowledge spring from the same causal antecedents and have the same eVects. But this is because no investigation had directly explored the issue. Consider first the existing evidence on antecedents. Three primary causes of attitude importance had been documented in the literature thus far: (1) the belief that the attitude object impinges on one’s material self‐interest, (2) identification with reference groups or individuals who attach importance to the attitude object or whose material interests are linked to the object, and (3) recognition of a link between the attitude and one’s core values (Boninger, Krosnick, &

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

19

Berent, 1995a). Attitude accessibility, the experienced ease of retrieving attitude‐relevant information from memory, and self‐esteem maintenance motives have also been posited to regulate attitude importance (Haddock et al., 1996, 1999; Pelham, 1991; Roese & Olson, 1994). Until recently, no research had tested whether these factors are causes of knowledge as well. The primary origins of knowledge documented in the literature thus far were: (1) direct experience with an attitude object (Fazio & Zanna, 1981) and (2) exposure and attention to information about the object provided by other people, through conversations or mass media (McGuire, 1986; Roberts & Maccoby, 1985). No studies had tested whether importance is enhanced by direct experience or exposure to information from informants; it is conceivable that these causal processes do indeed occur. With regard to the consequences of importance and knowledge, some overlap has been documented. Quantity of attitude‐relevant knowledge has been shown to be associated with greater consistency between attitudes and behavior, greater ability to encode new information about an object, reduced reliance on peripheral cues in evaluating persuasive messages, more extensive thinking about attitude‐relevant information, greater sensitivity to the quality of arguments in evaluating a persuasive message, and greater resistance to attitude change (Biek, Wood, & Chaiken, 1996; Davidson, 1995; Wilson, Dunn, Kraft, & Lisle, 1989; Wood, 1982; Wood & Kallgren, 1988; Wood et al., 1995). In line with three of these findings, importance has also been shown to be associated with greater attitude–behavior consistency (Budd, 1986; Parker et al., 1974; Rokeach & Kliejunas, 1972), more extensive thinking about attitude‐relevant information (Berent & Krosnick, 1993; Celsi & Olson, 1988; Howard‐Pitney et al., 1986), and greater resistance to attitude change (Fine, 1957; Gorn, 1975; Zuwerink & Devine, 1996). Until recently, other documented consequences of knowledge had not been investigated with regard to importance, raising the possibility that importance leads to these outcomes as well. And various documented eVects of importance (e.g., the motivation to acquire information about the attitude object, Berent & Krosnick, 1993; consistency between attitudes and core values, Jackman, 1977; Judd & Krosnick, 1989) had not yet been investigated as possible consequences of attitude‐relevant knowledge. 2. Distinct Antecedents and Consequences? However, when considered less mechanically and more conceptually, there are clearly reasons to expect that the causes and eVects of importance and knowledge will be diVerent. Because it reflects the degree of concern, caring, and significance an individual attaches to an attitude, attitude importance should serve to motivate people to use the attitude in processing

20

PENNY S. VISSER et al.

information, making decisions, and taking action. Consistent with this reasoning, attitude importance has been shown to inspire people to seek out attitude‐relevant information (Berent & Krosnick, 1993; Zaichkowsky, 1985) and to think carefully about that information (Berent, 1990). Importance also motivates people to use an attitude: more important attitudes have greater impact on judgments of liking for other people (Byrne et al., 1968; Clore & Baldridge, 1968; Granberg & Holmberg, 1986; Krosnick, 1988b; McGraw et al., 1990), on voting behavior in elections (Krosnick, 1988b; Schuman & Presser, 1981), and on trait inferences (Judd & Johnson, 1981). And importance has been shown to stimulate attitude‐ expressive behavior (Krosnick & Telhami, 1995; Schuman & Presser, 1981). Thus, importance appears primarily to be a motivator. In contrast, knowledge is not in and of itself motivational—it is simply a store of information in memory. And most of its eVects appear to be primarily ability‐based in character. Knowledge has been shown to enhance recall (Cooke, Atlas, Lane, & Berger, 1993; Fiske, Lau, & Smith, 1990; McGraw & Pinney, 1990; Schneider, Gruber, Gold, & Opwis, 1993), improve comprehension (Eckhardt, Wood, & Jacobvitz, 1991; Engle, Nations, & Cantor, 1990), increase the speed of judgments (Fiske et al., 1990; Paull & Glencross, 1997), improve cue utilization in decision tasks (Paull & Glencross, 1997), enable appropriate inferences (Pearson, Hansen, & Gordon, 1979), facilitate the objective processing of attitude‐relevant information (Biek et al., 1996), and the learning of new topic‐relevant information (Hansen, 1984; Kyllonen, Tirre, & Christal, 1991; Willoughby, Waller, Wood, & MacKinnon, 1993), and enable the generation of eVective counterarguments to a persuasive appeal (Wood, 1982; Wood et al., 1995). Thus, although knowledge seems to enable people to perform various relevant cognitive tasks more eVectively, we see no reason to suppose that it should, in and of itself, motivate people to engage in any behavior. These characterizations suggest that importance and knowledge are likely to have distinct eVects on thought and behavior. Importance and knowledge seem likely to be distinct in terms of their origins as well. Knowledge often accumulates simply as the result of exposure to information about an object. Simply being exposed to information is only likely to lead a person to attach importance to an attitude if that information makes a compelling case for the existence of a link between the object and a person’s self‐interest, reference groups or individuals, or values. Thus, knowledge acquisition is unlikely to have a uniform eVect on importance. On the other hand, information acquisition sometimes occurs intentionally—people sometimes seek out new knowledge about a particular object—and people who attach great personal importance to an object are likely to be motivated to gather information about it. Thus, importance may be a cause of knowledge accumulation.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

21

B. EVIDENCE In a series of studies conducted recently with both undergraduate samples and a large, nationally representative sample, Visser, Krosnick, and Norris (2004) compared the causes and consequences of attitude importance and attitude‐relevant knowledge and found many divergences. 1. Relation Between Importance and Knowledge Across several studies involving a variety of social and political issues (e.g., legalized abortion, global warming, capital punishment), the correlations between importance and knowledge were positive and moderately strong, ranging from .26 to .48. 2. Origins of Importance and Knowledge Consistent with previous research (Boninger et al., 1995a), Visser et al. (2004) found that self‐interest, the importance of the issue to reference groups and individuals, and value‐relevance each predicted unique variance in the importance that people attached to their attitudes toward legalized abortion. Exposure to news media, on the other hand, was unrelated to the importance people attached to this issue. In contrast, news media exposure was a significant (and indeed, the strongest) predictor of attitude‐relevant knowledge. The importance of the issue to reference groups and individuals was unrelated to knowledge. Interestingly, self‐interest and value‐relevance did predict a significant amount of variance in knowledge. Further analyses, however, revealed that the impact of self‐interest and value‐relevance on knowledge was mediated by attitude importance. Recognizing that material interests or cherished values are at stake in this issue led people to attach importance to their attitudes, which in turn motivated them to seek out relevant information about it and become more knowledgeable.3

3

Visser et al. (2004) also explored alternative mediational relations. For example, they assessed the possibility that the impact of self-interest on importance may have been mediated by knowledge: recognizing that an attitude object impinges on a person’s material interests may directly inspire him or her to gather information about the object. Having accumulated a great deal of such information, people may then come to decide that the attitude is important to them, perhaps in an eVort to rationalize having invested the eVort in information gathering or through inference processes (e.g., ‘‘If I know this much about an object, then it must be important to me.’’). Visser et al. (2004) found, however, that knowledge did not mediate the relations between any of the antecedents and attitude importance.

22

PENNY S. VISSER et al.

Fig. 1. Documenting the causes of attitude importance and knowledge.

These results are encapsulated in the causal model presented in Fig. 1, the parameters of which Visser et al. (2004) estimated using covariance structure modeling techniques and which fit the data well. As the coeYcients in Fig. 1 illustrate, self‐interest, the importance of the issue to significant others, and value‐relevance each led to increased attitude importance, and increases in attitude importance led to increased knowledge about the issue. Knowledge increased as a function of media use, but media use had no impact on attitude importance. These results suggest that importance and knowledge spring from largely distinct proximal sources. The trajectories of attitude importance and knowledge over time provide another source of evidence regarding the overlap in their causes. If these strength‐related attributes arise from a common set of causal antecedents, they should rise and fall together over time, reflecting the modulation of those shared antecedents. But if importance and knowledge arise from diVerent origins, they may rise and fall independently following the distinct ebbs and flows of their separate antecedents. To explore this issue, Visser et al. (2004) took advantage of a unique real‐world opportunity provided by the White House Conference on Global Climate Change on October 6, 1997, which drew a great deal of media attention and sparked a vigorous national debate about global warming. During the subsequent months, hundreds of stories on this issue appeared on television, in newspapers, on the radio, and in news magazines. Advertisements paid for by industry organizations and other advocacy groups further expanded the national discussion. The impact of this flood of information was explored by conducting telephone interviews with two nationally representative samples of American

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

23

adults. The first sample was interviewed just before media attention to global warming surged, and the second sample was interviewed several months later, after the media had turned their attention elsewhere. The same measures were used in both surveys to assess the importance that people attached to the issue of global warming and the amount of knowledge they possessed about this issue, permitting an examination of the changes in each construct over time. Between the first and second waves of data collection, the importance that people attached to the issue increased significantly. But the diversity of opinions expressed during the national discussion of global warming left people feeling no more knowledgeable on this issue: knowledge remained steady between the first and second waves of data collection. Importance and knowledge, then, exhibited diVerent trajectories over time, reinforcing the notion that they spring from diVerent origins that rise and fall independently. a. Consequences of Importance and Knowledge. Visser et al. (2004) next explored the degree to which importance and knowledge regulate the impact of attitudes on thought and behavior in the same ways. Specifically, these investigators conducted a series of studies examining the eVects of importance and knowledge on: (1) attitude polarization following exposure to mixed evidence, (2) perceptions of hostile media bias, (3) selective information gathering, and (4) attitude‐expressive behavior. b. Attitude Polarization. Following procedures developed by Lord, Ross, and Lepper (1979), and Visser et al. (2004) presented participants with summaries of two scientific studies, one yielding evidence of negative psychological consequences for women who obtained a legal abortion, and the other oVering evidence of positive psychological consequences. After reading this mixed set of evidence, participants answered questions measuring their attitudes toward abortion; participants had answered these same questions several weeks earlier as well, which permitted an assessment of attitude change. Participants also reported the degree to which they perceived their attitudes toward abortion to have changed, if at all, as a result of reading about the studies. Replicating Lord et al. (1979), Visser et al.’s (2004) participants perceived that the mixed evidence had polarized their attitudes: participants who were initially favorable toward legalized abortion perceived themselves to have become more favorable, and participants who were initially unfavorable toward legalized abortion perceived themselves to have become less favorable. And this perceived polarization was regulated by attitude importance: participants who attached more importance to the issue perceived greater polarization than did participants who attached less importance to their attitudes. In contrast, knowledge was unrelated to perceived attitude polarization.

24

PENNY S. VISSER et al.

Also replicating previous findings (Miller, McHoskey, Bane, & Dowd, 1993), Visser et al. (2004) found that participants’ perceptions of their own attitude changes were completely incorrect. Whereas participants perceived their attitudes to have become more extreme, the mixed evidence in fact caused attitude moderation: people who were initially favorable toward legalized abortion became less so after reading the mixed evidence, and people who were initially unfavorable toward abortion became less so as well. Like perceived attitude change, actual attitude change was regulated by attitude importance, but in a way opposite to its eVect on perceived polarization: participants who attached a great deal of importance to the issue exhibited less attitude change in response to the mixed evidence than did participants low in attitude importance. Interestingly, whereas importance was negatively associated with attitude change, knowledge was positively associated with change: people who were more knowledgeable about abortion exhibited more attitude moderation in response to the mixed evidence. Consistent with some prior research (Wood et al., 1995), more knowledge may have equipped people to objectively recognize the merits of the study that contradicted their own views and to see genuine flaws in the study that supported their views, making them more likely to temper their initial views. c. Hostile Media Bias. Visser et al. (2004) asked participants to evaluate the fairness of media coverage of global warming in an eVort to explore the impact of attitude importance and attitude‐relevant knowledge on the hostile media bias, which is the tendency to perceive that a balanced presentation of information on a controversial issue is biased against one’s own side of the issue (Vallone, Ross, & Lepper, 1985). Vallone et al. (1985) suggested that the hostile media eVect is driven at least partly by knowledge. They argued that when assessing the fairness of media coverage, people compare their own store of information about an issue to the information presented by the media. Because people tend to possess more attitude‐congruent than attitude‐incongruent information, even a balanced media presentation would appear to have omitted more of the former than of the latter, producing the perception of a bias against one’s own side of the issue. And indeed, Vallone et al. (1985) found that people with larger stores of knowledge about the issue manifested a stronger hostile media bias than did people with little knowledge. Consistent with these findings, Visser et al. (2004) found strong evidence of a hostile media bias in people’s perceptions of the news coverage of the existence of global warming that was regulated by knowledge: people who were highly knowledgeable about global warming perceived a much stronger hostile media bias than did people who were less knowledgeable about this issue. Importance, on the other hand, did not regulate the magnitude of the hostile media bias.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

25

d. Selective Information Gathering. Visser et al. (2004) next explored the hypothesis that attitude importance—but not attitude‐relevant knowledge— motivates people to selectively expose themselves to information that will permit them to use the attitude in a subsequent judgment. They told participants that they would receive information about 12 political candidates, each of whom they would later evaluate. For each candidate, participants were permitted to choose three out of six possible issues on which to learn the candidate’s positions. As expected, people who attached more importance to the issue of capital punishment requested candidates’ positions on that issue significantly more often. Similarly, people for whom legalized abortion was more important requested candidates’ positions on that issue more often. In neither case was attitude‐relevant knowledge related to information selection. Attaching importance to an issue apparently motivated participants to seek information that enabled them to use their attitudes when evaluating candidates, but possessing knowledge did not. e. Attitude‐Expressive Behavior. Performing an attitude‐expressive behavior requires suYcient motivation to do so and suYcient knowledge to plan and execute appropriate behavioral strategies. Importance and knowledge may provide such motivation and ability, respectively. To test this idea, Visser et al. (2004) asked undergraduates whether they had ever performed seven types of behaviors expressing their attitudes toward legalized abortion (e.g., contacting a public oYcial to express their views, wearing a button or t‐shirt indicating their views). Similar measures were included in a telephone survey of a representative national sample of American adults, asking about the issue of global warming. As predicted, importance and knowledge were positively associated with increase in attitude‐expressive behavior in both studies. And in both studies, importance and knowledge interacted significantly: the combination of high importance and high knowledge was associated with a pronounced surge in attitude‐expressive behavior. f. Negative AVect. Visser et al. (2004) reasoned that if importance motivates people to protect and express their attitudes, they should experience negative aVect when achieving these goals is blocked by impediments in the environment. For example, people who attach importance to a particular political issue should feel upset if the government enacts laws that are contrary to their position. People who simply possess a great deal of information about the issue, on the other hand, should be less likely to experience a negative aVective reaction of this sort. And indeed, Visser et al. (2004) found that people who attached a great deal of importance to the issue of legalized abortion reported that they would be very upset if the government enacted a law that contradicted their position on this issue, whereas knowledgeable people were no more likely than those with little knowledge to find this upsetting. Similarly, people who attached importance

26

PENNY S. VISSER et al.

to their abortion attitudes reported that they would find it distressing to learn that a close friend held a divergent viewpoint on this issue, whereas attitude‐ relevant knowledge was unrelated to this reaction. And attitude importance (but not knowledge) predicted the intensity of people’s negative aVective reactions to a compelling counterattitudinal persuasive message that was diYcult to refute: participants who attached great importance to their attitudes reported more negative emotions (e.g., anger, frustration, anxiety) than those who attached less importance to their attitudes. g. Biased Hypothesis Testing. Visser et al. (2004) next explored the hypothesis that people who attach importance to their attitudes are motivated to disconfirm counterattitudinal assertions, whereas people who simply possess a great deal of attitude‐relevant information have no particular motivation to do so. To explore this idea, Visser et al. (2004) used a modified version of the Wason (1966, 1968) selection task. The Wason task requires participants to test a particular hypothesis using a limited set of evidence available to them. In the original version of the task, for example, participants were presented with four cards. They were told that a letter is printed on the front of each card and a number is printed on the reverse side of the cards. The cards were arrayed in front of the participant such that the front of two of the cards were visible (revealing letters) and the back of the remaining two cards were visible (revealing numbers). Participants were presented with the assertion, ‘‘If a card has a vowel on one side, it has an even number on the other side.’’ Their task was to indicate which card(s) they would need to turn over to determine whether or not the assertion was true. Although the task is quite simple, participants do surprisingly poorly— in most studies, only about 20% of participants perform the task correctly. The most common errors reflect a confirmatory bias: a tendency to seek evidence that confirms the hypothesis one has set out to test and to neglect information that could potentially disconfirm it (Wason & Johnson‐ Laird, 1972). Recent evidence indicates that when participants are intrinsically motivated to disconfirm the hypothesis they are testing, they perform significantly better on the Wason selection task (Dawson, Gilovich, & Regan, 2002). Visser et al. (2004) constructed a version of the Wason selection task that required participants to test a counterattitudinal assertion. If attitude importance motivates people to protect and defend cherished attitudes, people who attach more importance to the target attitude should be more motivated to disconfirm the assertion, improving their performance on the task. Possessing a large store of attitude‐relevant knowledge, on the other hand, should not confer this motivation, suggesting that knowledge will be unrelated to task performance.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

27

And indeed, this is precisely what Visser et al. (2004) found. Participants who opposed capital punishment were asked to test the hypothesis that all states that use capital punishment have murder rates that are lower than the national average. Replicating past investigations, participants did quite poorly on this task: only about 30% of participants performed the task correctly, and as in past studies, the errors reflected a confirmatory bias. Consistent with predictions, however, task performance was significantly influenced by attitude importance: 53% of participants who considered capital punishment highly important performed the task correctly, whereas only 19% of participants who considered capital punishment to be unimportant did so. Also consistent with predictions, attitude‐relevant knowledge did not moderate task performance. The proportion of participants who performed the task correctly was virtually identical among participants low and high in knowledge: 31% and 32%, respectively.

C. CONCLUSIONS Across these various studies, importance and knowledge were both related to various indicators of attitude durability and impactfulness. But these relations were far from identical. In some cases, importance had an eVect when knowledge did not. For example, importance was associated with perceived attitude polarization following exposure to conflicting empirical evidence, whereas knowledge was not. Importance was also associated with negative aVective reactions when an attitude was threatened, whereas knowledge was not. In other cases, knowledge appeared to regulate an attitude eVect when importance did not. For example, more knowledgeable people perceived greater hostile media bias, whereas importance was not related to perceived media bias. And there were instances in which the two attributes related in opposite ways to attitude eVects. For example, more knowledge about an attitude object was associated with more attitude moderation in the face of conflicting empirical evidence, whereas attaching more importance to the attitude was associated with less moderation. And the two attributes sometimes interacted to produce an eVect: the combination of high importance and high knowledge was associated with a pronounced surge of attitude‐expressive behavior. These attributes also appeared to arise from distinct causal antecedents and fluctuate independently over time. And whereas importance seems to instigate the accumulation of attitude‐relevant knowledge, knowledge does not appear to lead to increased importance. Thus, even though these attributes have consistently loaded on the same factor in exploratory factor analyses, importance and knowledge seem better described as diVerent

28

PENNY S. VISSER et al.

constructs possessing distinct psychological properties, arising from diVerent origins, producing disparate outcomes, and apparently operating via diVerent causal processes.

V. Importance and Certainty In light of the findings that importance and certainty have often loaded on the same factor in exploratory factor analyses and have frequently been combined into indices, Visser, Krosnick, and Simmons (2003) explored whether this is a sensible strategy by comparing the cognitive and behavioral consequences of attaching importance to an attitude and of holding the attitude with certainty. We review their findings next.

A. HYPOTHESES Visser et al. (2003) examined whether importance and certainty regulate the degree to which Americans used their attitudes on government policy issues to choose between the candidates who ran for President of the United States in 1996. If importance motivates people to use an attitude, then greater importance attached to an issue such as abortion may have motivated individuals to choose between Bill Clinton and Bob Dole based on their attitudes toward abortion. That is, people who attached more importance to the issue of abortion may have been more likely to use the match between their own stand on the issue and the stands of Bill Clinton and Bob Dole to decide which of these candidates to support. Uncertainty may cause people to hesitate before using an attitude, so lower certainty may have inhibited people from using their preference on a particular policy issue to choose between the competing Presidential candidates. And an interaction might appear, such that especially powerful impact of a policy preference on candidate evaluations might occur when both importance and certainty are high. Visser et al. (2003) also explored the possibility that a person whose candidate preference is an expression of many important policy preferences may be more invested in that candidate preference. And if a person’s candidate preference is derived from policy preferences that he or she holds with little confidence, he or she may be only minimally invested in that candidate preference. Thus, high importance or high certainty regarding many policy preferences may lead to greater commitment to candidate preferences and therefore more unhappiness if one’s preferred candidate is

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

29

not elected, more eVorts to persuade others to vote for one’s preferred candidate, greater intention to vote on election day, higher likelihood of actually turning out to vote in the election, and greater frequency of other attitude‐expressive behaviors. Visser et al. (2003) also explored the possibility that observed diVerences in the ability of attitude importance and certainty to predict particular cognitive or behavior outcomes may be due to diVerences in the reliability with which the two constructs were measured. To do this, they used covariance structure modeling techniques to eliminate the distorting impact of random and systematic measurement error when assessing the relations of importance and certainty to four consequences of attitude strength.

B. EVIDENCE Visser et al. (2003) tested the first set of hypotheses using data from 1996 National Election Study conducted by the Center for Political Studies at the University of Michigan. This survey involved interviews with a large, nationally representative sample of American adults during the weeks immediately preceding the US Presidential election that year and again during the weeks following the election. 1. Relation Between Importance and Certainty Across five political issues (government spending on social services, defense spending, government assistant to Blacks, legalized abortion, and environmental protection), the correlations between importance and certainty were positive and moderate in magnitude, ranging from .29 to .35. 2. Impact of Policy Attitudes on Candidate Preferences Visser et al. (2003) first explored whether importance and certainty regulated the degree to which people used their attitudes toward a particular issue when formulating candidate preferences. To do so, they assessed the relative proximity of participants’ own attitudes on five diVerent political issues and the attitudes of President Clinton and Senator Dole on those same issues. They also assessed participants’ relative candidate evaluations, and tested the notion that relative proximity to the candidates on issues that are personally important or held with great certainty will have an especially pronounced impact on relative candidate preferences.

30

PENNY S. VISSER et al.

As expected, the more importance people attached to an issue, the more impact that issue had on candidate preferences. In addition, attitudes held with greater certainty had more impact on candidate preferences than those held with less certainty. These eVects were independent—importance and certainty each accounted for unique variance in candidate preferences—and there was no interaction between importance and certainty. 3. Strength of Candidate Preference Visser et al. (2003) next explored whether candidate preferences based on attitudes that are more important and/or held with more confidence are especially impactful. They constructed indices of total attitude importance and attitude certainty across a set of salient political issues and used these indices to predict various indicators of people’s commitment to their candidate preference. And indeed, they found interesting divergences in the consequences of importance and certainty. For example, the amount of importance people attached to a set of policy attitudes was unrelated to the degree to which they found one of their nonpreferred Presidential candidates acceptable, but people higher in certainty were significantly less likely to find any nonpreferred candidate acceptable. In contrast, people who attached a great deal of importance to their policy attitudes were more likely to try to convince other people how to vote, whereas people who held their policy attitudes with more certainty were no more likely to do so. And whereas importance and certainty were both positively related to preelection intentions to turn out to vote, only importance predicted whether people actually voted. 4. Distortions due to DiVerential Reliability? Visser et al. (2003) explored the relative impact of attitude importance and attitude certainty on four potential consequences of attitude strength: greater interest in obtaining information about the attitude object, greater attention to such information in the media, greater frequency of discussing the attitude object with friends, and greater eVort to obtain attitude‐relevant information for use in a subsequent judgment. And because these constructs had been measured with multiple items on several diVerent types of measurement scales, Visser et al. (2003) were able to estimate these relations after parsing out the potentially distorting impact of both random and systematic measurement error. They found that attitude importance was strongly and significantly related to all four of the consequences, whereas attitude

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

31

certainty was not associated with any of them. They also explored and found no evidence of an interaction between importance and certainty. 5. Attitude‐Expressive Behavior Using data from a large, representative sample of US adults, Visser et al. (2003) also explored the moderating impact of attitude certainty and attitude importance on attitude‐expressive behavior. They expected that people would be particularly likely to act in accordance with their attitudes when those attitudes were especially important to them and they were unconstrained by attitude uncertainty. And this is precisely what they found: attitude importance and attitude certainty interacted to predict whether people had performed behaviors such as writing a letter to a public oYcial to express their views or attending a meeting to discuss a particular issue. They also found that this two‐way interaction was further moderated by household income when they explored the predictors of attitude‐expressive financial contributions: among those who had suYcient resources, attitude importance and certainty interacted to predict giving, but among those who were under tight financial constraints, no such interaction emerged. C. CONCLUSIONS Taken together, these results argue against treating attitude importance and attitude certainty as reflections of a single underlying construct. Doing so would have obscured the fact that importance and certainty each predicted unique variance in the impact of a policy attitude on people’s candidate preferences and on their turnout intentions. And combining importance and certainty would have masked their distinct patterns of association with other outcomes: importance (but not certainty) predicted whether people turned out to vote on election day, whereas certainty (but not importance) predicted the degree to which people found a nonpreferred presidential candidate acceptable. Finally, combining measures of importance and certainty would have obscured the interaction between them in predicting attitude‐expressive behaviors. All of this suggests that there is utility in maintaining the distinction between attitude importance and attitude certainty in investigations of attitude strength. Furthermore, these results continue to reinforce the portrait of attitude importance as a motivator, because it appears to have inspired people to use their attitudes when evaluating candidates and to express those attitudes behaviorally. Uncertainty appears to have operated as a restraint, inhibiting people from using their attitudes to evaluate candidates or to express those

32

PENNY S. VISSER et al.

attitudes behaviorally. And uncertainty appears to have made people more open to the idea of supporting nonpreferred candidates.

VI. Importance and Accessibility Four past factor‐analytic studies included measures of both importance and accessibility, and in every case these two attributes loaded on diVerent factors. Furthermore, no past study we have uncovered has averaged importance and accessibility into a single index. Thus, it might appear that these measures are viewed as representing distinct constructs. But Roese and Olson (1994) argued that they may indeed amount to the same construct. We review work on this issue next. A. HYPOTHESES Highly accessible attitudes spring to mind spontaneously when an attitude object is encountered, without intent or cognitive control. Much theorizing about attitude accessibility has focused on the direct consequences of these cognitive processes, which unfold automatically and often nonconsciously (Fazio, 1995). However, a very diVerent literature has also considered accessibility to be consequential, but via perceptions of it in consciousness. For example, Tversky and Kahneman (1974) described the ‘‘availability heuristic’’ as a tool people use—the amount of diYculty a person has in trying to retrieve an instance of something from memory is taken to be diagnostic about the phenomenon being retrieved. Likewise, Schwarz (1998; Schwarz et al., 1991) has proposed that people use their own experience of the ease or diYculty of retrieving cognitive elements from memory as a basis for making inferences and judgments. In line with this latter perspective, Roese and Olson (1994) proposed that the experience of attitude accessibility may influence people’s judgments about the importance of their attitudes. As we described earlier, these investigators proposed that people’s internal cues regarding the personal importance of their attitudes are often weak and ambiguous. When asked to report the importance of an attitude, people cast about for cues on which to base this judgment. Roese and Olson (1994) suggested that one useful cue in such situations may be the speed with which one’s attitude comes to mind. If an attitude comes to mind quickly, people may infer that it must be important to them, whereas if an attitude comes to mind slowly, people may infer that it must not be very important to them. In this way, attitude importance may be an after‐the‐fact reflection of attitude accessibility.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

33

In contrast, Krosnick (1989) suggested that importance may be a cause of accessibility. Once a person decides to attach personal significance to an attitude, he or she is likely to seek out information relevant to it and to think deeply about that information and its implications for the attitude. As a result, the attitude is likely to become more accessible over time, springing to mind quickly and eVortlessly when an individual encounters the object. Thus, the eVect of importance on accessibility may be mediated by selective exposure and selective elaboration.

B. EVIDENCE To test their hypothesis, Roese and Olson (1994) manipulated the accessibility of attitudes and then measured the importance of those attitudes. Specifically, these investigators induced people to express some attitudes repeatedly, while not expressing other attitudes at all. Consistent with previous research (Fazio, Chen, McDonel, & Sherman, 1982), this manipulation increased the accessibility of the former attitudes. The manipulation also increased the degree of personal importance people said they attached to those attitudes. Roese and Olson (1994) attempted to test the notion that attitude accessibility caused attitude importance reports. These investigators reasoned that if attitude importance judgments are in fact derived from attitude accessibility, then accessibility should have mediated the impact of the repeated expression manipulation on importance reports. That is, repeated expression should have caused increased accessibility, which in turn caused increased importance ratings. Given their study design, testing this hypothesis required computing two within‐participants partial correlations: one correlation of the manipulation with importance controlling for accessibility, and another of the manipulation with accessibility controlling for importance. However, Roese and Olson (1994) instead computed between‐participants partial correlations (Roese, personal communication, October, 1995), so their results on this point are not informative with regard to causal impact. Therefore, it is possible that the observed increase in importance may not have resulted directly from the increase in accessibility. Rather, it may have resulted from greater thought about the repeatedly expressed attitudes, perhaps leading people to recognize genuine and legitimate reasons to consider the issues more important. Their study design manipulated repeated expression within participants across a set of issues. For some participants, attitudes on one set of issues were repeatedly expressed, whereas attitudes on another set of issues were not. For other participants, the first set of attitudes was not repeatedly expressed, and the other set of attitudes was. Thus, repeated attitude

34

PENNY S. VISSER et al.

expression varied across issues within participants; which issues were repeatedly expressed varied between participants. The appropriate statistical analysis would assess whether the within‐participants variation in repeated expression aVected importance while controlling for accessibility and vice versa. Implementing such a within‐participants analysis is computationally a bit tricky, and Roese and Olson (1994) accidentally carried out the analysis the wrong way. They unintentionally averaged all of the attitude measures for each participant together (thus combining attitudes that were and were not repeatedly expressed) and computed partial correlations across participants between which issues were repeatedly expressed and the average importance of all attitudes and the average accessibility of all attitudes. Roese (personal communication, October, 1995) graciously acknowledged this computational error that renders the mediational results Roese and Olson (1994) reported uninterpretable. Bizer and Krosnick (2001) conducted a set of studies aimed at resolving this ambiguity. In their first three studies, Bizer and Krosnick (2001) manipulated known antecedents of attitude accessibility and attitude importance, and observed the impact on both attributes. If both of these attributes reflect a common underlying construct, then any manipulation that influences one should also influence the other. But if the two attributes represent distinct constructs, then a cause of one will not necessarily influence the other. Finally, if both are influenced simultaneously by a manipulation, then the impact of the manipulation on one attribute may be mediated by the other. Bizer and Krosnick (2001) also examined naturally occurring changes in importance and accessibility via a panel survey to see whether one variable predicted subsequent changes in the other. Thus, these studies oVered opportunities to explore the latent structure of these attributes in a novel way. C. RELATION BETWEEN IMPORTANCE AND ACCESSIBILITY Across Bizer and Krosnick’s (2001) various studies, the naturally occurring associations between importance and accessibility were positive, though the magnitude of these associations was consistently quite small. The correlations ranged from .14 to .18. D. IMPACT OF REPEATED ATTITUDE EXPRESSION ON IMPORTANCE AND ACCESSIBILITY In two studies, participants repeatedly expressed two attitudes and did not repeatedly express two others. Later, attitude accessibility and attitude importance were assessed. As expected, repeated attitude expression rendered

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

35

participants’ attitudes more accessible in both studies. However, repeated expression did not increase importance ratings in either experiment. The manipulation had no eVect on importance ratings in Study 1, and it tended to decrease importance ratings in Study 2. Because importance did not increase in either study, there was no need to examine whether accessibility mediated the impact of the manipulation on importance. And the fact that accessibility increased without a parallel increase in importance demonstrates that the attributes arise from at least somewhat distinct sources.

E. IMPACT OF PERSONAL RELEVANCE ON ACCESSIBILITY AND IMPORTANCE Bizer and Krosnick (2001) next manipulated participants’ self‐interest in an issue and explored the impact of this manipulation on attitude importance and on attitude accessibility. Participants were given the opportunity to read news articles from a ‘‘computerized bulletin board service.’’ Two articles discussed policies that were going to be instituted at their own university, whereas two other articles discussed policies that had been rejected at a far‐ away university. Thus, two of the policies were related to participants’ own material outcomes and two were not. Participants were permitted to select which articles they wished to read from a list of headlines, and could spend as much time reading and thinking about the articles as they wanted. When they finished one article, participants could press a key to return to the list of articles, at which time they could select another article or end the reading portion of the experiment. Participants then reported their attitudes toward the four target policies, and a computer measured response latencies of these reports. Finally, participants completed a paper‐and‐pencil questionnaire that assessed their perceptions of the likelihood that each of the four policies would be enacted at their university and how important each issue was to them personally. As expected, policies that were described as personally relevant to participants were indeed perceived to be more likely to be implemented at participants’ own university than the other policies. Furthermore, attitudes toward the more relevant policies were more personally important and were reported more quickly than were attitudes on nonrelevant issues. To identify the causal processes responsible for the eVect of the relevance manipulation on importance and accessibility, Bizer and Krosnick (2001) estimated the parameters of the structural equation model shown in Fig. 2. In light of findings reported by Boninger et al. (1995a), Bizer and Krosnick (2001) expected that the personal relevance manipulation would influence perceptions of the likelihood of the policies being implemented at the

36

PENNY S. VISSER et al.

Fig. 2. Documenting the impact of manipulation of personal relevance on perceived likelihood of policy implementation, attitude importance, and accessibility.

participants’ university, which would in turn influence the importance they attached to each issue. Bizer and Krosnick (2001) also allowed for the possibility that the manipulation of personal relevance might impact accessibility. Reading a headline indicating that a story’s topic was personally relevant may have increased the likelihood that people would choose to read the story. And reading the story would have increased participants’ exposure to information about the target policy, which may have caused people to access their attitudes toward the policy. This would in turn increase the accessibility of those attitudes, even if some people ultimately concluded that the policy was not likely to be implemented at their own university or that the issue was not personally important to them for some other reason. Therefore, a direct influence of the manipulation on selective exposure was included in the model, and selective exposure could in turn aVect accessibility. According to the logic of instrumental variable analysis, this empirical context also aVords the opportunity to statistically separate the reciprocal eVects of importance and accessibility on one another (Kenny, 1979). In particular, Bizer and Krosnick (2001) tested two key hypotheses: that importance might cause accessibility and that accessibility might cause importance. The former eVect would occur if high importance causes people to think more about the policy and their attitude toward it, which would in turn enhance accessibility. The eVect of accessibility on importance could occur via the self‐perception processes outlined by Roese and Olson (1994). As the parameter estimates in Fig. 2 indicate, enhancing the personal relevance of a policy increased perceptions of the likelihood that it would be implemented at participants’ own university, which increased the amount of personal importance participants attached to their attitudes toward the policy. And as expected, enhancing the personal relevance of a policy in a story headline increased the likelihood that participants would choose to read about it, which in turn increased accessibility.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

37

Attitude importance exerted a positive eVect on attitude accessibility, but the eVect of accessibility on importance was not significant. Thus, there is no evidence here that people inferred that their attitudes were more important because they came to mind more quickly. Instead, these data are consistent with the notion that importance inspired thought about the target policies, and this additional thought increased the accessibility of participants’ attitudes toward the policies.

F. THE CAUSES OF NATURALLY OCCURRING CHANGES In a final study, Bizer and Krosnick (2001) analyzed the data from the global warming survey described earlier. During each interview, participants reported their attitudes toward global warming and reported how important this issue was to them personally. Using a technique developed by Bassili (1996b), interviewers marked the length of time between the completion of asking the attitude question and the beginning of participants’ answers, which was treated as a measure of attitude accessibility. The flow of information on global warming between the two interviews oVered Americans the opportunity to talk, think, and learn about the issue. During this time, people could have been selective in their exposure to and processing of this information, and we would expect people for whom the issue was more important to attend more to this information than people who initially attached little importance to the issue. Thus, high initial levels of importance may have led to increases in the subsequent accessibility of global warming attitudes. As the parameter estimates in Fig. 3 indicate, attitude importance evidenced a moderately high level of stability over time, and accessibility manifested a somewhat lower but nonetheless reliable level of stability. Furthermore, initially higher attitude importance predicted subsequent increases in accessibility, consistent with the notion that importance is a cause of accessibility. Interestingly, initial levels of attitude accessibility did not predict subsequent changes in importance. Thus, greater initial accessibility was not associated with subsequently increasing importance.

G. CONCLUSIONS Bizer and Krosnick’s (2001) first two studies pose a strong challenge to the general claim that people infer attitude importance from attitude accessibility, presuming that an object must be important to them if their attitude toward it came to mind quickly. Although repeated attitude expression

38

PENNY S. VISSER et al.

Fig. 3. Documenting the impact of attitude importance and accessibility on one another.

increased attitude accessibility in both studies, it had no impact on attitude importance in Study 1, and repeated expression tended to decrease importance in Study 2. In Study 3, increased accessibility did not cause increases in importance, though the reverse causal process did occur. And in Study 4, increases in importance led to increases in accessibility over time, whereas accessibility did not influence subsequent attitude importance. All of this is inconsistent with the notion that importance and accessibility reflect a single construct. Furthermore, these studies provide additional evidence that importance is a motivator of information exposure and elaboration. VII. Accessibility and Certainty As with attitude importance, some scholars have argued that judgments of the certainty with which people hold their attitudes are derived from the accessibility of those attitudes (Bassili, 1996a). The notion is that if an attitude comes to mind quickly and eVortlessly, people are likely to infer that they must hold the attitude with great certainty. In contrast, if an attitude requires time and eVort to call to mind, people may conclude that they are not very certain of their evaluation of an object or issue. Consistent with this view, attitude accessibility and attitude certainty have sometimes loaded on the same latent factor in exploratory factor analyses (Bassili, 1996a; Pomerantz et al., 1995). Berger (1992) explored this possibility more directly, and we describe her work next. A. HYPOTHESES Like Visser et al. (2003), Berger (1992) conceptualized uncertainty as an inhibitor, rendering individuals reluctant to rely on their attitudes when faced with the task of forming a judgment or choosing a behavior. Drawing on past research, Berger (1992) posited that attitude certainty is determined

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

39

at least in part by the volume and perceived reliability of attitude‐supportive information stored in memory. Specifically, she suggested that when an attitude is supported by a suYciently large base of reliable information, people will feel confident that their attitude is valid. In contrast, Berger (1992) speculated that exposure to attitude‐relevant knowledge does not, in and of itself, strengthen the link in memory between an individual’s representation of an attitude object and his or her evaluation of the object. People can and often do process new information about an object without retrieving from memory their stored attitude toward the object, leaving attitude accessibility unaVected by the new information. In particular, she proposed that the frequency with which people are exposed to a set of information will determine the likelihood that their attitudes will be retrieved. When the information is being encountered for the first time, people are less likely to retrieve their attitudes and instead will devote their cognitive resources to encoding the information. When they have encountered the information several times and have therefore had ample opportunity to encode it, people are more likely to activate their attitudes toward the object when confronted by the information. And indeed, Berger and Mitchell (1989) demonstrated that moderate increases in exposure to attitude‐relevant information do not increase attitude accessibility. Only when the frequency of exposure to new information becomes suYciently high that new information is no longer being encoded does additional exposure encourage the activation of the attitude, increasing attitude accessibility. Thus, Berger (1992) predicted that moderate increases in exposure to new information would increase attitude certainty but would have no aVect on attitude accessibility. In contrast, Berger (1992) expected that repeated attitude expression would increase the accessibility of people’s attitudes, and she did not expect repeated expression to regulate attitude certainty. Berger (1992) also anticipated distinct consequences of accessibility and certainty. Drawing on Fazio’s (1990) theoretical and empirical work, Berger (1992) anticipated that attitude accessibility and attitude certainty would regulate attitude–behavior correspondence in diVerent ways. In particular, following the tenets of the motivation and opportunity as determinants (MODE) model (Fazio & Towles‐Schwen, 1999), Berger (1992) predicted that attitude accessibility would regulate the attitude–behavior relation when behaviors were spontaneous. That is, attitudes that come to mind quickly and eVortlessly will guide spontaneous behaviors, whereas attitudes that are less accessible will have little impact on such behaviors. In contrast, she predicted that when people have the motivation and ability to thoughtfully contemplate a behavior, certainty (and not accessibility) will regulate attitude–behavior correspondence. That is, when they are carefully choosing a behavior, people will be relatively unaVected by the sheer accessibility of

40

PENNY S. VISSER et al.

their attitudes and will instead choose to behave in accordance with those attitudes that they hold with great certainty. B. EVIDENCE Berger (1992) enlisted research participants in the ostensible task of evaluating a set of five candy bars that were not currently available in their area but would be coming on the market in the near future. For each candy bar, participants were shown one print advertisement containing a photograph of the candy bar as well as product information. Some participants saw each advertisement one time, and others saw each advertisement three times. After they had seen the advertisements, participants expressed their attitudes toward each candy bar. Some participants accessed and expressed their attitudes toward each candy bar three times, whereas others accessed and expressed their attitudes toward the candy bars only once. Participants also rated the certainty with which they held their attitudes toward each candy bar. Participants next completed a computer task that included measures of the accessibility of participants’ attitudes toward the five candy bars. After they had completed this final task and had been paid for their participation, participants were directed to a large tray containing 10 each of the various candy bars and, out of the sight of the experimenter, were invited to help themselves to 7 candy bars as a parting gift. As expected, participants who repeatedly expressed their attitudes subsequently exhibited greater attitude accessibility than those who had expressed their attitudes only once. Repeated expression had no impact on attitude certainty, however. Also as expected, repeated exposure to information about the candy bars increased the certainty with which participants held their attitudes, but it did not render those attitudes any more accessible. Furthermore, attitude certainty and attitude accessibility were uncorrelated in each of the experimental conditions and in the sample as a whole. Finally, attitude certainty moderated the relation between attitudes and behaviors: people who held their attitudes with confidence were much more likely to act in accordance with those attitudes when selecting candy bars, whereas people who held their attitudes with less confidence exhibited weaker attitude–behavior correspondence. Attitude accessibility was entirely unrelated to attitude–behavior correspondence. C. CONCLUSIONS These findings dovetail with the work of Visser et al. (2003) reinforcing the notion that uncertainty may operate as an inhibitor, rendering people

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

41

hesitant to use their attitudes as guides to behavior. These findings are also in harmony with the work of Bizer and Krosnick (2001), suggesting that subjective judgments regarding strength‐related properties of one’s attitudes (e.g., importance, certainty) are not simply derivative byproducts of attitude accessibility. Instead, these results oVer compelling evidence that attitude certainty and attitude accessibility are distinct psychological constructs with at least partially unique antecedents and at least some distinct behavioral consequences.

VIII. Attitude Strength Composites and Their Constituents Several studies have compared the relations of a cognitive or behavioral consequence of attitudes with either individual strength‐related attributes or a composite index of those attributes. If a set of strength‐related attributes all relate in the same way to specific cognitive or behavioral consequences of attitudes, little is lost by combining the various attributes into an index. But evidence that diVerent attributes relate in diVerent ways to a cognitive or behavioral consequence would suggest that composite indices obscure the true structure and function of strength‐related attitude attributes. We next review evidence of this sort.

A. EASE OF RETRIEVAL EFFECTS ON IMPORTANCE, INTENSITY, AND CERTAINTY Haddock et al. (1996, 1999) and Wa¨nke, Bless, and Biller (1996) explored the impact of the ease of information retrieval on reports of attitude importance, intensity, and certainty. Specifically, these researchers manipulated the experienced ease of producing attitude‐congruent or attitude‐incongruent arguments: some respondents were asked to do a diYcult task (to list seven arguments supportive of or opposed to a particular policy), whereas others were asked to do an easier version of the same task (list only three arguments). Haddock et al. (1996, 1999) and Wa¨nke et al. (1996) expected that the experienced diYculty of generating arguments would influence respondents’ perceptions of attitude importance, intensity, and certainty. Having found it very diYcult to generate seven arguments consistent with their own opinion, people might reason, ‘‘If I had a strong opinion on this issue, I ought to have an easy time generating facts to back up my opinion. But because it was tough for me, maybe I don’t feel all that strongly on this issue, maybe I’m

42

PENNY S. VISSER et al.

not very certain about where I stand, and maybe the issue isn’t very important to me.’’ But if people find it easy to generate three supportive arguments, there would be no reason for self‐doubt in these regards. Likewise, the experience of easily generating three arguments challenging their own viewpoints might lead people to doubt the validity of their own opinions, thereby reducing perceived confidence, intensity, and importance. But if people have diYculty generating seven counter‐attitudinal arguments, they again have no reason for self‐doubt. To test these hypotheses, Haddock et al. (1996, 1999) averaged together measures of certainty, intensity, and importance to yield a single composite. This composite had an alpha reliability of .91, which was consistent with Haddock et al.’s (1999) finding from an exploratory factor analysis that the three dimensions loaded on a single factor. Further, this composite measure varied as expected according to the ease of argument generation. People who generated three attitude‐supportive arguments had higher composite scores than people who generated seven attitude‐supportive arguments, whereas people who generated seven counterattitudinal arguments had higher composite scores than people who generated only three counterattitudinal arguments. However, an interesting pattern emerged in Haddock et al.’s (1999) study when the three strength‐related attitude attributes were analyzed separately. Although the expected eVects appeared significantly for certainty and intensity ratings (Wa¨nke et al., 1996), importance ratings did not manifest the expected eVect of the argument generation manipulation significantly in either of the two studies. One possible explanation for the failure of importance to manifest the same eVects evident in certainty and intensity is that the latter two dimensions may have been measured more reliably than was the former. However, in Haddock et al.’s (1996, 1999) studies, each attribute was measured by two questions, and all six questions employed the same seven‐point ratings scale, ranging from ‘‘not at all’’ to ‘‘very.’’ Therefore, it seems unlikely that notable diVerences between the dimensions in measurement reliability were present. Interestingly, use of the same response scale for all items raises the possibility that the coeYcient alpha for the composite measure may have been inflated due to correlated measurement error shared across the six items (Brady, 1985; Green & Citrin, 1994). This study therefore oVers further reason to draw a distinction among strength‐related attitude attributes. Certainty and intensity ratings were identically influenced by the argument generation manipulation, which is consistent with the claim that they reflect a single underlying construct. However, the evidence suggesting that importance was not influenced by the manipulation indicates that it is a distinct construct. Collapsing the

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

43

strength‐related attributes into an index masked an interesting finding—that importance responded to a manipulation diVerently than did the other attributes.

B. EFFECTS OF IMPORTANCE, INTENSITY, AND CERTAINTY ON RESISTANCE TO ATTITUDE CHANGE AND OF ATTITUDE STABILITY Another study illustrating the same sort of disparity among constituents in their consequences was reported by Bassili (1996b). His composite of importance, knowledge, certainty, intensity, and other variables failed to predict attitudes’ resistance to change and stability over time in a series of tests. But when certainty, importance, and intensity were treated as separate and distinct predictors of resistance to change and persistence over time, all three variables had statistically significant eVects in the expected direction. That is, higher importance, intensity, and certainty were associated with more resistance to change and more stability over time, just as other studies had previously shown for importance and certainty individually (importance: Fine, 1957; Gorn, 1975; Zuwerink & Devine, 1996; certainty: Marks & Kamins, 1988; Swann & Ely, 1984; Swann, Pelham, & Chidester, 1988). Thus, creating a composite again clouded the appearance of eVects that were apparent when treating the constituents separately.

C. MODERATORS OF RESPONSE EFFECTS IN ATTITUDE MEASUREMENT A final set of studies explored whether eVects of the order in which questions are asked and the wordings of those questions are moderated by strength‐ related attitude features. Many scholars have presumed that such response eVects are most likely to appear in reports of weak attitudes (Cantril, 1944; Converse, 1974; Payne, 1951). The basis of this argument is the notion that question changes may alter people’s perceptions of their attitudes. If people’s internal psychological cues revealing their attitudes are weak, then those cues might be easily overwhelmed by aspects of the question encouraging certain responses. In one test of this notion, Krosnick and Schuman (1988) analyzed 27 experiments conducted in national surveys. These surveys included question manipulations known to cause response eVects. Surprisingly, measures of attitude importance, intensity, and certainty did not reliably predict the magnitude of the impact of most question wording, format, and order

44

PENNY S. VISSER et al.

manipulations on responses. In a series of similar experiments, Bishop (1990) found evidence supporting the same conclusion. However, for one type of response eVect, moderation did reliably appear. This response eVect involves oVering or omitting a middle alternative (e.g., ‘‘keep things as they are now’’) between two polar opposite viewpoints (e.g., ‘‘make divorce laws stricter than they are now’’ and ‘‘make divorce laws less strict than they are now’’). Krosnick and Schuman (1988) and Bishop (1990) found that people for whom attitudes were highly important or intense were relatively immune to whether the middle alternative was oVered or omitted in such questions. In contrast, people whose attitudes were low in importance or intensity were especially likely to be attracted to the middle alternative when oVered. Lavine et al. (1998) later presented evidence suggesting that a diVerent type of response eVect might be regulated by strength‐related attributes as well. In their study, people were first asked a series of context questions and a target attitude was then measured. The context questions were designed to promote either liberal or conservative responses to the target attitude question. Lavine et al. (1998) found that interattitudinal embeddedness—a construct not previously investigated—moderated the question order eVect. People who perceived strong implicational relations between the target attitude, and attitudes toward other social and political attitudes were less susceptible to the context manipulation than were people who perceived weaker interattitudinal links. More strikingly, an average of several strength‐related attitude attributes (including importance, certainty, and extremity) regulated the question order eVect as well. Lavine et al. (1998, p. 369) argued that this moderation appeared because their measure of embeddedness and aggregation of strength‐related constructs were more ‘‘broad in bandwidth’’ than were the measures used by Krosnick and Schuman (1988) and Bishop (1990). One can view Lavine et al.’s (1998) findings as validating the claim that when an attitude is weak, reports of it are especially likely to be susceptible to all sorts of response eVects. But another possibility is that Lavine et al.’s (1998) results are confined to the specific response eVect they examined—a particular type of question order manipulation. Lavine et al.’s (1998) target attitudes were toward welfare and the rights of people accused of committing crimes. Before expressing their views on these issues, participants first answered a series of questions designed to bring to mind considerations that supported either liberal or conservative stands on the target issues. These kinds of question order eVects seem likely to have occurred because the weights attached to various considerations in deriving the target attitudes were altered by the context questions (Tourangeau & Rasinski, 1988; Tourangeau et al., 1989). That would suggest that this particular type of

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

45

question order manipulation leads to real changes, at least temporarily, in people’s evaluations of the target attitude object. And attitudes that are tightly linked to a set of related attitudes may be invulnerable to this type of response eVect because people may recognize that changing such an attitude could create logical coherence problems in their system of related attitudes. More required movement in other cognitive elements should be a direct inhibitor of attitude change. The individual strength‐related attributes that composed Lavine et al.’s (1998) composite included importance, elaboration, certainty, extremity, ambivalence, and intensity. Each of these attributes has been shown to be correlated with resistance to attitude change. Thus, as expected, when each attribute was examined individually, Lavine et al. (1998) found strong and significant context eVects for both target issues among participants who were highly ambivalent, low in prior elaboration, low in attitude certainty, low in attitude extremity, and low in intensity. In contrast, context eVects were rarely significant for participants high on each of these strength‐related attributes. Nonetheless, with one exception, the overall interactions between each strength‐related attribute and the context manipulation were not statistically significant. As we reported above, however, aggregating these six attributes did yield a significant interaction between attitude strength and the context manipulation. This may suggest none of these attributes provided as reliable an assessment of attitudinal embeddedness as the direct measure that Lavine et al. (1998) developed, and the joint presence of importance and certainty and prior elaboration and the other constructs included in the composite also may have provided a better approximation of embeddedness. This is consistent with the notion that these attributes were each more indirect measures of the moderating construct at work. Other question wording and order manipulations may operate through diVerent processes, and may not be moderated by attitude embeddedness. For example, some question order eVects seem likely to be driven in large part by self‐presentational concerns rather than true changes in people’s attitudes (Schuman & Presser, 1981). People prefer to appear consistent in their attitudes and beliefs, so responses to initial questions may constrain subsequent responses as people strive to maintain consistency across a set of responses. Prior questions can also make salient particular norms to which people may want to appear to adhere. For example, people are more likely to say that reporters from communist countries should be permitted to come to the United States and file reports in their home countries if this question is directly preceded by a question about whether US reporters should be permitted to go into communist countries and file reports about that country back in the United States (Hyman & Sheatsley, 1950; Schuman & Presser, 1981). Apparently, answering the initial question about US reporters

46

PENNY S. VISSER et al.

invokes a norm of reciprocity or evenhandedness, making it uncomfortable for people to say that reporters from communist countries should be treated diVerently than US reporters. Question order eVects of this sort seem less likely to be moderated by interattitudinal embeddedness. Similarly, some question wording eVects may be driven not so much by real changes in people’s attitudes, but by changes in the perceived extremity of the response options. For example, people may perceive ‘‘forbidding’’ a particular behavior to be more extreme than ‘‘not allowing’’ that same behavior (Hippler & Schwarz, 1986). If this type of question‐wording eVect is driven by changes in the perceived extremity of the response options, it seems unlikely to be moderated by embeddedness. Instead, it may be moderated by attitude extremity or intensity, with people who hold moderate or less intense attitudes more likely to shy away from endorsing the extremely worded response. Interestingly, the response eVects investigated by Krosnick and Schuman (1988) and Bishop (1990) also seem unlikely to have resulted from the momentary alteration of the weights attached to specific attitude‐relevant considerations that presumably yielded Lavine et al.’s (1998) question‐order eVect (Krosnick, 1991). Therefore, interattitudinal embeddedness may not have been a relevant moderator, but other strength‐related attitude attributes may have been. All of this suggests that diVerent response eVects may be moderated by diVerent attitude attributes, depending upon the particular cognitive mechanism that is responsible for the response eVect. So the mere existence of a response eVect does not necessarily mean that the attitudes being measured are weak. And the particular attitude attribute that regulates one response eVect may not regulate another response eVect. Bassili and Krosnick (2000) set out to investigate these possibilities using a wide range of strength‐related attributes and a range of diVerent sorts of response eVects. In Bassili and Krosnick’s (2000) study, a random sample of University of Toronto students was interviewed by telephone to assess four types of response eVects. The first was a question order eVect involving abortion items. All respondents were asked whether a married woman who does not want any more children should be permitted to obtain a legal abortion. Some respondents were simply asked this question, whereas others were first asked whether it should be possible to obtain a legal abortion if there is a strong chance of a serious defect in the baby. Asking the birth defect question first renders people less likely to support the married woman’s right to a legal abortion (Schuman & Presser, 1981). Second, the impact of oVering or omitting a middle alternative was observed by measuring respondents’ attitudes regarding whether penalties for marijuana use should be made stricter or less strict. Some respondents were oVered a middle alternative (i.e., ‘‘keep laws about the same’’) and

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

47

others were not. People are more likely to select a middle option when it is presented to them than to volunteer this response when the response option is not presented explicitly (Schuman & Presser, 1981). Third, acquiescence response bias was measured by asking respondents whether they agreed or disagreed with one of two opposite statements about the primary cause of crime and lawlessness in America: ‘‘individuals’’ or ‘‘social conditions.’’ For respondents given one of the statements, blaming ‘‘individuals’’ for crime and lawlessness should have resulted in selecting the ‘‘agree’’ response option. For respondents given the other statement, blaming ‘‘social conditions’’ should also have resulted in an ‘‘agree’’ response. Because some people agree with just about any statement, regardless of its content, this design permitted assessment of the extent of such acquiescence (Schuman & Presser, 1981). Finally, to measure the eVects of tone of wording, respondents were asked whether antidemocracy speeches should be either allowed or forbidden. People are less likely to endorse ‘‘forbidding’’ a policy than to say it should not be allowed (Schuman & Presser, 1981). For each of the four experiments, seven strength‐related attitude attributes were assessed, including importance, knowledge, certainty, intensity, the perceived likelihood that one’s attitude will change over time, extremity, and accessibility (via response latencies). To test Lavine et al.’s (1998) claim about broadband coverage, a meta‐attitudinal aggregate was computed by averaging the measures of importance, knowledge, certainty, intensity, and perceived likelihood of change. No single strength‐related property regulated all four of the response eVects. Instead, one or two single attributes regulated each eVect (Fig. 4). For example, extremity regulated the question order eVect and acquiescence; certainty and knowledge regulated the middle alternative eVect; and intensity regulated the tone of wording eVect. The meta‐attitudinal aggregate only proved reliable in moderating response eVects for two of the four questions (question order and middle alternative); and in both of these cases, an individual attitude attribute evidenced just as much moderation. In each case, stronger attitudes manifested weaker eVects of question form, wording, or order, but diVerent strength‐related attributes moderated the various response eVects. These findings challenge the notion that broadband coverage is necessary to observe moderation by strength‐related attitude properties. Also, reaction time measures of accessibility did not predict the magnitude of any of the response eVects, disconfirming Bassili’s (1996a) implication that operative measures such as this will succeed in specifying attitude dynamics when meta‐attitudinal measures fail. Because individual strength‐related attributes predicted particular response eVects, and diVerent attributes regulated the various response eVects,

48

PENNY S. VISSER et al.

Fig. 4. Moderators of response eVects.

these results reinforce the general notion that strength‐related attributes are best viewed as distinct constructs. Furthermore, the failure of attribute aggregations to moderate any better‐than‐individual constituent properties challenges Lavine et al.’s (1998) claim that treating multiple attributes as surface manifestations of underlying constructs will improve the validity of results obtained. Indeed, in the case of the tone‐of‐wording eVect, individual properties successfully moderated eVects, but combining them with other properties into an index masked the eVects.

IX. General Discussion In their chapter in The Handbook of Social Psychology on ‘‘Attitude Structure and Function,’’ Eagly and Chaiken (1998, p. 291) reviewed the existing literature on the latent structure of strength‐related attitude properties and noted that factor analytic studies had suggested a distinction between cognitive dimensions of attitude strength and aVective dimensions. But these authors noted as well that ‘‘although several findings have . . . suggested the utility of distinguishing cognitive from aVective aspects of attitude strength, subsequent work may well yield other useful distinctions beyond, or within, these two broad dimensions (p. 291).’’ Eagly and Chaiken (1998) called upon researchers to ‘‘go beyond the question of strength’s dimensionality to the question of whether such distinctions matter. If all aspects of attitude strength produced the very same eVects, the theoretical importance

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

49

of distinguishing types of strength would be hollow (p. 291).’’ The current review was done in the spirit of Eagly and Chaiken’s (1998) recommendation, confirming their expectation. We have seen repeatedly that individual strength‐related attributes have distinct origins and diVerent eVects on thinking and behavior. More specifically, we have seen that antecedents of one attribute were often unrelated to other attributes. For example, perceiving that important others cared deeply about an issue was associated with greater attitude importance, but it was unrelated to attitude‐relevant knowledge. Similarly, in several studies, repeated attitude expression increased attitude accessibility, but repeated expression did not aVect attitude importance, nor did it aVect the certainty with which people held their attitudes. In contrast, repeated exposure to attitude‐relevant information increased attitude certainty but had no impact on attitude accessibility. And the ease or diYculty of generating attitude‐relevant arguments influenced attitude certainty and attitude intensity but was unrelated to attitude importance. In those few cases where an antecedent was related to two strength‐related attributes, diVerent mediators were shown to be at work. Finally, pairs of attributes fluctuated independently over time, reinforcing the notion that they have distinct antecedents. Strength‐related attributes also related to cognitive and behavioral outcomes in distinct ways. In many instances, pairs of attributes exerted uncoupled eVects. For example, importance was associated with perceived attitude polarization in response to conflicting empirical evidence, and with negative aVective reactions when an attitude was threatened, whereas knowledge was unrelated to both of these phenomena. In contrast, knowledge was positively associated with the hostile media bias, whereas importance was unrelated to this phenomenon. In addition, importance (but not certainty) predicted whether people turned out to vote on election day, whereas certainty (but not importance) predicted the degree to which people found a nonpreferred presidential candidate acceptable. And attitude certainty regulated the correspondence between attitudes and a deliberative behavior but attitude accessibility did not. In other cases, two attributes related in opposite ways to an outcome. For example, attitude‐relevant knowledge was positively associated with attitude moderation in the face of conflicting evidence, whereas attitude importance was negatively associated with moderation. And in still other cases, attributes interacted to predict a cognitive or behavioral outcome. For example, the combination of high importance and high knowledge was associated with an especially pronounced surge of attitude‐expressive behavior. This evidence sharply conflicts with the notion that two or more strength‐ related attributes can be treated as interchangeable manifestations of a more

50

PENNY S. VISSER et al.

general underlying construct. To the contrary, the evidence we have reviewed suggests that doing so will yield inaccurate characterizations of strength‐related processes, obscuring meaningful distinctions in the operation of the various strength‐related attributes. Instead, the extensive body of evidence that we have reviewed supports a view of strength‐related attributes as distinct constructs in their own right. A. META‐ATTITUDINAL INDICES OF ATTITUDE STRENGTH One particularly popular view regarding overlap among strength‐related attributes involves attributes that reflect people’s subjective judgments about their attitudes, which Basilli (1996a) dubbed ‘‘meta‐attitudinal’’ indices of attitude strength (e.g., importance, certainty, perceived volume of attitude‐ relevant knowledge stored in memory). Some scholars have suggested that these attributes are all constructions, built from blurry introspective glances, simply reflecting the extent to which an attitude seems to its holder to be mushy or firm (Bassili, 1996a; Haddock et al., 1996, 1999). Regardless of whether people are asked about the importance they attach to an attitude or their confidence in holding it or how much relevant information they possess or how strong their feelings are about the object, says this view, people look to internal psychological cues for any vague sense of attitudinal crystallization and use that sense to derive an answer to whatever question has been posed. If this is true, then self‐reports of such features are all manifestations of that vague introspective impression. The evidence that we have reviewed poses a strong challenge to that view. Much of this evidence involved explicit comparisons of meta‐attitudinal measures of importance and certainty, and we saw consistent evidence of unique antecedents and divergent cognitive and behavioral consequences. Although there is little doubt that these subjective judgments, like virtually all others, can sometimes be momentarily influenced by salient contextual features (Haddock et al., 1996, 1999; Wa¨nke et al., 1996), the results that we have reviewed indicate that people’s subjective judgments about their attitudes largely reflect psychologically meaningful variability in strength‐ related attitude properties.

B. NEVER CREATE COMPOSITE INDICES? Does all of this suggest that researchers should never create composite indices of strength‐related attitude factors by combining measures of diVerent strength‐related attributes? Not necessarily. Comprehensive comparisons

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

51

have not been conducted for every pair of strength‐related attributes, so it remains possible that some attributes do in fact arise from largely overlapping antecedents and set into motion essentially the same consequences via the same mechanisms. Indeed, commonalities were occasionally observed in the evidence reviewed here. For example, in two separate studies, manipulating the ease with which people generated pro‐ and counterattitudinal arguments aVected attitude certainty and intensity in comparable ways (Haddock et al., 1999; Wa¨nke et al., 1996). And Visser et al. (2003) found that importance and certainty were related to some of the same outcomes. Both regulated the impact of particular issues on people’s candidate preferences, for example, and both predicted their intentions to turn out to vote on election day. Thus, continuing the quest for parsimony in this domain may well be sensible and fruitful. It seems quite possible that combining sets of strength‐related attributes may sometimes be a prudent strategy. For example, there are only a limited number of distinct cognitive processes by which people can resist attitude change (e.g., generating counterarguments, bolstering one’s original attitude, derogating the source of a persuasive message). Because strength‐ related attributes appear to outnumber distinct processes of resistance, there is likely to be overlap in the mechanisms by which diVerent strength‐related attributes lead to resistance. It may therefore be possible to identify clusters of strength‐related attributes that lead to the same outcome through the same processes. This approach may yield a taxonomy of strength‐related attributes based not simply on covariation (the criterion that has been used in much prior work), but based instead on the impact that they have and on the mechanisms by which they operate. Thus, our position is not that researchers can never justify creating composite indices of strength‐related attributes. But correlations among those attributes do not provide adequate justification for doing so. Having said that, it is worth noting that although pair‐wise comparisons of all strength‐related attributes have not yet been conducted, the existing evidence seems to suggest that overlap in antecedents and consequences of attributes is the exception rather than the rule. As we have seen, even attributes that are quite strongly correlated and that reliably loaded on the same factor in exploratory factor analyses nonetheless exhibited very diVerent eVects on thought and behavior. It seems sensible, therefore, for researchers to adopt a starting assumption that the various attributes are distinct constructs until clear and compelling evidence of their overlap is documented. Importantly, this also suggests that we should be cautious when interpreting the results of past investigations of strength‐related processes that involved composite indices. The evidence we have reviewed suggests that such results may obscure sharp divergences in the consequences of the

52

PENNY S. VISSER et al.

individual attributes from which the composites were constructed. And indeed, this danger has been demonstrated in some of the investigations we have reviewed. For example, Pomerantz et al. (1995) combined measures of importance, knowledge, value‐relevance, and centrality to the self to yield an index of attitude strength and found that this index was not associated with attitude change in the Lord et al. (1979) paradigm. Visser et al. (2004) replicated this null result when they combined measures of importance and knowledge into a single index of attitude strength and used it to predict attitude change within the Lord et al. (1979) paradigm. But as we reported earlier, when Visser et al. (2004) decomposed the composite index, they found clear evidence that knowledge was unrelated to attitude change, whereas importance was strongly related to it. Similarly, Visser et al. (2003) demonstrated that whereas importance increased as the result of the national debate on global warming, a composite index constructed from measures of importance and certainty registered no change. This set of evidence suggests that claims about the antecedents or consequences of composite indices of attitude strength should be interpreted with caution (Bassili, 1996a; Bassili & Roy, 1998; Eagly et al., 2000; Hodson et al., 2001; Holland et al., 2002; Pomerantz et al., 1995; Prislin, 1996; Theodorakis, 1994; Thompson & Zanna, 1995; Verplanken, 1989, 1991).

C. FURTHER DISTINCTIONS? As it is premature to conclude that various pairs or sets of strength‐related attributes reflect a common underlying factor, it may also be premature to conclude that the various strength‐related attributes we have examined are themselves unidimensional constructs. And in fact, some scholars have challenged this notion. 1. Dimensionality of Attitude Importance Although generally treated as a unitary construct, attitude importance may be multidimensional, with multiple functional bases. Attitude importance that arises from the recognition of a connection between an attitude object and one’s core values may be distinct in terms of its phenomenology and its consequences from attitude importance that arises from the perception of a link between an attitude object and one’s material interests. And both may be distinct from attitude importance that arises from the perception that one’s reference groups or individuals view an attitude as important. Each may inspire discrete motivations: to protect the attitude that expresses one’s

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

53

core values, to hold the correct attitude toward the object that impinges on one’s self‐interest, and to remain in step with important others with regard to the attitudes they deem important. Johnson and Eagly (1989) explored such a distinction between value‐ relevant involvement (defined as the activation of attitudes that are linked to a person’s important values) and outcome‐relevant involvement, (defined as the activation of attitudes toward issues or objects that are relevant to a person’s currently important goals or outcomes) in meta‐analysis. They concluded that value relevance led to increased resistance to attitude change (particularly when the arguments were strong), whereas outcome relevance led to more attitude change when arguments were strong and less attitude change when arguments were weak. These results suggest that value relevance may have led to biased processing of the persuasive messages, whereas outcome relevance inspired objective processing (Petty & Cacioppo, 1986). There are a number of reasons to hesitate before accepting the conclusions drawn from this evidence, however. Several potentially significant confounds in the design of this investigation make it diYcult to know what to make of these findings. For example, almost all of the studies of outcome relevance were experimental and involved direct manipulations of outcome relevance, whereas the studies of value relevance were correlational (Petty & Cacioppo, 1990). Furthermore, most of the value‐relevance studies did not actually include measures of value relevance—instead, many of these studies were judged by the investigators to be value relevant based on the topic addressed in the study. Third, unlike the outcome‐relevance studies, almost none of the value‐relevance studies implemented an argument quality manipulation. Instead, participants in a separate study rated the strength of the arguments post hoc. This makes it diYcult to know with confidence what impact argument quality had, because a post hoc determination of argument may be susceptible to confounding with other aspects of the message such as satirical content, message discrepancy, and source credibility. Nevertheless, Lampron, Krosnick, ShaeVer, Petty, and See (2003) implemented an experiment that measured and manipulated both outcome relevance and value relevance and found results consistent with Johnson and Eagly’s (1989) conclusions. Because value relevance and outcome relevance as defined by Johnson and Eagly (1989) strongly resemble two of the primary causal antecedents of attitude importance, these results suggest that the route by which one comes to attach importance to an attitude may influence the psychological nature of attitude importance. And the nature of the importance one attaches to an attitude may determine its specific consequences. This raises the possibility that importance is a multidimensional construct. However, in an experimental investigation of this issue, Boninger et al. (1995a) found evidence inconsistent with the multidimensional view of

54

PENNY S. VISSER et al.

attitude importance. They found that manipulations of one of the causal antecedents of importance reverberated through participants’ cognitive structures, impacting other antecedents of importance. Specifically, Boninger et al. (1995a) found that increasing the degree to which an attitude impinged on participants’ material interests also led them to view the attitude as more closely linked to their core values. This suggests that the causal antecedents of attitude importance are related to one another and that changes in one can result in changes in others. This evidence challenges the notion that the causal underpinnings of attitude importance are discrete and lead cleanly to distinct ‘‘types’’ of importance. Instead, an attitude that is outcome relevant may also come to be seen as value relevant as well. Nevertheless, conclusions on this matter should be drawn with utmost caution, given the dearth of empirical evidence. Additional research addressing this issue clearly seems warranted. And here, too, we contend that the critical issue to be explored is not the factor structure of these various types of attitude importance but rather, whether these types of importance do in fact arise from distinct causal antecedents and whether they set into motion diVerent sorts of cognitive and behavioral consequences. To the extent that they do, diVerentiating among them would clearly seem warranted. 2. DiVerentiating Meta‐Attitudinal from Operative Measures Bassili (1996a) distinguished meta‐attitudinal from operative measures of strength‐related attitude features, and he presumed that each dimension is inherently either meta‐attitudinal or operative. This is certainly true for dimensions like importance and certainty, which are defined as a person’s perceptions of his or her attitude. But other dimensions, such as knowledge volume, accessibility, and ambivalence can be measured either meta‐attitudinally or operatively. That is, we can measure knowledge volume by asking a person how much he or she knows about an object, or we can ask him or her to list all such knowledge and count up the pieces, without ever telling the person we plan to do such counting. Likewise, we can measure response latency without ever saying we’re doing so, or we can ask a person how quickly his or her attitude comes to mind. And we can measure the conflict between the extents of people’s favorable and unfavorable reactions to an object, or we can ask them how ambivalent they feel toward it. Krosnick et al. (1993) reported confirmatory factor analysis tests that indicated meta‐attitudinal, and operative measures of knowledge volume tapped the same construct. But newer work by Holbrook (2003) showed that meta‐attitudinal and operative measures of knowledge volume, accessibility, and ambivalence are all associated with distinct cognitive and behavioral consequences that fit sensibly with faithful conceptualizations of these

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

55

measures. For both knowledge volume and ambivalence, the relation between the meta‐psychological and operative measures of the construct was moderate in size, but for accessibility, the relation was never significantly diVerent from zero, suggesting that people may not be aware of or able to accurately report the accessibility of their own attitudes and that their perceptions of accessibility come from a diVerent source. Furthermore, the eVects of the six measures (meta‐attitudinal and operative measures of knowledge volume, accessibility, and ambivalence) on the false consensus eVect, hostile media bias, the similarity‐attraction eVect, resistance to attitude change, and attitude–behavior consistency were never identical, suggesting that the meta‐attitudinal and operative measures tapped distinct constructs. This notion is reinforced by the work of Newby‐Clark and his colleagues (Newby‐Clark, McGregor, & Zanna, 2002). In several studies, they found modest correlations between meta‐attitudinal and operative measures of ambivalence (ranging from .18 to .39). They also found that the magnitude of this relation was moderated by the simultaneous accessibility of positive and negative reactions to an issue: when positive and negative reactions both came to mind quickly and eVortlessly, operative and meta‐attitudinal measures of ambivalence were quite strongly related, but when one or both reactions required time and eVort to retrieve, the two constructs were essentially independent. All of these findings suggest that it is worthwhile to abandon the notion that the meta‐attitudinal and operative measures of strength‐related attitude features are interchangeable and instead to build separate theories to account for the causes and consequences of perceived and actual knowledge volume, perceived and actual accessibility, and perceived and actual ambivalence.

D. ATTITUDE EXTREMITY: A SPECIAL CASE? Attitude extremity is unique among strength‐related dimensions, because whereas the other dimensions refer to attributes of or judgments about the attitude, extremity refers to the attitude itself (where, along a bipolar continuum, the attitude falls). Thus, extremity is the only dimension that is not independent of the content of the attitude: attitudes that diVer in extremity are, by definition, diVerent attitudes. Abelson (1995) has argued that extremity is conceptually rich, conveying: (1) the intensity of feeling a person experiences with regard to the attitude object, (2) the degree to which a person holds an unqualified position, (3) the lengths a person believes his or her group should go in defending its position, and (4) the lengths to which one would go in defending the position.

56

PENNY S. VISSER et al.

One might therefore wonder whether extremity may subsume some or all of the other strength‐related dimensions of attitude strength, thus providing a reasonable index of overall attitude strength. And indeed, some researchers have presumed this to be the case: attitude extremity has sometimes been used as an omnibus measure of attitude strength (Edwards & Smith, 1996). Of course, this line of thinking represents a return to the single‐factor view of attitude strength, and the evidence we reviewed suggests that no single dimension can adequately capture the dynamics of attitude strength. More specifically, a single dimension cannot account for the uncoupled, oppositely valenced, or interactive eVects that we routinely observed when pairs of strength‐related attitude attributes were examined simultaneously. Thus, attitude extremity appears unlikely to subsume any of the pairs of attributes that we have examined. Nonetheless, controlling for attitude extremity may well be a sensible strategy when exploring the causes and consequences of other strength‐ related attributes. This would seem to be especially important when the strength‐related attribute of interest is known to relate to attitude extremity. Attitudinal ambivalence, for example, is necessarily negatively correlated with extremity. When exploring the implications of ambivalence for cognitive and behavioral consequences, then, taking into account the eVect of attitude extremity would seem to be essential (Thompson et al., 1995). And in fact, in several of the investigations that we have reviewed, the original authors did replicate all of their primary analyses with attitude extremity included as a control variable (Berger, 1992; Holbrook, 2002; Visser et al., 2003, 2004), and this never meaningfully altered the results. This provides further evidence that attitude extremity cannot account for the associations between the various strength‐related attributes and the attitude eVects that we have reviewed. E. IMPLICATIONS Resolving the debate over the underlying structure of strength‐related attributes is not just a matter of ‘‘intellectual aesthetics.’’ That is, our motivation for the current review is not simply to clear up a technical dispute about how attitude strength ought to be measured. To the contrary, achieving clarity regarding the conceptualization of attitude strength is of fundamental importance for both basic and applied attitude research. 1. Conceptual Implications Assumptions about the underlying structure of strength‐related attitude attributes set the agenda for research in this domain. If clusters of

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

57

strength‐related attributes are assumed to reflect a small number of more general constructs, the primary objective within the attitude literature becomes identifying these general dimensions and exploring their origins and consequences. This assumption further suggests that there is no need to continue fine‐grained explorations of individual strength‐related attributes (Krosnick, 1988a,b; Wood, 1982). Instead, disparate lines of such research can be consolidated according to the assumed latent structure. A multidimensional view of attitude strength charts quite a diVerent course for attitude researchers in the future. Our primary objective would be to clarify the workings of each strength‐related attribute, alone and in combination. We would pay close attention to the bases of attitude strength and develop refined predictions regarding specific attitude eVects given the nature of particular strength‐related attitude attributes. The work reviewed here provides strong support for this latter conceptualization of attitude strength and illustrates the potential value of this approach. Individual lines of research fill in some details in the psychological portraits of various strength‐related attributes. For example, attitude importance appears to be, at its core, a motivation to protect and use one’s attitude, whereas knowledge appears to reflect a reservoir of ability, facilitating behavioral strategizing, critical analysis of new information, and more. We have also seen some illustrations of what a construct is not. For example, knowledge seems not to attenuate cognitive biases such as the false consensus eVect, whereas accessibility seems not to be a regulator of the magnitude of response eVects in attitude measurement. Lastly, we have seen evidence regarding the mechanisms by which eVects occur. For example, self‐interest and value relevance aVect knowledge only by inspiring increased importance. We hope that cataloging these sorts of findings along with others will eventually lead to a full and rich account of the origins and consequences of attitude strength. Future research on this topic might borrow two of the approaches employed here and apply them in new contexts. One design is that used by Bizer and Krosnick (2001): implement an experimental manipulation narrowly designed to alter just one strength‐related attribute (e.g., importance), and observe the consequences that follow for that attribute and various others. Other studies might employ the technique used by Bassili and Krosnick (2000) and Visser et al. (2003, 2004) whereby multiple dimensions are measured, and multivariate analysis is used to isolate their independent and interactive eVects. When many studies employing these and other approaches document the full range of causes and eVects of various attributes, we will be in a good position to build a general, integrative theory of attitude strength.

58

PENNY S. VISSER et al.

2. Practical Implications Clarifying the structure of strength‐related attributes also has important practical implications. Consider just one real‐world context in which the assumptions we make about the structure of attitude strength are likely to be tremendously important: public health. Public health oYcials have increasingly come to recognize that many of the leading causes of death in the United States could be drastically reduced if Americans would make a few simple changes in their behavior. In fact, an investigation recently published in the Journal of the American Medical Association concluded that approximately half of the deaths in the United States can be attributed to a small number of preventable behaviors such as smoking, inactivity, poor diet, and alcohol consumption (Mokdad et al., 2004). Because of this, public heath advocates have increasingly turned to the social and behavioral sciences for insight into behavior modification. In some cases, people already possess positive attitudes toward healthy behaviors and negative attitudes toward unhealthy behaviors. But these health‐positive attitudes do not always manifest themselves in the relevant health behaviors (Fisher & Fisher, 1992). The challenge for public health advocates, then, involves strengthening such existing attitudes so that they motivate and guide behavior and shape the way new information is processed, as well as resist change and persist over time. If many strength‐ related attributes reflect a common underlying construct, distinguishing among them would be unnecessary: interventions that brought about increases in any one attribute would result in increases in the others, so public health advocates could focus their eVorts on the attribute that can most easily be modified. The evidence reviewed here indicates that increasing one strength‐related attribute will not necessarily increase others. This implies that public health advocates should broaden their eVorts and target many strength‐related attributes. Such advocates should be sensitive to diVerences in the ways in which strength‐related attributes—alone and in combination—lead most reliably to attitude–behavior correspondence. The case of AIDS in the United States provides an excellent illustration. Initially, public health oYcials assumed that if they could educate people about the disease and how to avoid it, the appropriate behaviors would follow (Helweg‐Larsen & Collins, 1997). So they launched a massive public education campaign to increase the amount of knowledge people had about the disease (Fisher & Fisher, 1992). And it was tremendously successful—surveys show that virtually all US adults now know what AIDS is, have some sense of how it is transmitted, and know what steps can be taken to avoid exposure (DiClemente, Forrest, Mickler, & Principal Site Investigators,

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

59

1990; Rogers, Singer, & Imperio, 1993). Yet such educational campaigns often yielded virtually no reliable eVects on behavior (Mann, Tarantola, & Netter, 1992). Knowledge, in and of itself, wasn’t suYcient to instigate attitude‐congruent behavior. The research we reviewed suggests solutions to this problem. By recognizing knowledge as one of many distinct strength‐related attributes, the multidimensional view acknowledges multiple avenues through which the attitude–behavior link can be strengthened. This view encourages public health advocates to focus not simply on increasing knowledge levels, but on increasing other strength‐related attributes as well. And it suggests that particular combinations of strength‐related attributes may be especially eVective. In two studies involving diVerent attitude objects and diVerent subject populations, the combination of possessing a great deal of knowledge about an attitude object and attaching importance to the attitude was associated with a surge of attitude‐expressive behavior far exceeding the impact of either attribute alone (Visser et al., 2004). Other strength‐related attributes may similarly interact to produce stronger attitude–behavior links than any of the strength‐related attributes produce in isolation. There are also cases in which promoting healthy behavior requires changing existing attitudes that are counterproductive to healthy living. The multidimensional view acknowledges many avenues by which attitudes can be weakened, thereby facilitating persuasion. By reducing some of the strength‐related attributes of people’s attitudes, public health advocates may find subsequent eVorts to induce attitude change more eVective. Furthermore, by understanding the processes by which particular strength‐ related attributes confer resistance to change, promoters of public health may be better able to tailor persuasive campaigns to be maximally eVective within their target audience.

F. CONCLUSION Attitude strength has been a focus of serious empirical interest among psychologists for decades, yet our understanding of this complex construct, its constituents, and the causal processes in which it plays a part is at an early stage. Much work remains to be done to illuminate the nature, structure, and function of attitude strength. As this research is conducted, it will do much to enhance the field’s account of how, when, and why attitudes are the powerfully consequential psychological forces that fascinated Allport (1935).

60

PENNY S. VISSER et al.

Acknowledgments The authors wish to thank Richard Petty, Marilynn Brewer, GiVord Weary, Bill von Hippel, Timothy Brock, and Philip Tetlock for their very helpful suggestions regarding this research. The authors also wish to thank Jamie Franco and Alodia Velasco for their help in preparation of the manuscript.

References Abelson, R. P. (1988). Conviction. American Psychologist, 43, 267–275. Abelson, R. P. (1995). Attitude extremity. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 25–41). Mahwah, NJ: Lawrence Erlbaum Associates. Allport, G. W. (1935). Attitudes. In C. A. Murchinson (Ed.), A handbook of social psychology (Vol. 2, pp. 798–844). Worcester, MA: Clark University Press. Alwin, D. F., & Krosnick, J. A. (1985). The measurement of values in surveys: A comparison of ratings and rankings. Public Opinion Quarterly, 49, 535–552. Andrews, F. M. (1984). Construct validity and error components of survey measures: A structural modeling approach. Public Opinion Quarterly, 48, 409–442. Apsler, R., & Sears, D. O. (1968). Warning, personal involvement, and attitude change. Journal of Personality and Social Psychology, 9, 162–166. Bass, A. R., & Rosen, H. (1969). Some potential moderator variables in attitude research. Educational and Psychological Measurement, 29, 331–348. Bassili, J. N. (1996a). Meta‐judgmental versus operative indexes of psychological attributes: The case of measures of attitude strength. Journal of Personality and Social Psychology, 71, 637–653. Bassili, J. N. (1996b). The how and why of response latency measurement in telephone surveys. In N. Schwarz and S. Sudman (Eds.), Answering questions: Methodology for determining cognitive and communicative processes in survey research (pp. 319–346). San Francisco, CA: Jossey‐Bass Inc. Publishers. Bassili, J. N., & Krosnick, J. A. (2000). Does attitude strength moderate susceptibility to response eVects? New evidence using response latency, attitude extremity, aggregate indices, and continuous measures. Political Psychology, 21, 107–132. Bassili, J. N., & Roy, J. P. (1998). On the representation of strong and weak attitudes about policy in memory. Political Psychology, 19, 669–681. Bem, D. J. (1972). Self‐perception theory. Advances in Experimental Social Psychology, 6, 150–157. Bentler, P. M. (1969). Semantic space is (approximately) bipolar. Journal of Psychology, 71, 33–40. Berent, M. K. (1990). Attitude importance and the recall of attitude‐relevant information. Columbus, Ohio: Ohio State University, Unpublished Master’s Thesis. Berent, M. K., & Krosnick, J. A. (1993). Attitude importance and selective exposure to attitude‐ relevant information. Columbus, Ohio: Ohio State University, Unpublished Manuscript. Berger, I. E. (1992). The nature of attitude accessibility and attitude confidence: A triangulated experiment. Journal of Consumer Psychology, 1, 103–123. Berger, I. E., & Mitchell, A. A. (1989). The eVect of advertising on attitude accessibility, attitude confidence, and the attitude‐behavior relationship. Journal of Consumer Research, 16(3), 269–279.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

61

Biek, M., Wood, W., & Chaiken, S. (1996). Working knowledge, cognitive processing, and attitudes: On the determinants of bias. Personality and Social Psychology Bulletin, 22, 547–556. Bishop, G. F. (1990). Issue involvement and response eVects in public opinion surveys. Public Opinion Quarterly, 54, 209–218. Bizer, G. Y., & Krosnick, J. A. (2001). Exploring the structure of strength‐related attitude features: The relation between attitude importance and attitude accessibility. Journal of Personality and Social Psychology, 81, 566–586. Boninger, D. S., Krosnick, J. A., & Berent, M. K. (1995a). Origins of attitude importance: Self‐ interest, social identification, and value relevance. Journal of Personality and Social Psychology, 68, 61–80. Boninger, D. S., Krosnick, J. A., Berent, M. K., & Fabrigar, L. R. (1995b). The causes and consequences of attitude importance. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 159–190). Mahwah, NJ: Lawrence Erlbaum Associates. Borgida, E., & Howard‐Pitney, B. (1983). Personal involvement and the robustness of perceptual salience eVects. Journal of Personality and Social Psychology, 45, 560–570. Brady, H. E. (1985). Statistical consistency and hypothesis testing for nonmetric multidimensional scaling. Psychometrika, 50, 509–537. Brim, O. G. (1955). Attitude content‐intensity and probability expectations. American Sociological Review, 20, 68–76. Brown, D. W. (1974). Adolescent attitudes and lawful behavior. Public Opinion Quarterly, 38, 98–106. Budd, R. J. (1986). Predicting cigarette use: The need to incorporate measures of salience in the theory of reasoned action. Journal of Applied Social Psychology, 16, 633–685. Byrne, D. E., London, O., & GriYtt, W. (1968). The eVect of topic importance and attitude similarity‐dissimilarity on attraction in an intrastranger design. Psychonomic Science, 11, 303–304. Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait‐multimethod matrix. Psychological Bulletin, 56, 81–105. Cantril, H. (1944). Gauging public opinion. Princeton, NJ: Princeton University Press. Cantril, H. (1946). The intensity of attitude. Journal of Abnormal and Social Psychology, 41, 1–12. Celsi, R. L., & Olson, J. C. (1988). The role of involvement in attention and comprehension processes. Journal of Consumer Research, 15, 210–224. Chaiken, S., Pomerantz, E. M., & Giner‐Sorolla, R. (1995). Structural consistency and attitude strength. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 387–412). Mahwah, NJ: Lawrence Erlbaum Associates. Clore, G. L., & Baldridge, B. (1968). Interpersonal attraction: The role of agreement and topic interest. Journal of Personality and Social Psychology, 9, 340–346. Converse, P. E. (1974). Comment: The status of nonattitudes. American Political Science Review, 68, 650–660. Cooke, N. J., Atlas, R. S., Lane, D. M., & Berger, R. C. (1993). Role of high‐level knowledge in memory for chess positions. American Journal of Psychology, 106, 321–351. Costner, H. L. (1969). Theory, deduction, and rules of correspondence. American Journal of Sociology, 75, 245–263. Davidson, A. R. (1995). From attitudes to actions to attitude change: The eVects of amount and accuracy of information. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 315–336). Mahwah, NJ: Lawrence Erlbaum Associates. Dawson, E., Gilovich, T., & Regan, D. T. (2002). Motivated reasoning and performance on the Wason selection task. Personality and Social Psychology Bulletin, 28, 1379–1387.

62

PENNY S. VISSER et al.

DiClemente, R., Forrest, K., Mickler, S., & Principal Site Investigators. (1990). College students’ knowledge and attitudes about AIDS. AIDS Education and Prevention, 2, 201–212. Eagly, A. H., & Chaiken, S. (1998). Attitude structure and function. In D. T. Gilbert, S. T. Fiske, and G. Lindzey (Eds.), The handbook of social psychology (pp. 269–322). New York: McGraw Hill. Eagly, A. H., Kulesa, P., Brannon, L. A., Shaw, K., & Hutson‐Comeaux, S. (2000). Why counterattitudinal messages are as memorable as proattitudinal messages: The importance of active defense against attack. Personality and Social Psychology Bulletin, 26, 1392–1408. Eckhardt, B. B., Wood, M. R., & Jacobvitz, R. S. (1991). Verbal ability and prior knowledge: Contributions to adults’ comprehension of television. Communication Research, 18, 636–649. Edwards, K., & Smith, E. E. (1996). A disconfirmation bias in the evaluation of arguments. Journal of Personality and Social Psychology, 71(1), 5–24. Engle, R. W., Nations, J. K., & Cantor, J. (1990). Is ‘‘working memory capacity’’ just another name for word knowledge? Journal of Educational Psychology, 82, 799–804. Erber, M. W., Hodges, S. D., & Wilson, T. D. (1995). Attitude strength, attitude stability, and the eVects of analyzing reasons. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 433–454). Mahwah, NJ: Lawrence Erlbaum Associates. Fazio, R. H. (1990). A practical guide to the use of response latency in social psychological research. In C. Hendrick and M. S. Clark (Eds.), Research methods in personality and social psychology (Vol. 11, pp. 74–97). Newbury Park, CA: Sage Publications. Fazio, R. H. (1995). Attitudes as object‐evaluation associations: Determinants, consequences, and correlates of attitude accessibility. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 247–282). Mahwah, NJ: Lawrence Erlbaum Associates. Fazio, R. H., & Towles‐Schwen, T. (1999). The MODE model of attitude‐behavior processes. In S. Chaiken and Y. Trope (Eds.), Dual process theories in social psychology (pp. 97–116). New York: Guilford. Fazio, R. H., & Zanna, M. P. (1981). Direct experience and attitude‐behavior consistency. Advances in Experimental Social Psychology, 14, 161–202. Fazio, R. H., Chen, J., McDonel, E. C., & Sherman, S. J. (1982). Attitude accessibility, attitude‐ behavior consistency, and the strength of the object‐evaluation association. Journal of Experimental Social Psychology, 18, 339–357. Fine, B. J. (1957). Conclusion‐drawing, communicator credibility, and anxiety as factors in opinion change. Journal of Abnormal and Social Psychology, 5, 369–374. Fisher, J. D., & Fisher, W. A. (1992). Changing AIDS risk behavior. Psychological Bulletin, 111, 455–474. Fiske, S. T., Lau, R. R., & Smith, R. A. (1990). On the varieties and utilities of political expertise. Social Cognition, 8, 31–48. Gorn, G. J. (1975). The eVects of personal involvement, communication discrepancy, and source prestige on reactions to communications on separatism. Canadian Journal of Behavioral Science, 7, 369–386. Granberg, D., & Holmberg, S. (1986). Political perception among voters in Sweden and the US: Analyses of issues with specific alternatives. Western Political Quarterly, 39, 7–28. Green, D. P. (1988). On the dimensionality of public sentiment toward partisan and ideological groups. American Journal of Political Science, 32, 758–780. Green, D. P., & Citrin, J. (1994). Measurement error and the structure of attitudes: Are positive and negative judgments opposites? American Journal of Political Science, 64, 256–281. Green, D. P., Goldman, S. L., & Salovey, P. (1993). Measurement error masks bipolarity in aVect ratings. Journal of Personality and Social Psychology, 64, 1029–1041.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

63

Gross, S. R., Holtz, R., & Miller, N. (1995). Attitude certainty. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 215–246). Mahwah, NJ: Lawrence Erlbaum Associates. Guttman, L., & Schuman, E. A. (1947). Intensity and a zero point for attitude analysis. American Sociological Review, 12, 57–67. Haddock, G., Rothman, A. J., Reber, R., & Schwarz, N. (1999). Forming judgments of attitude certainty, intensity, and importance: The role of subjective experiences. Personality and Social Psychology Bulletin, 25, 771–782. Haddock, G., Rothman, A. J., & Schwarz, N. (1996). Are (some) reports of attitude strength context dependent? Canadian Journal of Behavioural Science, 28, 313–316. Hansen, J. (1984). The role of prior knowledge in content area learning. Learning and Learning Disabilities, 3, 66–72. Helweg‐Larsen, M., & Collins, B. E. (1997). A social psychological perspective on the role of knowledge about AIDS in AIDS prevention. Current Directions in Psychological Science, 6, 23–26. Hippler, H. J., & Schwarz, N. (1986). Not forbidding isn’t allowing: The cognitive basis of the forbid‐allow assymmetry. Public Opinion Quarterly, 50, 87–96. Hodson, G., Maio, G. R., & Esses, V. M. (2001). The role of attitudinal ambivalence in susceptibility to consensus information. Basic and Applied Social Psychology, 23, 197–205. Hoelter, J. W. (1985). The structure of self‐conception: Conceptualization and measurement. Journal of Personality and Social Psychology, 49, 1392–1407. Holbrook, A. L. (2003). Operative and meta‐psychological strength‐related attitude features: A study of knowledge volume, ambivalence, and accessibility. Unpublished doctoral dissertation, Ohio State University. Holland, R. W., Verplanken, B., & van Knippenberg, A. (2002). On the nature of attitude‐ behavior relations: The strong guide, the weak follow. European Journal of Social Psychology, 32, 869–876. Howard‐Pitney, B., Borgida, E., & Omoto, A. M. (1986). Personal involvement: An examination of processing diVerences. Social Cognition, 4, 39–57. Hyman, H. H., & Sheatsley, P. B. (1950). The current status of American public opinion. In J. C. Payne (Ed.), The teaching of contemporary aVairs: Twenty‐first yearbook of the national council of social studies (pp. 11–34). Washington, DC: National Council for the Social Studies. Jackman, M. R. (1977). Prejudice, tolerance, and attitudes toward ethnic groups. Social Science Research, 6, 145–169. Jackson, T. H., & Marcus, G. E. (1975). Political competence and ideological constraint. Social Science Research, 4, 93–111. Jarvis, W. B. G., & Petty, R. E. (1996). The need to evaluate. Journal of Personality and Social Psychology, 70, 172–194. Johnson, B. T., & Eagly, A. H. (1989). EVects on involvement on persuasion: A meta‐analysis. Psychological Bulletin, 106, 290–314. Judd, C. M., & Johnson, J. T. (1981). Attitudes, polarization, and diagnosticity: Exploring the eVect of aVect. Journal of Personality and Social Psychology, 41, 26–36. Judd, C. M., & Krosnick, J. A. (1989). The structural bases of consistency among political attitudes: EVects of political expertise and attitude importance. In A. R. Pratkanis, S. J. Breckler, and A. T. Greenwald (Eds.), Attitude structure and function (pp. 99–128). Hillsdale, NJ: Lawrence Erlbaum Associates. Katz, D. (1944). The measurement of intensity. In H. Cantril (Ed.), Gauging public opinion (pp. 51–65). Princeton, NJ: Princeton University Press. Kenny, D. A. (1979). Correlation and causality. New York: Wiley.

64

PENNY S. VISSER et al.

Kokkinaki, F. (1998). Attitudes toward European Monetary Union in Greece: Antecedents, strength and consequences. Journal of Economic Psychology, 19, 775–796. Krosnick, J. A. (1988a). The role of attitude importance in social evaluations: A study of policy preferences, presidential candidate evaluations, and voting behavior. Journal of Personality and Social Psychology, 55, 196–210. Krosnick, J. A. (1988b). Attitude importance and attitude change. Journal of Experimental Social Psychology, 24, 240–255. Krosnick, J. A. (1989). Attitude importance and attitude accessibility. Personality and Social Psychology Bulletin, 15, 297–308. Krosnick, J. A. (1991). Response strategies for coping with the cognitive demands of attitude measures in surveys. Applied Cognitive Psychology, 5, 213–236. Krosnick, J. A., & Abelson, R. P. (1992). The case for measuring attitude strength in surveys. In J. Tanur (Ed.), Questions about survey questions (pp. 177–203). New York: Russell Sage. Krosnick, J. A., & Petty, R. E. (1995). Attitude strength: An overview. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 1–24). Mahwah, NJ: Lawrence Erlbaum Associates. Krosnick, J. A., & Schuman, H. (1988). Attitude intensity, importance, and certainty and susceptibility to response eVects. Journal of Personality and Social Psychology, 54, 940–952. Krosnick, J. A., & Telhami, S. (1995). Public attitudes toward Israel: A study of the attentive and issue publics. International Studies Quarterly, 39, 535–554. Krosnick, J. A., Boninger, D. S., Chuang, Y. C., Berent, M. K., & Carnot, C. G. (1993). Attitude strength: One construct or many related constructs? Journal of Personality and Social Psychology, 66, 1132–1151. Krosnick, J. A., Jarvis, B., Strathman, A., & Petty, R. E. (1994). Relations among dimensions of attitude strength: Insights from induced variable models. Columbus: Ohio State University Unpublished Manuscript. Kyllonen, P. C., Tirre, W. C., & Christal, R. E. (1991). Knowledge and processing speed as determinants of associative learning. Journal of Experimental Psychology: General, 120, 57–79. Lampron, S. F., Krosnick, J. A., ShaeVer, E., Petty, R. E., & See, M. (2003). DiVerent types of involvement moderate persuasion (somewhat) diVerently: Contrasting outcome‐based and value‐based involvement. Paper presented at the Society for Personality and Social Psychology Annual Meeting, Los Angeles, California. Lastovicka, J. L., & Gardner, D. M. (1979). Components of involvement. In J. C. Maloney and B. Silverman (Eds.), Attitude research plays for high stakes (pp. 53–73). Chicago: American Marketing Association. Lavine, H., HuV, J. W., Wagner, S. H., & Sweeney, D. (1998). The moderating influence of attitude strength on the susceptibility to context eVects in attitude surveys. Journal of Personality and Social Psychology, 75, 359–373. Lemon, N. F. (1968). A model of the extremity, confidence and salience of an opinion. British Journal of Social and Clinical Psychology, 7, 106–114. Lord, C. G., Ross, L., & Lepper, M. R. (1979). Biased assimilation and attitude polarization: The eVects of prior theories on subsequently considered evidence. Journal of Personality and Social Psychology, 37, 2098–2109. Mann, J. M., Tarantola, D. J., & Netter, T. W. (1992). AIDS in the world: A global report. Cambridge, MA: Harvard University Press. Marks, L. J., & Kamins, M. A. (1988). The use of product sampling and advertising: EVects of sequence of exposure and degree of advertising claim exaggeration on consumers’ belief strength, belief confidence, and attitudes. Journal of Marketing Research, 25, 266–281.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

65

McDill, E. L. (1959). A comparison of three measures of attitude intensity. Social Forces, 38, 95–99. McGraw, K. M., Lodge, M., & Stroh, P. K. (1990). On‐line processing in candidate evaluation: The eVects of issue order, issue salience, and sophistication. Political Behavior, 12, 41–58. McGraw, K. M., & Pinney, N. (1990). The eVects of general and domain‐specific expertise on political memory and judgment. Social Cognition, 8, 9–30. McGuire, W. J. (1986). The myth of massive media impact: Savagings and salvagings. In G. Comstock (Ed.), Public communication and behavior (Vol. 1, pp. 173–257). San Diego, CA: Academic Press. Miller, A. G., McHoskey, J. W., Bane, C. M., & Dowd, T. G. (1993). The attitude polarization phenomenon: Role of response measure, attitude extremity, and behavioral consequences of reported attitude change. Journal of Personality and Social Psychology, 64, 561–574. Miller, N. E. (1965). Involvement and dogmatism as inhibitors of attitude change. Journal of Experimental Social Psychology, 1, 121–132. Mokdad, A. H., Marks, J. S., Stroup, D. F., & Gerberding, J. L. (2004). Actual causes of death in the U.S., 2000. Journal of the American Medical Association, 291, 1238–1245. Newby‐Clark, I. R., McGregor, I., & Zanna, M. P. (2002). Thinking and caring about cognitive inconsistency: When and for whom does attitudinal ambivalence feel uncomfortable? Journal of Personality and Social Psychology, 82, 157–166. Parker, H. A., Perry, R. W., & Gillespie, D. F. (1974). Prolegomenon to a theory of attitude‐ behavior relationships. Pakistan Journal of Psychology, 7, 21–39. Paull, G., & Glencross, D. (1997). Expert perception and decision making in baseball. International Journal of Sport Psychology, 28, 35–56. Payne, S. L. (1951). The art of asking questions. Princeton, NJ: Princeton University Press. Pearson, P. D., Hansen, J., & Gordon, C. (1979). The eVect of background knowledge on young children’s comprehension of explicit and implicit information. Journal of Reading Behavior, 11, 201–209. Pelham, B. W. (1991). On confidence and consequence: The certainty and importance of self‐ knowledge. Journal of Personality and Social Psychology, 60, 518–530. Petty, R. E., & Cacioppo, J. T. (1979). Issue involvement can increase or decrease persuasion by enhancing message‐relevant cognitive responses. Journal of Personality and Social Psychology, 37, 1915–1926. Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer‐Verlag. Petty, R. E., & Cacioppo, J. T. (1990). Involvement and persuasion: Tradition versus integration. Psychological Bulletin, 107, 367–374. Petty, R. E., Haugtvedt, C. P., & Smith, S. M. (1995). Elaboration as a determinant of attitude strength: Creating attitudes that are persistent, resistant, and predictive of behavior. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 93–130). Mahwah, NJ: Lawrence Erlbaum Associates. Petty, R. E., & Krosnick, J. A. (1995). Attitude strength: Antecedents and consequences. Mahwah, NJ: Lawrence Erlbaum Associates. Pomerantz, E. M., Chaiken, S., & Tordesillas, R. S. (1995). Attitude strength and resistance processes. Journal of Personality and Social Psychology, 69, 408–419. Powell, J. L. (1977). Satirical persuasion and topic salience. Southern Speech Communication Journal, 42, 151–162. Priester, J. R., & Petty, R. E. (1996). The gradual threshold model of ambivalence: Relating positive and negative bases of attitudes to subjective ambivalence. Journal of Personality and Social Psychology, 71(3), 431–449.

66

PENNY S. VISSER et al.

Prislin, R. (1996). Attitude stability and attitude strength: One is enough to make it stable. European Journal of Social Psychology, 26, 447–477. Raden, D. (1985). Strength‐related attitude dimensions. Social Psychology Quarterly, 48, 312–330. Roberts, D. F., & Maccoby, N. (1985). EVects of mass communication. In G. Lindzey and E. Aronson (Eds.), Handbook of social psychology (3rd., Vol. 2, pp. 539–598). New York: Random House. Roese, N. J., & Olson, J. M. (1994). Attitude importance as a function of repeated attitude expression. Journal of Experimental Social Psychology, 30, 39–51. Rogers, T. E, Singer, E., & Imperio, J. (1993). The polls: AIDS—an update. Public Opinion Quarterly, 57, 92–114. Rokeach, M., & Kliejunas, P. (1972). Behavior as a function of attitude‐toward‐object and attitude‐toward‐situation. Journal of Personality and Social Psychology, 22, 194–201. Schneider, W., Gruber, H., Gold, A., & Opwis, K. (1993). Chess expertise and memory for chess positions in children and adults. Journal of Experimental Child Psychology, 56, 328–349. Schuman, H., & Presser, S. (1981). Questions and answers in attitude surveys: Experiments on question form, wording, and context. New York: Academic Press. Schwarz, N. (1998). Accessible content and accessibility experiences: The interplay of declarative and experiential information in judgment. Personality and Social Psychology Review, 2, 87–99. Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer‐Schatka, H., & Simons, A. (1991). Ease of retrieval as information: Another look at the availability heuristic. Journal of Personality and Social Psychology, 61, 195–202. Stember, H., & Hyman, H. (1949–1950). How interviewer eVects operate through question form. International Journal of Opinion and Attitude Research, 3, 493–512. StouVer, S. A., Guttman, L., Suchman, E. A., Lazarsfeld, P. F., Star, S. A., & Clausen, J. A. (1950). Measurement and prediction. Princeton, NJ: Princeton University Press. Suchman, E. A. (1950). The intensity component in attitude and opinion research. In S. A. StouVer, L. Guttman, E. A. Suchman, P. F. Lazarsfeld, S. A. Star, and J. A. Clausen (Eds.), Measurement and prediction (pp. 213–276). Princeton, NJ: Princeton University Press. Swann, W. B., Jr., & Ely, R. J. (1984). A battle of wills: Self‐verification versus behavioral confirmation. Journal of Personality and Social Psychology, 46, 1287–1302. Swann, W. B., Jr., Pelham, B. W., & Chidester, T. R. (1988). Change through paradox: Using self‐verification to alter beliefs. Journal of Personality and Social Psychology, 54, 268–273. Tannenbaum, P. H. (1956). Initial attitude toward source and concept as factors in attitude change through communication. Public Opinion Quarterly, 20, 413–425. Tedin, K. L. (1980). Assessing peer and parental influence on adolescent political attitudes. American Journal of Political Science, 24, 136–154. Theodorakis, Y. (1994). Planned behavior, attitude strength, role identity, and the prediction of exercise behavior. The Sports Psychologist, 8, 149–165. Thompson, M. T., & Zanna, M. P. (1995). The conflicted individual: Personality‐based and domain‐specific antecedents of ambivalent social attitudes. Journal of Personality, 63, 259–288. Thompson, M. T., Zanna, M. P., & GriYn, D. W. (1995). Let’s not be indiVerent about (attitudinal) ambivalence. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 361–386). Mahwah, NJ: Lawrence Erlbaum Associates. Tourangeau, R., & Rasinski, K. A. (1988). Cognitive processes underlying context eVects in attitude measurement. Psychological Bulletin, 103, 299–314. Tourangeau, R., Rasinski, K. A., Bradburn, N., & D’Andrade, R. (1989). Carryover eVects in attitude surveys. Public Opinion Quarterly, 53, 495–524.

THE STRUCTURE OF STRENGTH‐RELATED ATTITUDE ATTRIBUTES

67

Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, 1124–1131. Vallone, R. P., Ross, L., & Lepper, M. R. (1985). The hostile media phenomenon: Biased perception and perceptions of media bias in coverate of the Beirut massacre. Journal of Personality and Social Psychology, 49, 577–585. Verplanken, B. (1989). Involvement and need for cognition as moderators of beliefs‐attitude‐ intention consistency. British Journal of Social Psychology, 28, 115–122. Verplanken, B. (1991). Persuasive communication of risk information: A test of cue versus message processing eVects in a field experiment. Personality and Social Psychology Bulletin, 17, 188–193. Visser, P. S. (1998). Assessing the structure and function of attitude strength: Insights from a new approach. Unpublished doctoral dissertation, Ohio State University. Visser., P. S., Krosnick, J. A., & Norris, C. M. (2004). Challenging the common‐factor model of strength‐related attitude attributes: Contrasting the antecedents and consequences of attitude importance and attitude‐relevant knowledge. Chicago, IL: University of Chicago, Unpublished manuscript. Visser, P. S., Krosnick, J. A., & Simmons, J. P. (2003). Distinguishing the cognitive and behavioral consequences of attitude importance and certainty: A new approach to testing the common‐factor hypothesis. Journal of Experimental Social Psychology, 39, 118–141. Wa¨nke, M., Bless, H., & Biller, B. (1996). Subjective experience versus content of information in the construction of attitude judgments. Personality and Social Psychology Bulletin, 22, 1105–1113. Wason, P. C. (1966). Reasoning. In B. M. Foss (Ed.), New horizons in psychology (pp. 135–151). Harmondsworth, Middlesex, UK: Penguin Books. Wason, P. C. (1968). Reasoning about a rule. The Quarterly Journal of Experimental Psychology, 20, 273–281. Wason, P. C., & Johnson‐Laird, P. N. (1972). Psychology of reasoning: Structure and content. Cambridge, MA: Harvard University Press. Willoughby, T., Waller, T. G., Wood, E., & MacKinnon, G. E. (1993). The eVect of prior knowledge on an immediate and delayed associative learning task following elaborative interrogation. Contemporary Educational Psychology, 18, 36–46. Wilson, T. D., Dunn, D. S., Kraft, D., & Lisle, D. J. (1989). Introspection, attitude change, and attitude‐behavior consistency: The disruptive eVects of explaining why we feel the way we do. Advances in Experimental Social Psychology, 22, 287–343. Wood, W. (1982). Retrieval of attitude‐relevant information from memory: EVects on susceptibility to persuasion and on intrinsic motivation. Journal of Personality and Social Psychology, 42, 798–810. Wood, W., & Kallgren, C. A. (1988). Communicator attributes and persuasion: Recipients’ access to attitude‐relevant information in memory. Personality and Social Psychology Bulletin, 4, 172–182. Wood, W., Rhodes, N., & Biek, M. (1995). Working knowledge and attitude strength: An information‐processing analysis. In R. E. Petty and J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 283–313). Mahwah, NJ: Lawrence Erlbaum Associates. Zaichkowsky, J. L. (1985). Measuring the involvement construct. Journal of Consumer Research, 12, 341–352. Zuwerink, J. R., & Devine, P. G. (1996). Attitude importance and resistance to persuasion: It’s not just the thought that counts. Journal of Personality and Social Psychology, 70, 931–944.

IMPLEMENTATION INTENTIONS AND GOAL ACHIEVEMENT: A META‐ANALYSIS OF EFFECTS AND PROCESSES

Peter M. Gollwitzer Paschal Sheeran

Holding a strong goal intention (‘‘I intend to reach Z!’’) does not guarantee goal achievement, because people may fail to deal eVectively with self‐ regulatory problems during goal striving. This review analyzes whether realization of goal intentions is facilitated by forming an implementation intention that spells out the when, where, and how of goal striving in advance (‘‘If situation Y is encountered, then I will initiate goal‐directed behavior X!’’). Findings from 94 independent tests showed that implementation intentions had a positive eVect of medium‐to‐large magnitude (d ¼ .65) on goal attainment. Implementation intentions were eVective in promoting the initiation of goal striving, the shielding of ongoing goal pursuit from unwanted influences, disengagement from failing courses of action, and conservation of capability for future goal striving. There was also strong support for postulated component processes: Implementation intention formation both enhanced the accessibility of specified opportunities and automated respective goal‐directed responses. Several directions for future research are outlined. I. Introduction Understanding what factors determine whether people succeed or fail in achieving desired outcomes is a fundamental concern in both basic and applied psychology. Most theories of motivation and self‐regulation converge on the idea that setting a behavioral or outcome goal is the key act of willing that promotes goal attainment (e.g., Ajzen, 1991; Atkinson, 1957; 69 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 38 DOI: 10.1016/S0065-2601(06)38002-1

Copyright 2006, Elsevier Inc. All rights reserved. 0065-2601/06 $35.00

70

PETER M. GOLLWITZER AND PASCHAL SHEERAN

Bandura, 1991; Carver & Scheier, 1998; Gollwitzer, 1990; Locke & Latham, 1990). The basic assumption is that the strength of a person’s intention determines respective accomplishments (Austin & Vancouver, 1996; Gollwitzer & Moskowitz, 1996; Oettingen & Gollwitzer, 2001; Sheeran, 2002). Although accumulated research supports this idea (e.g., Armitage & Conner, 2001; Sheeran, 2002; Sutton, 1998), there is also contrary evidence that gives credence to the proverb that ‘‘the road to hell is paved with good intentions’’ (Orbell & Sheeran, 1998; Sheeran, 2002). To address this issue, Gollwitzer (1993, 1996, 1999) proposed that successful goal achievement is facilitated by a second act of willing that furnishes the goal intention with an if–then plan specifying when, where, and how the person will instigate responses that promote goal realization. These plans are termed implementation intentions. Implementation intentions appear to be eVective at enhancing the likelihood of goal achievement. However, the eVectiveness of if–then planning has been reviewed only in narrative (e.g., Gollwitzer, 1999; Gollwitzer, Bayer, & McCulloch, 2005) and small‐scale quantitative (e.g., Koestner, Lekes, Powers, & Chicoine, 2002b; Sheeran, 2002) reports to date, and a comprehensive evaluation of implementation intention eVects and processes is overdue. The aim of this review is to quantify the overall impact of implementation intention formation on goal achievement using meta‐analytic techniques. In addition, this chapter tests the eVectiveness of implementation intentions in relation to diVerent self‐regulatory problems and goal domains and assesses potential moderators of implementation intention eVects. Finally, the impact of implementation intentions on theoretically specified component processes is examined to understand why implementation intentions may help people obtain outcomes that they desire.

II. Goal Intention Strength and Goal Achievement Goal intentions are self‐instructions to attain certain outcomes or perform particular behaviors and typically take the format of ‘‘I intend to reach Z!’’ They are derived from beliefs about the feasibility and desirability of actions and end states (e.g., Ajzen, 1991; Atkinson, 1957; Bandura, 1991, 1997; Brehm & Self, 1989; Carver & Scheier, 1998; Heckhausen, 1991; Locke & Latham, 1990; Vroom, 1964) and represent the culmination of the decision making process (Gollwitzer, 1990). Goal intentions signal the end of deliberation about what actions to perform or outcomes to reach; they imply a commitment to act that may vary in strength (Ajzen, 1991; Gollwitzer, 1990; Sheeran, 2002; Webb & Sheeran, 2005a).

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

71

In traditional theories of goal pursuit, goal intentions are construed as the most immediate and important predictor of attainment. For instance, preeminent accounts of goal‐directed behavior, such as control theory (Carver & Scheier, 1982, 1998), social cognitive theory (Bandura, 1991, 1997), and goal setting theory (Locke & Latham, 1990), models of attitude–behavior relations, such as the theories of reasoned action (Fishbein, 1980; Fishbein & Ajzen, 1975) and planned behavior (Ajzen, 1991), and the model of interpersonal behavior (Triandis, 1980), as well as theories of health‐related behavior, such as protection motivation theory (Rogers, 1983) and the prototype/ willingness model (Gibbons, Gerrard, Blanton, & Russell, 1998), each accord goal intentions a central role in their theorizing about action. Accordingly, research has been concerned for several decades with the factors that determine strong intentions—the assumption being that intention strength is a good predictor of intention realization. This assumption seems to be supported by meta‐analyses of correlational studies in which participants’ goal intentions (e.g., ‘‘I intend to perform behavior W !’’ or ‘‘I intend to achieve outcome Z!’’) are measured at one time‐point and behavior is measured at a later time‐point. For example, reviews of the theory of reasoned action (Kim & Hunter, 1993; Sheppard, Hardwick, & Warshaw, 1988; van den Putte, 1993), the theory of planned behavior (Armitage & Conner, 2001; Godin & Kok, 1996; Hausenblas, Carron, & Mack, 1997), and protection motivation theory (Floyd, Prentice‐Dunn, & Rogers, 2000; Milne, Sheeran, & Orbell, 2000), as well as meta‐analyses of particular behaviors (e.g., condom use, Sheeran & Orbell, 1998; physical activity, Hagger, Chatzisarantis, & Biddle, 2002), indicate that strength of intention typically explains 20–35% of the variance in goal achievement. To gain insight into the overall strength of intention–behavior consistency in this type of research, Sheeran (2002) conducted a meta‐analysis of 10 meta‐ analyses of the intention–behavior relation. Findings showed that intentions accounted for 28% of the variance in behavior, on average, across 422 studies involving 82,107 participants. According to Cohen’s (1992) power primer, R2 ¼ .28, constitutes a ‘‘large’’ eVect size, which suggests that intentions are ‘‘good’’ predictors of behavior—as traditional theories of goal pursuit have supposed. However, bivariate correlations between goal intentions and future behavior may overestimate the strength of intention–behavior relations because it is possible that future behavior and goal intentions are both determined by self‐perceptions of past behavior (Bem, 1972). The implication is that analyses should control for previous performance in order to determine to what extent goal intentions are associated with behavior change. Sutton and Sheeran (2003) conducted a meta‐analysis along these lines. Sampled‐weighted average correlations between past behavior, goal

72

PETER M. GOLLWITZER AND PASCHAL SHEERAN

intentions, and future behavior were computed from 51 studies involving 8166 participants and then used as inputs for a hierarchical regression analysis. Findings indicated that, not surprisingly, past behavior was a good predictor of future behavior on the first step of the equation and accounted for 26% of the variance. Entering goal intentions on the second step was associated with a significant increment in the variance explained in future behavior (R2change ¼ :07). These findings suggest that goal intentions have significant associations with future behavior even when previous performance is taken into account. However, the eVect size for goal intentions is small‐to‐medium rather than large. Even correlational analyses that statistically control for past behavior in estimating the goal intention–goal achievement relation are problematic, however, because it is always possible that a third variable is responsible for the observed associations. To eliminate this alternative explanation of intention–behavior consistency, it is necessary to experimentally manipulate goal intentions and then determine whether this manipulation produces a significant diVerence in subsequent goal attainment. Webb and Sheeran (in press a) tested this idea in a recent meta‐analysis. They identified 47 studies (N ¼ 8802) that (1) were successful at inducing statistically significant diVerences in goal intentions between experimental versus control participants and (2) followed up participants in order to measure diVerences in subsequent goal attainment. Findings showed that the mean diVerence in goal intention strength produced by the experimental manipulations had an eVect size of medium‐to‐large magnitude (d ¼ .66). Findings also indicated that manipulating goal intention strength engendered a significant diVerence in goal achievement. However, the eVect size was small‐to‐ medium only; d was .36 that equates to R2 ¼ .03. Thus, producing significant changes in goal intention strength only generates a modest change in goal achievement. This finding indicates that there is a substantial ‘‘gap’’ between people’s goal intentions and their subsequent attainment. A converging line of research has decomposed the intention–behavior relation in terms of a 2 (goal intention: to act vs. not to act)
2 (goal achievement: acted vs. did not act) matrix (McBroom & Reid, 1992; Orbell & Sheeran, 1998; Sheeran, 2002). This decomposition provides insight into the sources of consistency and discrepancy between intentions and action. Consistency is attributable to participants who intend to act and subsequently act (termed ‘‘inclined actors’’) and to participants who do not intend to act and do not act (‘‘disinclined abstainers’’). Discrepancies between intentions and action, on the other hand, can be attributed to participants who intend to act but do not act (‘‘inclined abstainers’’) and to participants who do not intend to act but end up acting (‘‘disinclined actors’’). A review by Sheeran (2002) found that inclined abstainers, rather than disinclined actors, were

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

73

principally responsible for the intention–behavior ‘‘gap.’’ The median proportion of participants who intended to but did not act was 47%, whereas the median proportion of participants who did not intend to act but subsequently acted was only 7%. These findings would seem to confirm that the proverbial road to hell is paved with good intentions—barely more than one‐half of people who intended to act were successful at translating those intentions into action. In sum, it appears that the single act of willing involved in forming a goal intention is not suYcient to ensure goal achievement. The implicit assumption in traditional models of goal pursuit—that goal intentions fashioned from appropriate evaluation of feasibility and desirability considerations satisfactorily account for the intensity of goal striving—is not strongly supported by the evidence. Clearly, some additional psychological concepts are needed (1) to understand why people often become inclined abstainers rather than inclined actors and (2) to develop self‐regulatory strategies to help people ‘‘bridge’’ the gap between their intentions and their behavior.

III. Self‐Regulation of Goal Striving Recent research on goals has demonstrated that variables other than strength of goal intention aVect the intensity of goal striving and rate of goal attainment (Gollwitzer & Moskowitz, 1996; Oettingen & Gollwitzer, 2001). Some goal theories focus on the implications of particular goal contents and structural features. For instance, people who set themselves learning goals rather than performance goals are better at dealing with failure experiences and, consequently, show more persistent and successful goal pursuit (Dweck, 2000). Higgins (2000) demonstrates that people who pursue their goals using means that have a natural fit to the content of the goal have a better chance of goal attainment. For example, people with promotion goals (that focus on gain and achievement) are more likely to realize those goals using eagerness means whereas prevention goals (that focus on safety and security) are more likely to be realized by vigilance means. Other important distinctions between types of goals have been drawn by Locke and Latham (1990) (e.g., specific vs. ‘‘do your best’’ goals), Bandura (1991) (e.g., proximal vs. distal goals), and Deci and Ryan (1991) (e.g., goals based on needs for autonomy, competence, and social integration vs. goals based on other needs). All of these theories construe features related to the content and structure of set goals as critical in determining the likelihood of goal achievement. Other goal theories assume that setting a goal (of whatever kind) is only a first step en route to goal realization. A key impetus for self‐regulation

74

PETER M. GOLLWITZER AND PASCHAL SHEERAN

research on goals is the model of action phases (Gollwitzer, 1990; Heckhausen, 1991; Heckhausen & Gollwitzer, 1987) that construes goal attainment in terms of solving a number of consecutive tasks. Goal setting is viewed as merely the first of these tasks—with planning how to achieve the goal, getting started, and successfully completing goal striving as equally important subsequent tasks. The model of action phases seeks to provide a comprehensive temporal account of goal pursuit. Four diVerent consecutive action phases are postulated by the model. The first, predecisional, phase starts from the assumption that people have many more wishes and desires than they can possibly realize. Here people’s task is to deliberate about the desirability and feasibility of their various wishes in order to choose which ones will be turned into binding goals. The model agrees with classic motivational notions (e.g., Atkinson, 1957; Fishbein & Ajzen, 1975; Lewin, 1926) that people commit to those goals in which attainment is perceived as both highly desirable and feasible. However, the model of action phases also states that goal attainment is not yet secured by the act of goal setting (i.e., by having formed strong goal intentions). Rather, goal accomplishment requires in addition that the individual eVectively regulates the actual striving for the goal (i.e., engages in eVective goal implementation). Once a person has committed to a goal, she makes the transition to the second action phase, preactional. Here the goal‐relevant task is to initiate goal‐directed behaviors successfully. This may be straightforward when the respective actions have become routinized through frequent and consistent performance in stable situational contexts. However, matters are likely to be more complex when people are unfamiliar with, or imprecise about, the respective goal‐directed actions and contexts of performance. In these circumstances, people are likely to benefit from fashioning plans that spell out when, where, and how to implement goal‐directed behaviors. The initiation of actual performance of the respective goal‐directed behaviors marks the transition to the third, actional, phase. The task to be accomplished during this phase pertains to responding flexibly and adaptively to contextual threats to goal progress so that goal striving is not derailed prematurely. In other words, the key actional task is to bring the respective goal‐directed activity to a successful conclusion by shielding it from distractions and temptations that could potentially disrupt goal striving. In the final action phase, postactional, the task is to evaluate goal achievement both in terms of degree of attainment (‘‘Did I do as well as I had hoped?’’) and quality of attainment outcomes (‘‘Was it worth doing?’’). This process involves comparing what has been achieved with one’s original wishes and desires, and it may at times imply eVortful disengagement from the goal (if further striving is inappropriate). Thus, goal completion is likely

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

75

to provide valuable information that can feed back into evaluations of the feasibility and desirability of future courses of action; people return to the position of deliberating about their various wishes and desires from which they started.

A. PROBLEMS EN ROUTE TO GOAL COMPLETION The foregoing discussion suggests that merely forming a goal intention does not guarantee goal achievement as people often face problems en route to goal completion. So what are these challenges, and how can people tackle them successfully by using self‐regulatory strategies? We propose that the following four problems may prevent people from reaching their goals. 1. Failing to Get Started The first problem that can undermine goal attainment is failing to get started with goal striving. A number of factors militate against getting started on one’s goals. The first has to do with remembering to act. When a behavior is not part of one’s routine, or when one has to postpone acting until a suitable opportunity presents itself, one can easily forget to perform the intended behavior. This is because situational demands on attention and memorial resources may serve to reduce the activation level of a focal goal intention compared to other intentions (Einstein & McDaniel, 1996). Dealing with many things at once or becoming preoccupied by a particular task can make it diYcult to remember to act on one’s goals, especially when the intended behavior is new or unfamiliar. Empirical support for this explanation of intention–behavior discrepancies comes from retrospective reports by inclined abstainers. For example, 70% of participants who had intended to perform a breast self‐examination but failed to do so, oVered ‘‘forgetting’’ as their reason for nonperformance (Milne, Orbell, & Sheeran, 2002; Orbell, Hodgkins, & Sheeran, 1997). Similarly, meta‐analysis has shown that the longer the time interval between measures of goal intentions and goal achievement, the less likely it is that intentions are realized (Sheeran & Orbell, 1998). These findings speak to the idea that remembering to act can be a vital but diYcult task. Even if one remembers what one intends to achieve, there is a second problem that may need to be resolved, namely, seizing the opportunity to act. This problem is especially acute when there is a deadline for performing the behavior or when the opportunity to act is presented only briefly. In these circumstances, people may fail to initiate goal‐directed responses either because they fail to notice that a good time to get started has arrived or they

76

PETER M. GOLLWITZER AND PASCHAL SHEERAN

are unsure how they should act when the moment presents itself. For instance, Oettingen, Ho¨nig, and Gollwitzer (2000, Study 3) showed that considerable slippage can occur even when people have formed strong goal intentions to perform a certain behavior at a particular time. In one of their experimental conditions, participants were provided with diskettes containing four arithmetic tasks and formed goal intentions to perform these tasks on their computers at a particular time each Wednesday morning for the next 4 weeks. The program on the diskette recorded the time that participants started to work on the task from the clock on participants’ computers. Findings indicated that the mean deviation from the intended start time was 8 hours, that is, a discrepancy of 2 hours on average for each specified opportunity. Similar findings were obtained by Dholakia and Bagozzi (2003, Study 2) using a ‘‘short fuse behavior’’ paradigm in which participants’ task was to evaluate a website that could be accessed only during a short time window. Here, only 37% of participants who formed goal intentions were successful at accomplishing the task (see also Gollwitzer & Brandsta¨tter, 1997). In sum, people may not get started with goal striving because they fail to seize suitable opportunities to act. Third, there are also many instances in which people remember their ‘‘good’’ intention (e.g., to order a low fat meal) and recognize that an opportune moment is upon them (e.g., it is lunchtime at one’s usual restaurant) but, nonetheless, they fail to initiate action (e.g., because ‘‘I just didn’t fancy the low fat meal!’’). This problem has to do with overcoming an initial reluctance to act. Initial reluctance is likely to arise when people have decided to initiate a behavior that involves a trade‐oV between attractive long‐term consequences versus less attractive short‐term consequences. For example, a strong goal intention to order the low fat meal might have been formed on the basis of longer‐term cognitive considerations (e.g., the low fat meal is perceived as ‘‘healthy’’ or ‘‘beneficial’’); however, one might not have anticipated how the short‐term aVective considerations would occupy attention at the moment of action (e.g., the low fat meal is perceived as ‘‘unsatisfying’’ or ‘‘tasteless’’ at the critical juncture). Such dilemmas between the head and the heart are commonplace (e.g., Loewenstein, Weber, Hsee, & Welch, 2001; Metcalfe & Mischel, 1999; Trafimow & Sheeran, 2004). Evidence for this explanation of intention–behavior discrepancies comes, for instance, from the field of sexual health in which findings show that young people may, in ‘‘the heat of the moment’’ of a sexual encounter, have problems overcoming reluctance to practice safer sex (e.g., Abraham et al., 1999; Sheeran, White, & Phillips, 1991; Wight, 1992). Overcoming initial reluctance is also a significant problem in several other domains (e.g., environmental, consumer, and academic goals).

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

77

2. Getting Derailed The goals of interest to social and health psychologists are not usually discrete one‐shot actions but sequences of action that require continuous striving and repeated behavioral performance to be accomplished. The problem with such striving is that many situational contexts or self‐states are not conducive to intention realization but instead hold the potential to derail an ongoing goal pursuit. Thus, the person’s self‐regulatory task is to shield goal striving from unwanted influences. Shielding one’s goal striving is necessary under the following circumstances. First, shielding is called for when conflicting attention and behavioral responses could make people stray oV course. For example, despite making good initial progress with one’s goal intention to finish a report, one may find one’s attention wandering and feel compelled to join colleagues whom one hears gathering around the water cooler. In these instances, spontaneous attention to distracting stimuli may have to be suppressed in order to complete the goal. Such suppression may not be easy when the distractions are vivid, arousing, or highly valenced because, as the literature on cravings has shown (Kavanagh, May, & Andrade, 2005), people are liable to elaborate desirable stimuli through mental imagery. However, failing to control the attention paid to enticing stimuli (opportunities related to competing goal pursuits) can greatly undermine achievement of the focal (task) goal—as was demonstrated by Mischel and Patterson’s (1978; Patterson & Mischel, 1976) classic studies on resistance to temptation (see also Gollwitzer & Schaal, 1998). Of course, it may not be enough to suppress unwanted attention responses to appealing distractions in order to reach one’s goal. Often, it will be necessary to suppress behavioral responses. For example, the person who succeeded in enacting her goal intention to order the low fat meal at lunchtime still has to forego the chocolate dessert after dinner if the superordinate goal intention is to lose weight. If an unwanted behavior possesses features of automaticity, it should be especially diYcult to control (Aarts & Dijskterhuis, 2000a,b; Sheeran et al., 2005a; Verplanken & Aarts, 1999; Wood, Quinn, & Kashy, 2002). Keeping such behavioral responses in check merely by forming the respective goal intention may not be suYcient, as research on weight loss and smoking cessation has shown (e.g., COMMIT Research Group, 1995; Garner & Wooley, 1991). Findings from a recent meta‐analysis (Ouellette & Wood, 1998) are also consistent with this idea. Goal intentions emerged as much poorer predictors of future action when antagonistic behaviors had been performed frequently and consistently in relevant contexts (see also Verplanken, Aarts, van Knippenberg, & Moonen, 1998). In sum, controlling interfering unwanted attention and behavioral responses makes an important

78

PETER M. GOLLWITZER AND PASCHAL SHEERAN

diVerence to whether one’s goal‐directed eVorts warrant the designation ‘‘inclined actor’’ versus ‘‘inclined abstainer.’’ There is a second circumstance in which shielding an ongoing goal pursuit becomes crucial (Gollwitzer et al., 2005). So far, our discussion of controlling unwanted attention and behavioral responses has assumed that people have some knowledge and awareness of what sorts of obstacles (distractions, temptations, barriers) the environmental context is likely to present, when those obstacles are likely to arise, and what kind of unwanted responses those obstacles typically generate. Knowing what might happen, when it might happen, and how it might aVect us thus appears to be a prerequisite for the successful use of any strategy of suppression. However, there is a route to eVectively shielding an ongoing goal pursuit that does not require the anticipation of a situational threat or its impact on goal striving. The existence of such an alternative strategy is crucial as, more often than not, we are not in a position to consciously anticipate the occurrence of obstacles and the working of habits, or in what form and intensity these will threaten ongoing goal pursuits. The following three social psychological phenomena exemplify what we have in mind when we speak of obstacles to an ongoing goal pursuit that are not anticipated by the individual: deindividuation eVects on social loafing, the impact of loss frames on negotiation outcomes, and nonconscious priming of antagonistic goals. Social loafing eVects occur when people are asked to work in groups in which performance outcomes cannot be checked at an individual level (Karau & Williams, 1993; Latane´, Williams, & Harkin, 1979). Under these circumstances, people show reduced eVort and performance compared to situations in which individual outcomes can be identified. The problem is that people are unlikely to have insight into the negative impact of the group setting and task instructions on their performance. Not surprisingly, therefore, having the goal intention to perform well is not suYcient to overcome social loafing (Gollwitzer & Bayer, 2000). A similar issue arises in relation to the impact of loss versus gain frames on negotiation outcomes (De Dreu, Carnevale, Emans, & van de Vliert, 1995; Neale & Bazerman, 1985). In a typical experiment, pairs of participants are asked to negotiate the distribution of some finite resource (e.g., land on an island). The framing of the negotiation is manipulated by either providing participants with information about how many points they lose by giving up elements of the resource (loss frame) or by telling participants how many points they gain by receiving these elements (gain frame). Findings indicate that cognitive loss frames lead to comparatively unfair agreements about resources and also hinder integrative solutions. Participants are not aware of the negative impact of loss frames on their negotiation behavior.

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

79

Again, merely forming a goal intention to engage in fair and cooperative negotiation with one’s partner fails to be suYcient to overcome the negative impact of loss framing (Tro¨tschel & Gollwitzer, 2004). Finally, people are unaware of the fact that their behavior is often guided by goals that have become activated directly by the situational context at hand. Auto‐motive theory (Bargh, 1990; Bargh & Gollwitzer, 1994) proposes that goals that have a history of being acted upon in a particular situation have the potential to become directly activated by this critical situation without the need for conscious intent. Studies have used priming techniques to show that activated chronic goals have predictable eVects on the intensity of goal striving. For example, participants who completed scrambled sentences designed to prime achievement goals performed better on a word puzzle task as compared to controls (Bargh, Gollwitzer, Lee‐Chai, Barndollar, & Tro¨tschel, 2001). Moreover, extensive debriefing indicated that participants had no awareness of either the activation of the goal or its impact on respective performance. Direct goal activation has serious implications for realizing one’s goal intentions when the situational context activates a goal that is antagonistic to the focal goal. Consistent with this idea, Gollwitzer, Sheeran, Tro¨tschel, and Webb (2004d) found that participants who had formed a goal intention to drive carefully in a driving simulator exhibited greater speed and more errors when they had been primed with the auto‐motive of ‘‘moving fast’’ compared to when the primed auto‐ motive was to ‘‘move slow.’’ Clearly, therefore, blocking the adverse contextual threat posed by situationally activated antagonistic goals constitutes an important challenge in shielding an ongoing goal pursuit. The discussion so far only refers to derailments of goal striving by unanticipated unwanted influences that originate in the environment. But such unanticipated unwanted influences can also originate within the person (Gollwitzer et al., 2005). This is the third circumstance in which the shielding of an ongoing goal pursuit is needed—when detrimental self‐states threaten goal attainment. The negative consequences of the following three self‐states on goal striving may serve as examples: the eVects of mood on stereotyping, the influence of self‐definitional incompleteness on social sensitivity, and the impact of ego‐depletion on subsequent task performance. Being in a good mood signals to the self that one’s current situation is unproblematic, and thus information processing is less elaborate or systematic as compared to being in a bad mood (Schwarz, 1990; Schwarz, Bless, & Bohner, 1991). Consequently, people are more liable to stereotyping when they are in a good mood than when they are in a bad mood (Bless, 1997; Bless & Fiedler, 1995). The impact of positive mood on stereotyping target persons is diYcult to anticipate by the layperson and, thus,

80

PETER M. GOLLWITZER AND PASCHAL SHEERAN

should be diYcult to control. Indeed, Gollwitzer and Bayer (2000) found that merely having the goal intention to form nonstereotypical impressions did not attenuate the good‐mood‐eVect on increased stereotyping. Symbolic self‐completion theory (Wicklund & Gollwitzer, 1982) proposes that when people who are highly committed to an identity goal (e.g., becoming a lawyer) obtain negative feedback about their accomplishments in the respective domain, they experience a sense of self‐definitional incompleteness. This is a highly aversive self‐evaluative state that is associated with compensatory eVorts to show oV alternative symbols or indicators of the aspired to identity in front of other people (e.g., by wanting to talk about one’s achievements). Consequently, incomplete individuals tend to become absorbed in self‐symbolizing activities and thus neglect the thoughts and feelings of an audience; their interactions with others exhibit social insensitivity (Gollwitzer & Wicklund, 1985). Gollwitzer and Bayer (2000) observed that this eVect cannot be ameliorated by explicitly assigning participants the goal of taking the perspective of their interaction partners. Finally, ego‐depletion refers to the phenomenon that exerting self‐control on an initial task produces a temporary reduction in people’s capacity for self‐control that is reflected in poor performance on a subsequent task (Baumeister, Bratlavsky, & Muraven, 1998; Muraven & Baumeister, 2000). For example, Baumeister et al. (1998, Experiment 1) showed that participants who had to eat radishes instead of tempting chocolate during an initial task, persisted for less time on a subsequent unsolvable puzzles task than did participants who were allowed to eat the chocolate during the initial task (these participants did not have to exert self‐control). Apparently, ego‐ depletion can undermine task performance even when people have strong goal intentions to perform well. Consistent with this idea, Webb and Sheeran (2003) found that ego‐depleted participants and nondepleted controls exhibited substantive diVerences in puzzle task performance. However, both groups reported devoting equivalent eVort to the puzzle task and had equivalent desire to quit. Thus, ego‐depletion would seem to be an important factor in reducing the intensity of goal striving and one in which appropriate goal intentions are not necessarily an eVective defense. 3. Not Calling a Halt Initiating and shielding goal striving from unwanted influences are crucial for successfully reaching goal completion. However, there is a third problem that needs to be resolved, namely, disengaging from goal striving that has become unproductive (Wrosch, Scheier, Carver, & Schulz, 2003). Disengagement may be straightforward when goal monitoring indicates satisfactory progress, or attainment of desired outcomes. However, a good deal

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

81

of research indicates that it is very diYcult to disengage from an ongoing goal pursuit when self‐defensive concerns are activated. Researchers have studied such failure to disengage under varying labels, such as sunk costs (e.g., Arkes & Blumer, 1985), entrapment (e.g., Brockner, Rubin, & Lang, 1981), and escalation of commitment (e.g., Tan & Yates, 2002). However, the basic conceptualization of the phenomenon is similar (Bragger, Hantula, Bragger, Kirnan, & Kutcher, 2003). Again, mere goal intentions to halt a failing course of action are often insuYcient as has been shown by work using standard escalation paradigms (Henderson, Gollwitzer, & Oettingen, 2004). 4. Overextending Oneself There is a fourth problem in goal striving that has to do with the fact that people have to pursue multiple goals (e.g., Austin & Vancouver, 1996; Carver & Scheier, 1998; Gollwitzer & Moskowitz, 1996). Thus, overextending oneself in an ongoing goal pursuit is likely to jeopardize the achievement of subsequent important goals. Accordingly, eVective self‐ regulation of goal striving needs to conserve the person’s capability to successfully engage in subsequent goal pursuit once striving for the initial goal has ended. However, action control by goal intentions makes people vulnerable to overextension. A good example is the phenomenon of ego‐depletion in which assigning participants goal intentions to perform well on an initial task that requires self‐control is associated with reduced self‐regulatory capability (and diminished performance) on a subsequent task (Baumeister et al., 1998). The well‐known ironic eVects of mental control (Wegner, 1994) constitute another instance in which goal intentions can produce overextension on an initial task and thereby diminish future capability. For example, Macrae, Bodenhausen, Milne, and Jetten (1994) assigned participants the goal intention of forming a nonstereotypical impression of a homeless person (or not) and asked them to provide a written statement of their impression. After a 5‐minute filler task, participants were asked to evaluate homeless people in general on semantic diVerential scales that included five stereotypical adjective pairs (e.g., drunk–sober, busy–lazy). Findings indicated that goal intentions were successful in producing less stereotypical impressions of the person on the initial task compared to controls. However, on the subsequent rating task, goal intention participants gave more stereotypical evaluations of homeless people in general. That is, goal intentions to suppress stereotypes produced a rebound eVect. In sum, achieving desired outcomes on the basis of mere goal intentions has costs in terms of undermining the success of subsequent goal pursuits.

82

PETER M. GOLLWITZER AND PASCHAL SHEERAN

B. FORMING IMPLEMENTATION INTENTIONS: A STRATEGY FOR EFFECTIVE SELF‐REGULATION OF GOAL STRIVING The idea tested in the present meta‐analysis is that implementation intentions (i.e., if–then plans) facilitate eVective self‐regulation of goal striving. Implementation intentions should enhance people’s ability to initiate, maintain, disengage from, and undertake further goal striving and thereby increase the likelihood that strong goal intentions are realized successfully. In other words, this form of planning is expected to bridge the intention– behavior gap. Implementation intentions are if–then plans that connect good opportunities to act with cognitive or behavioral responses that are eVective in accomplishing one’s goals. Whereas goal intentions specify what one wants to achieve (i.e., ‘‘I intend to reach Z!’’), implementation intentions specify both the behavior that one will perform in the service of goal achievement and the situational context in which one will enact it (i.e., ‘‘If situation Y occurs, then I will initiate goal‐directed behavior X!’’). Thus, goal intentions and implementation intentions can easily be distinguished on the basis of their content and structure; a goal intention refers to what one intends to achieve, whereas an implementation intention specifies when, where, and how one intends to achieve it. To form an implementation intention, the person must (1) identify a response that will promote goal attainment and (2) anticipate a suitable occasion to initiate that response. For instance, a possible implementation intention in the service of the goal intention to do more exercise would link an appropriate behavior (e.g., take the stairway instead of the elevator) to a suitable situational context (e.g., standing in front of the entrance to the elevator at work). As a consequence, a strong mental link is created between the critical situation of waiting for the elevator and the goal‐directed response of walking upstairs. Selecting suitable opportunities to enact goal‐directed responses entails that people anticipate situations in which it would be fitting to execute goal‐ directed responses. The critical situation specified in one’s plan can involve an internal cue (e.g., a strong feeling) or an external cue (e.g., a particular place, object, person, or point in time). The cues can either be related to good opportunities to act (i.e., it is easy to perform actions that are instrumental for reaching the goal) or to anticipated obstacles to goal striving. Thus, cue selection can focus on initiating and stabilizing the goal striving at hand or on shielding it from particular anticipated obstacles. Forming an implementation intention also involves the selection of an eVective goal‐directed behavior. In line with the theory of goal systems (Kruglanksi et al., 2002; Shah, Kruglanksi, & Friedman, 2003), it is assumed

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

83

that for any given goal, various routes to goal attainment are available. Accordingly, the specification of the then‐component of an implementation intention can take many diVerent forms. For instance, not only can an implementation intention specify one of the many behaviors that lead to goal attainment, it can also specify the suppression of one of the many responses that prevent goal attainment. In addition, the specification of the goal‐directed responses can either focus on the initiation or the maintenance of goal striving. Finally, the then‐component of an implementation intention may specify ignoring those stimuli that have the potential to instigate unwanted attention or behavior responses that could derail an ongoing goal pursuit (see Appendix 1 for sample implementation intentions).

C. COMPONENT PROCESSES OF IMPLEMENTATION INTENTIONS The mental links created by implementation intentions are expected to facilitate goal attainment on the basis of psychological processes that relate both to the anticipated situation (specified in the if‐component of the plan) and the specified response (the plan’s then‐component). As forming implementation intentions implies the selection of a critical future situation, it is assumed that the mental representation of this situation becomes highly activated (Gollwitzer, 1999). The person who forms an implementation intention selects a situation that is ripe for action to achieve the goal; the person is therefore perceptually ready to encounter this critical situation. This idea implies that processing information about the critical situation is highly proficient (Gollwitzer, 1993; Gollwitzer, Bayer, Steller, & Bargh, 2004a; Webb & Sheeran, 2004). That is, compared to those who merely form a respective goal intention, people who form implementation intentions should exhibit increased accessibility of the critical cue, and thus should be better able to detect the cue and discriminate the cue from other similar stimuli. For instance, Webb and Sheeran (2004) used a classic illusion paradigm from the psychology of language to investigate cue detection by goal intentions as compared to implementation intentions. Participants were asked to form the goal intention to count the instances of the letter f in the following piece of text: ‘‘Finished files are the/result of years of scientific/ study combined with the/experience of years.’’ (Line breaks are marked by back slashes.) The illusion resides in the fact that most people count only three fs because they miss the f in the three instances of the word ‘‘of.’’ However, when participants furnished the goal intention with a respective implementation intention (i.e., ‘‘And as soon as I see the letter f, then I’ll add one more to my count!’’) their detection of the diYcult‐to‐identity fs

84

PETER M. GOLLWITZER AND PASCHAL SHEERAN

improved. Additional experiments showed that this improved detection of critical cues did not have costs in terms of false alarms or reduced performance on identifying noncritical stimuli, even when these stimuli were quite similar to the critical cue (i.e., ambiguous cues). Increased accessibility of the specified situation should also facilitate spontaneous attention to the cue and engender better recall of the cue (Gollwitzer et al., 2004a). Spontaneous attention was demonstrated in a dichotic listening experiment. When the critical cues specified in implementation intentions were presented in the nonattended channel, participants’ shadowing performance (i.e., repeating the words presented in the attended channel) declined. Gollwitzer et al. (2004a) also showed in a cued recall experiment that the situations specified in if–then plans are better remembered compared to alternative good opportunities to act. In sum, forming an implementation intention should induce heightened sensitivity to the critical situation at each stage of information processing such that people are better able to detect, attend to, and remember specified cues when these cues are encountered later. Specifying that one will perform a particular goal‐directed response in the then‐component of a plan, at the critical moment stipulated in the if‐ component of the plan, involves a strategic abdication of eVortful action control. This is because forming an implementation intention delegates control of behavior from the self to specified situational cues that directly elicit action (i.e., implementation intentions create ‘‘instant habits’’) (Gollwitzer, 1999). Forming an if–then plan means that the person commits herself in advance to acting as soon as certain contextual constraints are satisfied. Once that situation is encountered, action initiation should proceed swiftly and eVortlessly and without requiring the person’s conscious intent. Accordingly, the execution of a behavior specified in an implementation intention should exhibit features of automaticity as identified by Bargh (1992, 1994). Automaticity commonly characterizes highly overlearned activities (e.g., driving a car, typing) including the operation of habits (Aarts & Dijksterhuis, 2000a,b; Sheeran et al., 2005a; Wood et al., 2002). Action control by implementation intentions seems to exhibit three features of automatic processes: immediacy, eYciency, and lack of conscious intent. Immediacy has been tested by means of response latencies (e.g., Webb & Sheeran, 2004) and the temporal proximity of actual performance to the time of performance specified in the implementation intention (e.g., Gollwitzer & Brandsta¨tter, 1997; Oettingen et al., 2000, Experiment 3). For instance, Gollwitzer and Brandsta¨tter had research participants watch a video presentation of a presumed Nazi who expressed racial slurs and mark good opportunities to speak up. Participants all formed the goal intention to counterargue at opportune moments when watching the video

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

85

a second time. A subset of participants also formed implementation intentions by mentally linking these critical situations with respective counterarguments. Only having marked critical opportunities (mere goal intention condition) was less eVective in promoting the immediate initiation of counterarguments compared to having also formed implementation intentions (i.e., if–then plan participants were much faster in using the marked opportunities). The eYciency of implementation intention eVects is supported by studies that varied cognitive load either through selection of the sample (e.g., schizophrenic patients, heroin addicts under withdrawal) or by experimental manipulations using dual task paradigms (e.g., Brandsta¨tter, Lengfelder, & Gollwitzer, 2001; Lengfelder & Gollwitzer, 2000). For instance, Brandsta¨tter et al. (2001) assigned heroin addiction patients the task of writing a curriculum vitae within a set time period. Forming implementation intentions that specified exactly when and where to get started with this task helped not only control participants (i.e., heroin users who were no longer experiencing withdrawal symptoms) to meet this task but also those participants who still showed withdrawal symptoms. Apparently, the eVect of implementation intentions on task achievement did not interact with the cognitive load (drug urge) experienced by the participants. Evidence for eYciency of action control by implementation intentions was also observed in experiments in which participants had to perform two tasks at the same time. The secondary task in these studies was always a Go/No Go task, whereas the primary task was either a memorization task or a tracking task (Brandsta¨tter et al., 2001, Studies 3 and 4; Lengfelder & Gollwitzer, 2001, Study 2). The implementation intention was linked to performing the Go/No Go task and the diYculty of the primary task was varied (easy vs. diYcult). The beneficial eVects of implementation intentions on performance in the secondary task were not qualified by an interaction with the primary task indicating that the operation of implementation intentions is eYcient (i.e., independent of cognitive load). Moreover, better performance was observed in the primary task during those phases of the secondary task that were guided by implementation intentions (i.e., a transfer of freed resources). Finally, there is evidence that the eVective operation of implementation intentions does not require that people be consciously aware of either the anticipated critical situation or the respective goal intention (e.g., Bayer, Moskowitz, & Gollwitzer, 2004; Sheeran, Webb, & Gollwitzer, 2005c). Bayer et al. (2004) demonstrated that conscious awareness of the specified situation was redundant in two experiments that used subliminal priming of respective cues. In one study, participants were asked to classify a series of geometric figures (e.g., circles, ellipses, squares) as rounded or angular objects by left‐ or right‐button–press responses. All participants formed the goal intention to classify the objects as fast and accurately as possible.

86

PETER M. GOLLWITZER AND PASCHAL SHEERAN

Implementation intention participants were in addition asked to make the following plan: ‘‘And if I see a triangle, then I will press the respective button immediately!’’ This implementation intention led to faster classification responses for triangles. Importantly, classification performance on all angular figures was facilitated when these figures were preceded by a subliminal triangle prime compared to a control prime (the percentage symbol, %). No such eVects were observed for goal intention participants. Moreover, Sheeran et al. (2005c) showed that people need not be consciously aware of the underlying goal intention for implementation intention eVects to occur. All participants formed the conscious goal intention to solve puzzles from the Wechsler Adult Intelligence Scale (WAIS III) as accurately as possible. Half of the participants also formed an implementation intention in relation to another dimension of performance, namely, to solve the puzzles as quickly as possible. This implementation intention manipulation was then crossed with a priming procedure that activated the goal of responding quickly outside of awareness. Speed and accuracy of responses to the puzzles was then measured. Even though participants reported no awareness of the primed goal during debriefing, findings indicated that responses were fastest when participants were primed to respond quickly and had formed respective implementation intentions. This study shows that conscious intent is not required to observe implementation intention eVects on performance. In sum, the evidence on component processes suggests that people can enhance rates of goal completion obtained by conscious and eVortful guidance of behavior (action control by goal intentions) by strategically switching to automated self‐regulation of goal striving (action control by implementation intentions).

IV. Present Review Accumulated research indicates that there is a substantial gap between people’s goal intentions and their goal achievement. This is because forming a goal intention does not prepare people suYciently for dealing with self‐ regulatory problems in initiating, maintaining, disengaging from, or overextending oneself in goal striving. Forming an implementation intention, on the other hand, spells out the when, where, and how of goal striving in advance. If–then plans are therefore thought to enhance the accessibility of the specified critical situation and induce automatic execution of the specified response. The consequence is that people should remember to act, seize good opportunities, overcome initial reluctance, suppress unwanted

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

87

responses, block detrimental self‐states and adverse contextual influences, and successfully disengage from goals without costs to self‐regulatory capability. Goal striving should be regulated eVectively, and goal achievement should thereby be facilitated. The present review tests these ideas using meta‐analysis. First, we assess the overall impact of implementation intention formation on goal achievement. We evaluate potential moderators of implementation intention eVects and test whether implementation intentions are eVective in promoting performance in diVerent domains of attainment. Second, we test the eVectiveness of implementation intentions in overcoming self‐regulatory problems that have to do with initiating goal striving, shielding goals from unwanted influences, disengaging from failing goals, and conserving self‐regulatory capability. Finally, we calibrate the eVect sizes for the component (if–then) processes of implementation intentions. A. METHOD 1. Sample of Studies Several methods were used to generate the sample of studies: (1) computerized searches were conducted on social scientific and medical databases (PsychINFO, Social Science Citation Index and Conference Papers Index [Web of Knowledge], Medline, Index Medicus, and Dissertation Abstracts International Online) from January 1990 to December 2003 using the keywords implementation intention(s) and plan(s), (2) references in each article identified above were evaluated for inclusion, and (3) authors were contacted and requests were made for unpublished studies and studies in press. Studies were included in the review if (a) the implementation intention formed by participants specified the performance of a goal‐directed response upon encountering an internal or external critical cue and (b) a statistical association between the formation of an implementation intention and an outcome variable could be retrieved (or obtained). Using these criteria, 94 tests of the relationship between implementation intentions and goal achievement could be included in the meta‐analysis. The focal goal and eVect size for each test are presented in Table I. The 94 independent tests come from 63 reports (these reports are preceded by an asterisk in the reference list). 2. Meta‐Analytic Strategy The eVect size estimate used here was d, which is the diVerence between the means for two groups divided by a pooled standard deviation and corrected for small sample bias (Hedges & Olkin, 1985). We subtracted the control

TABLE I STUDIES OF

THE IMPACT OF IMPLEMENTATION INTENTION

FORMATION

ON

GOAL ACHIEVEMENT Author(s) Aarts, Dijksterhuis, and Midden (1999) Ajzen, Czasch, and Flood (2002) Armitage (2004) Bagozzi, Dholakia, and Basuroy (2003) Bamberg (2000) Bamberg (2002) Bayer, Jaudas, and Gollwitzer (2002) Bayer, Moskowitz, and Gollwitzer (2004) Study 1 Bayer et al. (2004) Study 2 Brandsta¨tter et al. (2003) Brandsta¨tter, Lengfelder, and Gollwitzer (2001) Study 1 Brandsta¨tter et al. (2001) Study 2 Brandsta¨tter et al. (2001) Study 3 Brandsta¨tter et al. (2001) Study 4 Bru¨wer, Bayer, and Gollwitzer (2002) Bulgarella, Oettingen, and Gollwitzer (2003) Chasteen, Park, and Schwarz (2001) Study 1 Dewitte, Verguts, and Lens (2003) Study 1 Dewitte et al. (2003) Study 2 Dewitte et al. (2003) Study 3 Dholakia and Bagozzi (2003) Study 1 Dholakia and Bagozzi (2003) Study 2 Dholakia and Bagozzi (2003) Study 3 Dholakia and Bagozzi (2002) Study 1 Dholakia and Bagozzi (2002) Study 2 DieVendorf and Lord (2003)

Goal Collect a coupon/latencies to cues Rate TV newscasts Eat a low‐fat diet Personal goals Public transport use Organic food purchase Task switch Retaliation behavior Classification of geometric shapes Initiation of vocational retraining Write a curriculum vitae Go/No Go task Go/No Go task Go/No Go task Simon eVect task Numerical judgment dilemma Prospective memory task Ten personal goals Ten personal goals Ten personal goals Reading assignment Visit website Visit website Product/service purchase Personal goals Human resource task

N

d

40 102 126 153 90 160 40 61 61 126 41 61 68 33 34 34 68 15 14 16 102 138 179 131 169 170

.80 .47 .30 .68 .45 .32 .85 .98 .49 .72 1.32 .93 .63 1.10 2.20 .80 .82 .10 .10 .18 .56 .43 .82 .49 .41 .41

Einstein, McDaniel, Williford, Pagan, and Dismukes (2003) Gillholm, Ettema, Sellart, and Garling (2004a) Study 1 Gillholm et al. (1999) Study 2 Gollwitzer and Bayer (2004a) Study 1 Gollwitzer and Bayer (2004a) Study 2 Gollwitzer and Bayer (2004a) Study 3 Gollwitzer, Bayer, Steller, and Bargh (2002) Study 1 Gollwitzer et al. (2002) Study 2 Gollwitzer et al. (2002) Study 3 Gollwitzer and Brandsta¨tter (1997) Study 1 Gollwitzer and Brandsta¨tter (1997) Study 2 Gollwitzer and Brandsta¨tter (1997) Study 3 Gollwitzer, Sheeran, and Seifert (2004c) Study 3 Gollwitzer, Sheeran, Tro¨tschel, and Webb (2004d) Study 1 Gollwitzer et al. (2004d) Study 2 Gollwitzer et al. (2004d) Study 3 Gollwitzer et al. (2004d) Study 4 Gollwitzer, Tro¨tschel, Bayer, and Sumner (2004e) Study 1 Gollwitzer et al. (2004e) Study 2 Henderson, Gollwitzer, and Oettingen (2004) Study 1 Henderson et al. (2004) Study 2 Henderson et al. (2004) Study 3 Holland, Aarts, and Langendam (in press) Koestner, Downie, Horberg, and Hata (2002a) Koestner, Lekes, Powers, and Chicoine (2002b) Study 1 Koestner et al. (2002b) Study 2 Koole and Van’t Spijjker (2000) Lengfelder and Gollwitzer (2001) Study 2 Lippke and Ziegelmann (2002) Milne, Orbell, and Sheeran (2002) Milne and Sheeran (2002a)

Prospective memory task Mundane activities Mail response forms Degree of perspective taking Social loafing task Mood‐induced gender stereotyping Cue detection Dichotic listening task Recall of specified cues Personal goals Complete a written report Counter arguments to racist remarks Solving law cases Driving simulation dilemma Accessibility of drinking behavior Helping behavior dilemma Cooperation in resource dilemma Stereotype rebound task Anagram performance Escalation of commitment task Escalation of commitment task Escalation of commitment task Recycling behaviors Personal goals Personal goals New Year resolutions Write a report Go/No Go task Exercise Exercise Testicular self‐examination

48 28 48 34 42 54

.24 1.01 .02 1.16 .75 .80

46 55 79 70 36 60 55 69 72 60 60 30 31 87 96 187 54 106 106 38 80 67 88 248 432

.82 .72 .87 .43 .90 .52 1.10 .98 .49 1.25 .90 1.81 1.12 .52 .54 .43 1.42 .39 .39 .32 .75 .80 .18 1.25 .43 continues

TABLE I Continued Author(s) Milne and Sheeran (2002b) Milne and Sheeran (2002c) MurgraV, White, and Phillips (1996) Oettingen, Ho¨nig, and Gollwitzer (2000) Study 2 Oettingen et al. (2000) Study 3 Orbell, Hodgkins, and Sheeran (1997) Orbell and Sheeran (1999) Study 3 Orbell and Sheeran (1999) Study 4 Orbell and Sheeran (2000) Prestwich, Lawton, and Conner (2003a) Prestwich, Lawton, and Conner (2003b) Rise, Thompson, and Verplanken (2003) Schaal (1993) Sheeran and Milne (2003) Sheeran and Orbell (1999) Study 1 Sheeran and Orbell (1999) Study 2 Sheeran and Orbell (2000) Sheeran and Silverman (2003) Sheeran and Webb (2003) Sheeran, Webb, and Gollwitzer (2005c) Study 1

Goal Persistence with boring task Visit study skills website Binge drinking Compose a CV Complete a project Breast self‐examination Reduce snack food consumption Reduce snack food consumption Recovery of functional activities Personal goals Exercise Exercise and recycling Arithmetic task Reduce snack good consumption Vitamin supplement use Vitamin supplement use Cervical cancer screening Attendance at workplace safety training Stroop performance Independent study

N 66 183 102 20 25 155 111 93 64 79 58 112 40 129 78 37 104 271 48 85

d 1.35 .87 .68 1.25 .80 1.22 .61 .45 .70 .54 .68 1.58 .93 .49 .35 .59 .58 .52 1.62 .35

Sheeran et al. (2005c) Study 2 Sniehotta, Scholtz, and Schwarzer (2002a) Sniehotta, Scholtz, and Schwarzer (2002b) Steadman and Quine (2000) Steadman and Quine (in press) Stephens and Conner (1999) Tro¨tschel and Gollwitzer (2004) Study 1 Tro¨tschel and Gollwitzer (2004) Study 2 Verplanken and Faes (1999) Webb and Sheeran (2003) Study 1 Webb and Sheeran (2003) Study 2 Webb and Sheeran (2004) Study 1 Webb and Sheeran (2004) Study 2 Webb and Sheeran (2004) Study 3 Webb and Sheeran (in press, b) Study 1 Webb and Sheeran (in press, b) Study 2 Williams (2003)

Performance on puzzle task Cardiac rehabilitation exercise/training Physical activity Vitamin supplement use Testicular self‐examination Resist taking up smoking Framing eVects in negotiation Framing eVects in negotiation Healthy eating Persistence with unsolvable puzzles Stroop performance Letter identification Number identification Number identification Personal goals Academic performance Return postcard

40 74 65 174 75 124 43 57 102 32 57 54 42 53 646 129 60

.70 .61 .70 .32 .54 .37 .82 .87 .47 .87 .87 .63 .75 .82 .75 .56 .56

92

PETER M. GOLLWITZER AND PASCHAL SHEERAN

group mean from the mean for the implementation intention group so that a positive d value indicates the benefit in performance conferred by forming an implementation intention. The average eVect size was computed by averaging the d values with each d weighted by the reciprocal of its variance. As a test of significance, 95% confidence intervals were computed around each mean. Where studies reported r, t, F,
2, or contingency tables, we transformed values into ds using the formulas supplied by Hedges and Olkin (1985) and Hunter (1990). Homogeneity of eVect sizes was tested by means of the Q statistic that has an approximate chi‐square distribution with k‐1 degrees of freedom, k being the number of eVect sizes. When Q is significant ( p < .05), eVect sizes are heterogeneous. 3. Multiple Measures and Multiple Tests Several papers contained data from more than one sample or reported eVect sizes for multiple measures of an independent variable or multiple measures of the dependent variable. We tried to take advantage of the richness of these data without violating the assumption of independence that underlies the validity of meta‐analysis. Data from independent samples were, therefore, treated as separate units. In the case of multiple measures of independent or dependent variables, the average d within each study was the unit of analysis. Where studies contained multiple nonindependent samples, we used the conservative strategy of computing the weighted average eVect size and using the smallest N in the analysis in order to determine the overall eVect size for that study (Sheeran, Abraham, & Orbell, 1999). For example, Holland, Aarts, and Langendam (in press) examined the impact of forming implementation intentions on objective measures of recycling old paper (N ¼ 54, d ¼ 1.32) and recycling plastic cups (N ¼ 109, d ¼ 1.50). The eVect size used to represent this study is the weighted average of the two eVects (d ¼ 1.42), and the sample size is 54.

B. RESULTS 1. Overall EVect Size The overall impact of forming implementation intentions on goal achievement was d ¼ .65 based on k ¼ 94 tests that involved 8461 participants. This eVect had a 95% confidence interval from .60 to .70. According to Cohen’s (1992) power primer, d ¼ .20 is a ‘‘small’’ eVect, d ¼ .50 is a ‘‘medium’’‐sized eVect, whereas d ¼ .80 is a ‘‘large’’ eVect. Thus, the eVect size that characterizes the impact of if–then planning on goal achievement is of medium‐to‐large magnitude.

93

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

a. Moderators of Implementation Intentions EVects. We examined methodological moderators of the relationship between implementation intentions and goal achievement in order to ensure that the eVect sizes for if–then plans were not exaggerated by weaker methods (e.g., correlational rather than experimental designs). The homogeneity test encouraged a search for moderators as there was significant variability in the eVect sizes obtained in individual studies, Q(93) ¼ 173.46, p < .001. Moderator analyses were conducted for three methodological factors: type of sample, study design (correlational vs. experimental), and measurement of goal attainment (self‐report vs. objective). Table II shows that most tests of implementation intention eVects were conducted among university students (k ¼ 79) though eight tests sampled members of the public and there were two tests of children/young people. Four tests were conducted with physically ill people, and there were three tests among people with psychological problems (schizophrenic patients, frontal lobe patients, and heroin addicts). Findings showed that, excluding

IMPACT

OF

TABLE II METHODOLOGICAL FACTORS ON EFFECT SIZES

Factor Sample General public Children/young adults People with physical illness People with psychological problems Schizophrenic patients Brain‐injured patients Heroin addicts University students Design Correlational Experimental Measurement Self‐report Objective Publication status Unpublished Published

FOR IMPLEMENTATION INTENTIONS

N

k

d

95% CI

Q

1076 144 291

8 2 4

.58 .47 .52

[.45, .70] [.14, .85] [.28, .77]

14.09* 2.38 3.66

20 34 41 6855

1 1 1 79

1.01 .87 1.32 .65

[.61, .70]

147.93***

1688 6773

11 83

.70 .65

[.61, .82] [.61, .70]

20.23* 151.59***

4488 3973

36 58

.63 .67

[.58, .70] [.61, .74]

80.96*** 92.32**

3759 4702

46 48

.67 .65

[.61, .72] [.59, 70]

75.13*** 98.23***

*p < .05, **p < .01, ***p < .001. Note: The sum of k equals 96 because data from two diVerent samples were disaggregated in two studies (Lengfelder & Gollwitzer, 2001, Study 2; Brandsta¨tter et al., 2001, Study 2). N ¼ sample size; k ¼ number of independent eVects; d ¼ eVect size; CI ¼ confidence interval; Q ¼ homogeneity statistic.

94

PETER M. GOLLWITZER AND PASCHAL SHEERAN

people with psychological problems, eVect sizes were of medium size and equivalent magnitude across samples, Q(3) ¼ 4.01, p > .25. Implementation intentions appeared to have stronger eVects for people with psychological problems compared to the other groups; this diVerence proved significant when subgroup comparisons were conducted (ds ¼ 1.10 and .66, respectively), Q(1) ¼ 4.54, p < .04. This finding suggests that forming implementation intentions is especially beneficial to goal attainment among people who have diYculties with regulating their behavior. Findings indicated that implementation intentions were similarly eVective whether the study design was correlational or experimental (ds ¼ .70 and .65, respectively), Q(1) ¼ 1.93, p > .16. Moreover, the impact of implementation intentions on goal achievement was not exaggerated by overreliance on self‐report measures of behavior. Implementation intentions had similar eVects whether or not the outcome was measured objectively (d ¼ .67) or by self‐report (d ¼ .63), Q(1) ¼ 2.18, p ¼ .14. We also examined whether publication status was associated with the strength of observed implementation intention eVects. Forty‐nine percent of the eVects that could be included in the review were unpublished (k ¼ 46), and it is possible that unpublished tests may be of poorer methodological quality than are published tests (Rosenthal, 1984). This could mean that the overall estimate of eVect size is inflated. However, there was no diVerence in eVect sizes from published versus unpublished tests (ds ¼ .65 and .67, respectively), Q(1) ¼ 1.53, p ¼ .22. b. EVect Sizes for DiVerent Goal Domains. To test the generality of implementation intention eVects, values of d were computed for diVerent goal domains. We drew on the classification of domains used in Kim and Hunter’s (1993) comprehensive meta‐analysis of the impact of topic on attitude–behavior relations in order to categorize the goals (Canary & Seibold, 1984). EVects were available for seven of the domains identified by Kim and Hunter (health, academic, consumer, environmental, prosocial, antiracist, and laboratory tasks). An eighth category was personal goals because several studies asked participants to nominate their own desired outcomes that were then furnished with implementation intentions (or not). Findings indicated that implementation intentions had medium or large eVects for all domains (Table III). The goals examined most frequently related to laboratory tasks (k ¼ 38) and health (k ¼ 23). There were large eVects for antiracist, prosocial, and environmental behaviors (ds ¼ .87, 1.01, and 1.12, respectively). There were medium‐to‐large eVects for laboratory tasks and academic achievement (ds ¼ .70 and .72, respectively) and medium‐sized eVects for consumer behaviors, health behaviors, and personal goals. Overall, Table III indicates that implementation intentions have reliable eVects for a wide range of goal domains.

95

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

META‐ANALYSIS

TABLE III EFFECTS

OF IMPLEMENTATION INTENTION

FOR

DIFFERENT GOAL DOMAINS

Goal domain

N

k

d

95% CI


2

Consumer Environmental Antiracist Prosocial Academic Personal Health Laboratory

291 256 144 254 836 1391 2861 2428

2 3 3 5 9 11 23 38

.41 1.12 .87 1.01 .72 .58 .59 .70

[.16, .65] [.85, 1.42] [.52, 1.25] [.72, 1.28] [.56, .87] [.47, .70] [.52, .67] [.61, .79]

0.50 17.07*** 5.30 1.91 7.42 14.11 47.92*** 59.90*

*p < .05, **p < .01, ***p < .001. Note: N ¼ sample size; k ¼ number of independent eVects; d ¼ eVect size; CI ¼ confidence interval; Q ¼ homogeneity statistic.

2. EVect Sizes for DiVerent Self‐Regulatory Problems Table IV presents the eVect sizes for implementation intentions for self‐ regulatory problems associated with initiating goal striving, shielding goals from unwanted influences, disengaging from failing goals, and conserving self‐regulatory capability. Three problems that militate against action initiation are remembering to act, seizing opportunities, and overcoming initial reluctance. There were k ¼ 11, 20, and 21 tests of these problems, respectively. For all three problems, eVect sizes for implementation intentions were of medium‐to‐large magnitude (ds ¼ .54, .61, and .65, respectively). These findings indicate that implementation intentions help to ensure that people (1) do not forget to perform intended actions, (2) do not miss good opportunities to initiate action, and (3) do not fail to act because they are swayed by short‐term considerations. The overall eVect size was d ¼ .61 indicating that implementation intention formation makes an important diVerence to whether or not people initiate goal striving successfully. We identified three self‐regulatory tasks in relation to shielding goals from unwanted influences, namely, suppressing unwanted responses, blocking detrimental self‐states, and blocking adverse contextual influences. Implementation intentions proved beneficial for all three tasks. First, implementation intentions had a large eVect on suppressing unwanted attention responses (d ¼ .90) and had a medium eVect on suppressing unwanted behavioral responses (d ¼ .54). Second, implementation intentions had a large eVect on goal achievement even when participants were in a detrimental self‐state (d ¼ 1.10). Implementation intentions promoted performance

96

PETER M. GOLLWITZER AND PASCHAL SHEERAN

META‐ANALYSIS

TABLE IV EFFECTS FOR DIFFERENT SELF‐REGULATORY TASKS IN GOAL STRIVING

OF IMPLEMENTATION INTENTION

Self‐regulatory tasks

N

k

d

983 2270 2588 5841

11 20 21 52

.54 .61 .65 .61

Shielding Goal Striving from Unwanted Influences Suppressing unwanted responses Attention responses 184 Behavioral responses 559 Blocking detrimental self‐states 248 Blocking adverse contextual influences 405 Overall 1396

3 5 5 8 21

.90 .54 1.10 .93 .77

[.61, 1.25] [.35, .70] [.80, 1.39] [.70, 1.16] [.67, .87]

Initiating Goal Striving Remembering to act Seizing opportunities Overcoming initial reluctance Overall

95% CI

[.45, [.52, [.56, [.56,

.72] .70] .72] .67]

Q

16.79 23.75 72.82*** 114.21***

5.75 1.59 4.40 2.22 27.60

Disengaging from Futile Goal Striving

370

3

.47

[.26, .70]

0.22

Conserving Self‐Regulatory Capability

93

3

1.28

[.77, 1.76]

3.08

*p < .05, **p < .01, ***p < .001. Note: N ¼ sample size; k ¼ number of independent eVects; d ¼ eVect size; CI ¼ confidence interval; Q ¼ homogeneity statistic.

when participants had incomplete self‐definitions (d ¼ 1.12, k ¼ 2), were ego‐depleted (d ¼ 1.22, k ¼ 2), or were in a good mood and therefore liable to stereotyping (d ¼ .80, k ¼ 1). Third, a large eVect was obtained for implementation intentions when goal achievement was blocked by adverse contextual influences (d ¼ .93). Forming an implementation intention meant that participants were able to overcome the characteristic impacts of deindividuation on social loafing (d ¼ .80, k ¼ 1), loss frames on negotiation (d ¼ .85, k ¼ 2), and situational activation of goals that were antagonistic to the focal goal striving (d ¼ 0.98, k ¼ 5). In sum, implementation intentions are eVective in blocking adverse contextual influences. As well as initiating goal striving and shielding ongoing goal pursuits from unwanted influences, people must also disengage from goal striving when such striving is no longer productive. Three studies tested the eYcacy of implementation intentions in helping people disengage from failing goals. Findings indicated that implementation intention eVects were of approximately medium size (d ¼ .47). The final self‐regulatory problem is whether implementation intention formation conserves people’s capability for future goal striving. Findings showed that, even when the experimental settings involved two phases and the initial task was known to engender performance deficits on the subsequent task, implementation intentions still had a large eVect on performance

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

97

(d ¼ 1.28). Participants who formed implementation intentions to control initial performance did not exhibit ego‐depletion or stereotype rebound. In both cases, eVect sizes for implementation intentions were positive and large (ds ¼ .87, and 1.81, respectively). 3. Component Processes of Implementation Intentions Forming implementation intentions should activate the mental representation of the specified cues (if‐component) and automate responding to these cues (as specified in the then‐component). Table V shows that implementation intentions had large eVects on the detection, discrimination, and accessibility of critical cues (ds ¼ .72, .82, and .95, respectively) and on the attention paid to, and memory for, those cues (ds ¼ .72 and .87, respectively). The overall eVect size for processes related to the if‐component of the plan was large (d ¼ .80) indicating that implementation intentions are associated with highly proficient processing of critical cues. There were 7, 7, and 3 tests, respectively, of the immediacy and eYciency of, and redundancy of conscious intent for action control by implementation intentions. Implementation intentions showed large eVects for each of these three key features of automaticity. If–then plans produced more immediate responding (d ¼ .77), were eYcient with respect to cognitive resources (d ¼ .85), and proceeded without the need for conscious intent (d ¼ .72). These findings provide strong support for the postulated automaticity of action control induced by implementation intentions.

META‐ANALYSIS

OF

TABLE V COMPONENT PROCESSES OF IMPLEMENTATION INTENTIONS

Component process

N

k

d

95% CI

Q

If‐component Cue detection Cue discrimination Cue accessibility Attention to cue Memory for cue Overall

100 53 40 55 79 327

2 1 1 1 1 6

.72 .82 .95 .72 .87 .80

[.30, 1.16]

0.23

[.56, 1.04]

0.89

Then‐component Immediacy EYciency Lack of intent

363 278 122

7 7 3

.77 .85 .72

[.49, .98] [.58, 1.12] [.39, 1.07]

2.09 26.45 1.62

*p < .05, **p < .01, ***p < .001. Note: N ¼ sample size; k ¼ number of independent eVects; d ¼ eVect size; CI ¼ confidence interval; Q ¼ homogeneity statistic.

98

PETER M. GOLLWITZER AND PASCHAL SHEERAN

C. DISCUSSION Findings from 94 studies involving more than 8000 participants indicated that the eVect size associated with the impact of implementation intention formation on goal attainment is d ¼ .65, an eVect of medium‐to‐large magnitude (Cohen, 1992). This eVect size is impressive because d ¼ .65 represents the diVerence in goal achievement engendered by furnishing a goal intention with a respective implementation intention compared to the formation of a goal intention on its own. The implication is that if–then planning substantially increases the likelihood of attaining one’s goals. Several features of the meta‐analysis serve to underline the eVectiveness of implementation intentions in promoting goal achievement. First, it is unlikely that the review suVers from the ‘‘file drawer problem’’ (e.g., Rosenthal, 1984) as 49% of the included tests were unpublished. Moreover, publication status had no impact on the eVect size obtained for implementation intentions. Second, 88% of tests involved experimental designs (i.e., random assignment of participants to implementation intention formation) that increase confidence in the findings. It was also the case that the eVect sizes obtained for experimental versus correlational studies were equivalent (unlike meta‐analyses of the impact of goal intentions on goal achievement in which experimental tests show much weaker eVects compared to correlational tests; Webb & Sheeran, in press a). Third, the composition of the sample generally did not moderate implementation intention eVects. If–then plans were similarly eVective in promoting goal achievement among students, members of the general public, and people with physical illness. Fourth, when we used Kim and Hunter’s (1993) system to classify domains of attainment, implementation intentions were shown to have medium or large eVects for a wide variety of goals. Finally, the eVectiveness of implementation intentions was not exaggerated by overreliance on self‐report measures of behavior. The eVect size for implementation intentions was of equivalent magnitude in studies in which objective measures of performance were used. In sum, implementation intentions seem to have benefited goal achievement no matter how one looks at the data. 1. Implementation Intentions and Self‐Regulatory Tasks in Goal Striving Whereas traditional theories of goal pursuit assumed that strong goal intentions are a suYcient determinant of goal achievement (e.g., Ajzen, 1991; Atkinson, 1957; Fishbein, 1980; Locke & Latham, 1990; Rogers, 1983), the present research started from the position that there is a substantial gap between intentions and action as there are numerous problems of goal striving that need to be solved even if people hold strong binding goals (Gollwitzer,

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

99

1990, 1993, 1996, 1999; Heckhausen & Gollwitzer, 1987; Gollwitzer et al., 2005; Orbell & Sheeran, 1998; Sheeran, 2002). We analyzed the intention– behavior gap in terms of four key self‐regulatory tasks: initiating goal striving, shielding ongoing goal pursuit from unwanted influences, disengaging from unproductive goal striving, and conserving self‐regulatory capability. Findings indicated that if–then planning facilitated initiation of goal striving no matter whether getting started was an issue of remembering to act, seizing good opportunities, or overcoming initial reluctance. Although fewer studies were conducted on the issue of shielding goal striving, the beneficial eVects of implementation intentions were also strong. Forming if–then plans helped with diVerent problems of maintaining an ongoing wanted (focal) goal pursuit. The implementation intentions used were geared either at suppressing unwanted attention and behavioral responses, or toward spelling out the focal goal striving and thereby blocking detrimental self‐states and adverse contextual influences. It is worth noting that various detrimental self‐ states and adverse contextual influences have been scrutinized and the awareness of their presence varied between studies, as did the awareness of their potential negative impact on the person’s goal striving. Three studies investigated disengagement from failing courses of action (Henderson et al., 2004) using standard escalation of commitment paradigms (i.e., escalation of commitment was induced by instigating the justification motive; e.g., Bobocel & Meyer, 1994). Even though it is well established that it is very diYcult to overcome strong self‐justification concerns and thus halt escalation of commitment, implementation intention formation was eVective in doing so and produced an eVect of approximately medium size. Thus, implementation intentions provide a useful means for successfully bringing futile goal striving to a close. The final self‐regulatory task in goal striving is conserving capability for pursuing subsequent goals once an initial goal has been completed. Whereas action control by goal intentions has been shown to generate ironic rebound and ego‐depletion eVects for subsequent task performance, action control by implementation intentions did not produce such costs. In other words, controlling goal striving with implementation intentions allows people to move on to subsequent goal striving without these self‐regulatory handicaps; compared to self‐regulation by goal intentions, self‐regulation by implementation intentions conserves rather than diminishes capability for further goal striving. 2. Psychological Processes Underlying Implementation Intention EVects Several studies explored the if‐ and then‐component processes of implementation intentions. Findings strongly support the postulated mechanisms (Gollwitzer, 1999). Apparently, specifying a situational cue in the if‐component

100

PETER M. GOLLWITZER AND PASCHAL SHEERAN

of an implementation intention creates a heightened activation of the respective mental representation of the situation. The implied ease of accessibility could be observed in respective lexical decision, perceptual detection, cue discrimination, and memory performances. Moreover, implementation intention participants attended to specified critical cues even when the cues were presented on the nonattended channel in a dichotic listening task. Although studies used diVerent paradigms to assess heightened activation, the eVect sizes obtained were uniformly large. These findings strongly suggest that having selected a situational cue for acting toward one’s goal, it is hard for the person to overlook this opportunity. This contrasts with the predicament of the person who has only formed a goal intention and thus needs to actively search for and identify good opportunities to act (Sheeran, Milne, Webb, & Gollwitzer, 2005b). At the same time, specifying an eVective goal‐directed response in the then‐ component of the plan endows the control of this response with features of automaticity. The three features of automaticity that have been analyzed in various diVerent studies are immediacy, eYciency, and lack of awareness (i.e., conscious intent is not required). For all three features, eVect sizes were large. Apparently, furnishing goal intentions with implementation intentions switches the mode of goal‐directed behavior from hesitant to immediate, from eVortful to eYcient, and from a conscious intent to act to direct response elicitation by the situation. Whereas the person who has only formed a goal intention still has to deliberate in situ about what goal‐directed response to undertake and/or energize the self to perform it, forming an implementation intention means deciding these issues in advance, thereby delegating the control of goal‐directed behavior to specified situational cues. Once these cues are encountered, action initiation is triggered automatically. One might wonder whether a change in motivational factors (strength of respective goal intentions and/or self‐eYcacy) due to if–then plan formation may explain implementation intention eVects on goal attainment—in addition to, or even instead of, the postulated component processes. However, at least two lines of research contradict this idea. First, studies that measured strength of goal intentions (commitment) or self‐eYcacy both before and after respective implementation intention inductions found no evidence that if–then plan formation increased scores on these variables in either within‐ participants analyses (diVerences within the if–then plan group over time) or between‐participants analyses (diVerences between the if–then plan and control group at either time‐point) (e.g., Brandsta¨tter et al., 2001, Study 1; Milne et al., 2002; Oettingen et al., 2000, Study 2; Orbell et al., 1997; Sheeran & Orbell, 1999; Sheeran et al., 2005c, Study 1). Second, implementation intention formation enhanced rates of goal attainment even when participants had extremely high scores on goal intention and self‐eYcacy prior to

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

101

plan formation. For instance, Sheeran and Orbell (2000) found that if–then planning increased attendance for cervical cancer screening even though the preintervention means for the if–then plan group were 4.60 and 4.63, respectively, on 1–5 scales. It is implausible to attribute the observed 33% improvement in attendance behavior among implementation intention participants to postmanipulation increases in goal intentions or self‐eYcacy (for equivalent findings see also Sheeran & Orbell, 1999; Verplanken & Faes, 1999). These results, together with findings showing that implementation intention eVects do not exhibit the temporal decline of motivational interventions (e.g., Sheeran & Silverman, 2003; Sheeran et al., 2005b) and actually show stronger eVects for diYcult‐to‐implement as compared to easy‐to‐implement goals (e.g., Gollwitzer & Brandsta¨tter, 1997, Study 1), all indicate that increases in goal intention strength and self‐eYcacy as a consequence of if–then plan formation cannot explain implementation intention eVects on goal achievement. 3. Implementation Intentions in Everyday Life Two findings help to clarify when implementation intention formation is likely to especially benefit goal attainment. First, if–then planning has a significantly larger eVect size among people who are known to have problems with action control (e.g., frontal lobe patients, schizophrenics). Second, implementation intentions exhibit a noticeably large eVect size in tasks that are known to overextend people’s capability to regulate their behavior (i.e., in ego‐depletion and ironic rebound paradigms; d ¼ 1.28). These findings speak to the idea that the presence of problems in goal striving is an important determinant of the strength of implementation intention eVects. If the set goal is extremely easy to initiate and pursue, then simply forming the respective goal intention could satisfactorily facilitate goal achievement; in such instances, it is possible that forming an implementation intention may confer little additional benefit. If, on the other hand, person characteristics or task features make it diYcult to execute goal‐directed behaviors, then it is especially advantageous to engage in if–then planning. That is, forming implementation intentions is most likely to benefit goal achievement when regulating the behavior is diYcult or people have chronic diYculties in regulating their behavior. In the light of this analysis, and the overall support obtained in this review for the beneficial impact of implementation intentions on goal achievement, can we conclude that if–then planning will facilitate such attainment under any circumstances? In other words, is forming implementation intentions a foolproof self‐regulatory strategy of goal striving? In everyday life, people may fail to form eVective implementation intentions due to unfortunate

102

PETER M. GOLLWITZER AND PASCHAL SHEERAN

specifications of opportunities and goal‐directed responses. For instance, a person may identify an opportunity that hardly ever arises (e.g., when one rarely has the choice between walking vs. taking the elevator), or an opportunity in which it turns out to be impossible to act toward one’s goal (e.g., one’s boss insists that you ride the elevator together to discuss work). Similarly, a person may specify a behavior that has limited instrumentality with respect to reaching the goal (e.g., taking the stairs instead of the elevator is unlikely, on its own, to achieve the superordinate goal of reducing weight) or a behavior that, in reality, proves impossible for the person to perform (e.g., walk up 60 flights of stairs to one’s oYce). In addition, if–then plans may not be very eVective because opportunities and responses are not specified precisely. For example, a plan that specifies ‘‘eat healthily’’ in the then‐component and ‘‘tomorrow’’ in the if‐component has hardly spelled out an unambiguous opportunity to act or a specific goal‐ directed response to initiate—the person still has to identify a particular behavior to perform in a particular situation to facilitate goal achievement (e.g., order a salad at lunch time tomorrow in my usual restaurant). Having to thus deliberate about when, where, and what to do in situ means that the person is unlikely to garner much benefit from the enhanced activation of critical cues or automation of responding conferred by forming precise if–then plans; the person seems no better oV than having merely formed the goal intention to ‘‘eat healthily tomorrow.’’ In sum, implementation intention formation should prove useful in promoting goal achievement provided components of the plan are precise (i.e., deliberation about appropriate opportunities and responses is not required in situ), viable (i.e., the specified situation will be encountered, the specified response can be performed), and instrumental (i.e., the specified situation permits action, and the specified response facilitates goal achievement). How often the if–then plans fashioned in people’s everyday lives satisfy these conditions is an empirical issue. Finally, it is important that people who specify obstacles in the if‐component of their implementation intentions select those barriers and distractions that most hinder goal completion. In other words, it matters that people specify those obstacles that do indeed undermine goal striving. Research has demonstrated that the mental exercise of juxtaposing the desired future with the present negative reality (i.e., mental contrasting) is a particularly eVective strategy for discovering powerful barriers and hindrances that stand in the way of realizing desired outcomes (Oettingen, 2000; Oettingen et al., 2001). Accordingly, inviting people to engage in mental constrasting prior to if–then planning should ensure that people gear their implementation intentions to precisely those obstacles that present the greatest obstruction to goal attainment.

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

103

4. Do Implementation Intentions Engender Rigid Goal Striving? Assuming that implementation intentions create strong links between anticipated situations and goal‐directed behaviors, does this mean that implementation intention formation undermines performance when flexible goal striving is called for? The idea that implementation intentions could engender costs in terms of rigidity has at least three aspects. First, action control by implementation intentions could be rigid in the sense that goal striving no longer takes into account the state (activation, strength) of participants’ goal intentions. Research does not support this concern, however. Several studies have shown that goal intentions moderate the impact of implementation intentions on goal attainment such that strong eVects of if–then plans only emerge when participants hold strong respective goal intentions (e.g., Koestner et al., 2002b; Orbell et al., 1997; Sheeran et al., 2005c, Study 1). Similarly, studies that either activated (Bayer et al., 2004; Sheeran et al., 2005c, Study 2) or deactivated (Seehausen, Bayer, & Gollwitzer, 1994, cited in Gollwitzer, 1996) relevant goal intentions indicate that implementation intentions only aVected performance when the respective goal intention was activated. For example, Sheeran et al. (2005c) showed that an if–then plan to enhance speed of responding on a puzzle task only aVected response times when the goal intention to respond quickly had been primed in the situation. Thus, action control by implementation intentions does not involve a mechanistic elicitation of action in the presence of environmental cues but rather respects the presence versus absence of activated strong goal intentions. Apparently, the automaticity instigated by if–then plans is goal‐dependent (Bargh, 1992, 1994)—concerns that if–then plans could engender rigid adherence to a course of action that does not serve a person’s goals seem unfounded. The second aspect of rigidity concerns the possibility that implementation intentions could facilitate one aspect of goal striving but do so at the expense of other aspects of goal striving. That is, forming an if–then plan to promote one dimension of performance could consume self‐regulatory resources and thereby engender inflexible performance on other dimensions; as a consequence, overall goal attainment might be compromised. Again, evidence seems to contradict this idea. For example, although Sheeran et al. (2005c) found that implementation intentions enhanced response times on a puzzle task, accuracy of responses was not compromised. Similarly, Gollwitzer and Bayer (2000) showed that implementation intentions not only increased the number of solutions generated in a creativity task but also enhanced the conceptual variety of those solutions. These findings indicate that action control by implementation intentions does not induce rigidity in terms of inevitable trade‐oVs between dimensions of performance (speed vs. accuracy, quantity vs. quality). Rather, the automation of one aspect of goal striving

104

PETER M. GOLLWITZER AND PASCHAL SHEERAN

seems to free up cognitive capacity such that other aspects of the focal striving are not compromised, and even can be enhanced (Brandsta¨tter et al., 2001). The third aspect of potential rigidity concerns whether implementation intention participants refrain from using alternative good opportunities to act toward the goal by insisting on acting only when the critical situation specified in the if‐part of the implementation intention is encountered. Several features of if–then plans suggest that such rigid adherence to specified opportunities is unlikely. Because implementation intentions respect the activation and strength of participants’ superordinate goal intentions, participants who have formed if–then plans should still be sensitive to the issue of identifying good opportunities to act. Moreover, because action control by implementation intentions is eYcient and conserves self‐regulatory capability, if–then planners should be in a good position to eVectively process information about alternative opportunities, and to seize those opportunities judged suitable for execution of behavior. In sum, implementation intentions do not seem to engender rigid self‐regulation in terms of mechanistic situational control, performance trade‐oVs, or neglecting suitable alternative opportunities to move toward the goal. Finally, there may be a further fourth issue related to rigidity, this one having to do with how people deal with having acted on a faulty if–then plan. We do not know yet what happens when people recognize that they have formed an if–then plan that failed to lead to goal attainment (or even produced negative outcomes). Do people stubbornly adhere to the faulty if–then plan, or readily modify the if‐ and then‐components of that plan, or do they even completely refrain from forming if–then plans? Also one wonders how the explicitness of the failure feedback and the strength of the respective goal intention aVect whether people will adhere to or modify the plan, or stay away from planning altogether. 5. Future Research on If–Then Plans Although 94 independent tests of implementation intention eVects on goal achievement were examined in this chapter, further research is warranted to exploit the benefits of implementation intentions in facilitating goal attainment and to enhance understanding of this mode of action control. Findings from 52 and 21 studies, respectively, showed that implementation intentions facilitated initiation of goal striving and eVectively shielded ongoing goal pursuits from unwanted influences. However, there were fewer studies that addressed self‐regulatory problems in disengaging from futile goal striving and conserving capability for future goal striving. Even considering the 21 tests to do with the problem of getting derailed, additional studies

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

105

would help to corroborate the eYcacy of if–then plans in dealing with the various aspects of the respective self‐regulatory tasks (i.e., suppressing unwanted responses, blocking detrimental self‐states, and blocking adverse contextual influences). The same reasoning applies to research in diVerent goal domains and using diVerent samples. Most studies to date used laboratory tasks, and there have been relatively few applications to consumer, environmental, antiracist, and prosocial behaviors. Similarly, the 23 tests in relation to health goals predominantly concerned the initiation of health‐protective behaviors (e.g., exercise, cancer screening). However, health‐risk behaviors, such as smoking, excess alcohol consumption, and poor diet, are major contributors to mortality and morbidity in Western societies (Belloc, 1973; Breslow & Enstrom, 1980). How well implementation intentions can help people to assiduously avoid these actions constitutes an important avenue for future investigation. Previous studies also mainly used undergraduate samples, and although sample type did not generally moderate implementation intention eVects, further tests among more representative groups would enhance the generality of the present analysis. The finding that people with chronic problems in action control (e.g., schizophrenics) were especially likely to benefit from implementation intention formation is encouraging and provides grounds for further rigorous tests of if–then planning interventions among other clinical samples (e.g., ADHD children, depressed individuals). More generally, although the present meta‐analysis shows that implementation intentions are eVective in enabling people to translate their ‘‘good’’ intentions into action, the review also reveals considerable scope for further tests in relation to long‐standing self‐regulatory problems (e.g., control of pain or stress), under‐researched samples (e.g., people with physical illness), and new domains of application (e.g., educational, organizational, and clinical settings). In whatever context people’s goal intentions are found to fall short of their goal achievement, applied psychologists might do well to consider deploying if–then plans to promote eVective self‐regulation of goal striving. There is also room for further theoretical integration of the concept of implementation intentions with theories of motivation (e.g., Bandura, 1997) and willpower (e.g., Metcalfe & Mischel, 1999). For instance, with respect to motivation, future studies may want to explore whether implementation intentions can be used to elevate self‐eYcacy beliefs (e.g., ‘‘And if I run into problems with any of my homework, then I will tell myself ‘I can do it!’ ’’). With respect to willpower, implementation intentions can be used to turn oV the hot system and activate the cool system when self‐control is needed. For example, a person who wants to cope better with unpleasant social encounters could use implementation intentions to reduce feelings of frustration

106

PETER M. GOLLWITZER AND PASCHAL SHEERAN

and anger (e.g., ‘‘And if I run into an obnoxious person, then I will try to understand this person as if I was a therapist!’’). What distinguishes this approach from past research on implementation intentions is the fact that the then‐component of the if–then plan does not specify one particular goal‐ directed response, but rather focuses on changing motivation‐relevant beliefs and/or self‐regulatory systems that can ultimately facilitate the performance of multiple and various goal‐directed responses. The present review obtained strong support for the component processes postulated to underlie implementation intention eVects. Implementation intentions showed large eVects on processes to do with heightened activation of the critical situation (accessibility, detection, discrimination, attention, memory) and automation of the goal‐directed response (immediacy, eYciency, redundancy of intent). However, it would be valuable to conduct mediation analyses to explore whether these processes are indeed responsible for the positive eVects of implementation intention formation on rates of goal achievement. One study that conducted this type of analysis measured the accessibility of situational cues specified in participants’ if–then plans in a lexical decision task and subsequently measured rates of goal attainment (Aarts et al., 1999). Findings indicated that cue accessibility mediated the impact of implementation intention formation on goal completion. Recently, Webb and Sheeran (2005c) extended this paradigm to investigate the mediational role of both cue accessibility and the strength of cue–response links forged by if–then planning. In one experiment, participants had the goal intention to collect a coupon from a specified location as part of a series of laboratory tasks. A subset of participants also formed an implementation intention that specified the location for collecting the coupon in the if‐component and the action of coupon collection in the then‐ component. Subsequently, an ostensibly unrelated lexical decision task had to be performed that assessed the accessibility of the critical cues (location words) and the accessibility of the target behavior when subliminally primed by the critical cues (i.e., the word ‘‘collect’’ preceded by location words). Findings indicated that implementation intention formation increased the rate of coupon collection (goal achievement) as well as the accessibility of both location cues and location‐primed target behavior (i.e., the strength of the link between the if‐ and then‐components of the plan). Most important, implementation intention eVects on goal attainment were mediated by cue accessibility as well as the strength of respective cue–response links. These findings are consistent with the postulated theoretical mechanisms. Further tests are needed to explore the mediational role of the other hypothesized processes (e.g., immediacy, eYciency, and redundancy of conscious intent), however. Moderators of implementation intention eVects also warrant investigation. There are two aspects to moderation here. First, individual diVerences could

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

107

either enhance or reduce the impact of implementation intentions on goal achievement. Individuals with personal attributes that make regulating their behavior more diYcult, for instance, might especially benefit from implementation intention formation. Thus, people who score highly on measures of procrastination, distractability, or self‐defensiveness may show higher rates of goal attainment when they form if–then plans compared to people who obtain low scores on these measures. On the other hand, individuals who spontaneously form implementation intentions may garner less advantage from inductions designed to prompt plan formation. Individual diVerences in conscientiousness, planfulness, or need for cognition could predict spontaneous if–then planning. It is also possible that individual diVerence variables could be identified that render if–then planning counterproductive. For instance, people who are poor at reality monitoring could form plans that are antithetical to eVective goal striving. Similarly, people who set too much store by adherence to plans (e.g., perfectionist individuals) may be prone to self‐evaluative ruminations that undermine the eVective operation of their plans. Thus, standard individual diVerence variables could have an important influence on whether and how well implementation intentions are formed and how much of an eVect they have on goal achievement. If one conceives of personality in terms of ‘‘intra‐individually stable, if. . .then. . ., situation–behavior relations’’ (Mischel & Shoda, 1995, p. 248), the question of how personality and if–then planning work together in the self‐regulation of goal striving may get even more interesting. Let us assume that a person has the goal to reduce aggression in relating to others, and he also knows about his respective situation–behavior profile (i.e., he knows what kind of social situations elicit aggressive responses in him and which social situations allow him to stay calm and collected). Given this goal and knowledge, the person can now tailor his implementation intentions to those critical situations specifying any of the following goal‐directed responses: ‘‘. . .then I will not get aggressive!’’ or ‘‘. . .then I will stay calm and collected!’’ or ‘‘. . .then I will ignore this situation!’’ Thus, it seems possible that people could maximize the self‐regulatory benefits of forming implementation intentions by taking into account their unique chronic if–then (situation–behavior) profiles and specify implementation intentions exactly where they are needed. Exploring interactions between chronic and strategic situation–behavior links constitutes a promising direction for future studies. The second aspect of moderation concerns degree of plan formation and refers both to the activation level of the if‐ and then‐components of the plan and to the strength of the mental link between the if‐component and the then‐component of the plan. These features of implementation intentions are responsible for the enhanced identification of specified contextual cues and

108

PETER M. GOLLWITZER AND PASCHAL SHEERAN

automated action control in the presence of these cues, thus determining how well if–then plans facilitate goal attainment. The implication of variations in degree of plan formation is that procedures that enhance the activation level of critical cues or the strength of cue–response associations should thereby increase the impact of implementation intentions on goal striving and goal completion. To date, only a small number of studies have tested this aspect of moderation. For instance, Gollwitzer et al. (2004a) manipulated the strength of participants’ commitment to their implementation intentions presuming to thereby strengthen cue–response links. Findings from a cued recall paradigm showed that the high commitment group had superior memory for selected opportunities compared to the low commitment group. Similarly, Milne and Sheeran (2002c) manipulated cognitive rehearsal by having some participants concentrate on the cue–response link during plan formation; participants wrote down their plan to visit a particular website twice, and were instructed to concentrate on the link between the situation and action when they were writing the plan the second time. Another implementation intention group also wrote down their plan twice, but were instructed to take the second write‐up of the plan with them and put it in a prominent place at home as a reminder. Findings indicated that participants who rehearsed the cue–response link were more likely to act on their plans compared to both participants who wrote their implementation intention on a reminder note and a control group who did not form implementation intentions (rates of visiting the Web site were 87%, 40%, and 20%, respectively). Future studies should examine the eVectiveness of strategies to aid encoding of if–then plans (e.g., diVerent types of cognitive rehearsal, surprise recall tasks or plan reminders) and strategies to increase commitment to these plans (e.g., inducing anticipated regret about not following one’s plan or making one’s commitment public) in order to ensure that opportunities are highly accessible and opportunity–action links are strong. Some individuals are likely to be more in need of such strategies than others because people diVer in their ability to generate strong if–then links when asked to form implementation intentions (Gollwitzer, Grant, & Oettingen, 2004b). But even if people’s original if–then links are weak, it seems possible to strengthen these links by having people act repeatedly on their implementation intentions. Research by Orbell and Verplanken (2005) observed that whenever participants performed repeated actions (e.g., flossing one’s teeth) on the basis of an implementation intention, they reported experiencing features of habitual action control (e.g., I do it without thinking, I start doing it before I realize I’m doing it, I do it automatically, I do it without having to think consciously, It would require eVort not to do it, . . .) more so than participants who performed the repeated action on the basis of a mere goal intention. Further research along these lines would be valuable in order to

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

109

ensure that implementation intentions are as eVective as possible in facilitating the realization of goal intentions for particular people.

V. Conclusions Goal intentions are not always successfully translated into behavior because merely making a commitment to attain a goal does not necessarily prepare people for dealing eVectively with self‐regulatory problems in goal striving. This chapter tested the idea that goal striving could benefit from a second act of willing—the formation of if–then plans—that focuses on the enactment of goal intentions. A meta‐analysis of 94 studies showed that forming an implementation intention makes an important diVerence to whether or not people achieve their goals. This finding was robust across variations in study design, outcome measurement, and domains of goal attainment. Moreover, if–then planning facilitated goal striving no matter what self‐regulatory problem was at hand. Medium‐to‐large eVects were obtained in relation to initiating goal striving, shielding goals from unwanted influences, disengaging from failing goals, and preserving self‐regulatory capability for future goal striving. There was also strong support for the if–then component processes. People who form implementation intentions are in a good position to recognize opportunities to act and respond to these opportunities swiftly and eVortlessly. Thus, this chapter shows that the concept of implementation intentions is valuable both in understanding the processes of goal attainment and in providing a self‐regulatory strategy to help people reach their goals. Notwithstanding the self‐regulatory benefits of implementation intentions demonstrated here, there is considerable scope for further research to exploit the potential of if–then planning and to understand how implementation intentions can best be deployed to facilitate intention realization. Such research would seem to be a worthwhile goal pursuit for both basic and applied psychologists.

Acknowledgments We thank Icek Ajzen, Christopher J. Armitage, Paul Dholakia, Rob Holland, Richard Koestner, Sonia Lippke, Andrew Prestwich, Falko Sniehotta, Liz Steadman, and Jochen Philipp Ziegelman for access to unpublished findings, and Tom Webb, Ute Bayer, Anja Achtziger, Inge Schweiger Gallo, Alex Jaudas, Georg Odenthal, Caterina Gawrilow, and Tanya Faude for comments on an earlier draft of the manuscript.

110

PETER M. GOLLWITZER AND PASCHAL SHEERAN

Preparation of this chapter was supported in part by National Institute of Health grant (R01–67100) and the Center for Research on Intentions and Intentionality at the University of Konstanz, Germany.

Appendix I EXAMPLES OF POSSIBLE IMPLEMENTATION INTENTIONS GEARED AT RESOLVING THE FOUR PROBLEMS OF GOAL STRIVING 1. Failing to get started a. Remembering to act To achieve the goal intention of sending a birthday card on time: And if I walk by the institute’s mail box, then I will drop in my card! b. Seizing opportunities To achieve the goal intention of complaining about poor service: And if I see the manager walk into the restaurant, then I will go over to him and complain about the poor service! c. Overcoming initial reluctance To achieve the goal intention of completing course work on time: And if it is Saturday morning at 10 a.m., then I will sit down at my computer and make an outline for my essay! 2. Getting derailed a. Suppressing unwanted attention responses To achieve the goal intention of behaving calmly in the face of scary spider pictures: And if I see a spider, then I will ignore it! b. Suppressing unwanted behavioral responses To achieve the goal intention of behaving calmly in the wake of being insulted: And if I feel my anger rise, then I will tell myself to stay calm and not aggress back! c. Blocking detrimental self‐states To block the negative influence of ego‐depletion on solving diYcult anagrams: And if I have solved one anagram, then I will immediately move onto the next one! d. Blocking adverse contextual influences To block the negative influence of loss framing on negotiation outcomes when having to share an attractive commodity (e.g., a fictitious island in the Lake of Constance): And if I receive a proposal on how to share the island, then I will oVer a cooperative counterproposal!

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

111

3. Not calling a halt To prevent escalation of commitment to a certain strategy of performing a general knowledge test: And if I receive disappointing feedback, then I will switch to a diVerent strategy! 4. Overextending oneself To prevent the emergence of ego‐depletion in the wake of controlling one’s emotions, such as not laughing at amusing cartoons: And if an amusing scene is presented, then I will tell myself ‘‘these are just stupid, silly jokes!’’

References Papers included in the meta‐analysis are preceded by an asterisk. Aarts, H., & Dijksterhuis, A. (2000a). Habits as knowledge structures: Automaticity in goal‐ directed behavior. Journal of Personality and Social Psychology, 78, 53–63. Aarts,, H, & Dijksterhuis, A. (2000b). On the automatic activation of goal‐directed behavior: The case of travel habit. Journal of Environmental Psychology, 20, 75–82. *Aarts, H., Dijksterhuis, A., & Midden, C. (1999). To plan or not to plan? Goal achievement or interrupting the performance of mundane behaviors. European Journal of Social Psychology, 29, 971–979. Abraham, C., Sheeran, P., Norman, P., Conner, M., Otten, W., & de Vries, N. (1999). When good intentions are not enough: Modelling post‐intention cognitive correlates of condom use. Journal of Applied Social Psychology, 29, 2591–2612. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211. *Ajzen, I., Czasch, C., & Flood, M. G. (2002). From intentions to behavior: Implementation intention, commitment, and conscientiousness. Manuscript under review. Arkes, H. R., & Blumer, C. (1985). The psychology of sunk cost. Organizational Behavior and Human Performance, 35, 129–140. *Armitage, C. J. (2004). Implementation intentions and eating a low‐fat diet: A randomized controlled trial. Health Psychology, 23, 319–323. Armitage, C. J., & Conner, M. (2001). EYcacy of the theory of planned behaviour: A meta‐ analytic review. British Journal of Social Psychology, 40, 471–499. Atkinson, J. W. (1957). Motivational determinants of risk taking behavior. Psychological Review, 64, 359–372. Austin, J. T., & Vancouver, J. B. (1996). Goal constructs in psychology: Structure, process, and content. Psycological Bulletin, 120, 338–375. *Bagozzi, R. P., Dholakia, U. M., & Basuroy, S. (2003). How eVortful decisions get enacted: The motivating role of decision processes, desires, and anticipated emotions. Journal of Behavioral Decision Making, 16, 273–295. *Bamberg, S. (2000). The promotion of a new behavior by forming an implementation intention: Results of a field experiment in the domain of travel mode choice. Journal of Applied Social Psychology, 23, 573–587.

112

PETER M. GOLLWITZER AND PASCHAL SHEERAN

*Bamberg, S. (2002). Implementation intention versus monetary incentive comparing the eVects of interventions to promote the purchase of organically produced food. Journal of Economic Psychology, 23, 573–587. Bandura, A. (1991). Self‐regulation of motivation through anticipatory and self‐reactive mechanisms. In R. A. Dienstbier (Ed.), Nebraska symposium on motivation: Perspectives on motivation (Vol. 38, pp. 69–164). Lincoln, NE: University of Nebraska Press. Bandura, A. (1997). Self‐eYcacy: The exercise of control. New York: W. H. Freeman. Bargh, J. A. (1990). Auto‐motives: Preconscious determinants of social interaction. In E. T. Higgins and R. M. Sorrentino (Eds.), Handbook of motivation and cognition: Foundations of social behavior (Vol. 2, pp. 93–130). New York: Guilford Press. Bargh, J. A. (1992). The ecology of automaticity: Towards establishing the conditions needed to produce automatic processing eVects. American Journal of Psychology, 105, 181–199. Bargh, J. A. (1994). The four horsemen of automaticity: Awareness, eYciency, intention, and control in social interaction. In R. S. Wyer, Jr. and T. K. Srull (Eds.), Handbook of social cognition (2nd ed., pp. 1–40). Hillsdale, NJ: Erlbaum. Bargh, J. A., & Gollwitzer, P. M. (1994). Environmental control over goal‐directed action. Nebraska Symposium on Motivation, 41, 71–124. Bargh, J. A., Gollwitzer, P. M., Lee‐Chai, A., Barndollar, K., & Tro¨tschel, R. (2001). The automated will: Nonconscious activation and pursuit of behavioral goals. Journal of Personality and Social Psychology, 81, 1014–1027. Baumeister, R. F., Bratlavsky, E., Muraven, M., & Tice, D. M. (1998). Ego‐depletion: Is the active self a limited resource? Journal of Personality and Social Psychology, 74, 1252–1265. *Bayer, U. C., Jaudas, A., & Gollwitzer, P. M. (2002, July). Do implementation intentions facilitate switching between tasks?. Poster presented at the International Symposium on Executive Functions. Konstanz, Germany. *Bayer, U. C., Moskowitz, G. B., & Gollwitzer, P. M. (2004). Implementation intentions and action initiation without conscious intent. Manuscript under review. Belloc, N. B. (1973). Relationship of physical health status and health status. Preventive Medicine, 2, 67–81. Bem, D. J. (1972). Self‐perception theory. Advances in Experimental Social Psychology, 6, 1–62. Bless, H. (1997). Stimmung und Denken: Ein Modell zum Einflu von Stimmung auf Denkprozesse (Mood and reasoning: A model on the impact of mood on cognitive processes). Bern: Hans Huber. Bless, H., & Fiedler, K. (1995). AVective states and the influence of activated general knowledge. Personality and Social Psychology Bulletin, 21, 766–778. Bobocel, D. R., & Meyer, J. P. (1994). Escalating commitment to a failing course of action: Separating the roles of choice and justification. Journal of Applied Psychology, 79, 360–363. Bragger, J. D., Hantula, D. A., Bragger, D., Kirnan, J., & Kutcher, E. (2003). When success breeds failure: History, hysteresis, and delayed exit decisions. Journal of Applied Psychology, 88, 6–14. *Brandsta¨tter, V., Heimbeck, D., Malzacher, J. T., & Frese, M. (2003). Goals need implementation intentions: The model of action phases tested in the applied setting of continuing education. European Journal of Work and Organizational Psychology, 12, 37–59. *Brandsta¨tter, V., Lengfelder, A., & Gollwitzer, P. M. (2001). Implementation intentions and eYcient action initiation. Journal of Personality and Social Psychology, 81, 946–960. Brehm, J. W., & Self, E. (1989). The intensity of motivation. Annual Review of Psychology, 40, 109–131. Breslow, L., & Enstrom, J. E. (1980). Persistence of health habits and their relationship to mortality. Preventive Medicine, 9, 469–483.

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

113

Brockner, J., Rubin, J. Z., & Lang, E. (1981). Face‐saving and entrapment. Journal of Experimental Social Psychology, 17, 68–79. *Bru¨wer, J., Bayer, U. C., & Gollwitzer, P. M. (2002). Intentional control of the Simon eVect. Poster presented at the International Symposium on Executive Functions. Konstanz, Germany. *Bulgarella, C., Oettingen, G., & Gollwitzer, P. M. (2003, February). Reflexive competition and its self‐regulation. Poster presented at the meeting of the Society for Personality and Social Psychology. Los Angeles, CA. Canary, D. J., & Seibold, D. R. (1984). Attitudes and behavior: A comprehensive bibliography. New York: Praeger. Carver, C. S., & Scheier, M. F. (1982). Control theory: A useful conceptual framework for personality‐social, clinical, and health psychology. Psychological Bulletin, 92, 111–135. Carver, C. S., & Scheier, M. F. (1998). On the self‐regulation of behavior. Cambridge, UK: University Press. *Chasteen, A. L., Park, D. C., & Schwarz, N. (2001). Implementation intentions and facilitation of prospective memory. Psychological Science, 12, 457–461. Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. COMMIT Research Group (1995). Community intervention trial for smoking cessation (COMMIT): I. Cohort results from a four‐year community intervention. American Journal of Public Health, 85, 183–192. De Dreu, C. K. W., Carnevale, P. J. D., Emans, B. J. M., & van de Vliert, E. (1995). Outcome frames in bilateral negotiation: Resistance to concession making and frame adoption. In W. Stroebe and M. Hewstone (Eds.), European review of social psychology (Vol. 6, pp. 97–125). Chichester, UK: Wiley. Deci, E. L., & Ryan, R. M. (1991). A motivational approach to self: Integration in personality. Nebraska Symposium on Motivation, 38, 237–288. *Dewitte, S., Verguts, T., & Lens, W. (2003). Implementation intentions do not enhance all types of goals: The moderating role of goal diYculty. Current Psychology, 22, 73–89. *Dholakia, U. M., & Bagozzi, R. P. (2002). Mustering motivation to enact decisions: How decision process characteristics influence goal realization. Journal of Behavioral Decision Making, 15, 167–188. *Dholakia, U. M., & Bagozzi, R. P. (2003). As time goes by: How goal and implementation intentions influence enactment of short‐fuse behaviors. Journal of Applied Social Psychology, 33, 889–922. *DieVendorV, J. M., & Lord, R. G. (2003). The volitional and strategic eVects of planning on task performance and goal commitment. Human Performance, 16, 365–387. Dweck, C. S. (2000). Self‐theories: Their role in motivation, personality, and development. Philadelphia, PA: Psychology Press. Einstein, G. O., & McDaniel, M. A. (1996). Retrieval processes in prospective memory: Theoretical approaches and some new empirical findings. In M. Bradimonte, G. O. Einstein, and M. A. McDaniel (Eds.), Prospective memory: Theory and applications (pp. 115–141). Mahwah, NJ: Erlbaum. *Einstein, G. O., McDaniel, M. A., Williford, C. L., Pagan, J. L., & Dismukes, R. K. (2003). Forgetting of intentions in demanding situations is rapid. Journal of Experimental Psychology: Applied, 9, 147–162. Fishbein, M. (1980). Theory of reasoned action: Some applications and implications. In H. Howe and M. Page (Eds.), Nebraska symposium on motivation, 1979 (pp. 65–116). Lincoln, NB: University of Nebraska Press. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison‐Wesley.

114

PETER M. GOLLWITZER AND PASCHAL SHEERAN

Floyd, D. L., Prentice‐Dunn, S., & Rogers, R. W. (2000). A meta‐analysis of research on protection motivation theory. Journal of Applied Social Psychology, 30, 407–429. Garner, D. M., & Wooley, S. C. (1991). Confronting the failure of behavioral and dietary treatments of obesity. Clinical Psychology Review, 6, 58–137. Gibbons, F. X., Gerrard, M., Blanton, H., & Russell, D. W. (1998). Reasoned action and social reaction: Willingness and intention as independent predictors of health risk. Journal of Personality and Social Psychology, 74, 1164–1180. *Gillholm, R., Ettema, D., Selart, M., & Garling, T. (1999). The role of planning for intention‐ behavior consistency. Scandinavian Journal of Psychology, 40, 241–250. Godin, G., & Kok, G. (1996). The theory of planned behavior: A review of its applications in health–related behaviors. American Journal of Health Promotion, 11, 87–98. Gollwitzer, P. M. (1990). Action phases and mindsets. In E. T. Higgins and J. R. M. Sorrentino (Eds.), The handbook of motivation and cognition (Vol. 2, pp. 53–92). New York: Guilford. Gollwitzer, P. M. (1993). Goal achievement: The role of intentions. In W. Stroebe and M. Hewstone (Eds.), European Review of Social Psychology (Vol. 4, pp. 141–185). New York: Wiley. Gollwitzer, P. M. (1996). The volitional benefits of planning. In P. M. Gollwitzer and J. A. Bargh (Eds.), The psychology of action: Linking cognition and motivation to behavior (pp. 287–312). New York: Guilford. Gollwitzer, P. M. (1999). Implementation intentions: Strong eVects of simple plans. American Psychologist, 54, 493–503. *Gollwitzer, P. M., & Bayer, U. C. (2000, October). Becoming a better person without changing the self. Paper presented at the Self and Identity Pre‐conference of the Annual Meeting of the Society of Experimental Social Psychology. Atlanta, Georgia. Gollwitzer, P. M., Bayer, U. C., & McCulloch, K. C. (2005). The control of the unwanted. In R. R. Hassin, J. S. Uleman, and J. A. Bargh (Eds.), The new unconscious (pp. 485–515). New York: Oxford University Press. *Gollwitzer, P. M., Bayer, U. C., Steller, B., & Bargh, J. A. (2004a). Delegating control to the environment: Perception, attention, and memory for pre‐selected behavioral cues. Manuscript under review. *Gollwitzer, P. M., & Brandsta¨tter, V. (1997). Implementation intentions and eVective goal pursuit. Journal of Personality and Social Psychology, 73, 186–199. Gollwitzer, P. M., Grant, H., & Oettingen, G. (2004b, February). An individual diVerence measure of the tendency to form implementation intentions. Poster presented at the 5th Annual Meeting of SPSP. Austin, Texas. Gollwitzer, P. M., & Moskowitz, G. B. (1996). Goal eVects on action and cognition. In E. T. Higgins and A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 361–399). New York: Guilford. Gollwitzer, P. M., & Schaal, B. (1998). Metacognition in action: The importance of implementation intentions. Personality and Social Psychology Review, 2, 124–136. *Gollwitzer, P. M., Sheeran, P., & Seifert, E. (2004c). Intentions as symbols: How self‐completion aVects the translation of intentions into action. Manuscript under review. *Gollwitzer, P. M., Sheeran, P., Tro¨tschel, R., & Webb, T. L. (2004d). The control of behavior priming eVects by implementation intentions. Manuscript under review. *Gollwitzer, P. M., Tro¨tschel, R., Bayer, U. C., & Sumner, M. (2004e). Potential costs of self‐ regulation by implementation intentions: Rebound and ego‐depletion eVects? Manuscript under review. Gollwitzer, P. M., & Wicklund, R. A. (1985). Self‐symbolizing and the neglect of others’ perspectives. Journal of Personality and Social Psychology, 56, 531–715.

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

115

Hagger, M. S., Chatzisarantis, N. L. D., & Biddle, S. J. H. (2002). A meta‐analytic review of the theories of reasoned action and planned behavior in physical activity: Predictive validity and the contribution of additional variables. Journal of Sport and Exercise Psychology, 24, 3–32. Hausenblas, H. A., Carron, A. V., & Mack, D. E. (1997). Application of the theories of reasoned action and planned behavior to exercise behavior: A meta‐analysis. Journal of Sport and Exercise Psychology, 19, 36–51. Heckhausen, H. (1991). Motivation and action. Heidelberg: Springer‐Verlag. Heckhausen, H., & Gollwitzer, P. M. (1987). Thought contents and cognitive functioning in motivational versus volitional states of mind. Motivation and Emotion, 11, 101–120. *Henderson, M. D., Gollwitzer, P. M., & Oettingen, G. (2004). Using implementation intentions to disengage from a failing course of action: The eVects of planning on escalation of commitment. Manuscript under review. Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta‐analysis. New York: Academic Press. Higgins, E. T. (2000). Making a good decision: Value from fit. American Psychologist, 55, 1217–1230. *Holland, R., Aarts, H., & Langendam, D. (in press). Breaking and creating habits on the workfloor: A field experiment on the power of implementation intentions. Journal of Experimental Social Psychology. Hunter, J. E. (1990). Methods of meta‐analysis: Correcting error and bias in research findings. Newbury Park, CA: Sage. Karau, S. J., & Williams, K. D. (1993). Social loafing: A meta‐analytic review and theoretical integration. Journal of Personality and Social Psychology, 65, 681–706. Kavanagh, D. J., Andrade, J., & May, J. (2005). Imaginary relish and exquisite torture: The elaborated intrusion model of desire. Psychological Review, 112, 446–467. Kim, M. S., & Hunter, J. E. (1993). Attitude‐behavior relations: A meta‐analysis of attitudinal relevance and topic. Journal of Communication, 43, 101–142. *Koestner, R., Downie, M., Horburg, E., & Hata, S. (2002a). Self‐construals and the pursuit of personal goals: The role of implementation intentions and self‐concordance. Manuscript in preparation. Quebec, Canada: McGill University. *Koestner, R., Lekes, N., Powers, T. A., & Chicoine, E. (2002b). Attaining personal goals: Self‐ concordance plus implementation intentions equals success. Journal of Personality and Social Psychology, 83, 231–244. *Koole, S., & Van’t Spijker, M. (2000). Overcoming the planning fallacy through willpower: EVects of implementation intentions on actual and predicted task‐completion times. European Journal of Social Psychology, 30, 873–888. Kruglanksi, A. W., Shah, J. Y., Fishbach, A., Friedman, R., Chun, W. Y., & Sleeth‐Keppler, D. (2002). A theory of goal systems. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 331–378). San Diego, CA: Academic Press. Latane´, B., Williams, K., & Harkins, S. (1979). Many hands make light work: The causes and consequences of social loafing. Journal of Personality and Social Psychology, 37, 822–832. *Lengfelder, A., & Gollwitzer, P. M. (2001). Reflective and reflexive action control in patients with frontal brain lesions. Neuropsychology, 15, 80–100. *Lippke, S., & Ziegelmann, J. P. (2002). Self‐regulation and exercise: A study on stages of change and successful ageing. Unpublished manuscript. Germany: Free University of Berlin. Lewin, K. (1926). Vorsatz, wille, und bedu¨rfnis. Psychologische Forschung, 7, 330–385. Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance. Englewood CliVs, NJ: Prentice Hall.

116

PETER M. GOLLWITZER AND PASCHAL SHEERAN

Loewenstein, G. F., Weber, E. U., Hsee, C. K., & Welch, N. (2001). Risk as feelings. Psychological Bulletin, 127, 267–286. Macrae, C. N., Bodenhausen, G. V., Milne, A. B., & Jetten, J. (1994). Out of mind but back in sight: Stereotypes on the rebound. Journal of Personality and Social Psychology, 67, 808–817. McBroom, W. H., & Reid, F. W. (1992). Towards a reconceptualization of attitude‐behavior consistency. Social Psychology Quarterly, 55, 205–216. Metcalfe, J., & Mischel, W. (1999). A hot/cool‐system analysis of delay of gratification: Dynamics of willpower. Psychological Bulletin, 106, 3–19. Milne, S., Sheeran, P., & Orbell, S. (2000). Prediction and intervention in health‐related behavior: A meta‐analytic review of protection motivation theory. Journal of Applied Social Psychology, 30, 106–143. *Milne, S., Orbell, S., & Sheeran, P. (2002). Combining motivational and volitional interventions to promote exercise participation: Protection motivation theory and implementation intentions. British Journal of Health Psychology, 7, 163–184. *Milne, S., & Sheeran, P. (2002a, June). Combining motivational and volitional interventions to prevent testicular cancer. Testing the interaction between protection motivation theory and implementation intentions. Paper presented at the 13th general meeting of the European Association of Experimental Social Psychology. San Sebastian, Spain. *Milne, S., & Sheeran, P. (2002b). Unpublished raw data. UK: University of Bath. *Milne, S., & Sheeran, P. (2002c, October). Making good implementation intentions: Comparing associative learning and prospective memory in remembering intentions. Paper presented at the 16th conference of the European Health Psychology Society. Lisbon, Portugal. Mischel, W., & Patterson, C. J. (1978). EVective plans for self‐control in children. In W. A. Collins (Ed.), Minnesota symposium on child psychology (Vol. 11, pp. 199–230). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Mischel, W., & Shoda, Y. (1995). A cognitive‐aVective system theory of personality: Reconceptualiszing situations, dispositions, dynamics, and invariance in personality structure. Psychological Review, 102, 246–268. Muraven, M., & Baumeister, R. F. (2000). Self‐regulation and depletion of limited resources. Does self‐control resemble a muscle? Psychological Bulletin, 126, 247–259. *MurgraV, V., White, D., & Phillips, K. (1996). Moderating binge drinking: It is possible to change behaviour if you plan it in advance. Alcohol and Alcoholism, 6, 577–582. Neale, M. A., & Bazerman, M. H. (1985). The eVects of framing and negotiator overconfidence on bargaining behaviors and outcomes. Academy of Management Journal, 28, 34–49. Oettingen, G. (2000). Expectancy eVects on behavior depend on self‐regulatory thought. Social Cognition, 18, 101–129. Oettingen, G., & Gollwitzer, P. M. (2001). Goal setting and goal striving. In A. Tesser and N. Schwarz (Eds.), Intraindividual processes. Volume 1 of the Blackwell Handbook in Social Psychology. M. Hewstone and M. Brewer (Eds.) (pp. 329–347). Oxford: Blackwell. *Oettingen, G., Ho¨nig, G., & Gollwitzer, P. M. (2000). EVective self‐regulation of goal attainment. International Journal of Educational Research, 33, 705–732. Oettingen, G., Pak, H., & Schnetter, K. (2001). Self‐regulation of goal‐setting: Turning free fantasies about the future into binding goals. Journal of Personality and Social Psychology, 80, 736–753. *Orbell, S., Hodgkins, S., & Sheeran, P. (1997). Implementation intentions and the theory of planned behavior. Personality and Social Psychology Bulletin, 23, 945–954. Orbell, S., & Sheeran, P. (1998). ‘‘Inclined abstainers’’: A problem for predicting health‐related behavior. British Journal of Social Psychology, 37, 151–165.

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

117

*Orbell, S., & Sheeran, P. (1999). Volitional strategies and the theory of planned behavior. Unpublished manuscript. UK: University of SheYeld. *Orbell, S., & Sheeran, P. (2000). Motivational and volitional processes in action initiation: A field study of the role of implementation intentions. Journal of Applied Social Psychology, 30, 780–797. Orbell, S., & Verplanken, B. (2005). Habituation of goal‐directed behavior by implementation intentions. Manuscript under review. Ouellette, J. A., & Wood, W. (1998). Habit and intention in everyday life: The multiple processes by which past behavior predicts future behavior. Psychological Bulletin, 124, 54–74. Patterson, C., & Mischel, W. (1976). EVects of temptation‐inhibiting and task‐facilitating plans on self‐control. Journal of Personality and Social Psychology, 33, 209–217. *Prestwich, A., Lawton, R., & Conner, M. (2003a). Disguising the strength of implementation intentions: The eVect of spontaneously arising plans. Unpublished manuscript. UK: University of Leeds. *Prestwich, A., Lawton, R., & Conner, M. (2003b). The use of implementation intentions and the decision balance sheet in promoting exercise behavior. Psychology and Health, 18, 707–722. *Rise, J., Thompson, M., & Verplanken, B. (2003). Measuring implementation intentions in the context of the theory of planned behavior. Scandinavian Journal of Psychology, 44, 87–95. Rogers, R. W. (1983). Cognitive and physiological processes in fear appeals and attitude change: a revised theory of protection motivation. In B. L. Cacioppo and L. L. Petty (Eds.), Social psychophysiology: A sourcebook (pp. 153–176). London: Guildford. Rosenthal, R. (1984). Meta‐analytic procedures for social research. Beverley Hills, CA: Sage. *Schaal, B. (1993). Impulskontrolle—Wie Vorsa¨tze beherrschtes Handeln erleichtern. [Impulse control—How implementation intentions facilitate the control of behavior]. Unpublished Master’s Thesis. Munich, Germany: Ludwig‐Maximilians‐Universita¨t. Schwarz, N. (1990). Feelings as information: Informational and motivational functions of aVective states. In E. T. Higgins and R. M. Sorrentino (Eds.), Handbook of motivation and cognition: Foundations of social behavior (Vol. 2, pp. 527–561). New York: Guilford. Schwarz, N., Bless, H., & Bohner, G. (1991). Mood and persuasion: AVective states influence the processing of persuasive communications. In M. P. Zanna (Ed.), Advances in Experimental Social Psychology (Vol. 24, pp. 161–199). San Diego: Academic Press. Shah, J. Y., Kruglanksi, A. W., & Friedman, R. (2003). Goal systems theory: Integrating the cognitive and motivational aspects of self‐regulation. A goal systems approach to self‐ regulation. In S. J. Spencer, S. Fein, M. P. Zanna, and J. M. Olson (Eds.), Motivated social perception: The Ontario Symposium (Vol. 9, pp. 247–275). New Jersey: Erlbaum. Sheeran, P. (2002). Intention‐behavior relations: A conceptual and empirical review. In W. Stroebe and M. Hewstone (Eds.), European Review of Social Psychology (Vol. 12, pp. 1–30). New York: Wiley. Sheeran, P., Aarts, H., Custers, R., Rivis, A. J., Webb, T. L., & Cooke, R. (2005a). The goal‐ dependent automaticity of drinking habits. British Journal of Social Psychology, 44, 47–64. Sheeran, P., Abraham, C., & Orbell, S. (1999). Psychosocial correlates of heterosexual condom use: A meta‐analysis. Psychological Bulletin, 125, 90–132. *Sheeran, P., & Milne, S. (2003). Unpublished raw data. UK: University of SheYeld. *Sheeran, P., Milne, S., Webb, T. L., & Gollwitzer, P. M. (2005b). Implementation intentions. In M. Conner and P. Norman (Eds.), Predicting health behavior: Research and practice with social cognition models (2nd ed.). Milton Keynes, UK: Open University Press. Sheeran, P., & Orbell, S. (1998). Do intentions predict condom use? A meta‐analysis and examination of six moderator variables. British Journal of Social Psychology, 37, 231–250.

118

PETER M. GOLLWITZER AND PASCHAL SHEERAN

*Sheeran, P., & Orbell, S. (1999). Implementation intentions and repeated behavior: Augmenting the predictive validity of the theory of planned behavior. European Journal of Social Psychology, 29, 349–369. *Sheeran, P., & Orbell, S. (2000). Using implementation intentions to increase attendance for cervical cancer screening. Health Psychology, 19, 283–289. *Sheeran, P., & Silverman, M. (2003). Evaluation of three interventions to promote workplace health and safety: Evidence for the utility of implementation intentions. Social Science and Medicine, 56, 2153–2163. *Sheeran, P., & Webb, T. L. (2003). Implementation intentions versus ego‐training as strategies for preventing ego‐depletion. Unpublished manuscript. UK: University of SheYeld. *Sheeran, P., Webb, T. L., & Gollwitzer, P. M. (2005c). The interplay between goal intentions and implementation intentions. Personality and Social Psychology Bulletin, 31, 87–98. Sheeran, P., White, D., & Phillips, K. (1991). Premarital contraceptive use: A review of the psychological literature. Journal of Reproductive and Infant Psychology, 9, 253–269. Sheppard, B. H., Hartwick, J., & Warshaw, P. R. (1988). The theory of reasoned action: A meta‐ analysis of past research with recommendations for modifications and future research. Journal of Consumer Research, 15, 325–342. *Sniehotta, F. F., Scholtz, U., & Schwarzer, R. (2002a). The eVects of planning on initiation and maintenance of physical activity in cardiac rehabilitation patients. Unpublished research report. Germany: Free University Berlin. *Sniehotta, F. F., Scholtz, U., & Schwarzer, R. (2002b). Implementation intentions and coping intentions. Unpublished research report. Germany: Free University Berlin. *Steadman, L., & Quine, L. (2000, August). Are implementation intentions useful for bridging the intention‐behavior gap in adhering to long‐term medication regimens? An attempt to replicate Sheeran and Orbell’s (1999) intervention to enhance adherence to daily vitamin C intake. Paper presented at the British Psychological Society Division of Health Psychology annual conference, University of Kent. Canterbury, UK. *Steadman, L., & Quine, L. (2004). Encouraging young males to perform testicular self‐examination: A simple, but eVective, implementation intentions interventions. British Journal of Health Psychology, 9, 479–487. *Stephens, A., & Conner, M. (1999, September). A smoking prevention intervention in adolescents. Paper presented at the British Psychological Society Division of Health Psychology annual conference. University of Leeds. Sutton, S. (1998). Predicting and explaining intentions and behaviour: How well are we doing? Journal of Applied Social Psychology, 28, 1317–1338. Sutton, S., & Sheeran, P. (2003). Meta‐analysis of the theory of planned behaviour and past behaviour. UK: University of Cambridge. Manuscript in preparation. Tan, H. T., & Yates, J. F. (2002). Financial budgets and escalation eVects. Organizational Behavior and Human Decision Processes, 87, 300–322. Trafimow, D., & Sheeran, P. (2004). A theory about the translation of cognition into aVect and behavior. In G. Haddock and G. R. Maio (Eds.), Contemporary perspectives on the psychology of attitudes (pp. 57–76). Hove, UK: Psychology Press. Triandis, H. C. (1980). Values, attitudes, and interpersonal behavior. In H. Howe and M. Page (Eds.), Nebraska symposium on motivation (Vol. 27, pp. 195–259). Lincoln, NB: University of Nebraska Press. *Tro¨tschel, R., & Gollwitzer, P. M. (2004). Implementation intentions and the control of framing eVects in negotiations. Manuscript under review. van den Putte, B. (1993). On the theory of reasoned action. Unpublished doctoral dissertation. The Netherlands: University of Amsterdam.

META‐ANALYSIS OF IMPLEMENTATION INTENTIONS

119

Verplanken, B., & Aarts, H. (1999). Habit, attitude, and planned behaviour: Is habit an empty construct or an interesting case of automaticity? In W. Stroebe and M. Hewstone (Eds.), European review of social psychology (Vol. 10, pp. 101–134). Chicester, England: Wiley. Verplanken, B., Aarts, H., van Knippenberg, A., & Moonen, A. (1998). Habit versus planned behaviour: A field experiment. British Journal of Social Psychology, 37, 111–128. *Verplanken, B., & Faes, S. (1999). Good intentions, bad habits, and eVects of forming implementation intentions on healthy eating. European Journal of Social Psychology, 29, 591–604. Vroom, V. H. (1964). Work and motivation. New York: Wiley. *Webb, T. L., & Sheeran, P. (2003). Can implementation intentions help to overcome ego‐ depletion? Journal of Experimental Social Psychology, 39, 279–286. *Webb, T. L., & Sheeran, P. (2004). Identifying good opportunities to act: Implementation intentions and cue discrimination. European Journal of Social Psychology, 34, 407–419. *Webb, T. L., & Sheeran, P. (2005). Integrating concepts from goal theories to understand the achievement of personal goals. European Journal of Social Psychology, 35, 69–96. *Webb, T. L., & Sheeran, P. (in press a). Does changing behavioral intentions engender behavior change? A meta‐analysis of the experimental evidence. Psychological Bulletin. *Webb, T. L., & Sheeran, P. (in press b). How do implementation intentions promote goal attainment? Tests of component processes. Journal of Experimental Social Psychology. Wegner, D. M. (1994). Ironic processes in mental control. Psychological Review, 101, 34–52. Wicklund, R. A., & Gollwitzer, P. M. (1982). Symbolic self‐completion. Hillsdale, NJ: Erlbaum. Wight, D. (1992). Impediments to safer heterosexual sex: A review of research with young people. AIDS Care, 4, 11–21. *Williams, S. (2003). Stand by your plan: A new approach to the planning fallacy. Unpublished doctoral thesis. UK: University of Sussex. Wood, W., Quinn, J. M., & Kashy, D. (2002). Habits in everyday life: The thought and feel of action. Journal of Personality and Social Psychology, 83, 1281–1297. Wrosch, C., Scheier, M. F., Carver, C. S., & Schulz, R. (2003). The importance of goal disengagement in adaptive self‐regulation: When giving up is beneficial. Self and Identity, 2, 1–20.

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

J. Nicole Shelton Jennifer A. Richeson

The term social interaction conjures up images that involve at least two people. These two people are likely to have beliefs about one another, beliefs about how the other person views them, and beliefs about the interaction. Moreover, these beliefs are likely to influence both individuals’ experiences during the interaction. Although interconnectedness of this type has been pursued in examinations of interpersonal interactions (e.g., Baldwin, 1992; Darley & Fazio, 1980), research on interracial interactions has tended to adopt a more individualistic approach. Similar to interpersonal interactions, individuals’ experiences in interracial interactions are often shaped by the beliefs individuals have about one another and their beliefs about how they will be perceived by their interaction partners. In this chapter we examine interracial interactions from a perspective that highlights the interconnectedness that is often at the core of interpersonal interactions between members of diVerent racial groups. This perspective highlights that there are two people involved in dyadic interracial interactions and these two people influence each other’s outcomes and experiences.

I. Conceptual Approaches to Study Interracial Interactions The nature of interracial interactions allows researchers to explore how interpersonal perceptions and behaviors can be shaped by the presence of group boundaries. As has been recently suggested, however, to be such an interpersonal phenomenon, an overwhelming abundance of research on 121 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 38 DOI: 10.1016/S0065-2601(06)38003-3

Copyright 2006, Elsevier Inc. All rights reserved. 0065-2601/06 $35.00

122

J. NICOLE SHELTON AND JENNIFER A. RICHESON

interracial contact has been quite intrapersonal in nature (Devine, Evett, & Vasquez‐Suson, 1996; Hebl & Dovidio, 2005; Shelton, 2000). There is considerable research that shows that when social psychologists investigate intergroup interactions, we tend to rely on an individualistic approach. We approach the interaction as if only one person is present in the situation. Moreover, the focus tends to be on one individual’s impressions of other people, with the majority of this research focusing on Whites’ beliefs about Blacks (Fiske, 1998; Stephan & Stephan, 1985). For example, researchers have examined the extent to which Whites’ racial beliefs influence their judgments about Blacks. Although the implicit assumption is that these racial beliefs and judgments have implications for Blacks’ experiences during the interaction, Blacks’ actual experiences are often left unexamined. The individualistic approach has undoubtedly been fueled by the important research on prejudice reduction (Pettigrew & Tropp, 2000). We venture to add that social psychologists’ emphasis on how cognitive underpinnings, specifically stereotypes, influence social perception and interactions has obscured our understanding of the interpersonal reality of interracial interactions. In this chapter we oVer a framework that is grounded in the belief that in order to obtain an understanding of interracial interactions that reflects social reality, it is necessary to take a more relational approach.

A. PARADIGM SHIFT Adopting a relational approach represents a paradigm shift in the way social psychology attempts to understand the dynamics of interracial contact. Specifically, research will need to extend beyond the typical processes associated with person perception (i.e., stereotyping) and begin to consider meta‐ perceptual processes. Such an extension is not trivial, given the diVerences between the person perception and meta‐perceptual frameworks. In person perception research, for instance, the focus is on individuals’ perceptions of and feelings about others (e.g., Whites’ beliefs about Blacks). The psychological analysis is primarily about intrasubjectivity, and the self (the perceiver) is conceptualized as separate and distinct from the other (often the target). In contrast, in meta‐perception research the focus is on individuals’ perceptions of and feelings about how others view them (e.g., Whites’ perceptions of Blacks’ beliefs about Whites). The psychological analysis is primarily about intersubjectivity, and both the self and other are conceptualized as interconnected in important ways. Consistent with Baldwin’s (1992, p. 468) research on relational schemas, our approach to interracial interaction maintains that ‘‘as well as observing the external behaviors of self and other, an individual in an interaction will be aware of his or her own internal states

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

123

and also quite likely will be inferring something about the internal state of the other person.’’ The proposed framework also requires a careful rethinking of the ways in which Whites and ethnic minorities are included as research participants. The majority on studies of interracial contact adopt a distinctly individualistic approach. That is, the dominant paradigm of this research includes only one individual. In its purest form, the relational approach is reflected in methodologies that include both Whites and ethnic minorities as participants in the same study. Our review of the literature revealed that this quintessential relational methodology has been underutilized (for the exception see Hyers & Swim, 1998; Ickes, 1984; Shelton, 2003; Vorauer & Kumhry, 2001; see also Hebl & Dovidio, 2005 for a similar argument), resulting in the neglect of interesting and important research questions that can only be examined with a relational lens. For example, Do Whites and ethnic minorities share similar experiences during interracial interactions? Whose beliefs, Whites’ or ethnic minorities’, are most influential for individuals’ aVective, cognitive, and behavioral outcomes during interracial interactions? Do Whites’ and ethnic minorities’ beliefs interact to influence individuals’ experiences in a manner that diVers from the independent eVects of each person’s beliefs? Although a two‐participant methodology aVords the most ecologically valid benefits of a relational approach, it can be a logistical nightmare, contributing to why researchers often shy away from employing it. Admittedly, we do not always utilize this methodology in our research. We argue, however, that taking a relational perspective to the study of interracial interactions is largely independent of the particular methodology employed. For instance, a relational perspective can easily be adopted through the use of confederates. The use of confederates allows for greater experimental control than studies with two naı¨ve participants, while continuing to allow for the intersubjectivity of the interaction to be highlighted and influential. When facing an interaction with an actual person of a diVerent race, albeit a confederate, in other words, participants are less likely to focus on their own perceptions of the other person exclusively, but, rather, to consider what the other person in the interaction might be thinking about them as well. Furthermore, even in situations in which contact with either actual naı¨ve participants or with confederates is not possible, a relational approach can still be adopted. For instance, the relational approach can be embodied in the types of questions asked of participants, irrespective of whether they interact with another individual. Specifically, a relational spirit is captured by the fact that participants are not only asked to evaluate an outgroup ‘‘target,’’ but they are also asked to think about how the outgroup ‘‘target’’ is likely to evaluate them. In other words, participants’ attention to the self as both a perceiver and target

124

J. NICOLE SHELTON AND JENNIFER A. RICHESON

of the outgroup emphasizes the notion that the participant is not in the interaction alone—the essence of a relational perspective.

B. CLASSICS REVISITED In some ways, this shift from an individualistic to a relational approach to studying interracial interactions could be considered a call for social psychology to revisit the classic analyses of stigma and race relations. GoVman (1963) and Jones et al. (1984) placed the management of stigma in social interactions at the forefront of their discussion of stigma. They emphasized the importance of considering the experiences of both nonstigmatized and stigmatized individuals in interactions. In addition, although the topic of interracial contact was tangential to the broader issue of attitude–behavior consistency, LaPiere’s (1934) classic study shows the utility of the relational approach when studying interactions. LaPiere discovered that when he asked in a written letter whether or not an establishment would provide service to a Chinese couple, the overwhelming majority of responses were a definitive no. However, when he and the Chinese couple arrived at the establishment, the couple was provided service more often than not. LaPiere eloquently noted that the written letter required people to respond to an entirely symbolic situation. What was missing in the symbolic situation was the intersubjectivity of the situation. In the symbolic situation, people failed to think about how concerned they would be about being evaluated by others, including outgroup members, in the situation. Additionally, in the symbolic situation, people did not have to consider characteristics of outgroup individuals that might alter their decision. Consequently, the symbolic situation was unable to map on to the realities of actual interracial contact. The classic study by Word, Zanna, and Cooper (1974) also emphasized a relational approach when exploring the dynamics of interracial interactions. Word et al. (1974) examined interactions between Whites and Blacks in the context of a job interview setting. In Study 1, they found that naı¨ve, White interviewers displayed less friendly nonverbal behaviors toward Black interviewees than toward White interviewees (the interviewees were trained confederates). In Study 2, Word et al. (1974) trained White interviewers (confederates) to display either friendly or unfriendly nonverbal behaviors toward naı¨ve, White interviewees. Word et al. (1974) found that the naı¨ve, White interviewees who were the target of unfriendly nonverbal behaviors performed worse during the interview than those who were the target of friendly nonverbal behaviors. This classic study shows the importance of taking a relational approach when studying interracial interactions by exploring the interaction from the perspective of both individuals in the interaction. Not only

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

125

was it important to demonstrate that Whites behaved in a more negative manner toward Blacks than Whites, but it was also important to demonstrate how these behaviors influenced the other interactant. The goal for exploring interracial interactions from a relational approach is not to reveal one specific psychological phenomenon. Instead the goal is to demonstrate that the approach may generate novel findings that can lead to a more ecologically valid understanding of the dynamics of interracial interactions. Because of diVerences in sociocultural perspectives, the psychological experience of an interracial interaction may diVer for Whites and ethnic minorities. It is only by examining both individuals’ perspectives that we are able to uncover, and begin to understand, these diVerences.

C. OVERVIEW OF CHAPTER What follows is a description of research from our and others’ laboratories on interracial interactions derived from, or consistent with, a relational approach. We focus on two issues to illustrate what a relational approach can contribute to the study of interracial interactions. First, we consider the ways in which individuals’ beliefs regarding how their interaction partner is likely to view them influence the avoidance of interracial contact. Second, we examine the dynamics of interracial interactions, focusing on the consequences that individuals’ beliefs have for their own and their partner’s experiences during the interaction. Before we turn to these two issues, however, we provide a general background about intergroup meta‐perceptions.

II. Intergroup Meta‐Perceptions Meta‐perceptions are individuals’ beliefs about another person’s impression of them (Kenny & DePaulo, 1993). Theories, such as symbolic interaction theory, attachment theory, and self‐verification theory, suggest that individuals give considerable thought to understanding others’ reactions to them (Vorauer, 2001). Consistent with research on general meta‐perceptions, Whites and ethnic minorities also think about how others might view them during interracial interactions, and these meta‐perceptions shape their interaction experiences. Indeed, research assessing cultural stereotypes shows that Whites and ethnic minorities are aware of the stereotypes associated with their ingroup and outgroup (Krueger, 1996; Sigelman & Tuch, 1997). For instance, Wout, Shih, and Jackson (2005) examined Blacks’ perceptions of how they are viewed by both Whites and Blacks. They found that Blacks expect Whites more than other Blacks to view Blacks as a group as

126

J. NICOLE SHELTON AND JENNIFER A. RICHESON

untrustworthy, athletic, aggressive, and not hardworking, and to apply those stereotypes during interracial contact experiences. The majority of research on meta‐perception in an intergroup context has been inspired by the study of Vorauer and her colleagues. In one study, Whites listed the stereotypes they thought that First Nations have about Whites (Vorauer, Main, & O’Connell, 1998). Results revealed that Whites believe that First Nations perceive Whites as being prejudiced, selfish, and closed‐minded. In subsequent research, Vorauer, Hunter, Main, and Roy (2000) found that when Whites imagined (Study 1) or anticipated (Study 2) having an interaction with a First Nations person, these meta‐stereotypes (i.e., prejudiced, selfish, closed‐minded) were activated as measured by a word fragment completion task and a lexical decision‐making task. Importantly, exposure to a First Nations person whom Whites did not expect to interact with was not associated with an increase in meta‐stereotype activation. Consistent with research employing the person perception model, such exposure was associated with increased activation of negative stereotypes about First Nations individuals. Taken together, research on meta‐perceptions suggests that the context of an interracial interaction often activates concerns about being judged negatively in both Whites and ethnic minorities—a relational facet of interracial contact that is unlikely to be revealed by more individualistic approaches to prejudice, stereotyping, and interracial contact. Contemporary models and theories of prejudice have begun to incorporate these meta‐perceptions. Largely, this research suggests that Whites and ethnic minorities wrestle with these meta‐perceptions during interracial interactions, and these meta‐perceptions manifest as interpersonal concerns that individuals have regarding interracial interactions. Most models and theories propose that Whites are often concerned with appearing prejudiced (Crandall & Eshleman, 2003; Dunton & Fazio, 1997; Monin & Miller, 2001; Monteith, 1993; Plant & Devine, 1998) and ethnic minorities are often concerned with being treated negatively because of prejudice during interracial interactions (Crocker, Major, & Steele, 1998; Major, Quinton, & McCoy, 2002; Miller & Meyers, 1998). Whites’ and ethnic minorities’ meta‐perceptions, and resultant interpersonal concerns, are similar and diVerent in important ways. The similarity lies in the fact that both individuals are concerned that outgroup members will perceive them in a way that threatens their true identity (Steele, Spencer, & Aronson, 2002) and outgroup members will ultimately reject them during interracial interactions. The fundamental diVerence between Whites’ and ethnic minorities’ meta‐perceptions is that Whites’ concerns during interracial interactions are primarily self‐focused, whereas ethnic minorities’ concerns focus on both the self and their White interaction partners. Specifically, Whites are often concerned about maintaining their self‐image as egalitarian, unbiased people during the interaction. Ethnic minorities do not want to be perceived in a

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

127

stereotypical manner, and they do not want to be the target of biased, unfair treatment from Whites. Essentially, therefore, both Whites and Blacks are concerned that the White interaction partner will behave in prejudiced ways during their interactions with one another. These concerns have implications for individuals’ experiences during actual interactions, and as we demonstrate in the next section, contribute to whether interracial interactions occur at all.

III. Avoiding Interracial Interactions On many occasions, people have control over the types of social interactions they engage in. When given the opportunity, people typically choose situations and interaction partners that are consistent with their personal preferences (Ickes, Snyder, & Garcia, 1997). Research shows that when choosing situations people tend to avoid interethnic interactions (Plant & Devine, 2003; Stephan & Stephan, 1985; Towles‐Schwen & Fazio, 2003). Building on the notion that Whites and ethnic minorities hold intergroup meta‐ perceptions, we began to explore how these meta‐perceptions might influence the extent to which individuals choose to avoid interacting with outgroup members. Specifically, we began by addressing the influence of meta‐perceptions on individuals’ explanations for why they and outgroup members avoid interracial interactions. Individuals are likely to avoid such contact in part because they are concerned with how they will be viewed by outgroup members; more specifically, they are concerned that outgroup members will reject them because of their social identity. To what extent, however, do people believe that outgroup members have similar interpersonal concerns about interracial interactions? Is it possible that individuals fail to recognize that outgroup members’ avoidance of interracial contact reflects the same interpersonal concerns—being perceived stereotypically and fears of rejection because of their social identity—as their own? We started to address this question by first illustrating that Whites and ethnic minorities believe they, as well as their ingroup, are more interested in engaging in interracial contact than outgroup members (Shelton & Richeson, 2005). In our first study, we asked White and Black students to indicate their interest in having more outgroup friends, as well as having more contact with outgroup individuals in general. In addition, we asked these same students to indicate to what extent they thought the average White and Black student was interested in having more outgroup friends, as well as having more contact in general with outgroup members. As predicted, both racial groups perceived that they wanted to have more outgroup friends and interracial contact than the average outgroup student.

128

J. NICOLE SHELTON AND JENNIFER A. RICHESON

For example, Whites reported that they wanted to have more contact with Blacks, but that Blacks did not want to have more contact with them (see Tropp & Anderson, 2003 for a similar pattern of results).

A. DIVERGENT ATTRIBUTIONS Given this discrepancy in individuals’ perception of their own and outgroup members’ interest in having contact, we reasoned that individuals may rationalize that lack of interest is a more likely explanation for why outgroup members avoid interracial contact than those outgroup members’ concerns about being rejected. Thus, in the next set of studies (Shelton & Richeson, 2005, Studies 3–5) we examined the extent to which Whites and Blacks make divergent attributions about their own and an outgroup member’s explanation for avoiding interracial interactions. We predicted that when explaining why interracial contact fails to occur, Whites and Blacks would attribute their own failure to initiate contact to concerns with being rejected because of their race, whereas they would attribute an outgroup member’s failure to initiate contact to lack of interest. For our initial test of this prediction we asked Whites and Blacks to imagine the following situation: You enter the dining hall for dinner. You are alone because your close friends are in a review session. As you look around the dining hall for a place to sit, you notice several White (Black) students who live near you sitting together. These students also notice you. However, neither of you explicitly makes a move to sit together. Approximately half of the White and Black participants imagined that the targeted students were White, and the other half imagined that the targeted students were Black. After imagining the situation participants answered the following questions: How likely is it that fear of being rejected because of your race would inhibit you from sitting with these students; how likely is it that your lack of interest in getting to know the students would inhibit you from sitting with them; how likely is that fear of being rejected because of their race would inhibit the students from inviting you over; and how likely is that the other students’ lack of interest in getting to know you would inhibit them from inviting you over? The findings supported our predictions, revealing that both racial groups believed diVerent psychological states were underlying their own and the outgroup members’ motivations for not initiating interracial contact (Fig. 1 contains data for Blacks). Specifically, Whites and Blacks indicated that fear of rejection because of their race would be a more likely explanation for their own inaction than for the outgroup members’ inaction. Conversely, both Whites and Blacks indicated that lack of interest would be a more likely

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

129

Fig. 1. Attributions for why Blacks avoid interracial contact.

explanation for the outgroup members’ inaction than for their own. Additionally, when considering the attributions behind the outgroup members’ failure to establish interracial contact, individuals believed that the outgroup members’ lack of interest was a more likely attribution than their fear of being rejected. Conversely, when considering the reasons behind their own failure to establish interracial contact, individuals tended to report that fear of rejection was a better attribution than lack of interest. B. DIVERGENT ATTRIBUTIONS AND RACIAL ATTITUDES In subsequent research, we assessed the extent to which these divergent attributions are moderated by Whites’ racial attitudes (under review). In order to examine this question, we administered the attitudes toward Blacks scale (Brigham, 1993) to a group of White students during a pretesting session and several weeks later invited them to participate in a second session on social perception and friendship development. During the second session, the experimenter informed participants that they would have an interaction with another student. Furthermore, in order to encourage a sense of interdependence, the experimenter explained that if the two students decided they could be friends, then they would work on additional tasks together. Immediately prior to the supposed interaction, the experimenter explained that in order to facilitate the discussion, participants would exchange background information and a picture with their partner. Participants received a same‐sex photograph of either a White or Black student (confederate). After the participants had the opportunity to view

130

J. NICOLE SHELTON AND JENNIFER A. RICHESON

the photograph and read the background information sheet, they completed a preinteraction questionnaire. Specifically, participants indicated how interested they were in being friends with their partner, how interested they thought their partner was in being friends, how concerned they were that the other student would accept them as a friend, and how concerned they thought the other student was about being accepted as a friend. Results revealed that all participants reported the self–other discrepancy in their attributions when their partner was Black but not when their partner was White. That is, all participants indicated that they were more interested in becoming friends and more concerned about being accepted than the other student. However, low‐prejudice Whites were especially likely to make these divergent attributions. Although this result may seem counterintuitive, it is consistent with previous research on meta‐stereotypes. Specifically, Vorauer et al. (1998) found that low‐prejudice Whites believe that racial minorities will contrast them against the meta‐stereotype that they are prejudiced, whereas high‐prejudice Whites believe that racial minorities will assimilate them to the meta‐stereotype. If low‐prejudice Whites do not expect Blacks to think that they are prejudiced, then they are also particularly likely to underestimate the extent to which a Black interaction partner might be concerned that they will reject them. Moreover, once low‐prejudice Whites take fear of rejection oV the table as a possible explanation for why a Black individual might not want to pursue a friendship with them, then they are left with the possibility that the Black individual is simply not interested in interacting with them. By denying the possibility that fear of rejection may motivate potential interaction partners, while simultaneously harboring concerns about being rejected themselves, low‐prejudice Whites exaggerate the self–other bias we obtained in our original study. Collectively, our findings suggest that individuals are concerned about how they will be viewed and treated by outgroup members during interracial interactions. In addition, perhaps because they are focused on themselves, individuals have a diYcult time realizing that outgroup members may also be concerned with how they will be viewed and treated. In fact, individuals tend to (mis)perceive outgroup members as not being interested in pursuing interracial contact.

C. IMPLICATIONS OF DIVERGENT ATTRIBUTIONS One implication of a self–other bias in which outgroup members are thought to avoid interracial contact because of lack of interest but ingroup members avoid contact because they are concerned about rejection is the

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

131

possibility that individuals will simply avoid interracial contact all together. We set out to explore this possibility empirically. We examined whether individuals’ proclivity to generate the self–other bias, when considering why interracial contact does not occur, influences the probability that they will engage in interracial interactions in the future (Shelton & Richeson, 2005, Study 6). At an initial testing session during the 2nd week of the academic year, White students responded to the interracial contact vignette (i.e., imagine you enter a dining hall and notice Black students) and the associated attribution questions described previously. In addition, we asked these students to report the approximate percentage of their social interactions that occurs with other White individuals and the approximate percentage that occurs with Blacks. Approximately 7 weeks later, we asked these same White students to report how much contact they had had with Whites and Blacks since the beginning of the school year. Consistent with predictions, the more Whites believed that fear of rejection because of their race was a better explanation for their own inaction in the scenario (i.e., their unwillingness to pursue interracial contact) compared to the Black students’ inaction, the less contact they had with Blacks over the course of the first semester of school. In other words, Whites’ disproportionate focus on their own susceptibility of being rejected by Blacks in a hypothetical scenario seemed to influence the frequency with which they engaged in interracial contact over time. Although our data speak specifically to the comparison of self and other regarding the avoidance of interracial contact, several other studies show that meta‐perceptions alone influence the extent to which individuals avoid interracial contact. Building on theory on rejection sensitivity in close relationships, Mendoza‐Denton, Purdie, Downey, Davis, and Pietrzak (2002) developed a measure of race‐based rejection sensitivity for Blacks in which participants are asked to rate the extent to which they would be concerned about and expect to be the target of racial prejudice across several scenarios. For example, participants are asked: Imagine you have just finished shopping, and you are leaving the store carrying several bags. It is closing time, and several people are filing out of the store at once. Suddenly, the alarm begins to sound, and a security guard comes over to investigate. Next, participants rate how concerned or anxious they would be that the guard might stop them because of their race/ethnicity. In addition, participants rate how much they would expect that the guard might stop them because of their race/ethnicity. According to the rejection‐sensitivity theory, Blacks who are high in race‐based rejection sensitivity anxiously expect to be rejected because of their race because of previous experiences with discrimination and awareness of stereotypes associated with their group. Mendoza‐Denton et al. (2002) found that Blacks high in race‐based rejection sensitivity had

132

J. NICOLE SHELTON AND JENNIFER A. RICHESON

fewer Whites friends and interacted with professors and teaching assistants less compared to students low in race‐based rejection sensitivity.1 Race‐ based rejection sensitivity, however, was unrelated to the number of Black friends participants reported having demonstrating that Blacks higher in race‐based rejection sensitivity are likely to avoid only those they feel are most likely to reject them on the basis of race. Research on a related construct, stigma consciousness, also provides support for the notion that individuals’ perceptions of how outgroup members will perceive them predict their avoidance of intergroup settings. Pinel’s (1999) research on stigma consciousness reveals that individuals are aware that their social identity is a pivotal factor in how people interact with them. People high in stigma consciousness expect to be viewed through the lenses of stereotypes associated with their group, and these expectations prompt them to avoid situations in which the possibility that they could be stereotyped is high. For example, Pinel (1999, Study 6) found that when women were required to have an interaction with men during a game of ‘‘jeopardy,’’ the higher women scored on an individual diVerence measure of stigma consciousness the more likely they were to avoid stereotypically male topics. For example, women high in stigma consciousness avoided selected topics such as automobile names and the military. Thus, although not in the context of interracial interactions, these findings corroborate our claims that individuals’ meta‐perceptions regarding outgroup members impact their wariness regarding intergroup settings in which there is the potential to become the victim of prejudice.

D. RECONVERGING THE ATTRIBUTIONS Considered as a whole, the research suggests that individuals’ meta‐ perceptions influence their own behavior, but individuals often fail to see how similar concerns might impact the behavior of outgroup members. As other researchers have argued with respect to self–other biases in general (Prentice & Miller, 1996; Pronin, Gilovich, & Ross, 2004; Vorauer, 2001), the divergent attributions regarding why people avoid interracial interactions are likely a result of a diVerence in individuals’ access to private self‐ knowledge. As actors, individuals have access to their inner feelings and beliefs. Because these feelings and beliefs are salient to them, actors assume they must be transparent to others as well. As observers, however,

1 The professors and teaching assistants are most likely to be White because the institution where this study was conducted is predominately White.

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

133

individuals are not privy to others’ inner feelings and beliefs. Instead, they only see others’ behavior, which they take as reflecting their true selves. Thus, as actors, Whites and Blacks are aware of their interpersonal concerns with prejudice and are likely to use them when making judgments about why they avoid interracial encounters. As observers, however, Whites and Blacks are not privy to how outgroup individuals are feeling; instead, they are only privy to the outgroup individuals’ behavior. Although access to private knowledge and feelings is certain to contribute to the divergent attributions, we oVer an additional possibility. Individuals might also make divergent attributions about why they and outgroup members avoid interracial contact because outgroup members are not included as part of their sense of self (Aron, Aron, & Smollan, 1992). Research suggests that individuals include ingroup members (but not outgroup members) in the self (Smith & Henry, 1996; Smith, Coats, & Walling, 1999; Tropp & Wright, 2001). Because of this inclusion of other in the self, individuals grow to feel and think as if the other person is a part of the self and vice versa. With the inclusion of other in the self, characteristics of the other are considered to be part of the self, and, furthermore, individuals’ self‐representations include the other. That is, there is a perception and feeling of interconnectedness between the self and other. Because of this interconnectedness, individuals are able to empathize with ingroup members’ concerns, fears, and problems, whereas they are relatively unable (or unlikely) to do so with outgroup members’ concerns, fears, and problems. If outgroup members were included in the self, however, individuals would be more likely to consider their concerns and fears and recognize that they are identical to their own. Specific to our study, if outgroup members were included in the self, then individuals should not make divergent attributions about the self and other because they should believe that the same forces driving their own behaviors are also driving the behaviors of outgroup members. We tested our belief that inclusion of the other in the self contributes to the self–other bias regarding the avoidance of interracial interactions (Shelton & Richeson, 2005, Study 7). Similar to our other studies on this topic, we asked half of the White students in our sample to imagine the dining hall scenario, described previously, in which neither they nor a group of Black (or White) students ‘‘make a move to sit together.’’ We asked the other half not only to imagine the same scenario, but also to imagine that, ‘‘You don’t know these students well, but your best friend enjoys hanging out with them.’’ Why ask participants to imagine that their best friend enjoys socializing with these outgroup members? According to the extended contact eVect (Wright, Aron, McLaughlin‐Volpe, & Ropp, 1997), when people are aware that an ingroup member has a close relationship with an outgroup

134

J. NICOLE SHELTON AND JENNIFER A. RICHESON

member, they tend to include the outgroup member in the self. Because individuals in a close relationship are perceived as a single representation (Sedikides, Olsen, & Reis, 1993) the outgroup member is now linked to the ingroup member, who is already included in the self. Therefore, by asking participants to imagine that their best friend, who is most likely to be an ingroup member, enjoys hanging out with the outgroup individuals in the scenario, we are encouraging them to include these outgroup members in the self. After reading the scenario, participants answered the following questions: How likely is it that fear of being rejected because of your race would inhibit you from sitting with these students; how likely is it that your lack of interest in getting to know the students would inhibit you from sitting with them; how likely is it that fear of being rejected because of their race would inhibit the students from inviting you over; and how likely is it that the other students’ lack of interest in getting to know you would inhibit them from inviting you over? Our findings revealed that for Whites who did not have knowledge of their best friend’s opinion, the now familiar pattern of divergent attributions emerged when the targeted students were Black, but not when the targeted students were White. That is, Whites reported that fear of rejection because of their race explained their inaction more than the Black students’ inaction, whereas lack of interest explained the Black students’ inaction more than their own. However, Whites who had information that their best friend enjoys socializing with the targeted Black students did not reveal the same pattern of divergent attributions. In fact, Whites who had information about their best friend’s opinion reported that the Black targeted students were just as likely to be inhibited by fears of rejection because of their race as they themselves were. Likewise, there was no diVerence in the extent to which Whites reported that lack of interest influenced their own and the targeted Black students’ behavior. A closer examination of the pattern of results revealed that appearances of outgroup acceptance by knowing about a cross‐race friendship reduce Whites’ fears of rejection. Taken together, these findings suggest that expanding the self to include outgroup members may be an eVective strategy for reducing biased perceptions of why ingroup and outgroup members avoid interacting with one another, and, by extension, may serve to increase the extent to which individuals actually engage in interracial interactions. These two explanations, one based on diVerential access to internal feelings and the other on the (lack of) inclusion of the other in the self, both oVer reasonable accounts for the divergent attributions individuals make regarding why they and members of outgroups avoid interracial interactions. Of course, future research is needed in order to explore additional factors that

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

135

lead to divergent attributions that undermine both the quality and quantity of interracial interactions. 1. Summary When examining why interracial contact occurs so rarely, researchers have focused on individuals’ beliefs about others. Research that has taken a more relational approach has shifted the focus from person perception to meta‐ perception. In doing so, researchers have learned that interracial interactions may occur so rarely because people are concerned with how they are being perceived by the outgroup not just because they dislike the outgroup.

IV. Dynamics of Interracial Contact: A View from Both Sides One of the original goals of the ‘‘contact hypothesis’’ was to understand the conditions under which contact between members of diVerent groups would improve intergroup attitudes (Allport, 1954). Although the improvement of attitudes remains an extremely important outcome for research on intergroup contact, recent research has begun to explore immediate experiences and outcomes of intergroup contact. In our relational approach to study interracial interactions, we have focused on how both Whites’ and ethnic minorities’ interpersonal concerns with prejudice and their racial attitudes influence aVective, cognitive, and behavioral outcomes for the self and one’s partner during interracial interactions.

V. Interpersonal Concerns with Prejudice and Interracial Contact A. WHITES’ PREJUDICE CONCERNS: IMPLICATIONS FOR THE SELF Given that Whites are often concerned with appearing prejudiced during interactions, we set out to document aVective, cognitive, and behavioral consequences of these concerns for both Whites and their ethnic minority partners during interracial interactions. We began by exploring how Whites’ concerns with appearing prejudiced influence cognitive functioning for the self. Specifically, we considered the extent to which harboring concerns about prejudice during an interracial interaction may deplete Whites of important cognitive resources, resulting in their underperformance on

136

J. NICOLE SHELTON AND JENNIFER A. RICHESON

subsequent cognitive tasks. This possibility stems from research showing that individuals carefully monitor their thoughts, feelings, and behaviors during interracial interactions in order to avoid being perceived as prejudiced (Devine et al., 1996; Monteith, 1993). Such regulation and monitoring of thoughts, feelings, and behavior are cognitively demanding, however, resulting in the temporary depletion of important cognitive resources (Engle, Conway, Tuholski, & Shisler, 1995; Muraven & Baumeister, 2000). Thus, this study suggests that concerns about prejudice will leave Whites cognitively exhausted. In order to examine this possibility, we activated the prejudice concerns of White participants just prior to an interracial interaction (Richeson & Trawalter, 2005, Study 1). After the interaction, these same participants’ extent of cognitive depletion was tested using the Stroop color‐naming paradigm. We predicted that compared to a control group of White participants for whom prejudice concerns were not heightened, participants whose concerns were triggered would reveal greater evidence of cognitive disruption after the interaction. Specifically, we asked participants to take part in a study presumably examining ‘‘serial cognition.’’ We told them that we were interested in the influence of one cognitive task on a subsequent task when there is a delay between the two. For the first cognitive task, participants completed the race version of the Implicit Association Test (IAT) (Greenwald, McGhee, & Schwartz, 1998), which is a measure of automatic racial bias. Immediately after completing the IAT, participants received one of two types of feedback. In the prejudice concern condition, the experimenter told participants, ‘‘Several studies have used this task to study racial bias. These studies show that most people are more prejudiced than they think they are.’’ In the control condition, performance concerns were activated by telling participants, ‘‘Several studies have used this task to study category associations. These studies show that most people perform worse than they think they did.’’ Based on previous research (Dutton & Lake, 1973; Monteith, Ashburn‐Nardo, Voils, & Czopp, 2002), the prejudice concern feedback, but not the performance concern feedback, was expected to prompt participants to engage in the regulation of their thoughts, feelings, and behaviors during the interracial interaction, and, ultimately to the depletion of cognitive resources. Immediately after receiving the feedback, participants went to a diVerent room where they engaged in an ostensibly unrelated session with either a Black or a White experimenter. During this ‘‘delay task,’’ the experimenter remained in the room and videotaped the participants for approximately 8 minutes providing their opinions on several topics, including one race‐ related topic (e.g., campus diversity). After this task, participants returned to the room with the original experimenter and completed the second cognitive

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

137

task—the Stroop color‐naming task, which measures inhibitory performance. The more cognitively depleted participants were, the worse they were expected to perform on the Stroop task. Consistent with our predictions, results revealed that in the interracial interaction (i.e., when the second experimenter was Black), Whites who received the prejudice feedback performed significantly worse on the Stroop task than Whites who received the general performance feedback. The feedback did not influence Whites’ performance on the Stroop task in the same‐race interaction (i.e., when the second experimenter was White). In a subsequent study, Whites’ concerns about appearing prejudiced were allayed rather than elevated (Richeson & Trawalter, 2005, Study 2). Specifically, Whites were asked to have an interaction with either a Black or White confederate during which they were videotaped responding to a question about the costs and benefits of racial profiling. For most Whites, prejudice concerns will be active during such an interaction, especially with the Black confederate. In order to reduce these concerns, however, the confederate provided half of the participants with a written response to the question that they were allowed to use as a guide during the interaction. Thus, Whites could attribute any of their comments that could be perceived as controversial or potentially biased to the script, rather than to their own opinion. Consequently, Whites who were provided with the script were expected to be less concerned about appearing prejudiced, and, therefore, less likely to engage in eVortful self‐regulation in order to maintain a nonprejudiced self‐image. After the interaction, participants completed the Stroop task in order to assess the extent to which the interaction left them depleted of cognitive resources. Consistent with predictions, participants who were provided with the script performed as well on the Stroop task as participants who interacted with the White confederate, all of whom performed better on the Stroop task than participants who interacted with the Black confederate without the script. Considered in tandem with the previous experiment, this study underscores the potential for prejudice concerns to result in relatively negative outcomes for the self. Whites’ interpersonal concerns with prejudice are likely to influence more than cognitive outcomes during interracial interactions. Research has illustrated that interracial contact fosters negative aVective reactions (e.g., Britt, Boniecki, Vescio, Biernat, & Brown, 1996; Stephan & Stephan, 1985). Whites’ concerns with appearing prejudiced may add to the aVective distress they experience. Therefore, our next goal was to examine whether harboring concerns with appearing prejudiced results in negative aVect for Whites during interracial interactions (Shelton, 2003). In this study, Whites and Blacks engaged in a get‐to‐know‐you interaction with one another. Prior to the interaction, the experimenter told all of the White participants that

138

J. NICOLE SHELTON AND JENNIFER A. RICHESON

they would be interacting with a Black student. In addition, the experimenter told half of the White participants that ‘‘previous psychological research has shown that during interracial interactions Whites’ impressions of Blacks are often biased by racial stereotypes, which often make them appear as prejudiced individuals. You should try not to be prejudiced during this interaction.’’ The experimenter did not provide this information to the other half of the White participants. After receiving this information, participants met their partner and engaged in an approximately 15‐minute interaction, which was videotaped. In order to facilitate the discussion, the experimenter gave participants eight discussion topics. The first four topics were race neutral (e.g., discuss your hobbies) and the last four were race relevant (e.g., discuss campus race relations). Next, participants completed various questionnaires, including items to assess how much anxiety they experienced during the interaction (e.g., nervous, anxious, tense, worry). As predicted, Whites who were instructed to try not to be prejudiced reported experiencing more anxiety compared to those who were not given these instructions. Additional research supports the notion that Whites’ concerns with prejudice yield negative aVective reactions during interracial encounters (Devine et al., 1996; Ickes, 1984; Plant & Devine, 1998, 2003). For instance, Plant and Butz (2004, Study 1) led non‐Black participants to be concerned with being perceived as prejudiced by manipulating their expectations about their self‐eYcacy for facilitating pleasant interracial interactions. They found that non‐Black participants who were provided feedback that they are likely to experience diYculty when interacting with Blacks experienced more anxiety about an anticipated interaction with a Black person, compared to participants who did not receive feedback. In a subsequent study, Plant and Butz (2004, Study 2) found that non‐Blacks who received negative feedback experienced more anxiety, anger, and general negative self‐directed emotions about the anticipated interaction with a Black person, compared to participants who received positive eYcacy feedback. Moreover, individuals who received the negative, rather than the positive, feedback were more anxious during the interaction, as measured by self‐report and coders’ ratings. In a related study, Vorauer et al. (1998, Study 2) found that the more Whites expected a First Nations interaction partner to view them as consistent with the stereotypes associated with their group (e.g., unfair, prejudiced, close‐minded), the less they expected to enjoy an interracial interaction and the more they expected to have negative feelings during the interaction. Moreover, research suggests that Whites’ perceptions regarding how they will be viewed in an interracial interaction can even have negative implications for their mental health. Specifically, the more Whites believed that their First Nations interaction partner perceived them in a stereotypical manner, the lower their self‐esteem and self‐concept clarity after the interaction

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

139

(Vorauer et al., 1998, Study 3). Together, these findings illustrate that Whites’ beliefs that they might be viewed negatively by their interaction partners can have deleterious consequences for their aVective experiences both in anticipation of, during, as well as after, interracial interactions.

B. WHITES’ PREJUDICE CONCERNS: IMPLICATIONS FOR PARTNER Given our relational approach to studying interracial interactions, it was important for us also to explore how Whites’ concerns about appearing prejudiced might impact ethnic minorities during interracial interactions. How might Whites’ concerns with prejudice influence their ethnic minority partner’s experiences during an interracial interaction? If the negative aVect and cognitive disruption resulting from Whites’ preoccupation with how they are being perceived is apparent in the interaction, then it is easy to imagine that ethnic minorities are likely to have a negative experience. It is also possible, however, that under certain conditions Whites who are concerned with appearing prejudiced will engage in strategies to facilitate a smooth interaction. Indeed, research finds that when individuals are aware that an interaction partner might perceive them negatively, they often pursue compensatory strategies designed to elicit more favorable responses from their interaction partners (Hilton & Darley, 1985; Ickes, Patterson, Rajecki, & Tanford, 1982; Swann, 1987; Swann & Ely, 1984; Swann & Read, 1981). Ickes et al. (1982), for example, illustrated that perceivers who were led to believe that a target who they would interact with was unfriendly, compared to perceivers without any expectations about the target, engaged in more aYliative behaviors (e.g., directed eye gaze, smiling, and laughing) during the interaction. More pertinent to this discussion, targets rated perceivers who had the unfriendly expectation as more trustworthy and sincere, as well as friendlier than perceivers who did not have any expectations. In addition, targets displayed more aYliative behaviors with perceivers who had the unfriendly expectations than perceivers without any expectations. Whites who are concerned about appearing prejudiced during interracial interactions face an analogous predicament. Essentially, Whites who are concerned with appearing prejudiced have a negative expectation—they expect ethnic minorities to perceive them negatively. As a result, Whites may engage in compensatory strategies to counter those expectations, which, in turn, could result in their ethnic minority interaction partners having relatively positive perceptions and experiences. We tested the prediction that ethnic minorities would have more favorable impressions of Whites who are concerned with appearing prejudiced than Whites who are not concerned. In

140

J. NICOLE SHELTON AND JENNIFER A. RICHESON

the Shelton (2003) study described previously, Whites and Blacks engaged in a ‘‘get‐to‐know‐you’’ interaction with one another in which they discussed four neutral and four racially sensitive topics. Recall that prior to the interaction half of the White participants were explicitly told to try not to be prejudiced. In addition to completing the anxiety measure noted before, participants indicated how much they liked their partner. Consistent with previous research on compensatory strategies, the Black participants in this study liked White partners who were instructed not to be prejudiced more than Whites who were not told to avoid prejudice (see also Wetzel, Blalock, & Bolger, 2004 for similar results). 1. Summary Whites’ concerns with prejudice can result in paradoxical findings for their own and ethnic minorities’ experiences during interracial interactions. The evidence is clear that Whites’ concerns with prejudice usurp their cognitive resources and negatively influence their aVective states during interracial interactions. However, it may be necessary for Whites to be concerned with appearing prejudiced, to a certain level, because this concern can create a more pleasant atmosphere for ethnic minorities, at least as indicated by ethnic minorities’ impressions of Whites. Moreover, these divergent experiences—that is, negative outcomes for Whites, but relatively positive outcomes for Blacks—could have only been revealed by a relational approach to the study of interracial contact.

C. ETHNIC MINORITIES’ PREJUDICE CONCERNS: IMPLICATIONS FOR SELF Ethnic minorities’ concerns that others harbor prejudices against their group are likely to influence their aVective, cognitive, and behavioral outcomes during interracial interactions. Indeed, recent research finds that placing ethnic minorities in a situation in which there is a heightened threat that they could be the target of prejudice negatively influences their aVective reactions about an anticipated interaction, and their feelings about future interethnic interactions. Tropp (2003) led Latino and Asian‐American participants to believe that they had been ‘‘randomly’’ assigned to have an interaction with a White student (confederate). Prior to the ostensible interaction, participants overheard a scripted conversation between the confederate and experimenter. Half of the participants heard the confederate tell the experimenter that he would prefer to switch partners because he would rather not interact with a Latino/Asian person. The other half heard the

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

141

confederate ask the experimenter about the length of the study because he did want to be late for class. Immediately after overhearing the conversation, participants completed an aVective measure regarding the anticipated interaction and a questionnaire regarding their expectations about future interactions with outgroup members. Results revealed that ethnic minorities who overheard the confederate say he would rather not interact with a Latino/ Asian person reported feeling more hostile and anxious about the anticipated interaction, and marginally less positive about interacting with outgroup members in general, compared to those who overheard the confederate make the race‐neutral comment. Thus, these results suggest that the expectation that one could be the target of prejudice has negative implications for individuals’ experiences in anticipation of interracial interactions. In our research, we extended these findings to actual interactions, as well as to the behavioral consequences of being concerned that others harbor prejudice against one’s group (Shelton, Richeson, & Salvatore, 2005). With respect to aVective outcomes, we expected to find results similar to Tropp (2003). With respect to behavioral outcomes, however, we expected to find a diVerent pattern of findings. Similar to Whites who harbor concerns about appearing prejudiced during interracial interactions, we predicted that ethnic minorities who expect to be the target of prejudice would engage in compensatory strategies to deflect the negative expectancy. Specifically, we expected ethnic minorities who are concerned about being the target of prejudice to display more positive, socially engaging behaviors than those who are not concerned. Consistent with this idea, Miller and Meyers (1998) reasoned that stigmatized individuals are able to reduce the threat posed by prejudice by engaging in behaviors that enable them to achieve desired outcomes in spite of their stigma. Specifically, Miller and her colleagues found that obese women who were visible to their interaction partners (and, thus, vulnerable to prejudice) behaved in a more socially skillful manner, compared to obese women who were not visible, during interactions with normal weight individuals in order to prevent the interaction from being negative (Miller, Rothblum, Barbour, Brand, & Felicio, 1990; Miller, Rothblum, Felicio, & Brand, 1995). Thus, we predicted that although ethnic minorities who expect to be the target of prejudice will have negative aVective experiences during interethnic interactions, this would occur in tandem with an increase in socially engaging behaviors, perhaps as a means to reduce the threat posed by the prejudiced tainted situation. In our first study (Shelton et al., 2005), we examined how ethnic minorities’ concerns with being the target of prejudice influence their aVect during daily interactions with a White or an ethnic minority roommate. During the first week of the academic year, we invited ethnic minority freshmen to participate in a study on freshmen roommates and their college

142

J. NICOLE SHELTON AND JENNIFER A. RICHESON

experiences. Approximately half of our participants had a White roommate whereas the other half had an ethnic minority roommate, though we did not inform them that the study was about interethnic roommate interactions. The students attended an orientation session in which they completed a variety of questionnaires, including the Stigma Consciousness Scale‐Race (Pinel, 1999), which assesses the extent to which individuals expect to be stereotyped because of their ethnicity. During the next 3 weeks, participants completed a daily questionnaire regarding the dynamics of their roommate interactions. On each questionnaire, participants responded to items that tapped into how much they liked their roommate (e.g., I feel less close/more negative toward my roommate today) and the extent to which they experienced negative aVect (e.g., tense, anxious, frustrated) during their interactions and/or when they thought about their roommate. We were also interested in how authentic ethnic minorities felt during their interactions with their roommate. Ethnic minorities who expect to be the target of prejudice are likely to be concerned that their behaviors are being perceived through the lenses of stereotypes. For example, the Black student who listens to rap music or the Mexican student who misses class may fear that his or her behaviors may be seen as stereotypical of their racial groups, rather than indicative of a personal preference. In essence, expecting others to harbor prejudiced beliefs about one’s group may inhibit ethnic minorities from feeling comfortable presenting their true self during social interactions. In fact, one compensatory strategy may be to change oneself to fit in with one’s White roommate so as not to be perceived in a stereotypical manner. To test this prediction, participants answered questions regarding how authentic they felt with their roommate each day (e.g., I felt I had to change myself to fit in with my roommate today). Finally, based on previous research on compensatory strategies, we surmised that self‐disclosure might be used by ethnic minorities who expect to be the target of prejudice as a means to facilitate harmony in social interactions. Indeed, self‐disclosure has been found to promote relational harmony and positive aVect within close relationships (Reis & Shaver, 1988). Ethnic minorities who expect to be the target of prejudice may use self‐disclosure to facilitate smooth interactions with their White roommates. To test this idea, participants answered questions regarding how much they self‐disclosed to their roommate each day (e.g., How much personal information about yourself did you disclose to your roommate today?). Disclosing information may also foster feelings of inauthenticity because individuals may feel that they can only select information that counteracts the stereotypes about their group. As a whole, the results from this study were consistent with our predictions. We found that the more ethnic minorities expected to be the target of

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

143

prejudice, the more negative aVect they experienced during interactions with a White roommate. However, the more ethnic minorities expected to be the target of prejudice, the less negative aVect they experienced during interactions with an ethnic minority roommate. In addition, the more ethnic minorities expected to be the target of prejudice, the less authentic they felt during interactions with a White roommate but not with an ethnic minority roommate. Finally, the more ethnic minorities expected to be the target of prejudice, the more they self‐disclosed during interactions with White, but not ethnic minority, roommates, reflecting, perhaps, the strategic employment of compensatory strategies. Unfortunately, inconsistent with our predictions, prejudice expectations were unrelated to how much participants liked their roommate, regardless of the roommate’s ethnicity. Nevertheless, taken together, these findings suggest that ethnic minorities’ prejudice concerns have negative implications for their aVective experiences and views of the self during interracial interactions. Conclusions drawn from this first study are tentative because of the limitations associated with correlational designs. In addition, the findings were limited to ethnic minorities who are dispositionally high in prejudice expectations. In a second study (Shelton et al., 2005, Study 2), we created a situation in which most ethnic minorities, regardless of their dispositional tendency, would expect to be the target of prejudice, then examined their behavior during, and aVect after, an interracial interaction. Specifically, we told participants that they would participate in two (ostensibly) unrelated studies. In the first study, participants read three short newspaper articles and answered questions about them. For half of the ethnic minority participants, two articles were prejudice‐neutral and one article discussed pervasive prejudice and discrimination directed toward ethnic minorities. For the other half (i.e., control participants), two articles were prejudice‐neutral and one article discussed pervasive prejudice and discrimination directed toward elderly individuals. After reading and answering questions about the articles, we asked participants if they would help us out with another study. We explained that the study was on first impressions and they would have a brief (10 minute) interaction with another participant who was down the hall and then complete some questionnaires about the interaction. After participants signed a new consent form, we took them to a new location where they met a White participant. The White participant had also read three newspaper articles prior to this interaction— the two neutral articles and the article about elderly prejudice. Immediately following the interaction, participants were taken to diVerent rooms, where they completed several questionnaires. Similar to the first study, participants indicated how much they liked their partner, how much negative aVect they experienced, how much they enjoyed

144

J. NICOLE SHELTON AND JENNIFER A. RICHESON

the interaction, and how authentic they felt during the interaction. In addition, as a measure of compensatory strategies, participants indicated how socially engaged they felt they were during the interaction. In addition, we videotaped each participant separately and had coders, who were blind to the experimental manipulation, evaluate participants’ verbal behaviors from the audiotape and nonverbal behaviors from the silent videotape of the interaction. For example, coders rated the extent to which participants leaned toward their partner, held their arms in an open/inviting manner, smiled, asked their partner questions about themselves, elaborated on their own thoughts/feelings, appeared engaged in the interaction, and talked. These behaviors have been found in previous study to facilitate smooth, harmonious social interactions. Consistent with predictions, compared to participants in the control condition, ethnic minorities who were primed to expect racial prejudice liked their partner less, experienced more negative aVect, and felt less authentic during the interaction. In addition, as indicated by self‐report measures as well as coders’ ratings, ethnic minorities who were primed to expect racial prejudice were more socially engaged during the interaction than control participants. Although the findings from Shelton et al. (2005) and Tropp (2003) provide converging evidence that ethnic minorities’ prejudice concerns result in negative aVective outcomes, Shelton (2003) found a diVerent pattern of results. Black participants engaged in an interaction with either a White participant who they were told ‘‘might be prejudiced against Blacks’’ or with a White participant about whom they were provided with no expectancy. Contrary to her predictions, as well as the Shelton et al. (2005) and Tropp (2003) studies, the prejudice expectation manipulation had no eVect on Blacks’ level of anxiety. More astonishingly, Blacks reported that they enjoyed the interaction more when they expected their partner to be prejudiced than when they did not. However, coders’ who rated Blacks’ nonverbal behaviors indicated that Blacks who expected their partner to be prejudiced fidgeted more than Blacks who did not have this expectation. If fidgeting is interpreted as a sign of discomfort, as it often is in research on prejudice, then Blacks’ ‘‘true’’ feelings in this study may be more consistent with the self‐report findings from Shelton et al. (2005) and Tropp (2003). However, it is possible that Shelton’s (2003) self‐report findings may diVer from the other two studies because of methodological diVerences (see section below on Methodological Considerations for additional issues). For example, it is possible that the specific expectation that one’s partner is prejudiced combined with an actual interaction, as used in the Shelton (2003) study, increases enjoyment for Blacks during interracial interactions because it reduces uncertainty in managing the interaction. In the other two studies, the operationalizations involved general primes that prejudice is

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

145

pervasive in the world (Shelton et al., 2005) or an anticipated interaction with a prejudiced person (Tropp, 2003). Determining the condition under which ethnic minorities’ prejudice concerns result in negative or positive implications for the self constitutes an interesting avenue for future inquiry.

D. ETHNIC MINORITIES’ PREJUDICE CONCERNS: IMPLICATIONS FOR PARTNER There are two types of reactions that Whites might have as a result of interacting with an ethnic minority who expects them to be prejudiced. First, Whites could consider the situation stressful, resulting in anxiety and discomfort. This stress may translate into Whites having negative experiences with ethnic minorities who expect them to be prejudiced. At first blush, this may seem like the most logical reaction. On further consideration, however, it becomes clear that the situation may not be too stressful for Whites who interact with ethnic minorities who expect them to be prejudiced. Recall that ethnic minorities who expect to be the target of prejudice often engage in strategies to create harmonious interactions. If their strategies are successful, then Whites may benefit and, ironically, have a more positive perception of interactions involving ethnic minorities who expect them to be prejudiced. We set out to test the prediction that ethnic minorities’ concerns with being the target of prejudice would yield positive aVective reactions for Whites during social interactions (Shelton, 2003; Shelton et al., 2005). Consistent with our prediction, Shelton (2003) found that Whites experienced less anxiety when their Black partner expected them to be prejudiced compared to when their partner did not have this expectation. In addition, Whites enjoyed the interaction more when their Black partner expected them to be prejudiced compared to when their partner did not have this expectation. Similarly, Shelton et al. (2005, Study 2) found that Whites who interacted with an ethnic minority individual who was primed to expect racial prejudice liked their partner more, experienced less negative aVect, and enjoyed the interaction more than Whites who interacted with an ethnic minority who was primed to expect prejudice against elderly individuals. Thus, as opposed to yielding a stressful situation for Whites and perpetuating intergroup tension, ethnic minorities’ prejudice concerns can cause Whites to have pleasant experiences during interethnic interactions. This is most likely the result of ethnic minorities’ positive engagement in the interaction—a compensatory strategy designed to undermine the likelihood of becoming the target of prejudice. Related research on intergroup contact involving gender dyads suggests, however, that stigmatized individuals’ prejudiced expectations can be

146

J. NICOLE SHELTON AND JENNIFER A. RICHESON

disruptive for intergroup interactions. Pinel (2002) had women, who were either high or low in stigma consciousness (i.e., dispositionally prone to expect to be the target of prejudice) work with males on a task in which they decided on a winner for a journalism prize. The women received information (ostensibly) from their male partner indicating that he was either sexist or nonsexist or received information irrelevant to their partner’s attitudes. Participants made evaluations of the candidates for the prize. Next, they had the opportunity to read one another’s evaluations and rate one another. Results revealed that women high in stigma consciousness who thought they were communicating with a sexist male rated the male’s argument more negatively than women in the other conditions. Men who were working with women high in stigma consciousness who thought they were sexist, in turn, rated their partner’s evaluations more negatively. On the surface, these findings suggest that stigmatized individuals’ prejudice expectations yield negative experiences for their partners—men provided the women with more critical ratings probably because they were upset—which contradicts our study as well as others’ (e.g., Hilton & Darley, 1985; Miller & Myers, 1998). On further reflection, however, it appears that facets of Pinel’s paradigm that diVer in important ways from our own are responsible for the seemingly contradictory pattern of results. The women in Pinel’s study never thought they were going to interact with their male partner, and the women in the sexist condition were led to believe that their partner had extremely unfavorable attitudes toward women. Given that women did not actually interact (nor did they anticipate interacting with their partner) and the cues in the situation were quite clear that their partners were sexist, it is not surprising that women high in stigma consciousness behaved in a negative manner toward their partners (as opposed to attempting to compensate in order to facilitate a smooth interaction). When participants expect that prejudice may be a possible, rather than a definite, factor in an interracial interaction, we believe that they are likely to adopt the compensatory approach, resulting in positive outcomes for their interaction partners. 1. Summary By examining interracial interactions from a relational approach, it is clear to see that ethnic minorities’ concerns with prejudice have implications for the self and partner during interracial interactions. The majority of research illustrates that prejudice expectancies result in negative aVective experiences and perceptions for ethnic minorities’ during interracial interaction. However, perhaps as a means of preventing the negative consequences associated with actually being a victim of prejudice, ethnic minorities who

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

147

expect to be the target of prejudice engage in intimacy building behaviors. Additional research suggests that Whites are sensitive to these behaviors and, in turn, have more positive aVective experiences when they interact with ethnic minorities who expect to be the target of prejudice. Collectively, these findings show that the positivity of Whites’ experiences during interracial interactions may, at times, rest on the shoulders of their ethnic minority interaction partners.

VI. Racial Attitudes and Interracial Contact Prejudice reduction has been the primary goal of the majority of research on racial attitudes and intergroup contact. A review of that literature is beyond the scope of this chapter; however, we briefly highlight recent research exploring the link between interracial contact and prejudice reduction later (Pettigrew & Tropp, 2000). The research we selected is that in which researchers have focused on the perspectives of both Whites and ethnic minorities. Then we turn to what has been at the center of our research—the implications of racial attitudes for immediate, often subtle, experiences during interracial interactions. In a recent longitudinal study of college students, van Laar, Levin, Sinclair, and Sidanius (2005) examined whether or not exposure to roommates of diVerent ethnic outgroups improves Whites’ and ethnic minorities’ racial attitudes about and behaviors towards outgroups. Consistent with contact theory, van Laar et al. (2005) found that ethnic heterogeneity of roommate contact—living with individuals from a range of diVerent ethnic groups—was positively related to intergroup attitudes and behaviors during the first year of college when students are randomly assigned to roommates. Specifically, for Whites, Asian Americans, Blacks, and Latinos, roommate ethnic heterogeneity increased positive aVect toward all four ethnic groups, increased perceived competence in interethnic interactions, decreased antimiscegenation attitudes, and tended to decrease symbolic racism. However, results revealed that for Whites and Blacks, exposure to Asian‐American roommates resulted in negative intergroup attitudes and behaviors. Thus, in general, these data suggest that intimate interethnic contact that occurs in living arrangements can improve intergroup relations. The ethnicity of the individuals involved, however, may have unique eVects that are important to explore. In addition, accumulating evidence suggests that Whites’ and ethnic minorities’ racial attitudes negatively predict how frequently individuals interact and develop friendships with outgroup members (Ellison & Powers,

148

J. NICOLE SHELTON AND JENNIFER A. RICHESON

1994; Levin, van Laar, Sidanius, 2003; Pettigrew, 1997; Powers & Ellison, 1995). Using data from majority group members in four Western European countries, Pettigrew (1997) found that the Europeans who had more friends of another nationality, race, culture, religion, and social class were less prejudiced toward the minority members in their country. Similarly, Powers and Ellison (1995) found that close interracial friendships led to more positive racial attitudes among African‐Americans in the United States. Both Pettigrew (1997) and Powers and Ellison (1995) used statistical analyses that demonstrated that the causal path from more interracial friendship to lower prejudice was stronger than the path from lower prejudice to more friendship. Building on this study, Levin et al. (2003) conducted longitudinal research with Whites and ethnic minorities to examine the friendship–prejudice link. They found that individuals who exhibited more negative racial attitudes at the end of their first year of college had fewer cross‐racial friends during their second and third years of college, even after controlling for precollege friendships and other background variables. Additionally, Levin et al. (2003) found that individuals with more cross‐racial friends during their second and third years of college exhibited more tolerant racial attitudes at the end of their fourth year in college, even after controlling for prior attitudes, precollege friendships, and background variables. Collectively, these studies illustrate the implications of interethnic interactions on racial attitudes from the perspective of both Whites and ethnic minorities.

A. WHITES’ RACIAL ATTITUDES: IMPLICATIONS FOR THE SELF A large proportion of research on interracial relations has focused on Whites’ racial attitudes (Fiske, 1998). Most relevant to our study on interracial interactions, scholars have focused on how Whites’ racial attitudes shape the everyday experiences of Whites’ aVective, cognitive, and behavioral outcomes. Outgroup members, compared to ingroup members, often evoke a variety of negative emotions (Mackie, Devos, & Smith, 2000), especially for individuals who hold negative racial attitudes. Consistent with this research, Vorauer and Kumhyr (2001) demonstrated the aVective implication of Whites’ racial attitudes for the self during interracial interactions. In this study, White Canadian participants, who varied in their racial attitudes, had a 15‐minute ‘‘get‐to‐know‐you’’ interaction with either another White Canadian or with a First Nations participant. Results showed that both lower‐ and higher‐prejudiced Whites experienced more negative feelings toward the self (e.g., self‐critical, angry at myself) and toward others

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

149

(e.g., angry at others, irritated with others) after interacting with a First Nations compared to a White partner. In addition, the researchers found that lower‐prejudiced Whites experienced higher levels of positive aVect (e.g., friendly, happy, satisfied) after interacting with the First Nations as compared to the White partner. Higher‐prejudiced Whites’ level of positive aVect, however, was not influenced by their partner’s ethnicity. In our research, we explored how Whites’ racial attitudes shape their cognitive performance after interracial interactions (Richeson & Shelton, 2003). Previous research has found that humans and animals suVer from cognitive impairments after being exposed to acute stressors (Cohen, 1980). Interacting with a Black person is considered to be a stressor for some Whites (Blascovich, Mendes, Hunter, Lickel, & Kowai‐Bell, 2001), especially those who have little experience with racial minorities or more negative racial attitudes. We predicted that, similar to other stressors, interracial interactions should impair Whites’ cognitive functioning. To examine this prediction, we had Whites with lower and higher levels of both explicit and automatic racial bias participate in either an interracial or intraracial interaction. Specifically, we told participants that the study examined ‘‘serial cognition,’’ which focused on the eVects of completing two cognitive tasks when there is a delay between them. First, participants completed the IAT as a measure of automatic racial bias. During the ‘‘delay’’ portion of the study, a White or Black experimenter videotaped the participants (approximately 5 minutes) providing their opinions on one race‐sensitive and one neutral topic. Then participants completed the Stroop color‐naming task, our measure of cognitive functioning. Consistent with predictions, we found that Whites performed worse on the Stroop task after an interracial interaction compared to after a same‐race interaction. Furthermore, Whites’ racial attitudes (both explicit and automatic) predicted the extent to which they were impaired on the Stroop task after interracial interactions. That is, the more negative their attitudes toward Blacks, relative to Whites, the worse they performed on the cognitive task (Fig. 2). In addition to influencing aVective and cognitive outcomes, Whites’ racial attitudes influence their spontaneous, nonverbal behaviors toward ethnic minorities during interracial interactions. Wilson, Damiani, and Shelton (1998), for example, demonstrated that Whites’ implicit racial bias toward Blacks was correlated with their nonverbal behavior during a game equivalent to tic‐tac‐toe. Wilson et al. (1998) found that when playing a game of tic‐tac‐toe with a Black confederate, higher‐prejudiced Whites were more likely to place the pencil on the table compared to handing it to the Black confederate. Although subtle, by placing the pencil on the table compared to handing it to their partner, Whites had less physical contact with their partner.

150

J. NICOLE SHELTON AND JENNIFER A. RICHESON

Fig. 2. Predicted Stroop interference as a function of race of interaction partner and implicit racial bias. Figure taken from: Richeson, J. A., & Shelton, J. N. (2003), when prejudice does not pay: EVects of interracial contact on executive function, Psychological Science, Blackwell Publishers.

Additional research on Whites’ racial attitudes and behavioral implications for interracial interactions suggests that the distinction between implicit and explicit racial attitudes is important to consider. Implicit racial attitudes have been found to predict behaviors that people are relatively less able to monitor and control, whereas explicit racial attitudes have been found to predict behaviors that people are able monitor and control. Consistent with the study of other researchers (Fazio, Jackson, Dunton, & Williams, 1995; McConnell & Leibold, 2001), Dovidio and colleagues have shown that Whites’ implicit racial attitudes influence their nonverbal behaviors toward Blacks whereas Whites’ explicit racial attitudes are unrelated to these behaviors. For example, Dovidio, Kawakami, Johnson, Johnson, and Howard (1997) had White participants who varied in their racial attitudes interact with a Black and a White interviewer in a sequential manner. Coders rated participants’ amount of eye contact and blinking, both of which are diYcult to monitor and control, from the videotaped interactions. Results revealed that Whites higher in implicit bias toward Blacks, as measured by a response‐latency priming technique, had lower levels of visual contact and higher rates of blinking with the Black compared to the White interviewer. Whites’ explicit racial bias, as measured by McConahay’s (1986) Old‐Fashioned Racism and Modern Racism scales, however, was unrelated to these nonverbal behaviors. Instead, Whites’ explicit level of prejudice was related to negative self‐report judgments about the Black interviewer.

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

151

In similar research (Dovidio, Kawakami, & Gaertner, 2002), White individuals participated in two race‐neutral discussions: one with a same‐sex Black partner and the other with a same‐sex White partner (both confederates). After each interaction, participants and confederates rated one another along traits related to friendliness (i.e., pleasant, cruel, unfriendly, unlikable, and cold). In addition, using these same traits, coders rated the White participants’ nonverbal and verbal behaviors from silent videotape and an audiotape of the interaction, respectively. Consistent with previous research, Whites’ implicit racial attitudes predicted nonverbal friendliness, such that Whites higher in implicit racial bias behaved in a less friendly nonverbal manner toward the Black compared to the White partner. Whites’ implicit racial attitudes, however, were unrelated to their verbal friendliness toward Blacks. In contrast, Whites’ explicit level of prejudice predicted verbal but not nonverbal friendliness. Whites higher in explicit racial bias had less favorable verbal behaviors toward the Black compared to the White partner. The distinction between the consequences of implicit and explicit racial attitudes becomes particularly important when studying interracial interactions from a relational approach. As we will illustrate more fully in a subsequent section of the chapter, the consequences of Whites’ racial attitudes for their ethnic minority partner’s experiences in the interaction may, at times, hinge on the distinction between implicit and explicit attitudes. Although the aforementioned research demonstrates that Whites’ racial attitudes are related to their aVective, cognitive, and behavioral outcomes, recent research suggests that their attitudes may interact with evaluative concerns—general evaluative concerns and specific concerns with appearing prejudiced—to influence Whites’ outcomes during interracial interactions. Towles‐Schwen and Fazio (2003) found that among Whites who were highly concerned about acting prejudiced, those with more negative racial attitudes reported anticipating greater comfort interacting with Blacks than did those with more positive attitudes. As the researchers noted, it is feasible that these findings are the result of social desirability concerns. That is, Whites with relatively negative racial attitudes but who are also concerned with acting prejudiced may attempt to overcorrect for their attitudes by indicating that they would enjoy interacting with Blacks. Vorauer and Turpie’s (2004) research, however, suggests that the reported comfort may be more likely than it appears on the surface. Building on theory and research on choking under pressure, Vorauer and Turpie (2004) examined the implications of Whites’ racial attitudes and evaluative concerns for the subtle behaviors expressed during interracial interactions. Vorauer and Turpie (2004) posited that evaluative concerns may result in a ‘‘choking eVect’’ for lower‐prejudiced Whites but in a

152

J. NICOLE SHELTON AND JENNIFER A. RICHESON

‘‘shining eVect’’ for higher‐prejudiced Whites. Choking under pressure occurs when the presence of an evaluative audience causes individuals to become extremely self‐focused, and this heightened attention to the self disrupts automatic, routinized responses, resulting in poorer performance than under normal circumstances (Baumeister, 1984; Lewis & Linder, 1997). Because of lower‐prejudiced Whites’ egalitarian beliefs, their automatic response is to behave in a positive manner toward ethnic minorities. When evaluative concerns are high, however, these automatic responses are disrupted for lower‐prejudiced Whites and it is harder for them to express positive behaviors toward ethnic minorities. When evaluative concerns are high, lower‐prejudiced Whites are focusing on how they are coming across in the situation (e.g., Did I just say something racist? Did my comment sound sincere? Did I remember to smile when I greeted him?). As Vorauer and Turpie (2004) note, the heightened self‐focus causes lower‐prejudiced Whites to become more cautious and they begin to censor their comments and behaviors, which ultimately interferes with their ability to behave in a pleasant manner during the interracial interaction. Vorauer and Turpie (2004) posit that higher‐prejudiced Whites, by contrast, are less likely to have automatic, routinized positive behaviors toward ethnic minorities. As a result, evaluative concerns will not interfere or disrupt positive behaviors for higher‐prejudiced Whites during interracial interactions. Moreover, because of external norms and pressures to behave in an egalitarian manner, evaluative concerns should encourage higher‐ prejudiced Whites to be extremely concerned with behaving in an unprejudiced manner. Thus, evaluative concerns, especially as they relate to issues of racial prejudice, should result in higher‐prejudiced Whites expressing more positive behaviors, compared to their normal display of negative behaviors. Consistent with predictions, Vorauer and Turpie (2004) found that among lower‐prejudiced Whites, those with lower evaluative concerns displayed a similar number of intimacy‐building behaviors with First Nations and White interaction partners. Lower‐prejudiced participants with higher evaluative concerns, however, displayed fewer intimacy‐building behaviors toward a First Nations, relative to a White, interaction partner. Among the higher‐ prejudiced Whites, those with lower evaluative concerns displayed fewer intimacy‐building behaviors with First Nations, relative to White, interaction partners; whereas higher‐prejudiced Whites with higher evaluative concerns displayed a similar number of intimacy‐building behaviors with First Nations and White interaction partners. Taken together, these findings show that evaluative concerns can disrupt individuals’ automatic behaviors toward outgroup members, such that lower‐prejudiced Whites appear less friendly and higher‐prejudiced Whites appear friendlier than one would predict from their racial attitudes alone.

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

153

B. WHITES’ RACIAL ATTITUDES: IMPLICATIONS FOR PARTNER Lower‐ and higher‐prejudiced Whites do not exist in interracial interactions in isolation. As a result, it is important to examine the impact of Whites’ racial attitudes on the perceptions and experiences of their ethnic minority partner in the interaction. If ethnic minorities are attuned to the behaviors displayed by Whites, then they should have diVerent perceptions of and experiences with lower‐ and higher‐prejudiced Whites during interracial interactions. Research shows that this is indeed the case (Dovidio et al., 2002; Vorauer & Kumhyr, 2001). Vorauer and Kumhyr (2001) examined ethnic minorities’ aVective reactions and impressions of lower‐ and higher‐prejudiced Whites during an interracial interaction. In this study, White Canadians had a 15‐minute ‘‘get‐to‐know‐you’’ interaction with either a White Canadian or a First Nations participant. Results suggested that First Nations partners experienced more discomfort (e.g., tense, frustrated, anxious) after interacting with a higher‐prejudiced White individual, compared to a lower‐prejudiced White individual. White participants did not feel more uncomfortable interacting with low‐ compared to high‐prejudiced White partners. In addition, after interacting with a high‐prejudiced White individual, First Nations partners experienced more negative self‐directed aVect (e.g., self‐critical, annoyed at myself) than White partners did. After interacting with lower‐prejudiced White individuals, by contrast, First Nations partners experienced less negative self‐directed aVect than White partners did. Although the researchers did not examine the actual behaviors of Whites during the interaction, these findings corroborate previous research that lower‐ and higher‐prejudiced Whites behave diVerently during interracial interactions. More relevant to this discussion, the findings reveal that ethnic minorities are somewhat sensitive to these diVerent behaviors and are influenced accordingly. However, being attuned to Whites’ behaviors did not lead ethnic minorities to label higher‐prejudiced Whites as prejudiced or have more negative feelings toward them. In similar research, Sekaquaptewa, Espinoza, Thompson, Vargas, and von Hippel (2003) examined whether Whites’ level of implicit stereotyping was related to their partner’s impression of them during an interracial interaction. In this study, White participants played a game equivalent to tic‐tac‐toe with either a White or a Black confederate. After the interaction, participants completed a measure of implicit stereotyping that assesses the extent to which individuals attribute Black counterstereotypic behavior to external forces (e.g., ‘‘Marcellus got a job at Microsoft because he knew someone there’’) or internal traits or abilities (e.g., ‘‘Marcellus got a job at

154

J. NICOLE SHELTON AND JENNIFER A. RICHESON

Microsoft because he’s good with computers’’). In addition, after the interaction, the confederate rated his impression of the participant (e.g., I liked the participant), as well as the behaviors the participant exhibited (i.e., The participant looked at me in the eye, spoke to me before or during the game, and maintained a closed posture by crossing his or her arms). The ratings were combined to create a social interaction score. Results revealed that Black confederates rated their partner more negatively when Whites provided external attributions for Black stereotype‐inconsistency, whereas they rated their partner more positively when Whites provided internal attributions for Black stereotype‐inconsistency. By contrast, White confederates’ perceptions were unaVected by the attributions that White participants made. Together with Vorauer and Kumhyr’s (2001) study, these findings illustrate that Whites’ racial attitudes and stereotypes do not go unrecognized by ethnic minorities during social interactions. Consistent with this research, Dovidio et al. (2002) also found evidence that ethnic minorities are able to distinguish between low‐ and high‐prejudice White interaction partners by focusing on nonverbal aspects of Whites’ behaviors. Recall that Dovidio et al. (2002) found that Whites’ implicit prejudice levels were related to observer ratings of Whites’ nonverbal friendliness during an interaction. Although the Black confederates had access to both what the White participants said as well as how they behaved, Dovidio et al. (2002) found that Blacks’ perceptions of the White participants’ friendliness were driven by White participants’ nonverbal, but not their verbal, communications. That is, Black confederates’ friendliness judgments were correlated with observer ratings of Whites’ nonverbal friendliness, but not with observer ratings of Whites’ verbal friendliness. This finding suggests that Blacks may perceive Whites’ verbal statements to be insincere, and, therefore, they focus their attention on Whites’ nonverbal behavior. However, Whites’ perceptions of their own friendliness were correlated with observers’ ratings of their verbal, but not their nonverbal, friendliness. Consequently, Whites’ perceptions of their friendliness were unrelated to Black’ perceptions. Dovidio et al.’s (2002) findings underscore the benefits of taking a relational approach to studying interracial interactions. By examining the relationships among Whites’ racial attitudes, their perceptions, and their behavior in tandem with Blacks’ perceptions of lower‐ and higher‐ prejudiced Whites, one can immediately see that Blacks and Whites can easily have completely diVerent experiences during interactions with one another. Given this reality, it becomes clear how interracial interactions can be fraught with communication problems, mistrust, and misperceptions. Thus far, we have noted the implications of Whites’ racial attitudes on ethnic minorities’ feelings and thoughts during interracial interactions. As noted previously, Whites’ racial attitudes can interact with evaluative

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

155

concerns to yield ironic eVects on their behavior during interactions with outgroup members (Vorauer & Turpie, 2004). Recall that when evaluative concerns are high, lower‐prejudiced Whites are likely to exhibit fewer intimacy‐building behaviors as compared to higher‐prejudiced Whites. What are the implications of Whites’ racial attitudes and their evaluative concerns for ethnic minorities’ perceptions during the interaction? If ethnic minorities are sensitive to Whites’ behaviors, then they should have a less favorable impression of lower‐prejudiced compared to higher‐ prejudiced Whites when Whites’ evaluative concerns are high. We set out to test this prediction in our research (Shelton, Richeson, Salvatore, & Trawalter, 2005). In our study, White participants had a 10‐minute conversation with another participant who was either White or Black. After the interaction, participants were taken to separate rooms where they rated how favorably they perceived their partner (e.g., How much do you like your partner?) and how engaged they perceived their partner to be during the interaction (e.g., How involved was your partner during the interaction?). In order to make Whites’ evaluative concerns high during the interaction, we introduced three factors. First, immediately before the interaction, Whites completed the race version of the IAT. Research by Frantz, Cuddy, Burnett, Ray, and Hart (2004) suggests that completing the IAT can cause Whites to feel threatened about appearing racist. Second, at the beginning of the interaction, the experimenter told the participants to select a topic from a basket in order to facilitate the discussion. Unbeknownst to the participants all of the topics were identical: Discuss your opinions about race relations (e.g., Discuss your attitudes about racial profiling. How do you feel about aYrmative action?). We surmised that the race‐sensitive topic would cause Whites to monitor their comments and behaviors because they could appear prejudiced. Finally, we believed the presence of the other person as someone who could provide feedback to the participants during the conversation increased participants’ evaluative concerns. Consistent with predictions, we found that Blacks had a less favorable impression of a White partner with lower levels of automatic racial bias compared to a White partner with higher levels of automatic racial bias during an interracial interaction. By contrast, Whites’ perceptions of their White partner were unrelated to their partner’s automatic racial bias. Furthermore, Black participants perceived White participants with higher automatic racial bias scores more positively than White participants did, but White and Black participants’ perceptions of lower‐bias White interaction partners did not diVer. Based on Vorauer and Turpie’s (2004) findings, we predicted that this ironic eVect of Blacks having a less favorable impression of lower‐prejudiced Whites than higher‐prejudiced Whites would be driven by lower‐prejudiced Whites’ relative disengagement during the interaction.

156

J. NICOLE SHELTON AND JENNIFER A. RICHESON

Consistent with our prediction, we found that Blacks perceived Whites with higher levels of automatic racial bias as being more engaged during the interaction compared to Whites with lower levels of bias. Moreover, the more Blacks perceived their White partners as being engaged during the interaction, the more positively they evaluated them. Blacks’ perceptions of their White partners’ engagement during the interaction mediated the relationship between Whites’ automatic racial bias and Blacks’ favorability ratings. Hence, when evaluative concerns are high, Whites’ automatic racial attitudes may influence the dynamics of interracial interactions in ironic ways. 1. Summary Research taking an individualistic approach to interracial interaction often focuses on prejudice reduction among Whites as the golden outcome of interactions. In contrast, research adopting a relational approach to interracial interaction focuses on how Whites’ racial attitudes influence the immediate experiences of both Whites and ethnic minorities during interracial interactions. In general, this research shows that Whites’ negative racial attitudes result in negative aVective, cognitive, and behavioral outcomes for Whites during interracial interactions. Similarly, Whites’ negative racial attitudes result in negative aVective experiences and perceptions for ethnic minorities. Whites’ evaluative concerns, however, may interact with their racial attitudes to result in ironic eVects for Whites and ethnic minorities during interracial interactions, especially when evaluative concerns are high. With a relational lens researchers will be in a better position to predict, as well as understand, even these ironic outcomes.

C. ETHNIC MINORITIES’ RACIAL ATTITUDES: IMPLICATIONS FOR SELF Although the majority of research on racial attitudes and interracial contact has focused on Whites’ attitudes, there has been a recent shift to examining ethnic minorities’ racial attitudes (Brigham, 1993; Johnson & Leci, 2003; Monteith & Spicer, 2000; Stephan et al., 2002). In our research, we have been interested in examining the relationship between ethnic minorities’ racial attitudes and their immediate experiences during interethnic interactions. We began by exploring the relationship between ethnic minorities’ racial attitudes and the quality of their experiences with their close friends (Shelton & Richeson, in press, Study 1). We asked a sample of ethnic minority participants to answer questions about the quantity and quality of their interactions with either a close, same‐sex White friend or a close, same‐sex Black

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

157

friend. First, participants indicated how much contact they had with their friend (e.g., How often do you spend leisure time with this friend?). Next, they answered questions that tapped into the general quality of the interactions with the friend (e.g., Relative to all of your friendships, how stressful is this friendship?). Finally, participants indicated how comfortable they were discussing personal information (e.g., I feel comfortable talking about my worst fears) as well as race‐related information (e.g., I feel comfortable discussing what it means to be an ethnic/racial minority). We assessed participants’ racial attitudes by combining their responses from the General Evaluation scale and the AVective Prejudice scale as designed by Pettigrew and Meertens (1995). Results revealed that the more negative ethnic minorities’ racial attitudes, the less contact they had with their White friend compared to another ethnic minority friend. In addition, the more negative their attitudes, the less positive they described their contact with their White friend compared to with their ethnic minority friend. Moreover, the more negative ethnic minorities’ attitudes toward Whites, the less comfortable they felt discussing both personal and ethnic‐minority focused information with White friends. In addition to examining the relationship between ethnic minorities’ racial attitudes and voluntary contact experiences (i.e., friendships), we examined this relationship between attitudes and involuntary contact experiences. We were able to examine involuntary contact situations by exploring the dynamics of freshmen roommate interactions at the beginning of the academic year (Shelton & Richeson, in press, Study 2). To investigate this issue, we invited students to participate in a study examining freshmen roommates. Ethnic minority participants completed a racial attitude measure at the beginning of the semester; then, they kept a daily diary of the quantity and quality of contact experiences with their roommate during the first 3 weeks of school. The roommate was either White or ethnic minority and had been randomly assigned by the university. We found that among ethnic minorities who had a White roommate, those who had negative attitudes toward Whites avoided their roommate more often, felt less close to their roommate, and had less positive aVect but more negative aVect during interactions with their roommate. Moreover, there was a significant increase in how often ethnic minority students with negative racial attitudes avoided their White roommate, and a significant decrease in how close they felt to their roommate over the 3‐week study period. As time went on, these students wanted less contact and felt less close to their roommate. None of these findings were reliable with ethnic minorities who had an ethnic minority roommate. Taken together, these findings suggest that not only was there a negative relationship between ethnic minorities’ racial attitudes and the quality of their roommate interactions, but some aspects of the relationship worsened

158

J. NICOLE SHELTON AND JENNIFER A. RICHESON

over time. In addition to these field studies, we have also examined the relationship between Blacks’ racial attitudes and their interracial interaction experiences in the laboratory. In one study, Blacks engaged in a ‘‘get to know you’’ interaction with a same‐sex White stranger (Shelton & Richeson, in press, Study 3). Consistent with our previous findings, we found that the more negative Blacks’ racial attitudes, the less they liked their White partner and the less they enjoyed interacting with him or her. Drawing on this result, and previous research examining cognitive outcomes for Whites after interracial contact, we explored whether Blacks’ racial attitudes may predict negative cognitive outcomes for the self (Richeson, Trawalter, & Shelton, 2005). If Blacks with negative attitudes toward Whites enjoy interracial interactions less than Blacks with more positive attitudes, then they might be susceptible to the negative cognitive performance eVect associated with exposure to acute stressors. In order to test this prediction, we randomly assigned Black participants to engage in a 6–8‐minute interaction with either a Black or a White confederate. Participants completed the IAT as a measure of automatic racial bias before the interaction, and they completed the Stroop color‐naming task to measure inhibitory control after the interaction. Consistent with predictions, the more negatively Blacks’ racial bias toward Whites, the worse they performed on the Stroop task after the interaction with the White confederate. Racial bias was not predictive of Stroop performance after interacting with the Black confederate, however. Collectively, converging evidence demonstrates that regardless of whether interethnic contact is voluntary contact with close friends, involuntary contact with first‐year roommates, or involuntary contact with relative strangers, the more negative ethnic minorities’ attitudes toward Whites, the less positive and the more cognitively depleting their interracial contact experiences.

D. ETHNIC MINORITIES’ RACIAL ATTITUDES: IMPLICATIONS FOR PARTNER Do ethnic minorities’ racial attitudes negatively influence Whites’ aVective experiences and perceptions during interracial interactions? Unfortunately, our review of the literature uncovered only one study that has addressed this question. Specifically, as a first step to address this question, Shelton and Richeson (in press, Study 3) had White participants interact with a Black participant who varied in his or her level of negative attitudes toward Whites. Results revealed that Blacks’ racial attitudes were unrelated to the

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

159

amount of negative aVect Whites experienced, how much Whites liked their Black partner, and how much Whites enjoyed the interaction. At first blush, we were surprised by these null findings. We assumed that, similar to the relationship between Whites’ racial attitudes and Blacks’ experiences, Blacks’ racial attitudes would negatively influence Whites’ experiences during the interaction. On further reflection, however, it is possible that because ethnic minorities encounter Whites on a regular basis that they would have discovered ways not to let their racial attitudes ‘‘leak’’ during their interactions in such a way that impacts their partner’s experience. That is, ethnic minorities, even those with negative racial attitudes, may believe that in order to have smooth daily interpersonal interactions they must learn not to let Whites know how they feel. If this is the case, then this may explain why Whites were not influenced by ethnic minorities’ racial attitudes. Perhaps ethnic minorities’ negative attitudes will ‘‘leak out’’ when ethnic minorities are stressed or otherwise under cognitive load. Nevertheless, future research is needed to explore the stability of our null findings. It is possible that ethnic minorities’ racial attitudes do impact Whites’ aVective and cognitive experiences during interracial interactions, but we were unable to detect the relationship in our study.

E. PARADOX OF DIVERGENT EXPERIENCES When considering the impact of individuals’ concerns with prejudice for the self and one’s partner, the research suggests that these concerns can sometimes lead to divergent experiences during interracial interactions. Accumulating evidence suggests that both concerns about prejudice and racial attitudes can often lead to negative consequences for the self. However, these same concerns with prejudice do not necessarily yield negative consequences for one’s partner. In fact, depending on the study, these concerns yield positive or negative consequences for, or have no eVect on, one’s partner. Thus, the evidence is consistent that interpersonal concerns with prejudice yield negative consequences for the self, but it is somewhat inconsistent regarding the impact of interpersonal concerns with prejudice for one’s partner. In our research, we have been particularly interested in the situations in which interpersonal concerns with prejudice result in Whites and ethnic minorities having divergent experiences in interracial interactions. What might account for these divergent experiences? As we have alluded to previously, self‐regulatory processes are likely to account for at least part of the diVerence. We noted previously, for example, that ethnic minorities who expect others to harbor prejudiced beliefs about their group often engage

160

J. NICOLE SHELTON AND JENNIFER A. RICHESON

in self‐regulatory processes to foster smooth interactions with Whites. Next, we provide concrete, experimental evidence to support the claim that individuals’ concerns with prejudice are associated with self‐regulatory processes, which may account for the divergent experiences Whites and ethnic minorities sometimes have during interracial interactions. There is ample research to suggest that White individuals often rely on self‐regulatory processes in order to avoid behaving in prejudiced ways (e.g., von Hippel, Silver, & Lynch, 2000). Monteith, Devine, and their colleagues (Devine, Plant, Amodio, Harmon‐Jones, & Vance, 2002; Monteith, 1993; Monteith, Sherman, & Devine, 1998; Monteith et al., 2002) have been at the forefront of suggesting that these self‐regulatory processes are integral to the prejudice reduction process. Within the context of an interracial interaction, we have been considering the role of self‐regulatory eVort in the manifestation of many of the ‘‘eVects for self’’ identified in this chapter. For instance, our data suggest that Whites with relatively high racial bias scores and/or Whites who harbor prejudice concerns engage in eVortful self‐regulation during interracial interaction, which is the actual culprit for the cognitive disruption eVect. In one demonstration of this link, we employed functional magnetic resonance imaging technology (fMRI) to measure the neural activity of a sample White students while they were presented with a series of photographs of Blacks and Whites (Richeson et al., 2003). On average, participants revealed heightened activity in areas of the brain that have been implicated in the exercise of control over automatic or otherwise dominant responses. In other words, the simple presentation of Black, but not White, faces was suYcient to trigger cognitive control—perhaps, as a way to control automatic emotional reactions to these targets. Furthermore, we had tested these same students on the IAT 2 weeks before the neural imaging session and were able to examine the extent to which this neural activity was modulated by racial bias scores. Indeed, racial bias scores were correlated with neural activity to Black, but not to White, faces. Specifically, the greater individuals’ level of automatic pro‐White racial bias, the greater the neural activity they revealed in these cognitive control regions. In other words, the brains of individuals with greater levels of racial bias seemed to be working harder to control their responses to the Black faces, compared to the brains of Whites with lower levels of automatic racial bias. The imaging results are interesting because they reveal the eVort that some Whites engage in when evaluating, perceiving, and, perhaps, interacting with Black individuals (see also Eberhardt, 2005 for a review of research on race and neural imaging). These neural imaging results also shed light on several of the behavioral studies that we presented previously. If higher‐biased Whites are working particularly hard to negotiate interracial interactions, then they should be

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

161

relatively more cognitively depleted after the interactions. Indeed, this is exactly what we found (Richeson & Shelton, 2003). Furthermore, in our imaging study, we found that the individuals who were the most disrupted on a cognitive task after an interracial interaction, were the same individuals who revealed the most neural activity in cognitive control brain regions (Richeson et al., 2003). In other words, the same individuals who were ‘‘working harder’’ in the MRI in response to Black faces were also the most cognitively disrupted after interacting with a Black individual. Similarly, our finding that higher‐biased White individuals seemed to be engaging in greater cognitive eVort in the imaging study is consistent with the results of Shelton et al. (2005), wherein high‐biased Whites were more engaged in the interaction compared to lower‐biased Whites. And, furthermore, the eVorts of these high‐biased Whites were not in vain; their Black interaction partners preferred them to low‐biased Whites. Taken together, this research suggests that self‐regulatory processes may very well be linked to negative outcomes for the self, but those outcomes may come in the service of fostering a positive interaction for one’s partner.

F. CREATING COMMON GROUND When studying interracial interactions from a relational approach, it becomes clear that Whites’ interpersonal concerns combined with ethnic minorities’ interpersonal concerns produce a unique set of aVective and behavioral results. It appears that ethnic minorities will have the most positive experience during interracial interactions when they are not concerned about being the target of prejudice, but their White partner is concerned about appearing prejudiced. The opposite is true for Whites; that is, Whites will have the most positive experience during interracial interactions when they are not concerned about appearing prejudiced but their ethnic minority partner is concerned about being the target of prejudice. When neither individual is concerned about issues of prejudice, the potential for each to have a positive experience may be undermined by their partner’s relatively lukewarm (i.e., noncompensatory) behavior. Thus, interpersonal concerns with prejudice may cause interracial interactions to end up being a akin to the tragedy of the commons; individuals are better oV if their partners are worried about how they are being perceived but if they, themselves, are not concerned with such evaluation. This tragedy of the commons—what is good for one individual is not good for the other or the larger collective, in this case, the interaction—is disheartening. Is it possible for both Whites and ethnic minorities to simultaneously have pleasant experiences during interactions when, for example, Whites are

162

J. NICOLE SHELTON AND JENNIFER A. RICHESON

concerned about appearing prejudiced? Some recent research from our lab suggests that modifying individuals’ goal pursuit strategies may oVer a resolution to this disheartening paradox. Specifically, the regulatory focus that individuals adopt in pursuit of maintaining a nonprejudiced image may influence these divergent outcomes. Similar to the pursuit of any goal, Whites who are concerned with appearing prejudiced may enter interracial interactions with a promotion focus or a prevention focus. Individuals with a promotion focus approach desired end‐states by engaging in goal‐directed thoughts and behaviors, whereas individuals with a prevention focus avoid undesired end‐states by avoiding, withdrawing from, and becoming vigilant to end‐state relevant behaviors (Higgins, 1997; Higgins, Roney, Crowe, & Hymes, 1994). Pertinent to interracial interactions, research shows that the strength of individuals’ chronic prevention focus predicts their tendency to avoid outgroup members (Shah, Brazy, & Higgins, 2004). In the majority of studies dealing with Whites’ concerns with prejudice, the experimental manipulations are likely to put participants in a prevention focus mindset (e.g., Whites were given explicit instructions to try not to be prejudiced; Whites were given feedback that they are prejudiced). These manipulations require participants to adopt a protective self‐presentational style by placing emphasis on preventing the expression of prejudice, which, as we have shown, is related to negative cognitive and aVective consequences for the self. But what would happen if Whites entered the interaction with a promotion focus? Would entering an interracial interaction with the goal of simply having a positive interracial exchange (as opposed to trying not to express prejudice) yield more positive cognitive and aVective outcomes for the self? In addition, would Black partners continue to have positive outcomes? We began to address these questions by examining how a prevention, compared to a promotion, focus influences the cognitive consequences of interracial interactions (Trawalter & Richeson, in press). In this study, a White experimenter asked participants to help create stimulus materials for a study examining race relations. The experimenter informed participants that they would be videotaped while discussing their opinions on several topics. In addition, all participants were informed that the researchers had noticed that interracial interactions seem relatively unfamiliar for students at the college and students therefore find it diYcult to engage in them. Moreover, participants in the ‘‘prevention focused’’ condition were told, ‘‘It is important to the study that you avoid appearing prejudiced in any way during the interaction.’’ By contrast, participants in the ‘‘promotion focused’’ condition were told, ‘‘It is important to the study that you approach the interaction as an opportunity to have an enjoyable intercultural dialogue.’’ A control group was not given any instructions regarding how to behave during the interaction. Next, the experimenter led

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

163

the participants to a diVerent room where they were asked to provide their opinions on several topics, including one race‐sensitive topic (e.g., campus diversity), while being videotaped by a Black experimenter. The videotaping session lasted approximately 7 minutes. After this session, participants completed the Stroop color‐naming task. We found that Whites who were given the promotion focus instructions for the interaction performed better on the cognitive task than participants who were given the prevention focus instructions or no instructions. Moreover, Whites who were given no instruction performed just as poorly on the Stroop task as participants in the prevention condition, suggesting that a prevention frame is the default strategy with which Whites typically enter interracial interactions. These data suggest, furthermore, that if Whites are concerned about appearing prejudiced, it may be in their best interest (at least cognitively) to translate their concern into a goal that emphasizes approaching positive outcomes associated with interracial contact, rather than avoiding negative outcomes such as the expression of prejudice. Additional findings provide some support that a promotion focus also enables Whites’ interaction partners to have a pleasant experience during the interaction. Trained coders rated how uncomfortable participants appeared in the videotapes of the interactions with the Black confederate. Specifically, two White coders who were blind to participants’ experimental condition rated how tense, uncomfortable, and relaxed (reversed) each participant was during the mid‐most 20 seconds of the interaction. Consistent with the cognitive disruption results, participants in the promotion focus condition seemed relatively more comfortable during the interaction, compared to participants in either the control condition or the prevention focus condition. To the extent that participants’ own comfort will be communicated successfully to their Black interaction partners, it is possible that shifting individuals’ goal pursuit strategy may result in positive outcomes for both the self and the partner.2 Consequently, these results are promising as 2 Additional research, however, is needed to assess the extent to which Whites’ comfort level will be communicated successfully to their Black interaction partner. Recall that Shelton (2003) found that although Whites who were instructed to try not to be prejudiced reported experiencing anxiety, their Black partner liked them, suggesting that Whites’ anxiety was either not communicated during the interaction, not detected by their Black partners, or, perhaps, was interpreted as evidence of eVort, interest, and/or concern rather than as reflecting negativity. In addition, it is important to keep in mind that naı¨ve coders judged the comfort level of White participants in the study by Trawalter and Richeson (in press), and that these participants engaged in an interaction with an experimenter, whereas naı¨ve Black participants gave their impression of their White participant interaction partners in the study by Shelton (2003). As noted in the section on Methodological Considerations, these diVerences could have significant influences on the results obtained.

164

J. NICOLE SHELTON AND JENNIFER A. RICHESON

they suggest that divergent experiences may not be the only option individuals face as they contemplate how to successfully negotiate interracial interactions. 1. Summary The empirical research revealing these divergent experiences highlights the importance of studying interracial interactions from a relational approach. These divergent patterns would not have been as evident from research taking an individualistic approach. If only one side of the interaction is examined (e.g., how Whites’ concerns with prejudice impact their experiences), then researchers may (erroneously) conclude that certain predictor variables (e.g., prejudice concerns) impact interracial interactions in a certain way (negatively). For instance, because Whites’ concerns with prejudice are often associated with negative cognitive outcomes for the self, it is tempting to conclude that such concerns are necessarily damaging to interracial contact. Such a conclusion would be premature, and misleading in some cases, given what we now know about the positive outcomes ethnic minorities often experience during interactions with Whites who are concerned about appearing prejudiced. Furthermore, subsequent research designed to understand the conditions under which these divergent experiences are likely to arise will also require researchers to adopt a relational perspective on interracial interactions.

VII. Methodological Considerations Conducting studies involving interracial interactions is diYcult to do. In addition, as with most research topics, researchers use diVerent paradigms to address their questions. In our review of the literature we have discovered three methodological factors that often vary across studies of interracial interactions that take a relational approach. These methodological diVerences are likely to play a role in researchers’ empirical findings. As a result, we encourage researchers to consider these issues as they design their studies.

A. WHO IS INVOLVED? It is important for researchers to pay considerable attention to who is involved in the interaction and who is coding the behaviors of the individuals participating in the interaction. When considering who is involved in the

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

165

interaction it is important to think about the ethnicity of the individual, or the person’s stigmatized status more generally. When considering who is coding the behaviors of the individuals in the interaction it is important to think about whether the coders are naı¨ve participants, naı¨ve observers, or trained observers. In the classic research on stigma and interactions, researchers often did not include stigmatized individuals in the studies. In some cases this occurred even when the focus was on how stigmatized individuals react to how others treat them. Instead, researchers created situations in the laboratory in which nonstigmatized individuals played the role of a stigmatized individual. For example, nonstigmatized individuals played the role of someone who had a facial scar (Kleck & Strenta, 1980), was physically disabled (Kleck, Ono, & Hastorf, 1966), homosexual (Farina, Allen, & Saul, 1968), or mentally disabled (Farina & Ring, 1965). A similar methodology has been used in studies on interracial interactions. Instead of including ethnic minorities as participants or confederates, the researchers used Whites for what was considered an interracial interaction. For example, in a study on expectancies and interracial interactions, Chidester (1986) led White participants to believe they were having a phone conversation with either a White or a Black target. The target, however, was always White and unaware that the participant did not know his or her actual race. In addition, the classic study by Word et al. (1974), which we described previously, relied on a similar methodology. In Study 1, Whites interviewed Black and White confederates who were trained to be interviewees. Word et al. (1974) found that the White interviewers displayed less friendly nonverbal behaviors toward Black interviewees than toward White interviewees. Next, Word et al. (1974) trained White confederates to display either friendly or unfriendly nonverbal behaviors toward naı¨ve interviewees in an attempt to understand how the nonverbal behaviors would influence the Black interviewees’ performance in Study 1. Because of a priority on experimental control, the naı¨ve interviewees in Study 2 were White. The results would have been even richer had naı¨ve, Black interviewees been involved.3 The problem with not using ethnic minorities in interethnic interaction research is that Whites have little experience negotiating interactions as an ethnic minority. As Miller and Myers (1998, p. 198) noted, ‘‘It is perhaps not surprising that the nonstigmatized people who were the recipients of 3 Word, Zanna, and Cooper originally planned to include naı¨ve, Black participants in Study 2 (Zanna, personal communications). But, given the fact that there were so few Blacks at Princeton in 1974, it was virtually impossible to do so. The fact that Zanna and colleagues obtained their results with White participants is quite impressive.

166

J. NICOLE SHELTON AND JENNIFER A. RICHESON

prejudice‐inspired behavior were at a loss to know what to do. They were being confronted, perhaps for the one of the few times in their lives, with what it is like to be stigmatized.’’ Of course, it is possible that using stigmatized individuals as the actual targets in these studies would have produced similar results. Our research, however, suggests that this may not always be the case (e.g., Shelton, 2003; Shelton et al., 2005). Ethnic minorities, compared to most Whites, are quite adept at coping with being the target of racial prejudice. As a result, they often bring a host of coping strategies with them to interethnic interactions. We acknowledge the diYculty of recruiting ethnic minorities for psychological research. Nevertheless, we encourage researchers to do so because it is only with their inclusion that we will move closer to an ecologically valid understanding of interethnic interactions. A second yet related issue to who is involved in research as participants is who is involved as observer and interpreter of participants’ behavior during the interaction. There is a tradition in research on interracial interactions to examine how individuals’ attitudes and motivations influence their verbal and nonverbal behaviors (e.g., Crosby, Bromley, & Saxe, 1980; Feldman & Donohoe, 1978; Rollman, 1978; Weitz, 1972; see Devine et al., 1996 for ambiguities associated with using nonverbal behavior as a measure of prejudice). When developing these studies, researchers need to give additional thought to who codes and interprets these verbal and nonverbal behaviors. Specifically, will the coders be the other individual in the interaction, untrained naı¨ve observers, or trained observers? The distinction is important because the individual in the interaction and outside observers often have access to diVerent pieces of information. Participants in the interaction often have access to both verbal and nonverbal behaviors whereas trained and untrained observers may only have access to one channel of the behavioral stream (e.g., audio, silent video, tone of voice). In addition, trained observers are often asked to focus on specific behaviors (e.g., number of eye blinks, eye gaze) performed by the individuals in the interaction, whereas interaction participants and naı¨ve observers rarely focus on single behaviors. Similarly, naı¨ve observers and trained coders are typically asked to assess the nonverbal behaviors of more than one participant, whereas the actual participants of studies typically only interact with one other individual, or sometimes two. Finally, participants in the interaction are (presumably) actively involved in the interaction whereas observers make judgments about the interaction from the perspective of an outsider. There is mixed evidence regarding the extent to which the status of the coders (i.e., participants versus observers) makes a diVerence in how nonverbal behaviors are interpreted. Some evidence suggests that the verbal and nonverbal behavior ratings of participants involved in the interaction, naı¨ve observers, and trained observers are highly correlated and have similar

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

167

predictive validity (e.g., Dovidio et al., 2002; McConnell & Leibold, 2001). For example, McConnell and Leibold (2001) had White participants engage in brief interactions, which were videotaped, with a White and a Black experimenter. Two trained observers coded the White participants’ behavior in terms of molar judgments that assessed interaction quality (e.g., participants’ friendliness) and specific social behaviors (e.g., number of smiles). In addition, the White and Black experimenters rated the White participants along similar behaviors after the interaction. McConnell and Leibold found that the trained observers’ molar judgments were significantly related to the experimenters’ molar ratings. In addition, the trained observers’ and the experimenters’ molar ratings were both predictive of the White participants’ level of implicit racial bias. Thus, these results suggest that naı¨ve participants and trained observers show reliable agreement in perceiving and interpreting Whites’ nonverbal behaviors during social interactions. By contrast, other evidence suggests that the verbal and nonverbal behavioral judgments of participants involved in the interaction, naı¨ve observers, and trained observers are not always similar. Additional evidence by McConnell and Leibold (2001), for example, revealed that the experimenters’ ratings of the White individual were only related to three of the specific social behaviors the trained judges made. The trained judges’ molar ratings, however, were related to four completely diVerent specific social behaviors. Specifically, the experimenters appeared to be sensitive to the extent to which the White participant faced him or her, how far away the participant sat, and how much time the participant talked during the interaction. The judges, on the other hand, appeared to be more sensitive to the openness of the White participant’s arms, amount of eye contact made, amount of laughter at jokes, and extemporaneous social comments. Taken together, these findings suggest that naı¨ve participants and trained observers are often focusing on diVerent aspects of nonverbal behavior during interracial interactions. Additional research is needed to explore the predictive utility of ratings made by diVerent types of observers for the vast array of important outcomes often examined during interracial contact studies. Nevertheless, these findings provide further support for the notion that it is important to think about the type of information one might be likely to glean from the judgments of participants themselves, compared to the perceptions of outside observers. Finally, when researchers are interested in coding the nonverbal behaviors of individuals in interracial interactions, it is worthwhile to consider the race of the coders—be they trained or naı¨ve observers. Recent research in our lab suggests that the race of the coder is an important factor (Richeson & Shelton, 2005; Rollman, 1978). We found that naı¨ve, Black observers were significantly better at predicting White targets’ racial bias than were White observers. Although training White and Black observers may have resulted in similar findings for both groups, this study suggests that, at the very least,

168

J. NICOLE SHELTON AND JENNIFER A. RICHESON

it may be worthwhile to include observers of both races when assessing nonverbal behavior in interracial interactions. The inclusion of Black coders, rather than just White coders, has been instrumental in research on Black mother–adolescent interactions (Campione‐Barr & Smetana, 2004; Gonzales, Cauce, & Mason, 1996). For example, non‐Black coders, compared to Black coders, are more likely to rate Black adolescent–parent interactions as containing more conflict and Black mothers as controlling (Gonzales et al., 1996). Taken together, these findings suggest that ethnic minority coders may bring an important perspective to the evaluation of both White and other ethnic minority individuals’ behavior during interracial interaction.

B. WHAT ARE THEY DOING? A second methodological issue for researchers to contend with when studying interracial interactions is: What are the individuals doing in the interaction? Why do the individuals think they are together? Will the interaction be unstructured or structured? The answers to these questions may make race salient and thus may play a role in shaping participants’ experiences. In unstructured interracial interactions, participants’ spontaneous behaviors are assessed without experimental manipulations and little interference of task demands. Oftentimes in this paradigm two participants are left together at the beginning of the study because the research assistant is presumably running late or attempting to locate an ostensibly misplaced questionnaire. During this time, the interaction between the two participants is videotaped. Once the research assistant arrives, he or she informs the participants that the study is half over and the second part of the study consists of completing questionnaires. One of the key features of the unstructured dyadic paradigm is that participants do not think the study has started. The unstructured dyadic paradigm has not been used often in research on interracial interaction (see Ickes, 1984 for an exception). Instead, researchers typically rely on more structured interactions in which they give the participants something to do or discuss during the interaction. In more structured interracial interactions, contextual factors are apt to have important implications on the quality of the interactions.4 One contextual factor that is particularly important is the likelihood that race will be 4

It is possible, of course, for contextual factors to have important implications for the quality of unstructured, dyad interactions. One can imagine, for example, that an experimenter may leave two participants unattended in a room that is decorated in such a way to make race salient (e.g., ethnic minority art, music, etc.). However, to our knowledge this methodology has not been used in previous studies.

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

169

salient. Our review of the literature revealed that the extent to which race is made salient in interracial interactions varies widely across studies. In addition, the salience of race is operationalized in various ways. For example, race is made salient by informing participants that the researchers are explicitly studying interracial interactions or the participants are going to have an interaction with an outgroup member. Other ways race is made salient is by varying the topic of discussion questions given to participants (e.g., discuss aYrmative action versus discuss your career goals), the comments made by ethnic minority confederates (e.g., confederates mention that he or she has been the target of ethnic discrimination), and the racial composition of the group when researchers move beyond dyadic interactions (i.e., solo status versus nonsolo status). When race is made salient in interracial interactions, individuals no longer feel and behave as individuals; instead, they become representatives of their racial group (Brewer & Miller, 1998). The salience of race in interracial interactions may help explain why studies purported to assess similar ideas have produced diVerent findings. This is evident in four recent studies in which researchers have examined Blacks’ perceptions of Whites during interracial interactions. In these studies, the White individual completes an implicit measure of racial bias immediately before interacting with a Black individual. After the interaction, the Black individual completes a questionnaire that assesses his or her impression of the White person. In three of these four studies (i.e., Dovidio et al., 1997, 2002; McConnell & Leibold, 2002), the topic of discussion during the interaction was race‐neutral (e.g., current dating issues) whereas in the other study (Shelton et al., 2005) the topic was race‐sensitive (i.e., race relations in America). When the discussion topic was race‐neutral, results revealed that the more Whites’ implicit racial scores showed a bias against Blacks, the less favorably Black, compared to White, interaction partners perceived them. Ironically, however, when the discussion topic was race‐sensitive, the results were the opposite—Blacks perceived Whites higher in implicit racial bias more favorably than lower‐bias Whites. As noted previously, perhaps the sensitive nature of a racial topic caused Whites high in implicit racial bias to be more concerned about appearing prejudiced and caused them to engage in strategies to appear more favorable to their partner. Undoubtedly, many other factors contributed to these diVerent findings. Nevertheless, the discussion topic, which caused race to be salient, is one factor that could account for the contrasting findings. As a result, it is important to take into consideration what participants are doing, particularly if it makes race salient, when designing interracial interaction studies. In addition to the issue of whether race is salient during the interaction, researchers should also pay some attention to a few other aspects of the interaction context. For instance, what roles do individuals hold during the

170

J. NICOLE SHELTON AND JENNIFER A. RICHESON

interaction? Is there a power diVerence, and, if so, is it consistent with the typical roles that members of those groups tend to hold in society? For instance, how might the dynamics of the interaction diVer between a White supervisor and her Black employee, compared to the dynamics between a Black supervisor and her White employee? Previous research finds that roles influence both automatic attitude activation (e.g., Lowery, Hardin, & Sinclair, 2001; Richeson & Ambady, 2003) and collective self‐esteem (Richeson & Ambady, 2001) during interracial interactions. Roles and status could also influence interracial interactions through the modulation of individuals’ interpersonal concerns. For instance, Black supervisors may be less concerned about being the target of prejudice from their White employees, compared to their level of concern with fellow White supervisors, or with their own superiors in the organization. Indeed, research suggests that powerful individuals are less concerned about the feelings, experiences, and realities of their subordinates (e.g., Fiske, 1993; Keltner, Gruenfeld & Anderson, 2003). In a similar vein, the scripts that individuals have for the interaction may make them less likely to result in negative self‐outcomes, insofar as social scripts typically allow for behavior to proceed rather eVortlessly (Schank & Abelson, 1977). To the extent that some of the compensatory strategies become automated for individuals, perhaps through practice, or the context itself provides clear guides for appropriate behavior, individuals should find interracial contact less cognitively and emotionally exhausting.

C. HOW LONG IS THE INTERACTION? Finally, researchers should not take lightly the length of the interaction when they design their studies. In some studies, the interactions have been brief (e.g., 3 minutes; Dovidio et al., 2002; McConnell & Leibold, 2001) whereas in other studies they have been considerably longer (e.g., 10–15 minutes; Shelton et al. 2005; Vorauer & Kumhyr, 2001). The length of the interaction is likely to impact individuals’ behaviors. For example, in our research we have found that individuals’ concerns with prejudice cause them to engage in relatively positive, intimacy building behaviors during the interaction. Our interactions, however, tend to be 15 minutes or shorter. The same individuals may become fatigued in longer interactions, which, in turn, could reduce their ability to engage in intimacy building behaviors. In addition, the length of the interaction impacts the amount of information that is available to individuals, which, in turn, can influence individuals’ perceptions of the interaction. Remarkably, Richeson and Shelton (2005) found that naı¨ve Black observers were quite accurate in predicting the explicit racial bias scores of Whites during a brief 20‐second silent interaction in

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

171

which, unbeknownst to the Black observers, the White target was engaged in an interracial interaction. Similarly, Dovidio et al. (1997, 2002) found the number of nervous nonverbal behaviors displayed by Whites during a 3‐minute interracial interaction was related to their level of implicit racial bias (for similar findings see Fazio et al., 1995; McConnell & Leibold, 2001). These findings are quite impressive because they reveal that a lot of information about someone can be leaked and detected in a short time frame (Ambady, Bernieri, & Richeson, 2000). One might assume that a longer interaction might yield even stronger findings. Vorauer and Kumhyr (2001), however, found that ethnic minorities who interacted with a White partner for 15 minutes were unable to detect the explicit prejudice levels of the White individual. It is possible that the more time available provides individuals with information that may counter, or even cloud, their initial impression.5 In relatively long interactions, individuals may stop viewing outgroup members through the lenses of the group stereotypes, and start viewing them more as individuals. Evidence from person perception research suggests that the time course of an encounter with outgroup members can have implications for the extent to which stereotypes are activated and applied to individuals (Kunda, Davies, Adams, & Spencer, 2002). Research finds that exposure (even as minimal as 15 seconds) to outgroup members can automatically activate stereotypes associated with the group (e.g., Bargh, Chen, & Burrows, 1996; Chen & Bargh, 1997; Macrae, Bodenhausen, & Milne, 1995). If applied to the outgroup member, the spontaneous activation of the stereotypes is assumed to have dire consequences for interracial interactions. Kunda et al. (2002) report, however, that with prolonged exposure to an outgroup individual (12 minutes as opposed to 15 seconds) the activated stereotypes dissipate. If the stereotypes have faded away, they are not likely to influence judgments of outgroup members. In other words, it may be inappropriate to extrapolate from brief, initial segments of an interaction to longer slices of the behavioral episode. Undoubtedly, however, as Kunda and her colleagues note, there are apt to be contextual factors that reactivate the stereotypes, which may cause problems during interracial interactions (Kunda & Spencer, 2003). For example, 5 Alternatively, it is possible that the diVerence between the studies could be because Vorauer and Kumhyr focused on participants’ perceptions of an interaction partner, rather than experimenters’ or outside observers’ judgments. Furthermore, unlike the Vorauer and Kumhyr study, most of this other work relied on experimenters’ or observers’ ratings of nonverbal behaviors associated with prejudice rather than simply making prejudice judgments (but see, again, Richeson & Shelton, 2005). This latter issue suggests that another methodological factor to consider is how judgments of prejudice are operationalized and assessed. See Devine et al. (1996) for further discussion on this issue.

172

J. NICOLE SHELTON AND JENNIFER A. RICHESON

a change in conversation from a race‐neutral to race‐sensitive topic could trigger the activation, and perhaps application, of stereotypes during a long interracial interaction. What is of most importance to our consideration of the length of the interaction, however, is that individuals’ perceptions of and behaviors during brief (less than 3 minutes) interracial interactions may diVer from their perceptions of longer (10 minutes or more) interracial interactions. In other words, the amount of time participants spend interacting with someone may be influential to the perceptions and judgments those individuals make about their interaction partners, and, possibly, their feelings about the interactions themselves. Although we have discussed these methodological factors—who is involved, what are they doing, and the length of the interaction—separately, they often covary across studies. This, of course, makes it diYcult to make a clean comparison across studies to determine how each factor may independently influence the results. Thus, we are far from a complete understanding of how these factors influence the subtle dynamics of interracial interactions, and hope that this chapter inspires additional research on the theoretical implications of these methodological considerations for studies of interracial contact.

VIII. Suggestions for Future Directions Social psychologists have only started to scratch the surface in understanding the dynamics of interracial interactions from a relational approach. Although this approach will shed additional insight into the nature of prejudice, considerable work remains ahead of us. In the following sections we present three points to guide future research.

A. BOUNDARY CONDITIONS One important avenue for future research is to identify the boundary conditions of the self‐regulatory eVorts Whites and ethnic minorities employ in order to facilitate pleasant interracial interactions. Two variables that are apt to be important are individuals’ levels of (1) motivation and (2) self‐ eYcacy. Whites and ethnic minorities must be motivated to engage in self‐regulatory behaviors and believe that they have the skills to do so. In contexts or situations in which the stakes are low, participants may not be motivated to engage in self‐regulation. For instance, during an encounter with a retail sales clerk, Black individuals who are concerned about being the

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

173

target of prejudice may not engage in self‐regulatory eVorts with the sales clerk, but, instead, simply exit the situation by leaving the store. The same individuals, however, may engage in compensatory behaviors out of concern about being the target of prejudice during a job interview. Furthermore, even if individuals are motivated to do something to foster a positive interaction, they may not have the skills to successfully engage in compensatory behaviors. For instance, a White individual may want to dispel his Black interaction partner’s concerns that he is prejudiced but may not know how to do so. Under these circumstances, the White individual may chose a diVerent coping mechanism, such as behaving in an avoidant, distant manner, remaining completely inactive, or exiting the interaction all together (Folkman, Lazarus, Gruen, & DeLongis, 1986; Miller & Kaiser, 2001). Unfortunately, these forms of coping are likely to be interpreted as prejudice by the Black partner. Furthermore, in order for individuals to engage in self‐regulatory eVorts during interracial encounters, they must believe that using such behaviors could result in the desired outcome—a harmonious interethnic encounter (see Plant & Butz, 2004 for a similar discussion). If, for instance, the Black individual in an interracial interaction feels that there is nothing she can do to dispel her White interaction partners’ stereotypical perceptions of her, and/or the White partner feels that she will not be able to convince the Black person that she is not prejudiced, then both individuals are unlikely to engage in compensatory, self‐regulatory eVort to foster a smooth interaction. Indeed, it is likely that both individuals would engage in rather negative nonverbal, if not verbal, behavior, yielding a relatively unpleasant encounter for both. Exploring the conditions under which individuals are motivated and feel able to successfully negotiate interracial interactions, coupled with the outcomes of those interactions for both White and ethnic minority individuals, may be a fruitful approach to explaining the divergent outcomes and experiences of interracial contact currently found in the literature.

B. RECIPROCAL DYNAMICS Few studies have addressed how prejudice concerns and beliefs from the perspective of both Whites and ethnic minorities simultaneously influence the dynamics of interracial interactions. Under what conditions might the interpersonal concerns and attitudes of Whites override those of ethnic minorities, and vice versa? Considering both Whites’ and ethnic minorities’ concerns in tandem may clarify the conditions under which interpersonal concerns result in positive or negative aVective, cognitive, and behavioral consequences.

174

J. NICOLE SHELTON AND JENNIFER A. RICHESON

In addition, researchers should examine if the similarities and diVerences regarding Whites’ and ethnic minorities’ interpersonal concerns and racial attitudes yield meaningful diVerences for the dynamics of interracial interactions. For example, at the beginning of the chapter, we suggested that Whites’ interpersonal concerns with prejudice are self‐focused whereas ethnic minorities’ interpersonal concerns with prejudice are both self‐ and other‐focused. Future research might address if this distinction has diVerent implications for the dynamics of interracial interactions. For example, what would be the aVective and behavioral implications of Whites being both self‐ and other‐ focused in interracial interactions? Does this distinction between Whites’ and ethnic minorities’ foci have diVerent aVective, cognitive, and behavioral implications for the self and other during interracial interactions? Exploring the meaningful distinctions in Whites’ and ethnic minorities’ racial concerns and attitudes may be critical to understand why Whites and ethnic minorities sometimes have divergent experiences in interracial interactions.

C. TEMPORAL FACTORS To date, our research, as well as the work of other researchers in this area, has focused on outcomes measured either in anticipation of, during, or after, an interracial interaction. One idea for future direction is for researchers to assess the temporal dynamics of interracial interactions as they occur in real time. Researchers could examine how Person A’s behavior at Time 1 influences Person B’s behavior at Time 2, which, in turn, influences Person A’s behavior at Time 3. More concretely, researchers could examine the degree to which Whites’ concerns with prejudice dissipate during the course of an interaction because of ethnic minorities’ nonverbal behaviors. Perhaps Whites’ concerns with appearing prejudiced at Time 1 lead ethnic minorities to display positive nonverbal behaviors at Time 2, which, in turn, lead Whites to be less concerned about appearing prejudiced at Time 3. By illustrating the ebb and flow of a single interracial encounter, this type of assessment would contribute greatly to the development of a nuanced, relational model of interracial interactions.

D. FINAL THOUGHTS In this chapter, we oVered a conceptualization of interracial interactions that illuminates the value of focusing on participants’ interconnectedness. Looking beyond person perception, contemporary perspectives are beginning to

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

175

consider how meta‐perceptions—which are inherently relational—transform individuals’ judgments about and experiences during interracial interactions. In addition, by exploring identical issues regarding interracial interactions from the perspectives of both Whites and ethnic minorities, one can reveal the true complexity and dynamic nature of interracial interactions. By examining interracial interactions from a relational approach, we do not contend that research that has utilized a more individualistic approach is flawed or unnecessary. Nor do we mean to minimize the significance of research based on an individualistic approach. In fact, we acknowledge that the study of prejudice, stereotyping, and intergroup relations has made great conceptual and empirical progress while relying almost exclusively on an individualistic approach. Instead, we regard the relational approach as complementary to, and building on, research from an individualistic approach. Moreover, we hope that approaching interracial interactions from a relational perspective will spark a number of theoretical and empirical advancements that will be fruitful in providing important insights about intergroup relations.

Acknowledgments The chapter was completed while the authors were Visiting Fellows at the Center for Comparative Studies in Race and Ethnicity at Stanford University. Preparation of this chapter was supported by grants from the Russell Sage Foundation, National Science Foundation, and the National Institute of Mental Health.

References Allport, G. (1954). The nature of prejudice. New York: Doubleday. Ambady, N., Bernieri, F. J., & Richeson, J. A. (2000). Toward a histology of social behavior: Judgmental accuracy from thin slices of the behavioral stream. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 32, pp. 201–271). San Diego CA: Academic Press. Aron, A., Aron, E., & Smollan, D. (1992). Inclusion of other in the self scale and the structure of interpersonal closeness. Journal of Personality and Social Psychology, 63, 596–612. Baldwin, M. W. (1992). Relational schemas and the processing of social information. Psychological Bulletin, 112, 461–484. Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct eVects on trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71, 230–244. Baumeister, R. F. (1984). Choking under pressure: Self‐consciousness and paradoxical eVects of incentives on skillful performance. Journal of Personality and Social Psychology, 46, 610–620.

176

J. NICOLE SHELTON AND JENNIFER A. RICHESON

Blascovich, J., Mendes, W. B., Hunter, S. B., Lickel, B., & Kowai‐Bell, N. (2001). Perceiver threat in social interactions with stigmatized individuals. Journal of Personality and Social Psychology, 80, 253–267. Brewer, M., & Miller, N. (1998). Intergroup relations. In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (Vol. 2, pp. 554–594). New York: McGraw‐Hill. Brigham, J. C. (1993). College students’ racial attitudes. Journal of Applied Social Psychology, 23, 1933–1967. Britt, T. W., Boniecki, K. A., Vescio, T. K., Biernat, M., & Brown, L. M. (1996). Intergroup anxiety: A person X situation approach. Personality and Social Psychology Bulletin, 22, 1177–1188. Campione‐Barr, N., & Smetena, J. G. (2004). In the eye of the beholder: Subjective and observer ratings of middle‐class African American mother‐adolescent interactions. Developmental Psychology, 40, 927–934. Chen, M., & Bargh, J. A. (1997). Nonconscious behavioral confirmation processes: The self‐ fulfilling consequences of automatic stereotype activation. Journal of Experimental Social Psychology, 33, 541–560. Chidester, T. R. (1986). Problems in the study of interracial interaction: Pseudo‐interracial dyad paradigm. Journal of Personality and Social Psychology, 50, 74–79. Cohen, S. (1980). AftereVects of stress on human performance and social behavior: A review of research and theory. Psychological Bulletin, 88, 82–108. Crandall, C. S., & Eshleman, A. (2003). A justification‐suppression model of the expression and experience of prejudice. Psychological Bulletin, 129, 414–446. Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (Vol. 2, pp. 504–553). Boston: McGraw Hill. Crosby, F., Bromley, S., & Saxe, L. (1980). Recent unobtrusive studies of Black and White discrimination and prejudice: A literature review. Psychological Bulletin, 87, 546–563. Devine, P. G., Evett, S. R., & Vasquez‐Suson, K. A. (1996). Exploring the interpersonal dynamics of intergroup context. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook of motivation and cognition: The interpersonal context (Vol. 3, pp. 423–464). New York: Guilford Press. Devine, P. G., Plant, E. A., Amodio, D. M., Harmon‐Jones, E., & Vance, S. L. (2002). The regulation of explicit and implicit race bias: The role of motivations to respond without prejudice. Journal of Personality and Social Psychology, 82, 835–848. Dovidio, J. F., Kawakami, K., Johnson, C., Johnson, B., & Howard, A. (1997). On the nature of prejudice: Automatic and controlled processes. Journal of Experimental Social Psychology, 33, 510–540. Dovidio, J., Kawakami, K., & Gaertner, S. L. (2002). Implicit and explicit prejudice and interracial interaction. Journal of Personality and Social Psychology, 82, 62–68. Dunton, B. C., & Fazio, R. H. (1997). An individual diVerence measure of motivation to control prejudiced reactions. Personality and Social Psychology Bulletin, 23, 316–326. Dutton, D., & Lake, R. (1973). Threat of own prejudice and reverse discrimination in interracial interactions. Journal of Personality and Social Psychology, 29, 65–71. Eberhardt, J. L. (2005). Imaging race. American Psychologist, 60, 181–190. Ellison, C., & Powers, D. (1994). The contact hypothesis and racial attitudes among Black Americans. Social Science Quarterly, 75, 385–400. Engle, R. W., Conway, A. R. A., Tuholski, S. W., & Shisler, R. J. (1995). A resource account of inhibition. Psychological Science, 6, 122–125. Farina, A., Allen, J., & Saul, B. (1968). The role of the stigmatized person in aVecting social relationships. Journal of Personality, 36, 169–182.

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

177

Farina, A., & Ring, K. (1965). The influence of perceived mental illness on interpersonal relations. Journal of Abnormal Psychology, 70, 47–51. Fazio, R. H., Jackson, J. R., Dunton, B. C., & Williams, C. J. (1995). Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline? Journal of Personality and Social Psychology, 69, 1013–1027. Feldman, R. S., & Donohoe, L. F. (1978). Nonverbal communication of aVect in interracial dyads. Journal of Educational Psychology, 70, 979–987. Fiske, S. (1998). Stereotyping, prejudice, and discrimination. In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (Vol. 2, pp. 357–411). New York: McGraw‐Hill. Fiske, S. T. (1993). Controlling other people: The impact of power on stereotyping. American Psychologist, 48, 621–628. Folkman, S., Lazarus, R. S., Gruen, R. J., & DeLongis, A. (1986). Appraisal, coping, health status, and psychological symptoms. Journal of Personality and Social Psychology, 50, 571–579. Frantz, C. M., Cuddy, A. J. C., Burnett, M., Ray, A., & Hart, A. (2004). A threat in the computer: The race implicit association test as a stereotype threat experience. Personality and Social Psychology Bulletin, 30, 1611–1624. GoVman, I. (1963). Stigma. New York: Simon & Shuster, Inc. Gonzales, N. A., Cauce, A. M., & Mason, C. A. (1996). Interobserver agreement in the assessment of parental behavior and parent‐adolescent conflict: African American mothers, daughters, and independent observers. Child Development, 67, 1483–1498. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual diVerences in implicit cognition: The implicit association task. Journal of Personality and Social Psychology, 74, 1464–1480. Hebl, M. R., & Dovidio, J. F. (2005). Promoting the ‘‘social’’ in the examination of social stigmas. Personality and Social Psychology Review, 9, 156–182. Higgins, E. T. (1997). Beyond pleasure and pain. American Psychologist, 52, 1280–1300. Higgins, E. T., Roney, C., Crowe, E., & Hymes, C. (1994). Ideal versus ought predilections for approach and avoidance: Distinct self‐regulatory systems. Journal of Personality and Social Psychology, 66, 276–286. Hilton, J. L., & Darley, J. M. (1985). Constructing other persons: A limit on the eVect. Journal of Experimental Social Psychology, 21, 1–18. Hyers, L., & Swim, J. (1998). A comparison of the experiences of dominant and minority group members during an intergroup encounter. Group Processes and Intergroup Relations, 1, 143–163. Ickes, W. (1984). Compositions in Black and White: Determinants of interaction in interracial dyads. Journal of Personality and Social Psychology, 47, 330–341. Ickes, W., Patterson, M. L., Rajecki, D. W., & Tanford, S. (1982). Behavioral and cognitive consequences of reciprocal versus compensatory responses to preinteraction expectancies. Social Cognition, 1, 160–190. Ickes, W., Snyder, M., & Garcia, S. (1997). Personality influences on the choice of situations. In R. Hogan, J. A. Johnson, & S. Briggs (Eds.), Handbook of personality psychology (pp. 165–195). San Diego, CA: Academic Press. Johnson, J. D., & Leci, L. (2003). Assessing anti‐White attitudes and predicting perceived racism: The Johnson‐Leci scale. Personality and Social Bulletin, 29, 299–312. Jones, E. E., Farina, A., Hastorf, A. H., Markus, H., Miller, D. T., & Scott, R. A. (1984). Social stigma: The psychology of marked relationships. New York: Freeman. Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhibition. Psychological Review, 110, 265–284.

178

J. NICOLE SHELTON AND JENNIFER A. RICHESON

Kenny, D. A., & DePaulo, B. M. (1993). Do people know how others view them? An empirical and theoretical account. Psychological Bulletin, 114, 145–161. Kleck, R. E., & Strenta, A. (1980). Perceptions of the impact of negatively valued physical characteristics on social interaction. Journal of Personality and Social Psychology, 39, 861–873. Kleck, R., Ono, H., & Hastorf, A. H. (1966). The eVects of physical deviance upon face‐to‐face interaction. Human Relations, 19, 425–436. Krueger, J. (1996). Personal beliefs and cultural stereotypes about racial characteristics. Journal of Personality and Social Psychology, 71, 536–548. Kunda, Z., Davies, P. G., Adams, B. D., & Spencer, S. J. (2002). The dynamic time course of stereotype activation: Activation, dissipation, and resurrection. Journal of Personality and Social Psychology, 82, 283–299. Kunda, Z., & Spencer, S. J. (2003). When do stereotypes come to mind and when do they color judgment? A goal‐based theoretical framework for stereotype activation and application. Psychological Bulletin, 129, 522–544. LaPiere, R. T. (1934). Attitudes vs. actions. Social Forces, 13, 230–237. Levin, S., van Laar, C., & Sidanius, J. (2003). The eVects of ingroup and outgroup friendships on ethnic attitudes in college: A longitudinal study. Group Processes and Intergroup Relations, 6, 76–92. Lewis, B. P., & Linder, D. E. (1997). Thinking about choking? Attentional processes and paradoxical performance. Personality and Social Psychology Bulletin, 23, 937–944. Lowery, B. S., Hardin, C. D., & Sinclair, S. (2001). Social influence eVects on automatic racial prejudice. Journal of Personality and Social Psychology, 81, 842–855. Mackie, D. M., Devos, T., & Smith, E. (2000). Intergroup emotions: Explaining oVensive action tendencies in an intergroup contact. Journal of Personality and Social Psychology, 79, 602–616. Macrae, C. N., Bodenhausen, G. V., & Milne, A. B. (1995). The dissection of selection in person perception: Inhibitory processes in social stereotyping. Journal of Personality and Social Psychology, 69, 397–407. Major, B., Quinton, W., & McCoy, S. (2002). Antecedents and consequences of attributions to discrimination: Theoretical and empirical advances. In M. P. Zanna (Ed.), Advances of Experimental Social Psychology (Vol. 34, pp. 251–330). New York: Academic Press. McConnell, A. R., & Leibold, J. M. (2001). Relations between the Implicit Association Test, explicit racial attitudes, and discriminatory behavior. Journal of Experimental Social Psychology, 37, 435–442. McConahay, J. G. (1986). Modern racism, ambivalence, and the modern racism scale. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 91–125). New York: Academic Press. Mendoza‐Denton, R., Downey, G., Purdie, V. J., Davis, A., & Pietrzak, J. (2002). Sensitivity to status‐based rejection: Implications for African‐American students’ college experience. Journal of Personality and Social Psychology, 83, 896–918. Miller, C. T., & Kaiser, C. R. (2001). A theoretical perspective on coping with stigma. Journal of Social Issues, 57, 73–92. Miller, C. T., & Myers, A. (1998). Compensating for prejudice: How heavyweight people (and others) control outcomes despite prejudice. In J. K. Swim & C. Stangor (Eds.), Prejudice: The target’s perspective (pp. 191–218). San Diego, CA: Academic Press. Miller, C., Rothblum, E., Barbour, L., Brand, P., & Felicio, D. (1990). Social interactions of obese and nonobese women. Journal of Personality, 58, 365–380.

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

179

Miller, C., Rothblum, E., Felicio, D., & Brand, P. (1995). Compensating for stigma: Obese and nonobese women’s reactions to being visible. Personality and Social Psychology Bulletin, 21, 1093–1106. Monin, B., & Miller, D. (2001). Moral credentials and the expression of prejudice. Journal of Personality and Social Psychology, 81, 33–43. Monteith, M. J. (1993). Self‐regulation of prejudiced responses: Implications for progress in prejudice‐reduction eVorts. Journal of Personality and Social Psychology, 65, 469–485. Monteith, M. J., Ashburn‐Nardo, L., Voils, C. I., & Czopp, A. M. (2002). Putting the breaks on prejudice: On the development and operation of cues for control. Journal of Personality and Social Psychology, 83, 1029–1050. Monteith, M. J., Sherman, J., & Devine, P. G. (1998). Suppression as a stereotype control strategy. Personality and Social Psychology Review, 2, 63–82. Monteith, M. J., & Spicer, C.,V (2000). Contents and correlates of Whites’ and Blacks’ racial attitudes. Journal of Experimental Social Psychology, 36, 125–154. Muraven, M., & Baumeister, R. F. (2000). Self‐regulation and depletion of limited resources. Does self‐control resemble a muscle? Psychological Bulletin, 126, 247–259. Pettigrew, T. (1997). Generalized intergroup contact eVects on prejudice. Personality and Social Psychology Bulletin, 23, 173–185. Pettigrew, T. F., & Meertens, R. (1995). Subtle and blatant prejudice in western Europe. European Journal of Social Psychology, 25, 57–75. Pettigrew, T. F., & Tropp, L. R. (2000). Does intergroup contact reduce prejudice? Recent meta‐ analytic findings. In S. Oskamp (Ed.), Reducing prejudice and discrimination (pp. 93–114). Mahwah, NJ: Erlbaum. Pinel, E. C. (1999). Stigma consciousness: The psychological legacy of social stereotypes. Journal of Personality and Social Psychology, 76, 114–128. Pinel, E. C. (2002). Stigma consciousness in intergroup contexts: The power of conviction. Journal of Experimental Social Psychology, 38, 178–185. Plant, E. A., & Butz, D. A. (2004). Expectancies and motivations regarding interracial interactions: Implications for intergroup emotions and behavior. Unpublished manuscript. Plant, E. A., & Devine, P. (1998). Internal and external motivation to respond without prejudice. Journal of Personality and Social Psychology, 75, 811–832. Plant, E. A., & Devine, P. G. (2003). The antecedents and implications of interracial anxiety. Personality and Social Psychology Bulletin, 29, 790–801. Powers, D. A., & Ellison, C. G. (1995). Interracial contact and Black racial attitudes: The contact hypothesis and selectivity bias. Social Forces, 74, 205–226. Prentice, D., & Miller, D. (1996). Pluralistic ignorance and the perpetuation of social norms by unwitting actors. In M. P. Zanna (Ed.), Advances in Experimental Social Psychology (Vol. 28, pp. 161–209). San Diego, CA: Academic Press. Pronin, E., Gilovich, T., & Ross, L. (2004). Objectivity in the eye of the beholder: Divergent perceptions of bias in self versus others. Psychological Review, 111, 781–799. Reis, H. T., & Shaver, P. (1988). Intimacy as an interpersonal process. In S. Duck (Ed.), Handbook of personal relationships (pp. 367–389). New York: Wiley. Richeson, J. A., & Ambady, N. (2001). When roles reverse: Stigma, status, & self‐evaluation. Journal of Applied Social Psychology, 31, 1350–1377. Richeson, J. A., & Ambady, N. (2003). EVects of situational power on automatic racial prejudice. Journal of Experimental Social Psychology, 39, 177–183. Richeson, J. A., Baird, A. A., Gordon, H. L., Heatherton, T.,F, Wyland, C. L., Trawalter, S., & Shelton, J. N. (2003). An fMRI investigation of the impact of interracial contact on executive function. Nature Neuroscience, 6, 1323–1328.

180

J. NICOLE SHELTON AND JENNIFER A. RICHESON

Richeson, J. A., & Shelton, J. N. (2003). When prejudice does not pay: EVects of interracial contact on executive function. Psychological Science, 14, 287–290. Richeson, J. A., & Shelton, J. N. (2005). Thin slices of racial bias. Journal of Nonverbal Behavior, 29, 75–86. Richeson, J. A., & Trawalter, S. (2005). Why do interracial interactions impair executive function? A resource depletion account. Journal of Personality and Social Psychology, 88, 934–947. Richeson, J. A., Trawalter, S., & Shelton, J. N. (2005). African American’s implicit racial attitudes and the depletion of executive function after interracial interactions. Social Cognition, 23, 336–352. Rollman, S. A. (1978). The sensitivity of Black and White Americans to nonverbal cues of prejudice. The Journal of Social Psychology, 105, 73–77. Schank, R. C., & Abelson, R. P. (1977). Scripts, plans, goals, and understanding: An inquiry into human knowledge structures. Hillsdale, NJ: Lawrence Erlbaum. Sedikides, C., Olsen, N., & Reis, H. T. (1993). Relationships as natural categories. Journal of Personality and Social Psychology, 64, 71–82. Sekaquaptewa, D., Espinoza, P., Thompson, M., Vargas, P., & von Hippel, W. (2003). Stereotypic explanatory bias: Implicit stereotyping as a predictor of discrimination. Journal of Experimental Social Psychology, 39, 75–82. Shah, J. Y., Brazy, P. C., & Higgins, E. T. (2004). Promoting us or preventing them: Regulatory focus and manifestations of intergroup bias. Personality and Social Psychology Bulletin, 30, 433–446. Shelton, J. N. (2000). A reconceptualization of how we study issues of racial prejudice. Personality and Social Psychology Review, 4, 374–390. Shelton, J. N. (2003). Interpersonal concerns in social encounters between majority and minority group members. Group Processes and Intergroup Relations, 6, 171–185. Shelton, J. N., & Richeson, J. A. (2005). Intergroup contact and pluralistic ignorance. Journal of Personality and Social Psychology, 88, 91–107. Shelton, J. N., & Richeson, J. A. (in press). Ethnic minorities’ racial attitudes and contact experiences with Whites. Cultural Diversity and Ethnic Minority Psychology. Shelton, J. N., Richeson, J. A., & Salvatore, J. (2005). Expecting to be the target of prejudice. Implications for interethnic interactions. Personality and Social Psychology Bulletin, 31(9), 1189–1202. Shelton, J. N., Richeson, J. A., Salvatore, J., & Trawalter, S. (2005). Ironic eVects of racial bias during interracial interactions. Psychological Science, 16, 397–402. Sigelman, L., & Tuch, S. A. (1997). Metastereotypes: Blacks’ perceptions of Whites’ stereotypes of Blacks. Public Opinion Quarterly, 61, 87–101. Smith, E. R., Coats, S., & Walling, D. (1999). Overlapping mental representations of self, in group, and partner: Further response time evidence and a connectionist model. Personality and Social Psychology Bulletin, 25, 873–882. Smith, E. R., & Henry, S. (1996). An ingroup becomes part of the self: Response time evidence. Personality and Social Psychology Bulletin, 22, 635–642. Steele, C. M., Spencer, S. J., & Aronson, J. (2002). Contending with group image: The psychology of stereotype and social identity threat. In M. P. Zanna (Ed.), Advances in Experimental Social Psychology (Vol. 34, pp. 379–440). San Diego, CA: Academic Press. Stephan, W. G., Boniecki, K. A., Ybarra, O., Bettencourt, A., Ervin, K. S., Jackson, L. A., McNatt, P. S., & Renfro, C. L. (2002). The role of threats in the racial attitudes of Blacks and Whites. Personality and Social Psychology Bulletin, 28, 1242–1254. Stephan, W., & Stephan, C. W. (1985). Intergroup anxiety. Journal of Social Issues, 41, 157–175.

INTERRACIAL INTERACTIONS: A RELATIONAL APPROACH

181

Swann, W. (1987). Identity negotiation: Where two roads meet. Journal of Personality and Social Psychology, 53, 1038–1051. Swann, W. B., & Ely, R. J. (1984). A battle of wills: Self‐verification versus behavioral confirmation. Journal of Personality and Social Psychology, 46, 1287–1302. Swann, W. B., & Read, S. J. (1981). Self‐verification processes: How we sustain our self‐ conceptions. Journal of Experimental Social Psychology, 17, 351–372. Towles‐Schwen, T., & Fazio, R. H. (2003). Choosing social situations: The relation between automatically activated racial attitudes and anticipated comfort interacting with African Americans. Personality and Social Psychology Bulletin, 29, 170–182. Trawalter, S., & Richeson, J. R. (in press). Regulatory focus and executive function after interracial interactions. Journal of Experimental Social Psychology. Tropp, L. R. (2003). The psychological impact of prejudice: Implications for intergroup contact. Group Processes and Intergroup Relations, 6, 131–149. Tropp, L. R., & Anderson, R. A. (2003). Valuing diversity and interest in intergroup contact. Paper presented at the annual convention of Society for Personality and Social Psychology. Tropp, L. R., & Wright, S. C. (2001). Ingroup identification as the inclusion of ingroup in the self. Personality and Social Psychology Bulletin, 27, 585–600. van Laar, C., Levin, S., Sinclair, S., & Sidanius, J. (2005). The eVects of university roommate contact on ethnic attitudes and behaviors. Journal of Experimental Social Psychology, 41(4), 309–345. von Hippel, W., Silver, L. A., & Lynch, M. E. (Hippel 2000). Stereotyping against your will: The role of inhibitory ability in stereotyping and prejudice among the elderly. Personality and Social Psychology Bulletin, 26, 523–532. Vorauer, J. D. (2001). The other side of the story: Transparency estimation in social interactions. In G. Moskowitz (Ed.), Cognitive social psychology: The Princeton symposium on the legacy and future of social cognition (pp. 261–276). New Jersey: Lawrence Erlbaum Associates. Vorauer, J., Hunter, A., Main, K., & Roy, S. (2000). Concerns with evaluation and meta‐ stereotype activation. Journal of Personality and Social Psychology, 78, 690–707. Vorauer, J., & Kumhyr, S. (2001). Is this about you or me? Self‐ versus other‐directed judgments and feelings in response to intergroup interaction. Personality and Social Psychology Bulletin, 27, 706–719. Vorauer, J., Main, K., & O’ Connell, G. (1998). How do individuals expect to be viewed by members of lower status groups? Content and implications of meta‐stereotypes. Journal of Personality and Social Psychology, 75, 917–937. Vorauer, J. D., & Turpie, C. (2004). Disruptive eVects of vigilance on dominant group members’ treatment of outgroup members: Choking versus shining under pressure. Journal of Personality and Social Psychology, 27, 706–709. Weitz, S. (1972). Attitude, voice, and behavior: A repressed aVect model of interracial interaction. Journal of Personality and Social Psychology, 24, 14–21. Wetzel, C., Blalock, C., & Bolger, B. (2004). Interracial interactions: Trying to avoid appearing prejudiced can improve interaction quality. Unpublished data. Wilson, T., Damiani, M., & Shelton, J. N. (1998). University of Virginia. Unpublished data. Word, C. O., Zanna, M. P., & Cooper, J. (1974). The nonverbal mediation of self‐fulfilling prophecies in interracial interaction. Journal of Experimental Social Psychology, 10, 109–120. Wout, D., Shih, M., & Jackson, J. (2005). University of Michigan. Unpublished data. Wright, S., Aron, A., McLaughlin‐Volpe, T., & Ropp, S. (1997). The extended contact eVect: Knowledge of cross‐group friendships and prejudice. Journal of Personality and Social Psychology, 73, 73–90.

THE PSYCHOLOGY OF SELF‐DEFENSE: SELF‐AFFIRMATION THEORY

David K. Sherman GeoVrey L. Cohen

I. Introduction In major league baseball, a hitter could have a long and productive career by maintaining a .300 average, that is, by getting a base hit 30% of the time. A great deal of money could be earned and fame accrued. Yet the other 70% of the time, this player would have failed. The vast majority of attempts to hit the ball would result in ‘‘making an out’’ and thus pose a potential threat to the player’s sense of personal worth and social regard. Like major league baseball players, people in contemporary society face innumerable failures and self‐threats. These include substandard performance on the job or in class, frustrated goals or aspirations, information challenging the validity of long‐held beliefs, illness, the defeat of one’s political party in an election or of one’s favorite sports team in a playoV, scientific evidence suggesting that one is engaging in risky health behavior, negative feedback at work or in school, rejection in a romantic relationship, real and perceived social slights, interpersonal and intergroup conflict, the misbehavior of one’s child, the loss of a loved one, and so on. In the course of a given day, the potential number of events that could threaten people’s ‘‘moral and adaptive adequacy’’—their sense of themselves as good, virtuous, successful, and able to control important life outcomes (Steele, 1988)—seems limitless and likely to exceed the small number of events that aYrm it. A major undertaking for most people is to sustain self‐integrity when faced with the inevitable setbacks and disappointments of daily life—the 70% of the time 183 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 38 DOI: 10.1016/S0065-2601(06)38004-5

Copyright 2006, Elsevier Inc. All rights reserved. 0065-2601/06 $35.00

184

DAVID K. SHERMAN AND GEOFFREY L. COHEN

‘‘at bat’’ when they do not get a base hit. How do individuals adapt to such threats and defend self‐integrity? Much research suggests that people have a ‘‘psychological immune system’’ that initiates protective adaptations when an actual or impending threat is perceived (Gilbert, Pinel, Wilson, Blumberg, & Wheatley, 1998). Psychological adaptations to threats include the various cognitive strategies and even distortions whereby people come to construe a situation in a manner that renders it less threatening to personal worth and well‐being. Many of these psychological adaptations can be thought of as defensive in nature, insofar as they alter the meaning of the event in a way that shields people from the conclusion that their beliefs or actions were misguided. Psychologists have documented a wide array of such psychological adaptations that help people to protect their self‐integrity in response to threat. Indeed, defensive adaptations are so stubborn and pervasive that Greenwald (1980) described the ego as ‘‘totalitarian’’ in its ambition to interpret the past and present in a way congenial to its desires and needs. People view themselves as a potent causal agent even over events that they cannot control (Langer, 1975); they view themselves as selectively responsible for producing positive rather than negative outcomes (Greenwald, 1980; Miller & Ross, 1975; Taylor, 1983). They resist change or—if they do change—become more extreme versions of what they were before (Lord, Ross, & Lepper, 1979). People dismiss health information suggesting that they are at risk for disease or should change their risky behavior (Jemmott, Ditto, & Croyle, 1986; Kunda, 1987). Students may disidentify with, or downplay the personal importance of, domains where they fail, thus sustaining self‐worth but precluding the opportunity for improvement (Major, Spencer, Schmader, Wolfe, & Crocker, 1998; Steele, 1997). People are overoptimistic in their predictions of future success and estimations of their current knowledge and competence (Dunning, GriYn, Milojkovic, & Ross, 1990; Kruger & Dunning, 1999). Indeed, these defensive adaptations may even benefit psychological and physical health (Taylor & Brown, 1988). Although we suspect that people can be more realistic and more self‐critical than this research suggests, and that their optimism and positive illusions may be magnified in certain contexts rather than others (see Armor & Taylor, 2002), the idea that people are ego defensive resonates both with psychological research and lay wisdom. An important question, then, concerns the circumstances under which people are less ego defensive and more open‐minded in their relationship with the social world. We see defensive responses as adaptations aimed at ameliorating threats to self‐integrity. The vast research on defensive biases testifies to their robustness and to the frequency with which people use them. Although these defensive responses are adaptive in the sense of protecting or enhancing an individual’s sense of self‐integrity, they can be maladaptive to the extent they forestall learning from important, though threatening, experiences and

SELF‐AFFIRMATION THEORY

185

information. Moreover, peoples’ eVorts to protect self‐integrity may threaten the integrity of their relationships with others (Cohen et al., 2005; Murray, Holmes, MacDonald, & Ellsworth, 1998). Yet, these normal adaptations can be ‘‘turned oV’’ through an altogether diVerent psychological adaptation to threat, an alternative adaptation that does not hinge on distorting the threatening event to render it less significant. One way that these defensive adaptations can be reduced, or even eliminated, is through the process of self‐aYrmation (Aronson, Cohen, & Nail, 1999; Sherman & Cohen, 2002; Steele, 1988). Steele (1988) first proposed the theory of self‐aYrmation. It asserts that the overall goal of the self‐system is to protect an image of its self‐integrity, of its moral and adaptive adequacy. When this image of self‐integrity is threatened, people respond in such a way as to restore self‐worth. As noted previously, one way that this is accomplished is through defensive responses that directly reduce the threat. But another way is through the aYrmation of alternative sources of self‐integrity. Such ‘‘self‐aYrmations,’’ by fulfilling the need to protect self‐integrity in the face of threat, can enable people to deal with threatening events and information without resorting to defensive biases. In this paper, we update the field on research conducted using self‐aYrmation theory as a framework. This research illuminates both the motivational processes underlying self‐integrity maintenance and the implications of such processes for many domains of psychology. We illustrate how self‐ aYrmation aVects not only people’s cognitive responses to threatening information and events, but also their physiological adaptations and actual behavior. The research presented has implications for psychological and physical health, education, social conflict, closemindedness and resistance to change, prejudice and discrimination, and a variety of other important applied areas. We also examine how self‐aYrmations reduce threats to the self at the collective level, such as when people confront threatening information about their groups. We then review factors that qualify or limit the eVectiveness of self‐aYrmations, including situations where aYrmations backfire, and lead to greater defensiveness and discrimination. We discuss the connection of self‐aYrmation theory to other motivational theories of self‐defense and review relevant theoretical and empirical advances. We conclude with a discussion of the implications of self‐aYrmation theory for interpersonal relationships and coping.

A. OVERVIEW OF SELF‐AFFIRMATION THEORY Self‐aYrmation theory (Aronson et al., 1999; Sherman & Cohen, 2002; Steele, 1988) begins with the premise that people are motivated to maintain the integrity of the self. Integrity can be defined as the sense that, on the

186

DAVID K. SHERMAN AND GEOFFREY L. COHEN

whole, one is a good and appropriate person. Cultural anthropologists use the term ‘‘appropriate’’ to refer to behavior that is fitting or suitable given the cultural norms and the salient demands on people within that culture. Thus, the standards for what it means to be a good person vary across cultures, groups, and situations (e.g., Heine, 2005). Such standards of integrity can include the importance of being intelligent, rational, independent, and autonomous, and exerting control over important outcomes. Such standards of integrity can also include the importance of being a good group member and of maintaining close relationships. Threats to self‐integrity may thus take many forms but they will always involve real and perceived failures to meet culturally or socially significant standards (Leary & Baumeister, 2000). Consequently, people are vigilant to events and information that call their self‐integrity into question, both in their own eyes and in the eyes of others. In such situations, people try to restore or reassert the integrity of the self. Thus, the goal of protecting self‐integrity, and the impact of that goal on psychology and behavior, becomes apparent when integrity is threatened. There are three categories of responses that people deploy to cope with such threats. First, they can respond by accommodating to the threat. That is, they can accept the failure or the threatening information and then use it as a basis for attitudinal and behavioral change. However, to the extent that the threatened domain concerns an important part of one’s identity, the need to maintain self‐integrity can make it diYcult to accept the threatening information and to change one’s attitude or behavior accordingly. A second response thus involves ameliorating the threat via direct psychological adaptations. While some direct adaptations preserve the fundamental informational value of the event while also changing one’s construal of that event (e.g., framing a failure as a learning opportunity; Dweck & Leggett, 1988), other direct psychological adaptations are defensive in nature in that they involve dismissing, denying, or avoiding the threat in some way. We refer to these responses as defensive biases (see Sherman & Cohen, 2002). Although a defensive bias can restore self‐integrity, the rejection of the threatening information can lessen the probability that the person will learn from the potentially important information. Self‐aYrmation theory proposes a third alternative, a diVerent kind of psychological adaptation—one that, under many circumstances, enables both the restoration of self‐integrity and adaptive behavior change. People can respond to threats using the indirect psychological adaptation of aYrming alternative self‐resources unrelated to the provoking threat. Such ‘‘self‐aYrmations’’ include reflecting on important aspects of one’s life irrelevant to the threat, or engaging in an activity that makes salient important values unconnected to the threatening event. Whereas defensive psychological adaptations directly address the threatening information, indirect psychological

SELF‐AFFIRMATION THEORY

187

adaptations, such as self‐aYrmation, allow people to focus on domains of self‐integrity unrelated to the threat. When self‐aYrmed in this manner, people realize that their self‐worth does not hinge on the evaluative implications of the immediate situation. As a result, they have less need to distort or reconstrue the provoking threat and can respond to the threatening information in a more open and evenhanded manner.

B. BASIC TENETS OF SELF‐AFFIRMATION THEORY Much research within the self‐aYrmation framework examines whether an aYrmation of self‐integrity, unrelated to a specific provoking threat, can attenuate or eliminate people’s normal response to that threat. If it does, then one can infer that the response was motivated by a desire to protect self‐ integrity. The self‐aYrmation framework encompasses four tenets, which are enumerated below: 1. People are Motivated to Protect the Perceived Integrity and Worth of the Self The most basic tenet of self‐aYrmation theory (Steele, 1988) is that people are motivated to protect the perceived integrity and worth of the self. As Steele observed, the purpose of the self‐system is to ‘‘maintain a phenomenal experience of the self . . . as adaptively and morally adequate, that is, competent, good, coherent, unitary, stable, capable of free choice, capable of controlling important outcomes . . . (p. 262).’’ These self‐conceptions and images making up the self‐system can be thought of as the diVerent domains that are important to an individual, or the diVerent contingencies of a person’s self‐worth (Crocker & Wolfe, 2001). Figure 1 presents a schematic of the self‐system. The self is composed of diVerent domains, which include an individual’s roles, such as being a student or a parent; values, such as being religious or having a sense of humor; social identities, such as membership in groups or organizations and in racial, cultural, and gender groups; and belief systems, such as political ideologies. The self is also composed of people’s goals, such as the value of being healthy or succeeding in school. The self‐system is activated when a person experiences a threat to an important self‐conception or image. Such threat poses a challenge to a desired self‐conception. Thus, failure feedback could threaten a person’s identity as a student, negative health information could threaten a person’s self‐ conception as a healthful individual, news about anti‐American sentiment could threaten a person’s patriotic identity, and evidence of social inequality could challenge a person’s belief in a just world (Lerner, 1980). All of these

188

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Fig. 1. Schematic representation of self‐system.

events are threatening because they have implications for a person’s overall sense of self‐integrity. 2. Motivations to Protect Self‐Integrity can Result in Defensive Responses When self‐integrity is threatened, people are motivated to repair it, and this motivation can lead to defensive responses. The defensive responses may seem rational and defensible, though they are more ‘‘rationalizing’’ than ‘‘rational’’ (Aronson, 1968; Kunda, 1990; Pyszczynski & Greenberg, 1987). They serve to diminish the threat and consequently, restore the perceived integrity of the self. These defensive responses can be automatic and even unconscious in nature, and indeed, the rapidity with which people respond to threats speaks to the importance of self‐integrity maintenance. 3. The Self‐System is Flexible People often compensate for failures in one aspect of their lives by emphasizing successes in other domains. Personality theorists, such as Allport (1961) and Murphy (1947), have advanced this notion of compensation, and self‐aYrmation theory is consistent with this claim (see also Brown & Smart, 1991). Because the goal of the self‐system focuses on maintaining the overall worth and integrity of the self, people can respond to threats in one domain by aYrming the self in another domain. This fungibility in the sources of self‐integrity is what can enable smokers, for example, to maintain a perception of worth and integrity despite the potentially threatening conclusion that they are acting in a maladaptive, harmful, and irrational

SELF‐AFFIRMATION THEORY

189

way (Steele, 1988). AYrmations satisfy the motivation to maintain self‐ integrity—thus, they reduce the normal psychological adaptations people engage to ameliorate a specific provoking threat. 4. People can be AYrmed by Engaging in Activities that Remind them of ‘‘Who They Are’’ (And Doing so Reduces the Implications for Self‐Integrity of Threatening Events) Those qualities that are central to how people see themselves are potential domains of self‐aYrmation. Such aYrmations can concern friends and family, making art or music, a charity, or the observance of one’s religion. In a diYcult situation, reminders of these core qualities can provide people with perspective on who they are and anchor their sense of self‐integrity in the face of threat. A ‘‘self‐aYrmation’’ makes salient one of these important core qualities or sources of identity. Operationally, self‐aYrmations are typically ideographic, in that people first report an important value or life domain, and then they are given the opportunity either to write an essay about it or to complete a scale or exercise that allows them to assert its importance (McQueen & Klein, 2005). When global perceptions of self‐integrity are aYrmed, otherwise threatening events or information lose their self‐threatening capacity because the individual can view them within a broader, larger view of the self. People can thus focus not on the implications for self‐integrity of a given threat or stressor, but on its informational value. When self‐aYrmed, individuals feel as though the task of proving their worth, both to themselves and to others, is ‘‘settled.’’ As a consequence, they can focus on other salient demands in the situation beyond ego protection.

II. Self‐AYrmation and Threats to the Individual Self A great deal of research has used self‐aYrmation theory to address a wide range of psychological phenomena, including biased information processing, causal attributions, cognitive dissonance, prejudice and stereotyping, stress and rumination. What connects these disparate areas of research is that they all address situations or events where people contend with a threat to a valued self‐image. We first review research on the impact of experienced threats to self‐identities, such as one’s political identity or one’s identity as a healthful or intelligent individual. We then review research on the impact of experienced threats to collective identities, such as one’s team or racial group.

190

DAVID K. SHERMAN AND GEOFFREY L. COHEN

A. MOTIVATED INFERENCES AND BIASED ASSIMILATION People often interpret new information in a way that reinforces their beliefs and desires. One striking example occurred during the 2000 presidential controversy in Florida. Evaluations of seemingly arbitrary events (e.g., beliefs about appropriate policies for dealing with hanging chads, whether to count votes from certain counties) were consistently and predictably aligned with people’s political identities. The Florida situation presented a series of novel issues where partisanship lines had not, at first, been clearly defined. Yet, in little time, people were able to determine their criteria for establishing voter intent, and in most cases, their criteria were consistent with their partisanship, and the position that they ultimately advocated served the interests of their candidate of choice (cf. Gerber & Green, 1999). It is interesting to imagine what would have happened during the Florida debacle if the situation were reversed, and if Vice President Albert Gore were leading rather than Governor George W. Bush as the vote was contested. That is, if the situation was ‘‘counterbalanced’’ like a proper psychology experiment, would Republican leaders have argued for the sanctity of the vote and the need to count each chad and would Democratic leaders have argued against them? Research on biased assimilation by Lord et al. (1979) suggests that they would. In this classic study, proponents and opponents of capital punishment evaluated two studies on the eYcacy of capital punishment as a deterrent. The two studies featured diVerent designs (a panel design comparing murder rates from states before and after the implementation of capital punishment policy, and a concurrent design comparing murder rates from states that either used capital punishment or did not). Unlike the situation in Florida, the study was completely counterbalanced. That is, two versions of each study were created, one that supported and one that refuted the deterrent value of capital punishment. Thus, the study allows an examination of whether and to what extent prior beliefs (or identities) bias the interpretation of information. When the study was consistent with participants’ prior beliefs, they thought it was better conducted and more convincing regardless of the specific design (concurrent versus panel) of the study. Moreover, reading a mixed bag of evidence—one study supporting and one study contradicting their beliefs—led participants to report becoming more confident in the validity of their beliefs about the deterrent eYcacy of the death penalty. The study provides an experimental analog to the hypothetical switch in the Florida election scenario. Regardless of what evidence people were presented, they evaluated it in a way consistent with their prior beliefs and, in turn, this biased assimilation appeared to strengthen the very prior beliefs that give rise to the bias in evaluation.

SELF‐AFFIRMATION THEORY

191

Lord et al. (1979) emphasized the role of cool cognitive inferential processes in producing biased assimilation. Like Bayesian theorists, people use their prior beliefs to evaluate the validity of incoming data. It is thus logically permissible for them to reject belief‐incongruent evidence and to accept with little scrutiny belief‐congruent evidence (Nisbett & Ross, 1980). In contrast, we argue that such biases issue from the motivation to maintain and protect identity (see also Munro & Ditto, 1997). That is, partisans in the Lord et al. (1979) study scrutinized the evidence in a way that would protect their identity either as a ‘‘law and order’’ conservative or as a ‘‘humanitarian’’ liberal (Cohen, 2003; Ellsworth & Ross, 1983). According to our analysis, the need to protect a valued identity or self‐ view is a major source of such biased processing and closed mindedness. Because long‐held beliefs are often tied to important identities, they may be given up only with great reluctance, and they may be embraced even when they conflict with the demands of fact, logic, or material self‐interest (e.g., Abelson, 1986; Sears & Funk, 1991). Yet, people possess other important identities and values that they can draw on when they encounter belief‐ threatening information. Providing them with an aYrmation of one of these alternative sources of self‐integrity should enable them to evaluate the threatening information in a less biased and defensive manner. We tested this logic in a study of biased assimilation in the domain of capital punishment (Cohen, Aronson, & Steele, 2000). In a study patterned after Lord et al. (1979), proponents and opponents of capital punishment read an article from the Journal of Law and Human Behavior titled ‘‘The Death Penalty: New Evidence Informs an Old Debate.’’ The article was fabricated, but contained facts, statistics, and arguments that waged a persuasive assault on participants’ attitudes toward capital punishment. Thus, proponents of capital punishment read an anti‐capital punishment report and opponents of capital punishment read a pro‐capital punishment report. Prior to reading the article on capital punishment, all participants completed a writing exercise that constituted our self‐aYrmation manipulation. Participants in the self‐aYrmation condition wrote an essay about a personal value that they had rated, during pretest, as personally important (such as their relationships with friends or sense of humor). Specifically, they were asked to describe three to four personal experiences where the value had been important to them and had made them feel good about themselves. The value they wrote about was, in all cases, unrelated to their political views. Participants in the no‐aYrmation condition wrote about a neutral topic. In the no‐aYrmation control condition, our findings mirrored those of Lord et al. (1979). Participants found flaws in the methodology of the studies that contradicted their political beliefs, they suspected bias on the part of the authors of the report, and they persisted in their attitudes about capital

192

DAVID K. SHERMAN AND GEOFFREY L. COHEN

punishment. By contrast, the responses of participants who aYrmed a valued self‐identity proved more balanced. That is, self‐aYrmed participants were less critical of the reported research and suspected less bias on the part of the authors of the report. Participants even changed their global attitudes toward capital punishment in the direction of the report they read. That is, proponents of capital punishment supported the death penalty less, and opponents of capital punishment supported it more (Cohen et al., 2000, Study 2). That both partisan groups showed the eVect attests to the power of the psychological mechanism. Not only did proponents come to privilege ‘‘life’’ over ‘‘law and order’’ more when self‐aYrmed than when not, but opponents of capital punishment, when self‐aYrmed, came to support state‐sanctioned execution to a greater extent (see also Jacks & O’Brien, 2004). A third study examined how people evaluate others who either agree or disagree with their beliefs (Cohen et al., 2000). Pro‐choice and pro‐life partisans were presented with a debate between two activists on opposite sides of the abortion dispute. Participants who did not receive a self‐aYrmation judged the activist who shared their convictions more favorably than the activist who did not (see also Hastorf & Cantril, 1954). In contrast, participants who were given a self‐aYrmation became more balanced in their evaluation of the two activists and asserted that they were relatively less confident of the validity of their abortion attitudes, relative to their nonaYrmed peers. The motivation to protect identity may prove especially consequential in the context of negotiation. In negotiation—between parties, nations, individuals—there exists a barrier to compromise, a barrier that often leads disputing factions to reject even mutually beneficial settlements and instead persist in mutually destructive conflict (Ross & Ward, 1995; Sherman, Nelson, & Ross, 2003). This barrier issues, in part, from a motivation to defend one’s political, national, or regional identity—a motivation that can result in intransigence and stalemate (De Dreu & Carnevale, 2003). To accept compromise entails acting in a way that could exact painful costs to self and social identity. Compromise entails that one must accept policies anathema to one’s past ideological commitments, or take courses of action contrary to the interests of one’s constituencies and group loyalties. Accordingly, negotiation should be facilitated when people’s partisan identities are made less vital to their sense of self‐integrity through the aYrmation of alternative sources of self‐worth. In one study, pro‐choice participants entered into a negotiation with a pro‐life advocate about appropriate federal abortion policy (Cohen et al., 2005). To make their partisan identity salient prior to the negotiation, participants were first asked to assert their ‘‘true beliefs’’ on abortion policy. Additionally, prior to the negotiation, half of the participants received a values aYrmation and half did not. It was found that aYrmation increased

SELF‐AFFIRMATION THEORY

193

the number of concessions that pro‐choice participants made to their pro‐life adversary (e.g., they were more likely to agree to the idea of parental notification). More dramatically, aYrmation also led participants to evaluate their adversary as more objective and trustworthy (i.e., as less influenced by self‐interest and ideology)—a finding with clear implications for real‐ world negotiation where the cultivation of trust is a critical step in the resolution of conflict. These studies raise an important question as to why people are more open‐ minded when aYrmed than when not. Correll, Spencer, and Zanna (2004) proposed that aYrmation leads to greater attitude change via a more careful consideration of the arguments, rather than through more superficial, heuristic processing. That is, aYrmation does not lead to change through more peripheral or ‘‘mindless’’ routes (e.g., by raising mood and thus promoting agreeableness) but through the more central route of more balanced information processing (Petty & Cacioppo, 1986; Petty & Wegener, 1998). These researchers presented people with pro‐ and counter‐attitudinal arguments that varied in quality. According to their analysis, what should prove most threatening is strong evidence against one’s position and weak evidence in favor of one’s position. Thus, self‐aYrmation should increase openness only to strong evidence against one’s position and increase rejection of weak evidence in favor of one’s position. This is what Correll et al. (2004) found. Self‐aYrmation led people to be more responsive to the intrinsic strength of the arguments rather than the concordance of those arguments with their prior beliefs. Also consistent with the claim that self‐aYrmation helps reduce the bias of personal beliefs in processing information, Cohen et al. (2005) found that self‐aYrmed participants exhibited more balanced cognitive and aVective responses to a counter‐attitudinal report than did their non‐aYrmed peers.

B. THREATENING HEALTH INFORMATION Defensive processing can be particularly costly when it leads people to reject important health information. Individuals often face information suggesting that they are engaging in behavior that puts their health at risk. Health information can threaten the self by suggesting that people have acted unwisely, for example, by smoking, drinking, or practicing unsafe sex. Although it would be optimal if people responded to personal health information by ceasing risky behavior, people often resist threatening health messages, and subsequently, persist in unhealthful or risky behaviors. For example, one study found that sexually active students who saw an AIDS educational message responded by lowering their perceived risk for sexually

194

DAVID K. SHERMAN AND GEOFFREY L. COHEN

transmitted diseases—a defensive response (Morris & Swann, 1996). Several studies have found that when a health message is of high personal relevance, people are more likely to scrutinize the message for fault than when a message is of no special relevance (Ditto & Lopez, 1992; Kunda, 1987; Liberman & Chaiken, 1992). People use diVerent criteria or probative thresholds when evaluating the validity of information that either supports or contradicts their desires (Ditto, Scepansky, Munro, Apanovitch, & Lockhart, 1998). That is, they use a ‘‘must I believe this information?’’ criterion when evaluating evidence that runs contrary to their preexisting health beliefs, whereas they use a ‘‘can I believe this information?’’ criterion when evaluating evidence consistent with those beliefs (Dawson, Gilovich, & Regan, 2002). People are biased in their assessments of threatening health information, we argue, because being a ‘‘healthy person’’ is an important part of how they see themselves. Making a behavioral change in a positive direction carries the self‐evaluative burden of acknowledging that one has engaged in maladaptive behaviors in the past. Thus, out of a desire to protect a self‐image of oneself as rational and healthy, people may ultimately persevere in irrational and unhealthful behavior. However, the logic of self‐aYrmation theory suggests that if individuals can reflect on an alternative source of identity, their overall sense of self‐integrity will be secured, and they may be more apt to consider information without resorting to defensive biases. Sherman, Nelson, and Steele (2000) examined defensive responses to threatening health information in the context of breast‐cancer prevention. Participants were women who were either coVee drinkers or noncoVee drinkers, and they reviewed a (fabricated) scientific report linking caVeine consumption to fibrocystic disease, a precursor to breast cancer. The report concluded by suggesting that women can reduce their risk for fibrocystic disease by reducing their caVeine consumption. Those in the no‐aYrmation control condition exhibited the pattern found in earlier research (Kunda, 1987; Liberman & Chaiken, 1992): the coVee‐drinking women were more critical of the scientific article and more resistant to the message than were the noncoVee drinking women (see Fig. 2). By contrast, coVee‐drinking women who had completed a scale that enabled them to assert the personal importance of a central value (e.g., their religious or political values) proved more open to the message than any other group and intended to reduce their coVee drinking accordingly. Because the motivation to protect self‐worth was satisfied via this self‐aYrmation, people who would have otherwise felt threatened by the health message proved more open and more willing to engage in adaptive behavior change. Other researchers have also documented the ‘‘de‐biasing’’ eVects of self‐ aVirmation. In one study by Reed and Aspinwall (1998), heavy and light

SELF‐AFFIRMATION THEORY

195

Fig. 2. Acceptance of article’s conclusions as a function of coVee‐drinking status and aVirmation status. From Sherman, D. A. K., Nelson, L. D., & Steele, C. M. (2000). Do messages about health risks threaten the self ? Increasing the acceptance of threatening health messages via self‐aVirmation. Personality and Social Psychology Bulletin, 26, 1046–1058. Adapted with permission.

caVeine consuming women were given information about the link between caVeine consumption and fibrocystic disease. Half of the participants were then given the opportunity to aYrm the self by reflecting on acts of kindness that they had performed. The heavy caVeine consumers who were aYrmed oriented more rapidly to the risk‐confirming information, that is, they looked at this information more quickly and viewed it as more convincing. Furthermore, the aYrmed heavy caVeine consumers recalled less risk‐disconfirming information at a one‐week follow‐up. If self‐aYrmation can reduce the biased processing of health information, does it have implications for actual behavior? One study examined whether aYrmation could lead to change in health behavior (Sherman et al., 2000). Sexually active undergraduates watched an AIDS‐educational video suggesting that their sexual behavior could put them at risk for HIV. Half received a self‐aYrmation prior to watching the video; the others did not. Although non‐aYrmed participants tended to resist the presented information (maintaining their perceived risk from pretest levels), aYrmed participants responded by increasing their perceived potential risk for contracting AIDS. The aYrmation not only aVected perceived risk but also aVected subsequent health behavior. Whereas 25% of nonaYrmed individuals

196

DAVID K. SHERMAN AND GEOFFREY L. COHEN

purchased condoms after viewing the video, 50% of aYrmed participants did so. Additionally, participants who completed the self‐aYrmation were more likely to take an AIDS‐educational brochure (78% did so) than participants in the no‐aYrmation condition (54% did so). The self‐aYrmation seemed to buVer individuals from the potential threat of the AIDS‐educational message, opened them to the possibility that they were at risk, and motivated them to undertake preventive behaviors. Do self‐aYrmation interventions produce long‐term eVects on increasing the acceptance of threatening health messages? For the eVects to have practical significance, they must persist beyond the experimental setting. From a theoretical standpoint, if self‐aYrmation promotes attitude change via more central route processing (as was found in the Correll et al., 2004 study), then the eVects of self‐aYrmation on attitude change should be durable rather than short‐lived. To date, few studies have examined the longer‐term eVects of self‐aYrmation. Harris and Napper (2005) specifically examined the longer‐term eVects of self‐aYrmation on health persuasion. Young women (light versus heavy alcohol consumers) were given a pamphlet describing the risks of developing breast cancer from drinking too much alcohol. The pamphlet informed young women that excessive drinking can lead to breast cancer and is thus dangerous. Prior to reading the pamphlet, all participants completed a self‐ aYrmation wherein they either wrote an essay about their most important value or a relatively unimportant value. The researchers assessed perceived risk at the experimental session as well as at one week and one month after the experimental session. Among the participants most at risk for health problems—that is, the women who drank heavily—those who completed the self‐aYrmation saw themselves as being at greater risk for breast cancer than those who were not aYrmed (as in Sherman et al., 2000), an eVect that did not diminish over time. The high‐risk self–aYrmed participants also reported that they found it easier to imagine themselves with breast cancer—an eVect that persisted one month later. Finally, the high‐risk participants who completed the self‐aYrmation had stronger intentions to reduce their alcohol consumption at the experimental session, although this was not associated with more positive health behaviors in the form of drinking reduction during the following month. Thus, the eVects of the self‐aYrmation proved durable (1 month) at aVecting risk perceptions, although the long‐term eVects on behavior have not been established. An interesting question concerns the factors that facilitate long‐term behavioral eVects. Long‐term eVects will be enhanced, we suspect, when situational cues remind individuals of the change in attitude and identity that a persuasive message had engendered in them. For example, Dal Cin, MacDonald, Fong, Zanna, and Elton‐Marshall (in press) provided evidence

SELF‐AFFIRMATION THEORY

197

that an information campaign to reduce unsafe sexual practices could have long‐term eVects on condom use (as assessed up to 8 weeks later) if participants are given an ‘‘identity cue’’ to wear that would remind them both of the ‘‘safe sex’’ persuasive video they had previously viewed and to their commitment to the cause of promoting safe sexual practices. Specifically, the impact of the safer sex message on behavior was greatly increased if participants were given a ‘‘friendship bracelet’’ that served to aYrm their personal concern for people suVering from sexually transmitted diseases (Dal Cin et al., in press). Pairing such identity cues with self‐aYrmation constitutes an exciting direction for future research.

C. STRESS If self‐aYrmation enables people to view otherwise threatening events, such as negative health information, as less threatening, then one intriguing possibility is that it could also reduce evaluative stress. Stress is the process by which environmental events are appraised as threatening, which in turn elicits emotional and physiological responses that can adversely aVect health and increase susceptibility to disease (Lazarus, 1993; McEwen, 1998). As Keough, Garcia, and Steele (1998) suggested, stress may arise, in large part, from threats to the perceived worth and integrity of the self (Creswell, et al., 2005; Keough, 1998). Consistent with this notion, common laboratory methods for inducing stress, such as the Trier Social Stress Task, have participants deliver a personally relevant speech and perform diYcult mental arithmetic in front of a hostile audience (Kirschbaum, Pirke, & Hellhammer, 1993). These tasks have reliably been shown to threaten the self, induce appraisals of threat, and elicit the stress hormone cortisol (Dickerson & Kemeny, 2004). The self‐aYrmation analysis also suggests that stress can be ameliorated through the aYrmation of alternative sources of self‐integrity. Specifically, self‐aYrmations can secure self‐worth in a domain unrelated to the stressor (Creswell et al., 2005; Keough, 1998), and may, as a result, encourage people to view the stressful event without feeling as though their self integrity is in question. A study examining self‐enhancement and physiological responses to stress provides suggestive evidence for this eVect. Those who saw themselves as being above average across a number of domains responded to a stressful situation (mental arithmetic challenge in front of a hostile audience) with reduced blood pressure and heart rate, relative to those who were less self‐enhancing (Taylor, Lerner, Sherman, Sage, & McDowell, 2003). This finding suggests that those who perceive themselves as having more resources and abilities experience a potentially threatening task in a less physiologically taxing manner because their overall feelings of self‐integrity

198

DAVID K. SHERMAN AND GEOFFREY L. COHEN

hinge less on their performance on that particular task (see also Seery, Blascovich, Weisbuch, & Vick, 2004). An influential model of how to conceptualize psychological resources in stressful situations was advanced by Hobfoll (1989). He defined resources as ‘‘those objects, personal characteristics, conditions, or energies that are valued by the individual or that serve as a means for attainment of these objects, personal characteristics, and energies (p. 516).’’ Psychological resources can thus encompass aspects of self that are unrelated to a given threat or stressor. If this is true, then enhancing the perception of self‐ resources via self‐aYrmation could reduce the physiological costs of stress. Consistent with this work, one study found that self‐aYrmation reduced perceived stress during a diYcult serial subtraction task (Keough et al., 1998). To test the stress‐buVering eVects of self‐aYrmation directly, one study had participants complete a self‐aYrmation procedure—a values scale for their most important value (versus an unimportant value in the control condition)—prior to engaging in the Trier Social Stress Task (Creswell et al., 2005). Thus, those in the self‐aYrmation condition indicated their agreement with an important value (e.g., religion) in a domain of their life unrelated to the stressful task. Participants were instructed to prepare a speech to be given to two speech evaluators concerning why they would be qualified for a job as an administrative assistant in a psychology department. The speech evaluators (i.e., confederates who were trained to act in a sullen manner) directed the 5‐minute speech task and then asked the participant to complete a 5‐minute mental arithmetic task by counting aloud backwards from 2083 by 13’s (a task where self‐aYrmations have been successful at improving performance; Schimel, Arndt, Banko, & Cook, 2004). Salivary cortisol measurements were assessed at baseline and at 20, 30, and 45 minutes poststress onset. As displayed in Fig. 3, during baseline there were no diVerences in cortisol levels between participants in the self‐aYrmation and control conditions. Yet, at the 20‐minute measurement, and persisting through 45 minutes after the stress task, the participants in the control condition had elevated cortisol levels, whereas the self‐aYrmation participants did not (Creswell et al., 2005). This finding provides the first experimental evidence that a self‐ aYrmation manipulation can buVer the self not only at a psychological level, but also at a physiological level. Threatening experiences are not only stressful while they occur, but also after the fact, as people engage in ruminative thinking, and the thoughts of the threatening event reoccur (Martin & Tesser, 1996; Nolen‐Hoeksema & Morrow, 1991). One study examined whether this ruminative thinking would be reduced among participants who completed a self‐aYrmation (Koole, Smeets, van Knippenberg, & Dijksterhuis, 1999). In this research,

SELF‐AFFIRMATION THEORY

199

Fig. 3. Salivary cortisol as a function of time and aYrmation status. From Creswell, J. D., Welch, W., Taylor, S. E., Sherman, D. K., Gruenewald, T., & Mann, T. (2005). AYrmation of personal values buVers neuroendocrine and psychological stress responses. Psychological Science, 16, 846–851. Adapted with permission.

participants were given failure feedback on an analogy test that was ostensibly an IQ test. Then, half of them completed a values scale concerning their most important value, whereas the other half completed a values scale of a value they considered to be relatively unimportant to them. The accessibility of the words from the failed IQ test (i.e., words from the analogy items) was assessed, either by recognition accuracy (Study 1) or the use of a lexical decision‐making test (Study 2) (Koole et al., 1999). Both measures assessed the extent to which participants were ruminating about the failure. Those participants who aYrmed an important aspect of the self by completing the values scale were less likely to ruminate after failure (Koole et al., 1999). Self‐aYrmation, then, can decrease ruminative thinking, and this may be one mechanism by which it makes stressors less stressful. If self‐aYrmations can reduce physiological responses to stress (Creswell et al., 2005), then it is plausible that repeated aYrmations might help people cope with daily stressors. As daily stressors have been shown to impair immune functioning (Glaser & Kiecolt‐Glaser, 1994), a self‐aYrmation intervention could lead to beneficial health outcomes. Keough et al. (1998) examined this hypothesis in a sample of undergraduates who wrote essays over winter break on one of four topics. Participants in the aYrmation condition wrote about the events of the day in terms of their most important value. Specifically, they integrated the day’s events into that value (e.g., by writing about how the happenings of the day were relevant to the importance of their relationships with friends). There were three control conditions: a writing control condition (where people simply recorded what had

200

DAVID K. SHERMAN AND GEOFFREY L. COHEN

happened that day), a positive events control group (where people wrote about the positive things that had happened to them that day), and a nonwriting control group. Compared to all other groups, those who wrote the aYrming essays over winter break reported the least amount of stress, and visited the health center significantly less often (when they returned to school). This eVect was particularly strong among those who reported experiencing the greatest number of daily hassles and were, presumably, the most stressed. Thus, self‐aYrmation interventions were associated with positive physiological and health outcomes with two relatively healthy college student samples (Creswell et al., 2005; Keough et al., 1998). An important question centers on whether similar eVects would be obtained in a sample of people confronting disease. A study of women with breast cancer by Creswell et al. (2006) examined the relation between self‐aYrming writing and recovery from breast cancer. This research was based on a reanalysis of data from a study by Stanton et al. (2002). Stanton and colleagues (Stanton et al., 2002) had initially recruited 60 early‐stage breast cancer survivors and randomly assigned them to write four essays on (1) their deepest thoughts and feelings about breast cancer (emotional processing); (2) their positive thoughts and feelings about breast cancer (benefit finding); or (3) the facts of their day (control). Findings from this initial study showed that women in both the emotional‐processing and benefit‐finding conditions had reduced physical symptoms and doctor visits 3 months after the study. In the follow‐up study (Creswell et al., 2006), all essays were coded for self‐aYrmation (whether the essay evidenced positive reflection on valued self‐domains), cognitive processing (whether the essays evidenced active thinking about the positive aspects of the breast cancer experience), and discovery of meaning statements (whether the essay evidenced eVorts to find a larger lesson or significance to their breast cancer), to test for potential mediating mechanisms for the writing eVects on breast‐cancer recovery. Interestingly, the number of aYrming statements made in the essays (e.g., ‘‘We have been married over 31 years and we are very lucky because we still love each other.’’) fully mediated the intervention eVects found on physical symptoms at 3‐month follow‐up, whereas (at least in this study) no mediating role was found for cognitive processing or discovery of meaning. Self‐ aYrming cognitions played an important mediating role regardless of whether the intervention itself focused on emotional processing or benefit finding. Additionally, the frequency of self‐aYrmation statements predicted enhanced coping and reductions in distress at the follow‐up session. Together with the earlier work demonstrating the eVects of self‐aYrmation on increasing the acceptance of threatening health information, the research reviewed in this section suggests that self‐aYrmation can have both direct

SELF‐AFFIRMATION THEORY

201

and indirect benefits for health. Directly, it seems to reduce physiological stress reactivity (Creswell et al., 2005), reduce daily stress perceptions (Keough et al., 1998), and potentially, aid in coping with distress in people with chronic disease (Creswell et al., 2006). Indirectly, it may promote more positive health behaviors by making people more open to information that they are engaging in risky health behaviors (Harris & Napper, 2005; Reed & Aspinwall, 1998; Sherman et al., 2000).

D. COGNITIVE DISSONANCE The notion in self‐aYrmation theory (Steele, 1988) of a flexible self‐system that seeks to maintain its perceived integrity shed new light on cognitive dissonance theory, and in particular, on the assumption that dissonance is aroused by psychological inconsistency (Aronson, 1968; Festinger, 1957). Specifically, the self‐aYrmation studies suggested that people are primarily motivated to maintain self‐integrity rather than psychological consistency. Thus, people should be able to tolerate psychological inconsistency—and to refrain from dissonance‐motivated cognitive distortions—if their self‐integrity is buttressed through the aYrmations of an alternative domain of identity. For example, Steele, Hopp, and Gonzales (1986; cited in Steele, 1988) had participants rank 10 record albums in order of preference. They then gave the participants a choice of their 5th‐ or 6th‐ranked album—a dissonance provoking choice (Brehm, 1956) that typically leads participants to accentuate the diVerences in the direction of their choice. Half of the participants then put on a lab coat in anticipation of a second study. For those who indicated that science was their most important value, wearing the lab coat aYrmed an important domain. For others who indicated that science was not an important domain, wearing a lab coat was not self‐aYrming. All students then re‐ranked the record albums. As predicted, with one exception, all groups showed the standard ‘‘spreading of alternatives’’ eVect. They inflated the value of the option chosen and denigrated the value of the option foregone—a garden‐variety rationalization eVect. The exception was those science‐minded students who wore the lab coat, and who thus had the opportunity to restore their self‐integrity through the aYrmation of an important self‐identity. These participants did not defensively change their attitudes to make them concordant with their choice (see also Steele & Liu, 1983). A full review of the theoretical debate concerning the circumstances under which self‐aYrmation is a viable explanation for cognitive dissonance phenomena is beyond the scope of this chapter (see Aronson et al., 1999; Simon, Greenberg, & Brehm, 1995; Stone & Cooper, 2001). However, there have been several important developments that merit review.

202

DAVID K. SHERMAN AND GEOFFREY L. COHEN

In various dissonance‐provoking situations, self‐aYrmations have been found to reduce psychological discomfort and defensive biases. For example, Matz and Wood (2005) found that individuals report experiencing an aversive state of dissonance when others in a social group disagree with them. However, this dissonance discomfort was eliminated among participants who completed a self‐aYrmation. Another study suggested that dissonance reduction results in overconfident judgments (Blanton, Pelham, Dettart, & Carvallo, 2001). That is, if one cares about the accuracy of one’s predictions then any cognitions that call that accuracy into question are likely to arouse dissonance. This dissonance is reduced by finding reasons to bolster one’s confidence in the accuracy of one’s predictions. Blanton et al. (2001) found that people who cared more about a prediction (involving correctly identifying Pepsi versus Coke based on a taste test) were relatively overconfident in the accuracy of their judgments. However, this overconfidence was eliminated among those who completed a self‐aYrmation. Other studies have examined the conditions moderating the eVects of self‐ aYrmation on dissonance‐reducing justifications (as in Steele & Liu, 1983). When the self‐aYrmations are in the same domain as the threatening information, they have been found to exacerbate cognitive dissonance (Blanton, Cooper, Skurnik, & Aronson, 1997) because they make salient the personal standards that are violated with the dissonant behavior. Consequently, when given a choice, people tend to choose to aYrm the self in a domain unrelated to the perceived threat in order to reduce the dissonance they are experiencing (Aronson, Blanton, & Cooper, 1995). However, even aYrmations in alternative domains of self‐worth are not impervious to disconfirmation, and when they are disconfirmed (e.g., being told you are not religious shortly after aYrming religion as a central value), cognitive dissonance can be reinstated (Galinsky, Stone, & Cooper, 2000). Several studies (e.g., Steele, Spencer, & Lynch, 1993; Stone & Cooper, 2001) have also examined the relationship between self‐esteem, self‐aYrmation, and cognitive dissonance processes—a topic to which we will return in our discussion of individual diVerences and their moderating influence on self‐aYrmation processes.

E. MOTIVATED DISTORTIONS IN SOCIAL PERCEPTION When people make judgments of others—such as whether a person is a good candidate for a job, or whether a person is above or below average—they have the opportunity to put themselves in a positive light through social comparison. Indeed, Dunning (2003) has argued that people are ‘‘zealous self‐aYrmers’’ in that their social judgments often reflect more about their own self‐evaluative needs than about the target of judgment. Whether

SELF‐AFFIRMATION THEORY

203

students evaluate a person who studies 19 hours a week as ‘‘studious’’ depends upon how many hours a week they study themselves (Dunning & Cohen, 1992)—that is, people egocentrically define what evidence constitutes a given trait or ability. When people define what characteristics are likely to make a person have a successful marriage, they emphasize those characteristics that they themselves possess rather than those they do not (Kunda, 1987). And they do so more after experiencing a threat to self‐worth, than after experiencing an aYrmation of self‐worth (Dunning & Beauregard, 2000; Dunning & Hayes, 1996; Dunning, Leuenberger, & Sherman, 1995). After experiencing a self‐threat, people may engage in any number of strategies to reaYrm self‐integrity via social judgment. These strategies include comparing the self with a clearly inferior other (Fein, Hoshino‐ Browne, Davies, & Spencer, 2003), gossiping negatively about a third party (Wert, 2004; Wert & Salovey, 2004), or harshly judging a political ingroup member who fails to demonstrate as much fervor for the cause as one personally does (Beauregard & Dunning, 1998). At other times, esteem‐ boosting social judgments take a more disturbing form; people may try to feel better about themselves by putting down members of a marginalized group (Fein & Spencer, 1997). Because stereotypes are ‘‘cognitively defensible’’ justifications for denigrating others, people may be especially apt to use them to restore self‐worth. Derogating an outgroup member would not only enhance one’s own self‐worth via downward comparison processes, but also enhance the integrity of a person’s ingroup more generally. In one study, participants who were threatened with negative feedback on an intelligence task showed more stereotyping in their judgments of a gay male than those who received neutral feedback (Fein & Spencer, 1997). Thus, self‐threatening feedback can exacerbate outgroup derogation and the use of stereotypes. This finding suggests that self‐image maintenance concerns can motivate prejudicial responses. Is the converse true as well— can an aYrmation of an individual’s self‐integrity reduce the need to stereotype an outgroup member? Fein and Spencer (1997) examined this possibility in one study where participants completed a self‐aYrmation through writing about an important value (versus a control condition where they wrote about a relatively unimportant value). They did so prior to evaluating a job candidate who was presented either as a member of a negatively stereotyped group (Julie Goldberg, who fit with a ‘‘Jewish American Princess’’ stereotype widely in circulation on campus at the time) or not (Maria D’Agostino, an Italian American). In the no‐aYrmation condition, participants made more negative evaluations of the candidate’s qualification for the job and viewed the personality of the Jewish woman more negatively than that of the Italian woman. In contrast, those participants who completed the self‐aYrmation rated the Jewish candidate as favorably as the

204

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Italian candidate. Thus, the extent to which a person is threatened or aYrmed will aVect whether or not they are likely to make prejudicial judgments of an outgroup member (cf. Shrira & Martin, 2005; Zarate & Garza, 2002). Indeed, in this study, self‐aYrmation did not simply attenuate prejudice and discrimination but eliminated it. Comparing oneself to another person who is faring worse is another way that people may aYrm self‐integrity via social perception. When people have a vulnerable or easily threatened self‐image, they generally respond with downward social comparisons (Taylor & Lobel, 1989; Wills, 1981). Shelley Taylor and her colleagues (Taylor, Wood, & Lichtman, 1983) found that women with breast cancer responded to their highly threatening situation by making downward social comparisons. They would assert that their illness was, at least, less severe than that of another patient, or that they were coping with their illness better than others. These downward comparisons, Taylor suggested, helped patients to maintain a sense of worth in a situation where they struggled to maintain their sense of control, predictability, and optimism. Such downward comparisons, while helping to sustain self‐ integrity, may by themselves limit an individual’s opportunities to learn from others who are more experienced or performing more optimally (indeed, such ‘‘upward’’ comparisons also have motivational benefits; see Lockwood & Kunda, 1997). If people engage in downward social comparisons to compensate for a threatened self‐image, then they should engage in them to a lesser extent when self‐aYrmed. Indeed, they may even be more prone to engage in upward comparisons that they might otherwise view as threatening (Spencer, Fein, & Lomore, 2001). In one study, college students completed a test of intelligence, and they were informed that they performed at the 47th percentile. This mediocre performance presumably threatened participants’ self‐image as intelligent college students. Half of the students then had the opportunity to aYrm the self by writing an essay about an important value, whereas the other half wrote about an unimportant value. Then the participants, in a separate task, were informed that another participant would interview them. For the ostensible purpose of preparing them for their interview, they then listened to excerpts of two previous interviews. In one of the excerpts, the interviewee made a terrible impression (e.g., unintentionally insulting the interviewer, speaking incoherently), whereas in the other the interviewee was smooth and articulate. The participants had a choice of which interview to listen to in full. This choice provided an opportunity to make an upward comparison (if they selected the superior interviewee) or a downward comparison (if they selected the inferior interviewee). Because all participants had been threatened by the feedback they had received on the intelligence task, the study allowed an examination of social

SELF‐AFFIRMATION THEORY

205

comparison under threat. In the no‐aYrmation condition, the participants generally made downward comparisons, as 83% chose to hear the inferior interviewee. By contrast, among those who completed the self‐aYrmation, 83% chose to hear the superior interviewee, making an upward social comparison. Once again, self‐aYrmation reduced threat and thereby encouraged people to expose themselves to an informative but potentially threatening learning opportunity. In summary, predictions derived from self‐aYrmation theory (Steele, 1988) have been supported in a wide range of situations involving self‐threat. When self‐integrity is aYrmed, people are less biased in their judgments of information related to their political identity (Cohen et al., 2000), their health (Reed & Aspinwall, 1998; Sherman et al., 2000), and their impressions of others (Fein & Spencer, 1997; Spencer et al., 2001). Self‐aYrmation inoculates people against threat, and thus makes them more open to ideas that would otherwise be too painful to accept. It also reduces the stressfulness of evaluative situations (Creswell et al., 2005) and the cognitive accessibility of threatening cognitions (Koole et al., 1999). Theoretically consistent eVects of self‐aYrmation have been found on self‐report measures, physiological responses, and behavior. When self‐integrity is secured, people seem less concerned with the self‐evaluative implications of social experiences and are more likely to engage their social world in a non‐defensive, open manner.

III. Self‐AYrmation and Responses to Collective Threats Originally, self‐aYrmation theory (Steele, 1988) focused on how people respond to information and events that threaten a valued self‐image, such as situations that provoke cognitive dissonance and defensive rationalization of counter‐attitudinal behavior (Steele & Liu, 1983), or situations that challenge a sense of personal control (Liu & Steele, 1986). The research detailed in the previous sections extends this theorizing to many other situations where people contend with events that challenge a personal identity. A major advance in self‐aYrmation theory concerns its relevance to the way people cope with threats to their social (i.e., group) identities. This work begins with the premise that social identities—such as aYliation with a sports team, membership in a gender or racial group, citizenship in a country, involvement in an organization—constitute important sources of identity (Abrams & Hogg, 1988; Deaux, 1996; Tajfel & Turner, 1986). Consequently, people will defend against threats to collective aspects of the self much as they defend against threats to individual or personal aspects of self. They may do so even when these events do not directly implicate

206

DAVID K. SHERMAN AND GEOFFREY L. COHEN

oneself (e.g., even when the threat involves the behavior of another group member rather than one’s own behavior; Cohen & Garcia, 2005; Norton, Monin, Cooper, & Hogg, 2003; Wiesenfeld, Brockner, & Martin, 1999). We argue that both personal and social (i.e., group) identities are fundamentally confounded (Cohen & Garcia, 2005), in that both types of identities contribute to the same overarching goal of maintaining self‐integrity (see also Gaertner, Sedikides, & Graetz, 1999; Sherman & Kim, 2005). When a fellow member of one’s racial group confronts the threat of a negative stereotype, it is threatening to self even when one personally faces no risk of being prejudiced against (Cohen & Garcia, 2005). When an athlete’s team loses, he or she will resist explanations that implicate the team because doing otherwise would threaten the image of a valued group. When one’s country is vilified in an editorial, a patriot will attack the credibility of the source to protect the perceived integrity of his or her national identity. The self‐aYrmation analysis of such collective threats, however, asserts that because social identities are only one part of a larger, flexible self‐ system, people can respond to threats to their group memberships or social identities indirectly. That is, they can maintain an overall self‐perception of worth and integrity by aYrming some other aspect of the self, unrelated to their group. This insight has applications to a wide range of phenomena related to group identity.

A. GROUP‐SERVING JUDGMENTS A popular aphorism is that ‘‘There is no I in team.’’ It conveys the notion that the goals of a team or a group should be important enough that individuals are willing to sacrifice their personal ambitions and self‐interests in the service of team success. Self‐serving aspirations are antithetical to the team’s progress. However, social groups are a central part of how people see themselves, and people are motivated to defend their social identities (Abrams & Hogg, 1988). Consequently, when explaining the success or failure of their group, people tend to be defensive and group serving. This phenomenon has been observed in several studies examining attributions made for success or failure in real world contexts. Letters to shareholders feature more internal attributions after successful years (e.g., stocks rose because of the work of the management team) than after disappointing years (Bettman & Weitz, 1983), and athletes’ explanations for success feature more internal attributions than do their explanations for failure (Lau & Russell, 1980; Winkler & Taylor, 1979). These biased judgments—encompassing both inflated internal attributions for success and attenuated internal attributions for failure—are group serving in the sense that they suggest that

SELF‐AFFIRMATION THEORY

207

group members are selectively responsible for causing the positive events that befall their group. In two field studies involving intramural sports team athletes as participants, Sherman and Kim (2005) examined the role of the self‐integrity motivations in group‐serving judgments. The intramural sports teams, consisting of voluntarily formed groups of friends, played their games where they either won or lost and then participated in the study. They completed a self‐aYrmation manipulation in which they filled out a values scale concerning either their most important value or a relatively unimportant value. Then, they assessed their attributions for the outcome of the game. They estimated how much their personal performance contributed to the outcome of the game, and how much their team’s play and teamwork contributed to the outcome of the game. Overall, the athletes were both self‐serving and group serving in their attributions. That is, winners thought that their personal performance contributed more to their team’s victory than losers did to their defeat (a self‐ serving bias). Additionally, winners thought that their team’s performance and teamwork contributed more to their team’s victory than losers did to their defeat (a group‐serving bias). Both self‐ and group‐serving biases, however, were eliminated among those who completed a self‐aYrmation. The reduction of group‐serving judgments is depicted in Fig. 4. Whereas in the no

Fig. 4. Group‐serving attributions as a function of game outcome and aVirmation status. From Sherman, D. K., & Kim, H. S. (2005). Is there an ‘‘I’’ in ‘‘team’’? The role of the self in group‐serving judgments. Journal of Personality and Social Psychology, 88, 108–120. Adapted with permission.

208

DAVID K. SHERMAN AND GEOFFREY L. COHEN

aYrmation condition, winners made much stronger internal team attributions than did losers, this diVerence was eliminated in the self‐aYrmation condition (Sherman & Kim, 2005). A second goal of this research focused on examining the relation between self‐ and group‐serving judgments. Because one’s self‐concept and one’s social identity are overlapping cognitive constructs (Smith & Henry, 1996; Smith, Coats, & Walling, 1999), people will tend to use the self‐concept as an anchor or evaluative base to form judgments of the group (Cadinu & Rothbart, 1996; Gramzow & Gaertner, 2005; Otten, 2002). People generalize their positive evaluation of self to their evaluation of their group. However, if the need to protect self‐integrity is satisfied via self‐aYrmation, then people should be able to evaluate the group independently of how they evaluate the self. We examined whether such self‐group anchoring would be attenuated by a self‐aYrmation (Sherman & Kim, 2005, Study 2), again with intramural athletes who had just won or lost a game. We replicated the basic findings that self‐aYrmation reduced both the self‐ and the group‐serving attributional biases. Additionally, we examined the correlations between the attributions to the self and the attributions to the group as a function of aYrmation condition. As predicted, in the no‐aYrmation condition, there was a strong correlation (r ¼ .60) between judgments about the self and judgments about the group, indicating that under normal conditions, the athletes generalized their self‐evaluations to their evaluations of their group. In contrast, when participants were self‐aYrmed in an alternative domain, they evaluated the group independently of their self‐evaluations, as the correlation was eliminated (r ¼
.10). Self‐aYrmation, then, allows people to evaluate their groups independently of the way they evaluate themselves.

B. DEFENSE OF COLLECTIVE IDENTITY Patriotism and pride in one’s national citizenship represent manifestations of another collective identity that people try to defend—particularly when their country is under threat. In the wake of September 11th, for example, two responses emerged. On the one hand, some people responded to the national threat by embracing their American identity uncritically, and by finding little if any problem or fault in America’s foreign policy in the Middle East. On the other hand, there were those who took a critical eye to American foreign policies and noted how these policies (such as its support of totalitarian regimes in the Middle East) may have sowed the conditions and sense of disenfranchisement among that populace, which contributed, in part, to the terrorist acts against the United States.

SELF‐AFFIRMATION THEORY

209

We were interested in how people’s self‐identities as American ‘‘patriots’’ (i.e., people who asserted that they were patriotic, and who felt that the US constituted a force of good in the world) or American ‘‘anti‐patriots’’ (i.e., people who asserted that they were unpatriotic, and who felt that the US constituted a force of harm in the world) would aVect their evaluation of a persuasive report entitled ‘‘Beyond the Rhetoric: Understanding the Recent Terrorist Attacks in Context’’ (Cohen et al., 2005). The report argued that Islamic terrorism can be understood in terms of the social and economic forces of the Middle East. It further underscored the role that US foreign policy had played in fostering some of the social—economic conditions in the Middle East that later led to the terrorist attacks of 9/11. The arguments were credible (drawn from the writings of several prominent analysts) and buttressed with factual evidence and historical analysis. To amplify the counter‐American tone of the report, it was ostensibly written by an author of Arab descent, ‘‘Babek Hafezi.’’ To make participants’ American identity situationally salient, the experimenter wore a small American flag pin on her lapel.

Fig. 5. Openness to anti‐U.S. foreign policy report among patriots and anti‐patriots as a function of aYrmation status.

210

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Figure 5 presents the relevant results. In the absence of aYrmation, there was a partisan divide in perception. Patriots proved far more critical of the report than did antipatriots. Indeed, the correlation between a measure of participants’ identification as an American patriot versus antipatriot (administered several weeks prior to the experiment), and their openness to the report was r ¼ .58. In other words, participants’ preexisting identity accounted for 34% of the variance in their openness to the report. By contrast, in a condition where participants self‐aYrmed prior to reading the report (by writing about an important value unrelated to their national identity), patriots became more open to the report, and anti‐patriots became more skeptical of it. Indeed, the correlation between prior identity and openness was reduced to nil (r ¼
.05). Collective identity ceased to aVect the assimilation of new information when participants had aYrmed an alternative source of self‐integrity.

C. STEREOTYPE THREAT AND PERFORMANCE Stereotype threat—the potential that one could be judged in light of a negative stereotype about one’s group—is a potent type of collective threat (Steele, 1997). Members of negatively stereotyped groups may experience elevated levels of stress when performing on tasks where they risk confirming a stereotype about their group in the eyes of others (Ben‐Zeev, Fein, & Inzlicht, 2005; Blascovich, Spencer, Quinn, & Steele, 2001; O’Brien & Crandell, 2003). This stress can in turn undermine performance. Several studies, both in the laboratory and in the field, have now examined whether self‐aYrmation can reduce stereotype threat and facilitate performance in conditions where people from negatively stereotyped groups have been shown to underperform (Spencer, Steele, & Quinn, 1999; Steele & Aronson, 1995). One study examined whether stereotype threat could be reduced among women taking a math test (Martens, Johns, Greenberg, & Schimel, 2006). To the extent that women care about performing well in such a situation, they may worry, should they perform poorly, that they could confirm the negative stereotype that women are worse than men at math. If so, then aYrmation of an alternative aspect of the self could reduce the threat and facilitate performance. Whereas other interventions to reduce stereotype threat directly refute the stereotype or its relevance (e.g., by portraying a test as gender‐fair; Cohen, Steele, & Ross, 1999; Spencer et al., 1999), this intervention was psychologically farther downstream—aimed at altering students’ threat/stress response to the perceived relevance of the stereotype.

SELF‐AFFIRMATION THEORY

211

The participants were female college students who took a diYcult math test that was described either as ‘‘diagnostic’’ of their math and reasoning abilities or as a reasoning task unrelated to their ability and still being developed for future research. Replicating the standard stereotype threat eVect (Steele, Spencer, & Aronson, 2002), women performed worse in the ability‐diagnostic condition (where they were aware that a potentially poor performance on their part could be viewed as evidence of lack of math ability and as ultimately a validation of the negative stereotype), worse than their female counterparts in the ability‐nondiagnostic condition, and worse than men overall. However, women in the diagnostic condition who completed a self‐aYrmation—where they wrote about an important value unrelated to math or to their gender—performed as well as women in the no‐threat condition and no diVerent from male college students. Given the evidence that self‐aYrmation appears to reduce stereotype threat among students in laboratory situations (Martens et al., 2006; Schimel et al., 2004), perhaps having students engage in aYrmation of non‐academic aspects of themselves in school settings could produce positive educational benefits by reducing the psychological impact of social identity threat and other sources of psychological distress. In such educational contexts, social identity threat may be exacerbated by the mistrust that minority students sometimes feel when confronted with negative or critical feedback from White evaluators and teachers (Cohen et al., 1999). Because they know that the stereotype could bias members of the outgroup, these students may come to doubt the intentions motivating critical feedback. Is it possible for aYrmation to mitigate these detrimental responses? To examine this question, a field study (Cohen, Garcia, Apfel, & Masters, 2006) was conducted that featured a randomized, double‐blind experimental design. Students completed a 20‐minute in‐class self‐aYrmation exercise in which they wrote about an important value and why it mattered to them, or, in a control condition, wrote about an unimportant value and why it might be important to someone else. In the control condition, minority students, unlike their majority peers, displayed a decline in trust in their teachers and school administrators over the course of the 7th‐grade school year. That is, they judged their grades and treatment in school as less fair and more biased at the end of the year than they had at the beginning. By contrast, aYrmed minority students remained constant in their relatively high levels of trust and perceived fairness over the course of the school year (see Fig. 6). Just as aYrmation increases trust across partisan lines in the context of negotiation (Cohen et al., 2000, 2005), it cultivates trust and reduces threat across racial lines as well.

212

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Fig. 6. Trust in academic authorities as a function of race, time, and aYrmation status.

D. PERCEPTIONS OF RACISM When individuals interact with members of stereotyped groups, they often want to appear nonprejudiced (e.g., Gaertner & Dovidio, 1986; Monteith & Voils, 1998). The very idea of prejudice itself could be threatening to an individual’s self‐integrity for many reasons. Consider a European American evaluating whether prejudice against minorities still constitutes a major factor in the diVerential academic and economic success of people from diVerent groups. If prejudice is a major factor, this suggests that one may have been unfairly privileged, that one may have benefited (indirectly or even directly) from the oppression of minority groups, and that one’s country fails, in some ways, to live up to the values of equality and egalitarianism. One way people from majority groups can reduce this threat is by minimizing the perceived frequency and impact of prejudice against minority groups. Building on this logic, Adams, Tormala, and O’Brien (in press) examined the eVect of self‐aYrmation on perceptions of prejudice against minorities. The participants were both European Americans and Latinos. Overall, Latinos perceived more prejudice against minorities in everyday life than did European American participants. However, this race eVect was attenuated by a self‐aYrmation. There were no group diVerences in perceived prejudice among those who aYrmed an important value (unrelated to race)

SELF‐AFFIRMATION THEORY

213

prior to assessing their perceptions of prejudice. Importantly, the eVect of self‐aYrmation was more evident among European Americans than Latino Americans. While Latino Americans showed a slight drop in their perception of prejudice against minority groups when self‐aYrmed, this drop was not significant. On the other hand, self‐aYrmation led European Americans to perceive significantly more racism against minorities in the United States (Adams et al., in press, Study 2). AYrmed European American participants also agreed, to a far greater extent than their non‐aYrmed peers, that European Americans in general tend to understate the impact of racism in daily life. Thus, the otherwise threatening idea of racism in America was more acceptable among those who were buVered by a self‐aYrmation. Indeed, these studies suggest that it is not just minority students who defensively exaggerate the role of prejudice to protect self‐worth (cf. Crocker & Major, 1989). It is the majority group members, as well, who defensively deny racism in order to protect their self‐esteem.

IV. Moderator Variables and Qualifying Conditions Across a variety of potentially threatening situations, self‐aYrmations reduced perceived threat and the likelihood of engaging in defensive adaptations to threat. Health information suggesting that one has acted in a risky manner, the defeat of one’s team, a stereotype directed at oneself or a fellow group member, a report attacking one’s political worldview, a failure at an intellectual or athletic task—these events threaten individuals’ self‐integrity. When the self is under threat, people engage defensive adaptations to ameliorate the threat. Motivated inferences about health information, group‐serving biases, biased assimilation of new information, and outgroup derogation are examples of such adaptations. These self‐protective strategies can be reduced and even eliminated, however, when people aYrm alternative sources of self‐integrity unrelated to the provoking threat. To understand the process of self‐aYrmation, it is important to examine factors that moderate when and how self‐aYrmation operates. Cultural diVerences may moderate the eVects of threat and aYrmation. Additionally, individual diVerences in self‐esteem and in identification with the domain of threat should moderate whether and to what extent self‐aYrmation is eVective. Situational diVerences in the salience of a particular identity may prove critical. Moreover, aYrmations may have diVerent eVects as a function of their form and content. While the majority of the research previously reviewed demonstrates how self‐aYrmations can lead to greater openness,

214

DAVID K. SHERMAN AND GEOFFREY L. COHEN

some types of aYrmations may backfire; they may decrease openness and promote defensive responses. We now review these moderating variables.

A. CULTURE Given the extensive research and theorizing on cross‐cultural diVerences in the conception and experience of selfhood (e.g., Heine, Lehman, Markus, & Kitayama, 1999; Kim & Markus, 1999; Markus & Kitayama, 1991; Triandis, 1989), an important question concerns the eVect of culture on self‐aYrmation processes. In individualist cultures (such as the US), the self is a more bounded, autonomous entity, whereas in collectivist cultures (such as in East Asia), the self is more interconnected and tied to important relationships. Members of collectivist cultures may thus be less motivated to protect self‐integrity because their culture places less emphasis on esteeming the self (Heine et al., 1999). Alternatively, they may be just as motivated to protect self‐integrity, but they may diVer in terms of the events and issues they find threatening versus aYrming. Heine and Lehman (1997) conducted one of the first studies on culture, self‐aYrmation, and cognitive dissonance. Among European Canadians, they replicated the Steele and Liu (1983) finding that aYrmation of personal values eliminated the spreading of alternatives in the free‐choice dissonance paradigm (Brehm, 1956). Yet among Japanese participants, no spreading‐ of‐alternatives eVect was found, perhaps because they did not experience the situation as self‐threatening (Heine et al., 1999). Consequently, there was no defensive response for the aYrmation to ameliorate. Thus, to identify how self‐aYrmation operates cross‐culturally, it is important to test the eVects of a situation that could theoretically threaten self‐ integrity and produce defensiveness among people in diVerent cultures. Research by Hoshino‐Browne and colleagues (Hoshino‐Browne et al., 2005) identified such conditions in their examination of the cultural moderation of dissonance and aYrmation processes (see also Kitayama, Snibbe, Markus, & Suzuki, 2004). First, they found that when a choice is potentially threatening to their relationships, members of collectivist ethnic groups (i.e., Asian Canadians) experience cognitive dissonance. When Asian Canadians made choices of food preferences, they rationalized their decisions more when they chose for their friends than for themselves (Hoshino‐Browne et al., 2005). That is, after making a choice between two comparable food options for a friend, they came to believe that the friend would find the chosen option far superior to the non‐chosen option. In contrast, European Canadians rationalized their decisions more when they made choices for themselves rather than for their friends (Study 1 and Study 2 of Hoshino‐Browne et al., 2005).

SELF‐AFFIRMATION THEORY

215

To examine whether this dissonance would be reduced by a self‐aYrmation, the researchers modified the standard self‐aYrmation procedure to contrast an independent with an interdependent self‐aYrmation. For the interdependent aYrmation, participants ranked values in terms of how important they were to ‘‘themselves and their family’’ and wrote a paragraph about why this value is shared among them and their families. For the independent self‐ aYrmation, participants ranked values in terms of personal importance and wrote an essay about why that value was personally important. The participants in the study, all of East Asian descent, then made choices for their friends and had the opportunity to rationalize these choices. As predicted, the interdependent aYrmation reduced dissonance‐motivated rationalization when participants had to make a choice for a friend, whereas the independent self‐aYrmation did not. Follow‐up studies also revealed that, for the European Canadians, the opposite eVect occurred: the independent aYrmation reduced dissonance‐motivated rationalization, but the interdependent aYrmation did not. Moreover, both aYrmations were eVective among bicultural Asian Canadians who identified with both Asian and Canadian culture (see also Hoshino‐Browne, Zanna, Spencer, & Zanna, 2004). The latter result suggests that bicultural identity can confer access to diVerent sources of aYrmation. This research sheds light on how culture moderates the eVectiveness of self‐aYrmation. Where the self is structured diVerently, what constitutes a threat and an aYrmation are diVerent as well. However, the general process whereby aYrmation reduces defensive responses to threats appears to be culturally invariant.

B. SELF‐ESTEEM If one way that people respond to threats is by aYrming the self in an alternative domain, then it stands to reason that those who have more positive views of the self will have more psychological resources with which to self‐aYrm. Steele et al. (1993) predicted that high self‐esteem individuals would have greater aYrmational resources and would thus be more resilient to threatening events than would low self‐esteem individuals. Consistent with this logic, they found that high self‐esteem people were less likely to rationalize a choice they had made (i.e., by inflating the value of a music album they had selected and/or by deflating the value of a desirable music album they had foregone). Interestingly, this eVect was found only among those high self‐esteem people who had first completed a self‐esteem scale that made their self‐resources cognitively accessible to them. Under this condition, those with greater aYrmational resources were more resilient to

216

DAVID K. SHERMAN AND GEOFFREY L. COHEN

threats to their self‐image and less likely to engage in defensive rationalization of their behavior (see also Nail, Misak, & Davis, 2004). There are two schools of thought about the relationship between self‐ esteem and the ability to tolerate dissonant acts or beliefs. Whereas the aYrmational resources view posits that people with high self‐esteem have more positive self‐concepts, making it easier for them to tolerate their dissonant actions (Spencer, Josephs, & Steele, 1993; Steele et al., 1993), the self‐consistency view posits that high self‐esteem people will view a dissonant act as more discrepant with their positive self‐image (Aronson, 1968; Thibodeau & Aronson, 1992). According to this latter view, high self‐esteem people are likely to feel more dissonance than low self‐esteem people and are more prone to engage in dissonance reduction strategies such as justification. For example, one study examined smokers who were participating in a smoking cessation clinic (Gibbons, Eggleston, & Benthin, 1997). The smokers who relapsed and began smoking again reduced the dissonance associated with this attitude‐discrepant behavior (discrepant because they were committed to quitting smoking) by reducing their self‐perceived risk of acquiring smoking‐related illnesses. However, this occurred only among high self‐esteem smokers (Gibbons et al., 1997). A third theory, the self‐standards model of cognitive dissonance (Stone & Cooper, 2001) proposes that people, at times, use self‐esteem as a resource (as in Steele et al., 1993). However, people may also use their level of self‐ esteem as a standard or expectancy that determines whether they view a particular behavior as dissonant. Once people act, they evaluate their behavior in light of a standard—but what standard comes to mind varies with situational cues. When the cues focus on normative standards for behavior (e.g., how most people act), no diVerences between low– and high–self‐ esteem individuals emerge. It is only when the cues focus people on their own personal standards that self‐esteem diVerences do, in fact, emerge. Whether low or high self‐esteem individuals rationalize more, in turn, depends on the nature of the self‐aYrmation. Self‐aYrmations related to the threatening act (e.g., writing about how compassionate one is after having behaved uncompassionately) increase the salience of personal standards of conduct, draw attention to the discrepancy between those standards and one’s previous behavior and thus lead high self‐esteem people to experience more dissonance (and rationalize more) than low self‐esteem people. By contrast, self‐aYrmations unrelated to the threatening act draw attention to alternative psychological resources, and thus lead high self‐esteem people to experience less dissonance (and rationalize less) than low self‐esteem people (Stone & Cooper, 2003). One variable that holds important implications for the relation between self‐esteem and cognitive dissonance is whether a person’s high self‐esteem is

SELF‐AFFIRMATION THEORY

217

secure and stable versus insecure, defensive, and fragile (Baumeister, Heatherton, & Tice, 1993; Deci & Ryan, 1995; Jordan, Spencer, Zanna, Hoshino‐Browne, & Cornell, 2003; see also Kernis & Paradise, 2002). Secure, high self‐esteem is operationalized as having high explicit self‐esteem (as assessed by a self‐report measure of self‐esteem such as the Rosenberg Self‐Esteem Scale; Rosenberg, 1965) and high implicit self‐esteem (i.e., nonconscious, automatic positive associations with the self, as assessed by the Implicit Associations Test; Greenwald, McGhee, & Schwartz, 1998). Such people have high self‐esteem both at a conscious level and at a deep, unconscious, and reflexive level. By contrast, defensive high self‐esteem is operationalized as having high explicit self‐esteem but low implicit self‐ esteem. Such people have high conscious self‐esteem but unconsciously, and reflexively, evaluate the self negatively. Jordan and colleagues find that those with secure high self‐esteem show little dissonance reduction (consistent with the aYrmational resources view). That is, those with secure high self‐esteem are less likely to rationalize their decisions to reduce dissonance, whereas those with defensive high self‐esteem are more likely to engage in dissonance reduction (consistent with the self‐consistency view) (Jordan, Spencer, & Zanna, 2003). It seems that people with defensive high self‐esteem may have fewer self‐resources that can be drawn upon when experiencing threat.

C. IDENTITY CENTRALITY AND SALIENCE Whether a potentially threatening domain is personally important to an individual or constitutes a part of their personal identity aVects whether or not they can experience self‐threat in that domain (Boninger, Krosnick, & Berent, 1995), and consequently, whether an aYrmation will prove eVective at ameliorating defensiveness. Ironically, it is the people who view an issue as important rather than unimportant who should prove the most open to aYrmation‐induced change. For example, a report describing the link between caVeine use and fibrocystic disease concerns a topic of greater importance for heavy caVeine consumers than for light caVeine consumers. Consequently, it is only heavy caVeine consumers who respond defensively to such a report (e.g., Kunda, 1987; Liberman and Chaiken, 1992) and who exhibit greater openness to the report when self‐aYrmed (Sherman et al., 2000). Adopting another approach, Correll et al. (2004) examined students’ responses to a debate over the merits of a tuition increase. They found that only participants who saw tuition increases as personally important were biased in their assessments of the quality of pro‐ versus counterattitudinal information on this topic. Further, it was only these participants for whom

218

DAVID K. SHERMAN AND GEOFFREY L. COHEN

aYrmation proved an eVective debiasing intervention. In a related vein, one study found that sports fans that watched their team win or lose were only biased in their attributions for the outcome of the game when being a fan was viewed by them as an important part of their identity (Sherman, Kinias, Major, Kim, & Prenovost, 2005). Moreover, this group‐serving bias could be eliminated when participants aYrmed a value of high importance to their group. Individuals may thus vary in the strength of their allegiance to particular sources of identity and in their resistance to new and challenging ideas that would force them to see the world in other than black and white terms (Correll et al., 2004; Zuwerink & Devine, 1996). Additionally, implicit and explicit situational cues may play a role in heightening these types of identity defense motivations that lead to closed‐mindedness and inflexibility. Situational contexts diVer in how much they highlight a particular identity and its importance to self‐integrity (Cohen et al., 2005). When a given context heightens the salience of an individual’s partisan identity and/or commitment to a given position on an identity‐relevant issue, that individual may feel relatively more obliged to consider the identity costs of openness and compromise. It is precisely those costs, in turn, that are reduced when an alternative source of self‐integrity is made salient and aYrmed. In one study, for example, political partisans with varying views regarding US foreign policy evaluated a report critical of that policy. They did so after a self‐aYrmation or threat was administered, under conditions where either their identities as self‐rated ‘‘patriots’’ (versus ‘‘anti‐patriots’’) or the goal of rationality was made salient. In the identity salient condition, the experimenter wore an American flag pin on her lapel, whereas in the rationality salient condition, the experimenter wore a white lab coat. A focus on rationality was expected to diminish attention to, and concern with, sources of identity that would otherwise make the prospect of belief change and compromise threatening. The results in the American flag condition were described previously: participants interpreted information critical of the United States in a manner consistent with their national identity—but a self‐aYrmation eliminated this bias. In contrast, when the experimenter wore the lab coat—minimizing concerns of identity maintenance—aYrmation had no eVect (Cohen et al., 2005).

D. WHEN AFFIRMATIONS BACKFIRE An interesting issue concerns classes of aYrmations that increase rather than decrease bias and resistance to change. One such class of aYrmations is those that are in the same domain as the threatening event or information.

SELF‐AFFIRMATION THEORY

219

Whereas diVerent‐domain aYrmations decrease bias, closed‐mindedness, and inflexibility, same‐domain aYrmations increase people’s sense of self‐ confidence, certainty, and impunity. They thus produce eVects opposite to those observed when the aYrmation targets a domain unrelated to the threat (Blanton et al., 1997). Prior to reviewing a counter‐attitudinal report, for example, people might aYrm a value related to the issue that figures in that report. Before reading a report critical of evidence supporting global warming, an environmentalist might reflect on the personal importance of the cause of environmentalism (Sherman et al., 2005). Alternatively, before reading a report arguing against allowing gays in the military, an egalitarian might assert the importance of tolerance (Jacks & O’Brien, 2004). In contrast to domain‐irrelevant aYrmations, such domain‐relevant aYrmations consistently increase resistance to change, presumably by highlighting a person’s commitment to the issue and the identity at stake. This research suggests that, if attitude and behavior change is the goal, policy‐makers and interventionists should resist lay intuitions to provide aYrmations targeting the domain of threat. It may do more harm than good to provide academically at‐risk students with aYrmations of their academic ability, risk‐seeking adolescents with aYrmations of their rationality, or negotiators with aYrmations of their ideological commitments. Instead, the more eVective strategy is the counterintuitive one—that of aYrming domains unrelated to the threat. An intriguing phenomenon involves the way in which, under some circumstances, aYrmations of moral worth can lead to a sense of personal impunity, in which people no longer feel obliged to prove themselves in the domain in question and thus feel licensed to act in ways that violate important moral principles. This seems to occur particularly in contexts where the ‘‘right’’ or ‘‘moral’’ course of action is ambiguous. For example, when people aYrm their lack of prejudice, they engage in more gender and racial discrimination, at least under some situations that render such discrimination ambiguous in its moral implications (Jacks & O’Brien, 2004; Monin & Miller, 2001). In one line of studies, people made their ‘‘moral credentials’’ clear—in our language, aYrmed their self‐concepts as ‘‘moral’’ and ‘‘egalitarian’’—by being led to select a clearly qualified ethnic minority for a job. Relative to those who did not aYrm their lack of prejudice, those who did later proved more prejudiced; that is, they were more willing to state that a particular job (one that could have posed problems for a minority due to its hostile work environment) was more appropriate for White workers than for Black ones. When people aYrm their lack of prejudice, they seem to become more lax in their adherence to egalitarian values. In an unusually disturbing series of studies, Brown (2000) further demonstrated how moral aYrmations not only lead to impunity but also license what many would view as immoral behavior. College students were given

220

DAVID K. SHERMAN AND GEOFFREY L. COHEN

positive feedback concerning the morality of individuals at their school— they were told that students at their school scored higher on standard moral development tests than did students attending a rival school. ‘‘Morally aYrmed’’ participants were subsequently more swayed by negative propaganda targeting foreign graduate students. The propaganda in question asserted that foreign graduate students posed a threat to national security due to the possibility that they might, for example, steal scientific secrets or engage in espionage. In response to this propaganda, morally aYrmed subjects proved more likely to endorse punitive and ‘‘proto‐genocidal’’ policies for dealing with the ‘‘foreign graduate student problem.’’ They advocated to a greater extent the policy of forcing foreign graduate students to carry identification papers with them at all times and even became more favorable to the proposal to banish all foreign students from studying in American universities. The moral aYrmation seemed to imbue participants with impunity to discriminate in the name of protecting national security. Clearly, more research needs to be done to determine the conditions under which such domain‐relevant aYrmations lead to open‐mindedness, closed‐mindedness, and moral impunity. Another class of potentially detrimental aYrmations involves those that activate a person’s sense of self‐perceived objectivity. When made to feel objective, people adopt an ‘‘I think it, therefore it’s true’’ mindset. That is, they assume that their own thoughts and beliefs—by virtue of being theirs— are objective, valid, and therefore worthy of being acted upon (Uhlmann & Cohen, 2005a). As a consequence, objectivity‐aYrmed evaluators may become more likely to act on any prejudicial thoughts and beliefs that they harbor, and whose influence they might have otherwise suppressed. Consistent with this analysis, in a series of studies, Uhlmann and Cohen (2005b) found that aYrming individuals’ sense of objectivity made them more likely to discriminate against women when deciding whom to hire for stereotypically male, high‐status jobs (e.g., corporate representative, police chief). Objectivity‐aYrmed subjects also proved more likely to act on stereotypes that had been nonconsciously primed in their social environment. In one study, participants were either primed on words related to negative stereotypes about women (e.g., pink, Barbie, make‐up, emotional) or not. Those who aYrmed their objectivity (i.e., by answering several easy‐to‐agree‐with questions concerning their personal objectivity) were more influenced by this prime. When primed on gender stereotypes, participants proved more likely to hire a man over a woman when they were objectivity‐aYrmed than when they were not (see Fig. 7 for the relevant results). Paradoxically, making people feel objective causes them to become more subjective. It confers a sense of impunity about using personal thoughts and beliefs that happen to be cognitively salient at the particular moment of a decision.

SELF‐AFFIRMATION THEORY

221

Fig. 7. Pro‐male gender discrimination as a function of gender stereotype prime and aYrmation status.

V. Underlying Processes and Superordinate Functions Given the wide range of studies reported in this review, it is clear that self‐ aYrmations exert eVects across many domains of psychological functioning. Yet, important questions remain concerning both the mechanisms underlying these eVects and the ultimate function of the self‐integrity system. These questions set the stage for exciting and important opportunities for future research. A. UNDERLYING MECHANISM OF SELF‐AFFIRMATION Through what mechanisms do self‐aYrmations produce their eVects? AYrmations, we suggest, lift people’s self‐evaluative concerns in the situation at hand and allow other motivations, such as a desire to be even‐handed, rational, or healthful, to predominate. But, what are the psychological mechanisms implicated in this process? What do the self‐aYrmation manipulations do to individuals that result in such eVects? One point seems important to acknowledge. For phenomena as rich and complex as defensive resistance to persuasion, stereotyping, and dissonance reduction, there is unlikely to be one mechanism or mediator through

222

DAVID K. SHERMAN AND GEOFFREY L. COHEN

which aYrmation produces the eVects it does. It seems likely that there are a number of aVective, cognitive, and motivational processes acting in concert to produce self‐aYrmation eVects. One way to address this question is to probe other theories of self‐defense for underlying commonalities. Tesser (2000) has argued that self‐esteem maintenance processes are essentially interchangeable with self‐aYrmation processes. One focus of self‐esteem maintenance theory (Tesser, 1988) is how people respond to potentially threatening social comparisons. In performance settings, they behave in ways that maintain their self‐esteem, for example, by sabotaging the performance of close others (rather than that of strangers) on important intellectual tasks (Tesser & Smith, 1980). Moreover, supporting the idea of interchangeability or fluidity in self‐defensive processes, aYrmation of an important value reduces people’s tendency to sabotage the performance of friends working on intellectual tasks (Tesser & Cornell, 1991). People are also more likely to spontaneously self‐aYrm, for example, by writing essays with more self‐aYrming content, after making a self‐threatening upward comparison (Tesser, Crepaz, Collins, Cornell, & Beach, 2000). Tesser (2000) argues that these studies support a basic interchangeability of self‐evaluative processes. Importantly, he asks whether the diVerent self‐related processes identified by self‐aYrmation theory and self‐esteem maintenance theory share the common mechanism of maintaining positive aVect or mood. From this perspective, the eVect of self‐aYrmation on behavior should be mediated by a reduction in negative aVect. Contrary to this expectation, self‐ aYrmation studies generally find that aYrmation of important values has no eVect on self‐reported mood (Fein & Spencer, 1997; Schmeichel & Martens, 2005; Sherman et al., 2000; Spencer et al., 2001) and that manipulations of mood fail to produce the same eVects as self‐aYrmations (e.g., Steele et al., 1993; but see Raghunathan & Trope, 2002). Moreover, if aYrmations were operating via positive mood, then one would expect aYrmed individuals to be accepting of persuasive arguments regardless of the strength or weakness of those arguments (Mackie & Worth, 1989). The finding that self‐aYrmation makes people more responsive to the strength of persuasive arguments (Correll et al., 2004) runs contrary to this expectation. On the other hand, an alternative possibility is that the common mechanism is aVect of which a person is unaware (Tesser, Martin, & Cornell, 1996). While it is clear that much cognition lies beneath conscious awareness (Bargh & Chartrand, 1999; Dijksterhuis, 2004; Wilson, 2002), an interesting question concerns whether aVect can lie beneath conscious awareness as well. The use of implicit measures of aVect (i.e., measures that circumvent conscious control) would help to address this issue. As one example, Koole et al. (1999) found that a measure of nonconscious aVect (i.e., a word

SELF‐AFFIRMATION THEORY

223

fragment completion task) mediated the eVect of self‐aYrmation on the reduction in rumination reported previously. That is, self‐aYrmation led to increases in implicit positive mood, which in turn was associated with decreases in ruminative thought. State self‐esteem is another mediational candidate. Self‐aYrmations might boost state self‐esteem and thus enable people to accept conclusions that might otherwise lower their state self‐esteem below an acceptable threshold. Consistent with this reasoning, Fein and Spencer (1997) gave people positive personality feedback and found that this self‐aYrmation not only reduced stereotyping but also raised state self‐esteem. Moreover, the increase in state self‐esteem mediated the reduction in stereotyping. However, in other studies, aYrmations of personal values have not been found to boost self‐esteem. Schmeichel and Martens (2005) extensively tested the impact of aYrmation (i.e., writing about an important personal value) on state self‐esteem. While they did find eVects on the primary outcomes featured in that study (worldview defense and death‐thought accessibility), they found no eVects on self‐esteem. It appears that self‐aYrmations can have eVects similar to self‐esteem eVects, but without influencing self‐esteem. One potential mediator that has not been tested, but that may hold promise, is self‐certainty. That is, reflecting on a core source of self‐integrity, such as one’s long‐held values, may make people more certain of their identity and their priorities. Such certainty could anchor individuals’ self‐ worth and make them less susceptible to self‐integrity threats. Future research examining this issue could adopt measures used to assess self‐concept clarity (Campbell, 1990; Markus, 1977), for example by having people indicate ‘‘me’’/‘‘not me’’ to a series of trait adjectives. To the extent that aYrmation makes people more certain of the self, they may respond more quickly to such questions when they involve self‐descriptive traits (see also McGregor & Marigold, 2003). To close, while diVerent individual studies have found evidence for potential mechanisms underlying self‐aYrmation eVects—including elevations in implicit positive aVect (Koole et al., 1999), in state self‐esteem (Cohen et al., 2000; Fein & Spencer, 1997), in collective self‐esteem (Sherman & Kim, 2005), and in message scrutiny (Correll et al., 2004)—few if any consistent mediators have emerged across multiple studies. Thus, this remains an important topic for future research. In future studies, researchers should also consider the proposal of Spencer, Zanna, and Fong (2005). In questioning whether experiments employing the mediational analyses suggested by Baron and Kenny (1986) are a proper ‘‘gold standard’’ in social psychology, they argue that designs that rely on mediational analyses should only be preferred when measurement of a psychological process is easy and manipulation of it is diYcult. In contrast, an alternative strategy in the

224

DAVID K. SHERMAN AND GEOFFREY L. COHEN

present context would be to experimentally manipulate these posited psychological mediators independently of self‐aYrmation and to assess whether the standard eVects of self‐aYrmation are achieved in their absence.

B. ARE PEOPLE AWARE OF SELF‐AFFIRMATION PROCESSES? One important question concerns whether people are aware of the self‐ aYrmation process or of the eVects of aYrmations on their behavior. Consider Al Franken’s character of Stuart Smalley (1992), who parodied the self‐ help movement with his so‐called daily aYrmations, ‘‘I’m good enough, I’m smart enough and doggone it, people like me.’’ It appears that his attempts at aYrmation fall short, in part, because he is well aware that he is desperately trying to boost his own self‐worth. Conversely, the process of self‐ aYrmation, as described in this chapter, tends to occur outside of awareness, in that participants are generally unaware that the goal of the procedure is to modify or enhance their feelings of self‐worth. (Moreover, in contrast to Stuart Smalley’s self‐help attempts, self‐aYrmation interventions usually entail aYrming a specific area of self‐integrity that is unrelated to the domain of threat.) We propose that awareness is a critical moderator of the eVects of self‐aYrmation. When one is made aware of the link between a self‐ aYrmation intervention and one’s potential responses to a threatening event, the eVectiveness of aYrmation may be diminished. In a series of studies, we examined the role of awareness in self‐aYrmation processes (Sherman et al., 2005). In one pair of studies, we adopted the methodology of Gilbert et al. (1998) who used two groups of participants—experiencers and forecasters—to compare the actual eVects of a stimulus or event with their anticipated eVect. In the first group of participants (the experiencers), we demonstrated that self‐aYrmation of an important value unrelated to identity‐threatening news made people more open to counterattitudinal information. In this case, environmentalists who had aYrmed an alternative value proved more open to the ostensibly ‘‘good’’ but identity‐challenging news that ecological catastrophe in the near future was unlikely. For a second group of environmentalists (the forecasters), we simply described the procedure of writing the value‐aYrming essay and presented them the identity‐threatening news report. Then we asked participants whether and how writing the essay would aVect their evaluation of the news report. Participants evidenced no awareness that the aYrmation exercise would aVect openness to the essay. The rated influence of the value‐aYrming essay was no greater than the rated influence of a neutral essay.

SELF‐AFFIRMATION THEORY

225

A more direct test of the role of awareness is to make people aware of the potential impact of the self‐aYrmation exercise and to assess the impact of such awareness on the eYcacy of the aYrmation. In one study, we examined optimistic illusions about invulnerability to disease; such illusions are notoriously diYcult to undo (Weinstein & Klein, 1995). We found that such illusions were reduced via a self‐aYrmation procedure. Participants who completed a values scale relevant to an important value proved more likely to acknowledge that they, relative to the average student, were at some risk for negative health outcomes. However, aYrmation had no impact on participants who had been informed that the value aYrmation was linked to their assessments of physical and health risk (i.e., that the researchers were examining the connection between personal values and health assessments). Merely suggesting the potential link between a self‐aVirmation and one’s subsequent response rendered the aVirmation impotent. A third study examined an implication of the argument that aYrmation processes occur beneath conscious awareness: that the self‐aYrmation process can proceed entirely without participants’ awareness either of the perceived connection between the aYrmation and the outcome or the self‐aYrming stimulus itself. Adapting the procedure of Bargh, Chen, and Burrows (1996), we primed participants to be in an aYrmed state and found that such implicit aYrmation can reduce defensiveness (Sherman et al., 2005). Participants failed a diYcult math test and then indicated their interest in pursuing various math‐related careers (e.g., physicist, mathematician). The typical, defensive response to such failure is to disidentify from domain‐ relevant careers—that is, to downplay one’s personal interest in pursuing math‐related professions (e.g., Major et al., 1998). Before failing the math test, participants completed a sentence‐unscrambling task. That is, they had to remove one word and unscramble the remaining words to make a sentence (e.g., ‘‘he grape intelligent is’’). In one condition, the words featured in some of the items involved self‐aYrming words (e.g., intelligent, triumphant, proficient), whereas in the other conditions the words involved either threatening words (e.g., unintelligent, defeat) or achievement‐related control words (e.g., progress, eVort). Participants who had completed the self‐ aYrming, sentence–completion task showed much less defensive disidentification from math‐related careers than did participants in either of the two other conditions. Indeed, the eVect size of this priming manipulation was equal in magnitude to the eVect size of participants’ sex in predicting their interest in math‐related careers. This was the case in spite of the fact that participants showed little, if any, awareness of the thematic content of the words used in the self‐aYrmation task and no awareness at all of the potential impact of this task on their interest in math careers.

226

DAVID K. SHERMAN AND GEOFFREY L. COHEN

These findings hold implications for how the process of aYrmation may operate in everyday life. AYrmations may operate with subtlety and without a mediating role for awareness or conscious intent. Consider an individual going to the doctor and receiving information that his or her cholesterol is higher than it should be and without significant changes in diet, medication may be necessary. An initial response might be skepticism, the desire to seek a retest or find fault in the information (Ditto & Lopez, 1992). Yet, if people are to profit from this type of diagnostic information, they need to accept it at some level. However, this acceptance need not be immediate. Perhaps the individual returns home and browses the Internet, checking for information on a coming election (aYrming a political identity), or examines the scores from last night’s game (aYrming a valued social identity). In such situations, people may think that they are procrastinating, but this procrastination may serve an important integrity‐reparative function. As a consequence, a second appraisal of the health‐risk information may cause less psychological pain and avoidance. Contrast this type of ‘‘non‐conscious’’ self‐aYrmation with the type oVered in an online self‐help webpage for people with bipolar disorder: Self aYrmation is you telling yourself positive things about yourself, making sure that you impart the same caring for yourself that you would provide for another . . . . Repeat aYrmations to yourself each and every day, and each and every time your ‘‘dragon’’ attacks you with a negative thought (bipolarworld.net, 2005).

Such aYrmations are intended to help people confront the threats of everyday life. However, we suspect that self‐aYrmation may prove more eVective—as suggested by the results reviewed above—if it occurs without conscious intention. Indeed, for people in chronically stressful situations, aYrmations may prove more eVective when they are delivered less frequently—the better to camouflage their intended purpose (Cohen et al., 2006). In summary, aYrmation appears to operate not in a self‐conscious, deliberate manner, but rather in a subtle, nonconscious manner. Future research could address this issue more directly by experimentally manipulating whether participants are informed of the psychological benefits of self‐aYrmation, and assessing whether such aYrmations are less eVective when people are unaware, rather than aware, of their beneficial eVects. C. QUESTIONS OF SUPERORDINATE FUNCTIONS What is the function of self‐integrity? And why do people go to such lengths to protect self‐integrity, to the point of even misrepresenting reality and misperceiving themselves? Although self‐aYrmation theory posits a self‐system

SELF‐AFFIRMATION THEORY

227

motivated to protect worth and integrity, it does not speak to the ultimate purpose of such a system. The benefits of positive self‐illusions for psychological and even physical health suggest the adaptive value, and perhaps even evolutionary advantage, of maintaining self‐integrity (cf. Colvin, Block, & Funder, 1995; Taylor & Brown, 1988; Taylor & Sherman, in press). To return to the baseball analogy noted previously, for a ‘‘good’’ hitter who bats .300 but fails nearly 70% of the time, it seems important to maintain a sense of self‐worth and eYcacy in order to take advantage of those few opportunities where one could get a ‘‘hit.’’ Maintaining self‐integrity in the face of threat could help sustain optimism and eVort in the sometimes long wait for success. Here, we oVer some additional, tentative speculations about the ‘‘superordinate’’ function of self‐integrity. One possibility is suggested by terror management theory—to wit, that the maintenance of self‐integrity serves to stave oV the existential terror that accompanies the knowledge of one’s own mortality (Solomon, Greenberg, & Pyszczynski, 1991). According to this theory, people cope with the anxiety of impending mortality by identifying with a cultural worldview—a worldview that provides meaning and the possibility of symbolic immortality through participation in culturally sanctioned projects. In the wake of September 11th, for example, American flags appeared in great abundance, and patriotism increased dramatically. After this vivid reminder of death, people found comfort in asserting their American identities, by identifying with their fellow citizens, and by supporting the symbolic leader of their culture, the President (Landau et al., 2004; Pyszczynski, Solomon, & Greenberg, 2003). Another related way that people cope with the knowledge of their mortality is through the cultivation of self‐esteem, which is achieved by being a good member of one’s culture. According to terror management theorists, maintaining self‐esteem (and perhaps self‐integrity more generally) serves to buVer people against the terror associated with the knowledge of their own mortality (Greenberg, Solomon, Arndt, & Schimel, 2004; Harmon‐Jones et al., 1997). If cultural worldviews and the maintenance of self‐esteem help to protect people from the terror associated with their mortality, then leading individuals to focus on their own mortality should increase both their endorsement of cultural worldviews and engagement in self‐esteem protective behaviors. Numerous studies support these hypotheses (see Greenberg, Solomon, & Pyszczynski, 1997 for a review). Research by Schmeichel and Martens (2005) specifically examined the relationship between self‐ aYrmation processes and terror management processes. They examined whether self‐aYrmation reduces the eVects of mortality salience. Writing about an important value, for example, not only has self‐aYrming eVects, but also serves to bolster one’s worldview (Schmeichel & Martens, 2005). As

228

DAVID K. SHERMAN AND GEOFFREY L. COHEN

a consequence, people might have less need to stave oV thoughts of their own mortality by resorting to various forms of worldview defense, such as derogation of outgroup members who hold views anathema to one’s own. In one study, a mortality salience manipulation—thinking about one’s death—led participants to defend their cultural worldview. Participants increased their support of a pro‐American essay writer over an anti‐American essay writer—a standard terror management eVect (e.g., Greenberg et al., 1990). However, this cultural worldview defense was eliminated among participants who had completed a self‐aYrmation—that is, among participants who had written about an important value unrelated to the threat. Thus, the aYrmation of self‐integrity seems to buVer people against the existential terror associated with the knowledge of their inevitable death. Such studies, however, do not address the question of motivational primacy. Does the maintenance of self‐integrity serve the ultimate purpose of managing existential terror? Such questions of ultimate purpose are, of course, diYcult to answer. However, it is clear that both terror management theory and self‐aYrmation theory provide complementary evidence concerning how the self‐integrity maintenance system copes with threats to the self, both existential and otherwise. From a more evolutionary perspective, an intriguing possibility is that the motivation to protect self‐integrity serves to increase social fitness. As Trivers (2000) has argued, people’s positive illusions and self‐deceptions may be in the service of social deception. Individuals who create a public persona as a person of integrity—as someone whose goodness, strength, and ability to control important outcomes are clear—will acquire material and even reproductive advantages in the social hierarchy. It thus serves the individual well to hide, rationalize, and compensate for personal flaws and mistakes that might otherwise compromise his or her status in the eyes of others. A similar idea is found in Nisbett and Cohen’s work (1996) on the Southern culture of honor: by projecting a social image of strength and willingness to retaliate, Southern men maintain their ‘‘honor’’ and consequently deter others from challenging their status or stealing their resources. The critical insight is this: the public posturing, social manipulations, and interpersonal deceptions needed to convince others of one’s integrity will prove far more plausible and persuasive if the individual in question actually believes in the persona he or she is trying to project (Trivers, 2000). As noted in the sociometer theory of self‐esteem (Leary & Baumeister, 2000), the ego‐ protective strategies people have evolved to maintain self‐integrity may thus help to promote their social fitness. Future research might test this account by manipulating whether self‐threats and self‐aYrmations are delivered in private or in public (such as in front of a socially significant audience). If the maintenance of self‐integrity serves a social fitness function, then

SELF‐AFFIRMATION THEORY

229

self‐threats may be more threatening and self‐aYrmations more aYrming, when delivered in the presence of others.

VI. Implications for Interpersonal Relationships and Coping Because people experience many potential threats to their self‐integrity, an important question concerns the extent to which people use aYrmational strategies in everyday life. To what extent do people spontaneously rely on this mechanism involved in the psychological immune system? Relatedly, what are the real‐world analogs to the self‐aYrmation manipulations that have been featured in the laboratory? Research on close relationships and on coping and resiliency suggests some answers and points the way to exciting directions for future research.

A. INTERPERSONAL RELATIONSHIPS When people rank their values in terms of their personal importance in self‐ aYrmation studies, they consistently rank relationships with friends and family highly. Additionally, in a writing intervention study designed to elicit thoughts and feelings among patients about their illness—a study that provided an opportunity for self‐aYrming without an explicit prompt— people most frequently invoked the importance of their close relationships (Creswell et al., 2006). Thus, personal relationships seem to be an important aYrmational resource that people draw on in times of stress. Several studies in the close‐relationships literature have examined this idea in some depth. People can use their relationships and their romantic partners as aYrmational resources by emphasizing the love they receive from their partners (Kumashiro & Sedikides, 2005; Murray, Bellavia, Feeney, Holmes, & Rose, 2001; Swann, De La Ronde, & Hixon, 1994). People with high self‐esteem in particular seem to respond to threats to their self‐image by aYrming the integrity of their relationships (Murray et al., 1998). Threats can also exist within relationships, as for example when one partner outperforms the other. Tesser (1988) has found that upward social comparisons within a relationship could constitute a threat, particularly if the domain is of high personal importance. One adaptive way to respond to such a threat is to draw upon the relationship as an aYrmational resource (and perhaps also bask in the glory of one’s partner). In a series of studies, people in close relationships were able to use the relationship as an aYrmation after engaging in a potentially threatening upward social comparison

230

DAVID K. SHERMAN AND GEOFFREY L. COHEN

with their partner (Lockwood, Dolderman, Sadler, & Gerchak, 2004). When dealing with this potential within‐relationship threat, those people in close relationships aYrmed their relationships by emphasizing the warmth and kindness of the relationship. Many questions remain for researchers interested in the how personal relationships serve such aYrming functions. For example, the aYrmation framework may be useful in understanding the relationship between possessing multiple roles and depression. Sociological studies have found that people who have multiple roles, such as family and career, tend to have reduced stress and depression—at least as long as the number of roles remains below a manageable threshold (e.g., Gore & Mangione, 1983). For women in particular, it has been found that although work and family can be great strains on resources, those who participate in multiple roles have reduced levels of depression (Kandel, Davies, & Raveis, 1985). These studies raise questions of causality, as those who have better mental health may be more able to handle multiple roles without becoming depressed (Kandel et al., 1985). This caveat notwithstanding, the self‐aYrmation framework provides a possible explanation for the observed phenomenon. That is, perhaps multiple roles function as aYrmational resources allowing people to use one domain of life to aYrm the self when threatened in the other. Adopting the methodological approaches of close relationships researchers, such as use of daily diary reports and experience sampling, is an exciting direction for researchers interested in how people use relationships and other aYrmational resources to cope with threats and manage stress on a day‐to‐day level.

B. COPING AND RESILIENCE The assumption that people need to work through their grief, and to experience negative emotions after facing loss or trauma, has been challenged by work on resilience—the ability to maintain a stable equilibrium after a loss (Bonanno, 2004; Wortman & Silver, 1989). The failure to experience negative emotions after trauma was thought to be associated with delayed grief and greater psychological problems in the future; however, evidence does not support this intuition that people must experience grief in order to ‘‘move on’’ (Bonanno & Field, 2001). By contrast, those who express positive emotions after a loss are more likely to exhibit resilience in subsequent years (Bonanno & Keltner, 1997). What are some of the factors that promote resiliency in the face of loss? One factor seems to be the ability to self‐enhance (Bonanno, 2004)—to perceive the self as possessing more positive attributes than the average person (Taylor & Brown, 1988). In support of this claim, measured levels

SELF‐AFFIRMATION THEORY

231

of self‐enhancement were associated with better adjustment among Bosnian civilians after the civil war in Sarajevo (Bonanno, Field, Kovacevic, & Kaltman, 2002). In another study involving people near the World Trade Center at the time of the 9/11 terrorist attacks, self‐enhancement (as assessed by endorsement of such statements as ‘‘I am fully in control of my own fate,’’ and ‘‘I always know why I do things’’; see Paulhus, 1984) prospectively predicted a reduction in both depression and in Post Traumatic Stress Disorder symptoms (Bonanno, Rennicke, Dekel, & Rosen, 2005). It seems plausible that self‐aYrmation is another mechanism by which people remain resilient after trauma. Just as the ability to self‐enhance may help people to muster valuable resources to grapple with life challenges, the ability to aYrm alternative domains of self‐integrity in the face of threat may facilitate better coping with the trauma or loss (Taylor & Sherman, in press). AYrmation of other aspects of self‐integrity—such as religion, relationships, work, or hobbies—may enable people to deal constructively with the threats to perceived control, meaning, and significance that issue from real‐world loss and trauma. Future research is needed to identify whether and how people use self‐ aYrmation to cope with trauma and loss. Just as self‐aYrmation can lead to greater acceptance and less defensiveness across a wide range of threatening situations featured in laboratory studies, it will be exciting to explore whether and when it can lead to more adaptive coping among those who have experienced trauma. VII. Conclusions In her seminal review on motivated cognition, Kunda (1990) observed: [M]otivated illusions can be dangerous when they are used to guide behavior and decisions, especially in those cases in which objective reasoning could facilitate more adaptive behavior. For example, people who play down the seriousness of early symptoms of severe diseases such as skin cancer and people who see only weaknesses in research pointing to the dangers of drugs such as caVeine or of behaviors such as drunken driving may literally pay with their lives for their motivated reasoning. Hopefully, once the mechanisms producing such biases are fully understood, it will be possible to help people overcome them. (Kunda, 1990, p. 495–496)

Self-aYrmation theory provides a framework to understand and overcome such biases. At both the individual and collective levels, important domains of functioning—health, political decision-making, conflict, relationships, academic performance—call forth the motivation to defend the

232

DAVID K. SHERMAN AND GEOFFREY L. COHEN

self. People defensively distort, deny, and misrepresent reality in a manner that protects self-integrity. The cost of doing so, of course, is that they miss potential opportunities for learning and growth that, if acted upon, could otherwise increase their adaptiveness in the long term. However, in the face of daily threats, people can also protect self-integrity through the aVirmation of alternative sources of self-identity. Doing so helps them to accept experiences and information that, although threatening, hold important lessons for self-change. By illuminating the psychology of self-defense, the research reviewed in this chapter oVers practitioners, teachers, clinicians, mediators, and interventionists more generally theory-driven strategies for overcoming self-defense and encouraging self-improvement.

Acknowledgments We would like to thank Nancy Apfel, Patricia Brzustoski, David Creswell, Jennifer Crocker, David Dunning, Julio Garcia, Heejung Kim, Leif Nelson, Steve Spencer, Claude Steele, and Mark Zanna for their thoughtful comments on earlier versions of this paper.

References Abelson, R. (1986). Beliefs are like possessions. Journal for the Theory of Social Behavior, 16, 223–250. Abrams, D., & Hogg, M. (1988). Comments on the motivational status of self‐esteem in social identity and intergroup discrimination. European Journal of Social Psychology, 18, 317–334. Adams, G., Tormala, T. T., & O’Brien, L. T. (in press). The eVect of self‐aYrmation on perceptions of racism. Journal of Experimental Social Psychology. Allport, G. W. (1961). Pattern and growth in personality. New York: Holt, Rinehart and Wonston. Armor, D. A., & Taylor, S. E. (2002). When predictions fail: The dilemma of unrealistic optimism. In T. Gilovich, D. GriYn, and D. Kahneman (Eds.), Heuristics and biases: The psychology of intuitive judgment (pp. 334–347). New York: Cambridge University Press. Aronson, E. (1968). Dissonance theory: Progress and problems. In R. P. Abelson, E. Arson, W. J. McGuire, T. M. Newcomb, M. J. Rosenberg, and P. H. Tannembaum (Eds.), Theories of cognitive consistency: A sourcebook (pp. 5–27). Chicago: Rand McNally. Aronson, J., Blanton, H., & Cooper, J. (1995). From dissonance to disidentification: Selectivity in the self‐aYrmation process. Journal of Personality and Social Psychology, 68, 986–996. Aronson, J., Cohen, G. L., & Nail, P. R. (1999). Self‐aYrmation theory: An update and appraisal. In E. Harmon‐Jones and J. Mills (Eds.), Cognitive dissonance theory: Revival with revisions and controversies (pp. 127–147). Washington, DC: American Psychological Association. Aronson, J., Fried, C. B., & Good, C. (2002). Reducing the eVects of stereotype threat on African American college students by shaping theories of intelligence. Journal of Experimental Social Psychology, 38, 113–125.

SELF‐AFFIRMATION THEORY

233

Bargh, J. A., & Chartrand, T. L. (1999). The unbearable automaticity of being. American Psychologist, 54, 462–476. Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct eVects of trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71, 230–244. Baron, R. M., & Kenny, D. A. (1986). The moderator‐mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Baumeister, R. F., Heatherton, T. F., & Tice, D. M. (1993). When ego threats lead to self‐ regulation failure: Negative consequences of high self‐esteem. Journal of Personality and Social Psychology, 64, 141–156. Beauregard, K. S., & Dunning, D. (1998). Turning up the contrast: Self‐enhancement motives prompt egocentric contrast eVects in social judgments. Journal of Personality and Social Psychology, 74, 606–621. Ben‐Zeev, T., Fein, S., & Inzlicht, M. (2005). Arousal and stereotype threat. Journal of Experimental Social Psychology, 41, 174–181. Bettman, J. R., & Weitz, B. A. (1983). Attributions in the boardroom: Causal reasoning in corporate annual reports. Administrative Science Quarterly, 28, 165–183. Bipolarworld.net. (2005). AYrmation. Retrieved July 1, 2005 from http://www.bipolarworld.net/ Treatments/Self%20Help/self_aYrmation.htm Blanton, H., Cooper, J., Skurnik, I., & Aronson, J. (1997). When bad things happen to good feedback: Exacerbating the need for self‐justification with self‐aYrmations. Personality and Social Psychology Bulletin, 23, 684–692. Blanton, H., Pelham, B. W., DeHart, T., & Carvallo, M. (2001). Overconfidence as dissonance reduction. Journal of Experimental Social Psychology, 37, 373–385. Blascovich, J., Spencer, S. J., Quinn, D., & Steele, C. (2001). African Americans and high blood pressure: The role of stereotype threat. Psychological Science, 12, 225–229. Bonanno, G. A., & Field, N. P. (2001). Evaluating the delayed grief hypothesis across 5 years of bereavement. American Behavioral Scientist, 44, 798–816. Bonanno, G. A., & Keltner, D. (1997). Facial expressions of emotion and the course of conjugal bereavement. Journal of Abnormal Psychology, 106, 126–137. Bonanno, G. A., Field, N. P., Kovacevic, A., & Kaltman, S. (2002). Self‐enhancement as a buVer against extreme adversity: Civil war in Bosnia and traumatic loss in the United States. Personality and Social Psychology Bulletin, 28, 184–196. Bonanno, G. A., Rennicke, C., Dekel, S., & Rosen, J. (2005). Self‐enhancement among high‐ exposure survivors of the September 11th terrorist attack: Resilience or social maladjustment? Journal of Personality and Social Psychology, 88, 984–988. Bonanno, G. A. (2004). Loss, trauma, and human resilience: Have we underestimated the human capacity to thrive after extremely aversive events? American Psychologist, 59, 20–28. Boninger, D. S., Krosnick, J. A., & Berent, M. K. (1995). Origins of attitude importance: Self‐interest, social identification, and value relevance. Journal of Personality and Social Psychology, 68, 61–80. Brehm, J. (1956). Postdecision changes in the desirability of alternatives. Journal of Abnormal and Social Psychology, 52, 384–389. Brown, B. R. (2000). When moral entitlement leads to immoral acts. Doctoral dissertation, Stanford University. Dissertation Abstracts International: Sciences and Engineering, 61, 1132. Brown, J. D., & Smart, S. A. (1991). The self and social conduct: Linking self‐representations to prosocial behavior. Journal of Personality and Social Psychology, 60, 368–375.

234

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Cadinu, M. R., & Rothbart, M. (1996). Self‐anchoring and diVerentiation processes in the minimal group setting. Journal of Personality and Social Psychology, 70, 661–677. Campbell, J. D. (1990). Self‐esteem and clarity of the self‐concept. Journal of Personality and Social Psychology, 59, 538–549. Cohen, G. L. (2003). Party over policy: The dominating impact of group influence on political beliefs. Journal of Personality and Social Psychology, 85, 808–822. Cohen, G. L., Aronson, J., & Steele, C. M. (2000). When beliefs yield to evidence: Reducing biased evaluation by aYrming the self. Personality and Social Psychology Bulletin, 26, 1151–1164. Cohen, G. L., & Garcia, J. (2005). ‘‘I am us’’: Negative stereotypes as collective threats. Journal of Personality and Social Psychology, 89, 566–582. Cohen, G. L., Garcia, J., Apfel, N., & Master, A. (2006). A social‐psychological intervention to reduce the racial achievement gap in school. Manuscript under review. Cohen, G. L., Sherman, D. K., McGoey, M., Hsu, L., Bastardi, A., & Ross, L. D. (2005). Bridging the partisan divide: Self‐aYrmation and the reduction of ideological bias. Manuscript under review. Cohen, G. L., Steele, C. M., & Ross, L. D. (1999). The mentor’s dilemma: Providing critical feedback across the racial divide. Personality and Social Psychology Bulletin, 25, 1302–1318. Correll, J., Spencer, S. J., & Zanna, M. P. (2004). An aYrmed self and an open mind: Self‐ aYrmation and sensitivity to argument strength. Journal of Experimental Social Psychology, 40, 350–356. Colvin, C. R., Block, J., & Funder, D. C. (1995). Overly positive self‐evaluations and personality: Negative implications for mental health. Journal of Personality and Social Psychology, 68, 1152–1162. Creswell, J. D., Lam, S., Stanton, A. L., Taylor, S. E., Bower, J. E., & Sherman, D. K. (2006). Does self‐aYrmation, cognitive processing, or discovery of meaning explain cancer‐related health benefits of expressive writing? Manuscript under review. Creswell, J. D., Welch, W., Taylor, S. E., Sherman, D. K., Gruenewald, T., & Mann, T. (2005). AYrmation of personal values buVers neuroendocrine and psychological stress responses. Psychological Science, 16, 846–851. Crocker, J., & Major, B. (1989). Social stigma and self‐esteem: The self‐protective properties of stigma. Psychological Review, 96, 608–630. Crocker, J., & Wolfe, C. T. (2001). Contingencies of self‐worth. Psychological Review, 108, 593–623. Dal Cin, S., MacDonald, T. K., Fong, G. T., Zanna, M. P., & Elton‐Marshall, T. E. (in press). Remembering the message: Using a reminder cue to increase condom use following a safer sex intervention. Health Psychology. Dawson, E., Gilovich, T., & Regan, D. T. (2002). Motivated reasoning and performance on the Wason selection task. Personality and Social Psychology Bulletin, 28, 1379–1387. De Dreu, C. K. W., & Carnevale, P. J. (2003). Motivational bases of information processing and strategy in negotiation and social conflict. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 35, pp. 235–290). San Diego: Academic Press. Deaux, K. (1996). Social identification. In E. T. Higgins and A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 777–798). New York: Guilford. Deci, E. L., & Ryan, R. M. (1995). Human autonomy: The basis for true self‐esteem. In M. H. Kernis (Ed.), EYcacy, agency, and self‐esteem (pp. 31–49). New York: Plenum. Dickerson, S. S., & Kemeny, M. E. (2004). Acute stressors and cortisol responses: A theoretical integration and synthesis of laboratory research. Psychological Bulletin, 130, 355–391.

SELF‐AFFIRMATION THEORY

235

Dijksterhuis, A. (2004). Think diVerent: The merits of unconscious thought in preference development and decision making. Journal of Personality and Social Psychology, 87, 586–598. Ditto, P. H., & Lopez, D. F. (1992). Motivated skepticism: Use of diVerential decision criteria for preferred and nonpreferred conclusions. Journal of Personality and Social Psychology, 63, 569–584. Ditto, P. H., Scepansky, J. A., Munro, G. D., Apanovitch, A. M., & Lockhart, L. K. (1998). Motivated sensitivity to preference‐inconsistent information. Journal of Personality and Social Psychology, 75, 53–69. Dunning, D. (2003). The zealous self‐aYrmer: How and why the self lurks so pervasively behind social judgment. In S. J. Spencer, S. Fein, M. P. Zanna, and J. M. Olson (Eds.), Motivated social perception: The Ontario symposium (Vol. 9, pp. 45–72). Mahwah, New Jersey: Lawrence Erlbaum Associates. Dunning, D., & Beauregard, K. S. (2000). Regulating impressions of others to aYrm images of the self. Social Cognition, 18, 198–222. Dunning, D., & Cohen, G. L. (1992). Egocentric definitions of traits and abilities in social judgment. Journal of Personality and Social Psychology, 63, 341–355. Dunning, D., GriYn, D. W., Milojkovic, J. H., & Ross, L. (1990). The overconfidence eVect in social prediction. Journal of Personality and Social Psychology, 58, 568–592. Dunning, D., & Hayes, A. F. (1996). Evidence for egocentric comparison in social judgment. Journal of Personality and Social Psychology, 71, 213–229. Dunning, D., Leuenberger, A., & Sherman, D. A. (1995). A new look at motivated inference: Are self‐serving theories of success a product of motivational forces? Journal of Personality and Social Psychology, 69, 58–68. Dweck, C. S., & Leggett, E. (1988). A social cognitive approach to motivation and personality. Psychological Review, 95, 256–273. Ellsworth, P. C., & Ross, L. (1983). Public opinion and capital punishment: A close examination of the views of abolitionists and retentionists. Crime and Delinquency, 29, 116–169. Fein, S., & Spencer, S. J. (1997). Prejudice as self‐image maintenance: AYrming the self through derogating others. Journal of Personality and Social Psychology, 73, 31–44. Fein, S., Hoshino‐Browne, E., Davies, P. G., & Spencer, S. J. (2003). Self‐image maintenance goals and sociocultural norms in motivated social perception. In S. J. Spencer, S. Fein, M. P. Zanna, and J. M. Olson (Eds.), Motivated social perception: The Ontario symposium (Vol. 9, pp. 21–44). Mahwah, NJ: Lawrence Erlbaum Associates. Festinger, L. A. (1957). A theory of cognitive dissonance. Evanston, IL: Row, Peterson. Franken, A. (1992). I’m good enough, I’m smart enough, and doggone it, people like me! Daily aYrmations by Stuart Smalley. New York: Dell Publishing. Gaertner, L., Sedikides, C., & Graetz, K. (1999). In search of self‐definition: Motivational primacy of the individual self, motivational primacy of the collective self, or contextual primacy? Journal of Personality and Social Psychology, 76, 5–18. Gaertner, S. L., & Dovidio, J. F. (1986). The aversive form of racism. In J. F. Dovidio and S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 61–89). San Diego: Academic Press. Galinsky, A. D., Stone, J., & Cooper, J. (2000). The reinstatement of dissonance and psychological discomfort following failed aYrmations. European Journal of Social Psychology, 30, 123–147. Gerber, A., & Green, D. (1999). Misperceptions about perceptual bias. Annual Review of Political Science, 2, 189–210. Gibbons, F. X., Eggleston, T. J., & Benthin, A. (1997). Cognitive reactions to smoking relapse: The reciprocal relation of dissonance and self‐esteem. Journal of Personality and Social Psychology, 72, 184–195.

236

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Gilbert, D. T., Pinel, E. C., Wilson, T. D., Blumberg, S. J., & Wheatley, T. P. (1998). Immune neglect: A source of durability bias in aVective forecasting. Journal of Personality and Social Psychology, 75, 617–638. Glaser, R., & Kiecolt‐Glaser, J. (1994). Handbook of human stress and immunity. New York: Academic Press. Gore, S., & Mangione, T. W. (1983). Social roles, sex role, and psychological distress: Addditive and interactive models of sex diVerences. Journal of Health and Social Behavior, 24, 300–312. Gramzow, R. H., & Gaertner, L. (2005). Self‐esteem and favoritism toward novel in‐groups: The self as an evaluative base. Journal of Personality and Social Psychology, 88, 801–815. Greenberg, J., Pyszczynski, T., Solomon, S., Rosenblatt, A., Veeder, M., Kirkland, S., & Lyon, D. (1990). Evidence for terror management theory II: The eVects of mortality salience on reactions to those who threaten or bolster the cultural worldview. Journal of Personality and Social Psychology, 58, 308–318. Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory of self‐esteem and cultural worldviews: Empirical assessments and conceptual refinements. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 29, pp. 61–139). San Diego: Academic Press. Greenwald, A. G. (1980). The totalitarian ego: Fabrication and revision of personal history. American Psychologist, 35, 603–618. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual diVerences in implicit cognition: The implicit association test. Journal of Personality and Social Psychology, 74, 1464–1480. Harmon‐Jones, E., Simon, L., Greenberg, J., Pyszczynski, T., Solomon, S., & McGregor, H. (1997). Terror management theory and self‐esteem: Evidence that increased self‐esteem reduces mortality salience eVects. Journal of Personality and Social Psychology, 72, 24–36. Harris, P. R., & Napper, L. (2005). Self‐aYrmation and the biased processing of threatening health‐risk information. Personality and Social Psychology Bulletin, 31, 1250–1263. Hastorf, A., & Cantril, H. (1954). They saw a game: A case study. Journal of Abnormal and Social Psychology, 49, 129–134. Heine, S. J. (2005). Constructing good selves in Japan and North America. In R. M. Sorrentino, D. Cohen, J. M. Olson, and M. P. Zanna (Eds.), Cultural and social behavior: The Ontario Symposium (Vol. 10, pp. 95–116). Mahwah, NJ: Lawrence Erlbaum Associates. Heine, S. J., & Lehman, D. R. (1997). Culture, dissonance, and self‐aYrmation. Personality and Social Psychology Bulletin, 23, 389–400. Heine, S. J., Lehman, D. R., Markus, H., & Kitayama, S. (1999). Is there a universal need for positive self‐regard? Psychological Review, 106, 766–794. Hobfoll, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist, 44, 513–524. Hoshino‐Browne, E., Zanna, A. S., Spencer, S. J., & Zanna, M. P. (2004). Investigating attitudes cross‐culturally: A case of cognitive dissonance among East Asians and North Americans. In G. Haddock and G. R. Maio (Eds.), Contemporary perspectives on the psychology of attitudes (pp. 375–397). New York: Psychology Press. Hoshino‐Browne, E., Zanna, A. S., Spencer, S. J., Zanna, M. P., Kitayama, S., & Lackenbauer, S. (2005). On the cultural guises of cognitive dissonance: The case of Easterners and Westerners. Journal of Personality and Social Psychology, 89, 294–310. Jacks, J. Z., & O’Brien, M. E. (2004). Decreasing resistance by aYrming the self. In E. S. Knowles and J. A. Linn (Eds.), Resistance and persuasion (pp. 235–257). Mahwah, NJ: Lawrence Erlbaum Associates.

SELF‐AFFIRMATION THEORY

237

Jemmott, J. B., III, Ditto, P. H., & Croyle, R. T. (1986). Judging health status: EVects of perceived prevalence and personal relevance. Journal of Personality and Social Psychology, 50, 899–905. Jordan, C. H, Spencer, S. J., & Zanna, M. P. (2003). ‘‘I love me. . .I love me not:’’ Implicit self‐ esteem, explicit self‐esteem, and defensiveness. In S. J. Spencer, S. Fein, M. P. Zanna, and J. M. Olson (Eds.), Motivated social perception: The Ontario symposium (Vol. 9, pp. 117–145). Mahwah, NJ: Lawrence Erlbaum Associates. Jordan, C.,H, Spencer, S. J., Zanna, M. P., Hoshino‐Browne, E., & Correll, J. (2003). Secure and defensive high self‐esteem. Journal of Personality and Social Psychology, 85, 969–978. Kandel, D. B., Davies, M., & Raveis, V. H. (1985). The stressfulness of daily social roles for women: Marital, occupational, and household roles. Journal of Health and Social Behavior, 26, 64–78. Keough, K. A. (1998). When the self is at stake: Integrating the self into stress and physical health research (Doctoral dissertation, Stanford University, 1998). Dissertation Abstracts International, 58, 3959. Keough, K. A., Garcia, J., & Steele, C. M. (1998). Reducing stress and illness by aYrming the self. Unpublished manuscript, Stanford University. Kernis, M. H., & Paradise, A. W. (2002). Distinguishing between secure and fragile forms of high self‐esteem. In E. L. Deci and R. M. Ryan (Eds.), Handbook of self‐determination research (pp. 339–360). Rochester, NY: University of Rochester Press. Kim, H., & Markus, H. R. (1999). Deviance or uniqueness, harmony or conformity: A cultural analysis. Journal of Personality and Social Psychology, 77, 785–800. Kirschbaum, C., Pirke, K., & Hellhammer, D. H. (1993). The ‘Trier Social Stress Test’: A tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology, 28, 76–81. Kitayama, S., Snibbe, A. C., Markus, H. R., & Suzuki, T. (2004). Is there any ‘‘free’’ choice? Self and dissonance in two cultures. Psychological Science, 15, 527–533. Koole, S. L., Smeets, K., van Knippenberg, A., & Dijksterhuis, A. (1999). The cessation of rumination through self‐aYrmation. Journal of Personality and Social Psychology, 77, 111–125. Kruger, J. M., & Dunning, D. (1999). Unskilled and unaware of it: How diYculties in recognizing one’s own incompetence lead to inflated self‐assessments. Journal of Personality and Social Psychology, 77, 1121–1134. Kumashiro, M., & Sedikides, C. (2005). Taking on board liability‐focused information: Close positive relationships as a self‐bolstering resource. Psychological Science, 16, 732–739. Kunda, Z. (1987). Motivated inference: Self‐serving generation and evaluation of causal theories. Journal of Personality and Social Psychology, 53, 636–647. Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480–498. Landau, M. J., Solomon, S., Greenberg, J., Cohen, F., Pyszczynski, T., Arndt, J., Miller, C. H., Ogilvie, D. M., & Cook, A. (2004). Deliver us from evil: The eVects of mortality salience and reminders of 9/11 on support for president George W. Bush. Personality and Social Psychology Bulletin, 30, 1136–1150. Langer, E. J. (1975). The illusion of control. Journal of Personality and Social Psychology, 32, 311–328. Lau, R. R., & Russell, D. R. (1980). Attributions in the sports pages. Journal of Personality and Social Psychology, 39, 29–38. Lazarus, R. S. (1993). From psychological stress to the emotions: A history of changing outlooks. Annual Review of Psychology, 44, 1–21. Leary, M. R., & Baumeister, R. F. (2000). The nature and function of self‐esteem: Sociometer theory. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 32, pp. 1–62). San Diego: Academic Press.

238

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Lerner, M. J. (1980). The belief in a just world: A fundamental delusion. New York: Plenum Press. Liberman, A., & Chaiken, S. (1992). Defensive processing of personally relevant health messages. Personality and Social Psychology Bulletin, 18, 669–679. Liu, T. J., & Steele, C. M. (1986). Attributional analysis as self‐aYrmation. Journal of Personality and Social Psychology, 51, 531–540. Lockwood, P., Dolderman, D., Sadler, P., & Gerchak, E. (2004). Feeling better about doing worse: Social comparisons within romantic relationships. Journal of Personality and Social Psychology, 87, 91–103. Lockwood, P., & Kunda, Z. (1997). Superstars and me: Predicting the impact of role models on the self. Journal of Personality and Social Psychology, 73, 91–103. Lord, C., Ross, L., & Lepper, M. R. (1979). Biased assimilation and attitude polarization: The eVects of prior theories on subsequently considered evidence. Journal of Personality and Social Psychology, 37, 2098–2109. Mackie, D. M., & Worth, L. T. (1989). Processing deficits and the mediation of positive aVect in persuasion. Journal of Personality and Social Psychology, 57, 27–40. Major, B., Spencer, S. J., Schmader, T., Wolfe, C., & Crocker, J. (1998). Coping with negative stereotypes about intellectual performance: The role of psychological disengagement. Personality and Social Psychology Bulletin, 24, 34–50. Markus, H. (1977). Self‐schemata and processing information about the self. Journal of Personality and Social Psychology, 35, 63–78. Markus, H., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224–253. Martens, A., Johns, M., Greenberg, J., & Schimel, J. (2006). Combating stereotype threat: The eVect of self‐aYrmation on women’s intellectual performance. Journal of Experimental Social Psychology, 42, 236–243. Martin, L. L., & Tesser, A. (1996). Some ruminative thoughts. In R. S. Wyer (Ed.), Advances in social cognition (Vol. 9, pp. 1–47). Hillsdale, NJ: Erlbaum. Matz, D. C., & Wood, W. (2005). Cognitive dissonance in groups: The consequences of disagreement. Journal of Personality and Social Psychology, 88, 22–37. McEwen, B. S. (1998). Protective and damaging eVects of stress mediators. New England Journal of Medicine, 338, 171–179. McGregor, I., & Marigold, D. C. (2003). Defensive zeal and the uncertain self: What makes you so sure? Journal of Personality and Social Psychology, 85, 838–852. McQueen, A., & Klein, W. (2005). Experimental manipulations of self‐aYrmation: A systematic review. Manuscript under review. Miller, D. T., & Ross, M. (1975). Self‐serving biases in the attribution of causality: Fact or fiction? Psychological Bulletin, 82, 213–225. Monin, B., & Miller, D. T. (2001). Moral credentials and the expression of prejudice. Journal of Personality and Social Psychology, 81, 33–43. Monteith, M. J., & Voils, C. I. (1998). Proneness to prejudiced responses: Toward understanding the authenticity of self‐reported discrepancies. Journal of Personality and Social Psychology, 75, 901–916. Morris, K. A., & Swann, W. B. (1996). Denial and the AIDS crisis: On wishing away the threat of AIDS. In S. Oskamp and S. Thompson (Eds.), Safer sex in the ‘90s: Understanding and preventing HIV risk behavior (pp. 57–79). New York: Sage. Munro, G. D., & Ditto, P. H. (1997). Biased assimilation, attitude polarization, and aVect in reactions to stereotype‐relevant scientific information. Personality and Social Psychology Bulletin, 23, 636–653. Murphy, G. (1947). Personality. New York: Harper and Brothers.

SELF‐AFFIRMATION THEORY

239

Murray, S. L., Holmes, J. G., MacDonald, G., & Ellsworth, P. (1998). Through the looking glass darkly? When self‐doubts turn into relationship insecurities. Journal of Personality and Social Psychology, 75, 1459–1480. Murray, S. L., Bellavia, G., Feeney, B., Holmes, J. G., & Rose, P. (2001). The contingencies of interpersonal acceptance: When romantic relationships function as a self‐aYrmational resource. Motivation and Emotion, 25, 163–189. Nail, P. R., Misak, J. E., & Davis, R. M. (2004). Self‐aYrmation versus self‐consistency: A comparison of two competing self‐theories of dissonance phenomena. Personality and Individual DiVerences, 36, 1893–1905. Nisbett, R. E., & Ross, L. D. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood CliVs, NJ: Prentice Hall. Nisbett, R. E., & Cohen, D. (1996). Culture of honor: The psychology of violence in the South. Boulder, CO: Westview Press. Nolen‐Hoeksema, S., & Morrow, J. (1991). A prospective study of depression and posttraumatic stress symptoms after a natural disaster: The 1989 Loma Prieta earthquake. Journal of Personality and Social Psychology, 61, 115–121. Norton, M. I., Monin, B., Cooper, J., & Hogg, M. A. (2003). Vicarious dissonance: Attitude change from the inconsistency of others. Journal of Personality and Social Psychology, 85, 47–62. O’Brien, L. T., & Crandall, C. S. (2003). Stereotype threat and arousal: EVects on women’s math performance. Personality and Social Psychology Bulletin, 29, 782–789. Otten, S. (2002). ‘‘Me and us’’ or ‘‘us and them’’? The self as heuristic for defining minimal ingroups. European Review of Social Psychology, 13, 1–33. Paulhus, D. L. (1984). Two‐component models of socially desirable responding. Journal of Personality and Social Psychology, 46, 598–609. Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood method of persuasion. In L. Berkowitz (Ed.), Advances in experimental social psychology. (Vol. 19, pp. 123–205). New York: Academic Press. Petty, R. E., & Wegener, D. T. (1998). Attitude change: Multiple roles for persuasion variables. In D. T. Gilbert, S. T. Fiske, and G. Lindzey (Eds.), The handbook of social psychology (4th ed., pp. 323–390). New York: McGraw‐Hill. Pyszczynski, T., & Greenberg, J. (1987). Toward and integration of cognitive and motivational perspectives on social inference: A biased hypothesis‐testing model. In L. Berkowitz (Ed.), Advances in experimental social psycology (Vol. 20, pp. 297–340). New York: Academic Press. Pyszczynski, T., Greenberg, J., Solomon, S., Arndt, J., & Schimel, J. (2004). Why do people need self‐esteem? A theoretical and empirical review. Psychological Bulletin, 130, 435–468. Pyszczynski, T., Solomon, S., & Greenberg, J. (2003). In the wake of 9/11: The psychology of terror. Washington, DC: American Psychological Association. Raghunathan, R., & Trope, Y. (2002). Walking the tightrope between feeling good and being accurate: Mood as a resource in processing persuasive messages. Journal of Personality and Social Psychology, 83, 510–525. Reed, M. B., & Aspinwall, L. G. (1998). Self‐aYrmation reduces biased processing of health‐risk information. Motivation and Emotion, 22, 99–132. Rosenberg, M. (1965). Society and the adolescent self‐image. Princeton, NJ: Princeton University Press. Ross, L., & Ward, A. (1995). Psychological barriers to dispute resolution. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 27, pp. 255–304). San Diego: Academic Press. Schimel, J., Arndt, J., Pyszczynski, T., & Greenberg, J. (2001). Being accepted for who we are: Evidence that social validation of the intrinsic self reduces general defensiveness. Journal of Personality and Social Psychology, 80, 35–52.

240

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Schimel, J., Arndt, J., Banko, K. M., & Cook, A. (2004). Not all self‐aYrmations were created equal: The cognitive and social benefits of aYrming the intrinsic (vs. extrinsic) self. Social Cognition, 22, 75–99. Schmeichel, B. J., & Martens, A. (2005). Self‐aYrmation and mortality salience: AYrming values reduces worldview defense and death‐thought accessibility. Personality and Social Psychology Bulletin, 31, 658–667. Sears, C. L., & Funk, C. L. (1991). Are we reaching undergraduates? A survey of course oVering in political psychology. Political Psychology, 12, 559–572. Seery, M., Blascovich, J., Weisbuch, M., & Vick, S. B. (2004). The relationship between self‐ esteem level, self‐esteem stability, and cardiovascular reactions to performance feedback. Journal of Personality and Social Psychology, 87, 133–145. Sherman, D. A. K., Nelson, L. D., & Steele, C. M. (2000). Do messages about health risks threaten the self? Increasing the acceptance of threatening health messages via self‐ aYrmation. Personality and Social Psychology Bulletin, 26, 1046–1058. Sherman, D. K., & Cohen, G. L. (2002). Accepting threatening information: Self‐aYrmation and the reduction of defensive biases. Current Directions in Psychological Science, 11, 119–123. Sherman, D. K., Cohen, G. L., Nelson, L. D., Bastardi, A., Nussbaum, A. D., & Ross, L. (2005). AYrmed yet unaware: The role of awareness in the process of self‐aYrmation. Manuscript under review. Sherman, D. K., & Kim, H. S. (2005). Is there an ‘‘I’’ in ‘‘team’’? The role of the self in group‐ serving judgments. Journal of Personality and Social Psychology, 88, 108–120. Sherman, D. K., Kinias, Z., Major, B., Kim, H. S., & Prenovost, M. (2005). Group identification, group aYrmation, and group‐serving judgments. Manuscript in preparation. Sherman, D. K., Nelson, L. D., & Ross, L. D. (2003). Naı¨ve realism and aYrmative action: Adversaries are more similar than they think. Basic and Applied Social Psychology, 25, 275–290. Shrira, I., & Martin, L. L. (2005). Stereotyping, self‐aYrmation, and the cerebral hemispheres. Personality and Social Psychology Bulletin, 31, 846–856. Simon, L., Greenberg, J., & Brehm, J. (1995). Trivialization: The forgotten mode of dissonance reduction. Journal of Personality and Social Psychology, 68, 247–260. Smith, E. R., Coats, S., & Walling, D. (1999). Overlapping mental representations of self, in‐ group, and partner: Further response time evidence and a connectionist model. Personality and Social Psychology Bulletin, 25, 873–882. Smith, E. R., & Henry, S. (1996). An in‐group becomes part of the self: Response time evidence. Personality and Social Psychology Bulletin, 22, 635–642. Solomon, S., Greenberg, J., & Pyszczynski, T. (1991). A terror management theory of social behavior: The psychological functions of self‐esteem and cultural worldviews. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, pp. 93–159). New York: Academic Press. Spencer, S. J., Fein, S., & Lomore, C. D. (2001). Maintaining one’s self‐image vis‐a‐vis others: The role of self‐aYrmation in the social evaluation of the self. Motivation and Emotion, 25, 41–65. Spencer, S. J., Josephs, R. A., & Steele, C. M. (1993). Low self‐esteem: The uphill struggle for self‐integrity. In R. Baumeister (Ed.), Self‐esteem: The puzzle of low self‐regard (pp. 21–36). New York: Plenum Press. Spencer, S. J., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and women’s math performance. Journal of Experimental Social Psychology, 35, 4–28. Spencer, S. J., Zanna, M. P., & Fong, G. T. (2005). Establishing a causal chain: Why experiments are often more eVective than mediational analyses in examining psychological processes. Journal of Personality and Social Psychology, 89, 845–851. Stanton, A. L., DanoV‐Burg, S., Sworowski, L. A., Collins, C. A., Branstetter, A., Rodriguez‐ Hanley, A., Kirk, S. B., & Austenfeld, J. L. (2002). Randomized, controlled trial of written

SELF‐AFFIRMATION THEORY

241

emotional expression and benefit finding in breast cancer patients. Journal of Clinical Oncology, 20, 4160–4168. Steele, C. M. (1988). The psychology of self‐aYrmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 261–302). New York: Academic Press. Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance. American Psychologist, 52, 613–629. Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69, 797–811. Steele, C. M., Hopp, H., & Gonzales, J. (1986). Dissonance and the lab coat: Self‐aVirmation and the free‐choice paradigm. Unpublished manuscript, University of Washington. Steele, C. M., & Liu, T. J. (1983). Dissonance processes as self‐aYrmation. Journal of Personality and Social Psychology, 45, 5–19. Steele, C. M., Spencer, S. J., & Aronson, J. (2002). Contending with group image: The psychology of stereotype and social identity threat. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 379–440). New York: Academic Press. Steele, C. M., Spencer, S. J., & Lynch, M. (1993). Self‐image resilience and dissonance: The role of aYrmational resources. Journal of Personality and Social Psychology, 64, 885–896. Stone, J., & Cooper, J. (2001). A self‐standards model of cognitive dissonance. Journal of Experimental Social Psychology, 37, 228–243. Stone, J., & Cooper, J. (2003). The eVect of self‐attribute relevance on how self‐esteem moderates dissonance processes. Journal of Experimental Social Psychology, 39, 508–515. Swann, W. B., Jr., De La Ronde, C., & Hixon, J. G. (1994). Authenticity and positivity strivings in marriage and courtship. Journal of Personality and Social Psychology, 66, 857–869. Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel and W. G. Austin (Eds.), Psychology of intergroup relations (pp. 33–48). Chicago: Nelson‐Hall. Taylor, S. E. (1983). Adjustment to threatening events: A theory of cognitive adaptation. American Psychologist, 38, 1161–1173. Taylor, S. E., & Brown, J. D. (1988). Illusion and well‐being: A social psychological perspective on mental health. Psychological Bulletin, 103, 193–210. Taylor, S. E., Lerner, J., Sherman, D. K., Sage, R., & McDowell, N. (2003). Are self‐enhancing cognitions associated with healthy or unhealthy biological profiles? Journal of Personality and Social Psychology, 85, 605–615. Taylor, S. E., & Lobel, M. (1989). Social comparison activity under threat: Downward evaluation and upward contacts. Psychological Review, 96, 569–575. Taylor, S. E., & Sherman, D. K. (in press). Self‐enhancement and self‐aYrmation: The eVects of positive self‐thoughts on motivation and behavior. Chapter to appear in J. Shah & W. Gardner (Eds.) Handbook of motivation science. New York: Guilford Publications. Taylor, S. E., Wood, J. V., & Lichtman, R. R. (1983). It could be worse: Selective evaluation as a response to victimization. Journal of Social Issues, 39, 19–40. Tesser, A. (1988). Toward a self‐evaluation maintenance model of social behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 181–227). San Diego, CA: Academic Press. Tesser, A. (2000). On the confluence of self‐esteem maintenance mechanisms. Personality and Social Psychology Review, 4, 290–299. Tesser, A., & Cornell, D. P. (1991). On the confluence of self‐processes. Journal of Experimental Social Psychology, 27, 501–526. Tesser, A., Crepaz, N., Collins, J. C., Cornell, D., & Beach, S. R. H. (2000). Confluence of self defense mechanisms: On integrating the self‐zoo. Personality and Social Psychology Bulletin, 26, 1476–1489.

242

DAVID K. SHERMAN AND GEOFFREY L. COHEN

Tesser, A., Martin, L. L., & Cornell, D. P. (1996). On the substitutability of self‐protective mechanisms. In J. A. Bargh (Ed.), The psychology of action: Linking cognition and motivation to behavior (pp. 48–68). New York: Guilford Press. Tesser, A., & Smith, J. (1980). Some eVects of task relevance and friendship on helping: You don’t always help the one you like. Journal of Experimental Social Psychology, 16, 582–590. Thibodeau, R., & Aronson, E. (1992). Taking a closer look: Reasserting the role of the self‐ concept in dissonance theory. Personality and Social Psychology Bulletin, 18, 591–602. Triandis, H. C. (1989). The self and social behavior in diVering cultural contexts. Psychological Review, 96, 506–520. Trivers, R. (2000). The elements of a scientific theory of self‐deception. Annals of the New York Academy of Sciences, 907, 114–131. Uhlmann, E., & Cohen, G. L. (2005a). Constructed criteria: Redefining merit to justify discrimination. Psychological Science, 16, 474–480. Uhlmann, E., & Cohen, G. L. (2005b). I think it, therefore it’s true: Self‐perceived objectivity licenses the expression of intergroup bias. Manuscript under review. Weinstein, N. D., & Klein, W. M. (1995). Resistance of personal risk perceptions to debiasing interventions. Health Psychology, 14, 132–140. Wert, S. R. (2004). Negative gossip as a response to threatened social self‐esteem. Doctoral dissertation, Yale University. Dissertation Abstracts International: Section B: The Sciences & Engineering, 65(3‐B), 1601. Wert, S. R., & Salovey, P. (2004). A social comparison account of gossip. Review of General Psychology, 8, 122–137. Wiesenfeld, B. M., Brockner, J., & Martin, C. (1999). A self‐aYrmation analysis of survivors’ reactions to unfair organizational downsizings. Journal of Experimental Social Psychology, 35, 441–460. Wills, T. A. (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90, 245–271. Wilson, T. D. (2002). Strangers to ourselves: Discovering the adaptive unconscious. Cambridge, MA: Harvard University Press. Winkler, J. D., & Taylor, S. E. (1979). Preference, expectations, and attributional bias: Two field studies. Journal of Applied Social Psychology, 9, 183–197. Wortman, C. B., & Silver, R. C. (1989). The myths of coping with loss. Journal of Consulting and Clinical Psychology, 57, 349–357. Zarate, M. A., & Garza, A. A. (2002). In‐group distinctiveness and self‐aYrmation as dual components of prejudice reduction. Self and Identity, 1, 235–249. Zuwerink, J. R., & Devine, P. G. (1996). Attitude importance and resistance to persuasion: It’s not just the thought that counts. Journal of Personality and Social Psychology, 70, 931–944.

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

Charles Stangor Scott P. Leary

I. Introduction The beliefs that human beings hold about social groups and the members of those groups (including their ingroup), represent among the most basic of all human knowledge. These beliefs help create our self‐concept and define our relationships with others. They influence our career aspirations and task performance (Stangor & Sechrist, 1998) and impact our mental and physical health (Stangor, 2003). Intergroup beliefs may also lead to local and worldwide hostility and violence and produce gender, racial, and class disparities (Staub, 1989). But intergroup beliefs also lead to intergroup cooperation and attempts to increase social harmony, tolerance, and equality (Moskowitz, Wasel, Schaal, & Gollwitzer, 1999). Because they are such a critical aspect of human existence, intergroup beliefs and attitudes have been actively studied by social psychologists, developmental psychologists, educational psychologists, sociologists, politicians, historians, and others. From the point of view of social psychology, intergroup beliefs are important not only because of their social relevance, but also because they involve a wide variety of basic psychological processes, including attitudes, conformity, person perception, heuristic processing, perceptual biases, and self‐evaluation. Intergroup attitudes are generally considered within the social psychology literature as stereotypes and prejudice. Over the past 50 years we can say that there have been two basic streams of research in this topic. One line has primarily addressed individual social cognition, focused on the basic processes 243 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 38 DOI: 10.1016/S0065-2601(06)38005-7

Copyright 2006, Elsevier Inc. All rights reserved. 0065-2601/06 $35.00

244

CHARLES STANGOR AND SCOTT P. LEARY

of social categorization and social judgment. This research has studied the development, maintenance, and change of stereotypes and prejudice through information processing biases that result from perception, and particularly, misperception of the social environment (Hamilton & Sherman, 1994; Stangor & Lange, 1994). These approaches have generally addressed the issue of stereotype development and change by studying how individuals directly perceive and interpret the social categories and the behaviors of others (Hewstone, 1996). This research includes such phenomena as illusory correlations (Hamilton, 1981), memory biases (Stangor & McMillan, 1992), categorization and recategorization (Gaertner, Dovidio, Anastasio, Bachman, & Rust, 1993), and the transmission of beliefs through the media (Ruscher, 2001). We can call these approaches bottom‐up, because they involve extracting meaning from direct observations of others. These approaches have generally focused on the cognitive, rather than the aVective or motivational components of group perceptions, although motivations are not irrelevant to the outcome of these processes (Bodenhausen, Kramer, & Sasser, 1994; Sinclair & Kunda, 2000). Indeed, although primarily cognitive in outlook, this approach is based on the assumption that intergroup beliefs serve the motivational goal of providing cognitive economy by helping provide clear, simple, and potentially useful information about group diVerences (Allport, 1954; Diehl & Jonas, 1991; Ford & Stangor, 1992; Oakes, Turner, & Haslam, 1991). A second stream of theorizing can be considered more top‐down in orientation, in the sense that it has investigated the function of stereotypes and prejudice for individuals and their social communication and exchange (Stangor & Schaller, 1996). In this literature the self‐concept, including not only its cognitive but also its aVective and motivational components, is perceived as an integral part of intergroup perceptions. Literature related to this latter stream includes, for instance, the concept of the Authoritarian Personality, which was based on the notion that prejudice served a psychodynamic function for individuals and was created by early family experiences (Adorno, Frenkel‐Brunswik, Levinson, & Sanford, 1950; Harding, Proshansky, Kutner, & Chein, 1969; Snyder & Miene, 1994). Other approaches within this stream include the suggestion that stereotypes function to preserve the social system or status quo (Jost & Banaji, 1994), that stereotypes develop to justify prejudice and discrimination (Allport, 1954), and that group beliefs allow enhancing the self and the ingroup (Allport, 1954; Tajfel & Turner, 2004; Wills, 1981). Still another line that exemplifies the top‐down approach is based on the idea that stereotypes and prejudice are the result of social norms (Allport, 1954; Crandall & Stangor, 2005). This approach proposes that stereotypes and prejudice are developed, maintained, and changed as a result of social communication and the perception of the beliefs of others, and in service of

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

245

the underlying social goal of conformity and acceptance. In short, according to this perspective, stereotypes and prejudice are learned more from other ingroup members than from the observation of outgroup members. In this chapter we review current literature related to the top‐down, social approaches to intergroup beliefs, with an emphasis on data from our lab. Our goal is to consider intergroup beliefs in their social context—as social and psychologically meaningful constructs for individuals. We will consider these processes by focusing on the acquisition and social sharing of stereotypes and prejudice. Our review is based on the expectation that individuals hold beliefs about social groups because these beliefs are an integral part of our social experience, they represent social norms, and allow us to meet fundamental social goals. It is our hope that such an analysis will be informative about intergroup attitudes in general and also help better tie these important concepts to social behavior.

II. What are Stereotypes and Prejudice? It is perhaps useful to clarify the terms stereotyping and prejudice. These terms are in a sense unfortunate because they are not that precise. One problem is that the definitions of stereotypes and stereotyping have changed over the years and also vary across research domain and even researcher. This variation creates a great deal of conceptual ambiguity. For example, historical definitions of stereotypes required that they be overgeneralized, inaccurate, and rigid (Allport, 1954), whereas modern definitions of stereotypes are generally broader, encompassing the beliefs—generally traits—that we hold about social groups and about members of those groups, regardless of the accuracy or rigidity with which these opinions are held (Hamilton & Sherman, 1994). And definitions of prejudice have similarly changed—at one point prejudice was considered as an attitude that was negative, unrealistic, incorrect, and unjustified, whereas today prejudice is more likely to be considered as a general negative evaluation toward outgroups. These changes in conceptualization are problematic because they create confusion about the meaning of the constructs—for instance prejudice can be considered by some as a natural cognitive process of favoring some groups over others and by others to be an irrational hatred of all members of an outgroup. One potentially useful approach to simplifying these constructs is to consider intergroup beliefs the way we consider other social constructs—in terms of an evaluation component and a belief component. Both evaluations (that ingroups are perceived as better than outgroups or members of a

246

CHARLES STANGOR AND SCOTT P. LEARY

specific ingroup or outgroup are disliked) as well as beliefs about the content of the categories (that boys are more likely to play with trucks than girls or that Italians are ‘‘expressive’’) are important and form the basis of theory and research. Furthermore, the two components are related to each other. For instance, when young children first recognize their membership in a group (knowledge), they quickly begin to believe that their ingroup has more positive characteristics than the outgroup (evaluation). In adults, the relationship between knowledge (stereotype) and evaluation (attitude) is weak, but present (Dovidio, Brigham, Johnson, & Gaertner, 1996). Another unfortunate aspect of the terms stereotyping and prejudice is that, with few exceptions, they apply primarily to beliefs about outgroups. Although the idea of the self‐stereotype has been used in some cases to refer to knowledge about the characteristics of the ingroup, and although group cohesion refers essentially to an evaluation of the ingroup (Hogg, 1992; Hogg, Hardie, & Reynolds, 1995), the focus of this literature is in large part on beliefs about outgroups (although the social identity literature has studied ingroup beliefs as an outcome of social identity—Tajfel & Turner, 2004; Turner, 1975). This has limited the study of stereotypes and prejudice as social phenomena for both empirical and theoretical reasons. One problem is that some measures that are designed to measure evaluations of the outgroup (e.g., stereotypes or prejudice) actually conflate ingroup and outgroup evaluations. For instance, research using the Implicit Association Test has demonstrated that individuals associate positive terms more with ingroups than outgroups and associate negative terms more with outgroups than ingroups. And this diVerence has been called a measure of prejudice (Rudman, Greenwald, Mellott, & Schwartz, 1999). However, it is not always clear from these findings whether the results indicate positive ingroup evaluation or negative outgroup evaluation. People may be classified as prejudiced if they assign positive characteristics with the ingroup, even if they do not associate negative characteristics with outgroups. A similar problem has been noted in measures of prejudice in children in which children are asked to choose, for instance, which person (e.g., a black or a white child) ‘‘plays fair’’ or ‘‘is bossy’’ (Doyle & Aboud, 1995; Williams, Best, Boswell, Mattson, & Graves, 1975), or which child they prefer (Clark & Clark, 1947). Indeed, most measures of racial and gender ‘‘prejudice’’ used in the child development literature take this format. The problem is that the choice may indicate either ingroup preference or outgroup devaluation, but it is not clear which (Cameron, Alvarez, Ruble, & Fuligni, 2001). DiVerentiating ingroup and outgroup attitudes, treating them as separate constructs, and considering the psychological importance of each is theoretically important (Brewer, 1999; Cameron et al., 2001) and this distinction must be part of any full theory of group beliefs. For one, recent research

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

247

has demonstrated that ingroup beliefs have a major impact on cognitions, aVect, and even behavior—for instance the literature on expectancies and stereotype threat (Steele & Aronson, 1995; Wheeler & Petty, 2001). Furthermore, a number of lines of research have found that group evaluations reflect primarily ingroup favoritism rather than outgroup devaluation (Brewer, 1999). That is, we tend to assign more positive than negative traits to our own groups than to other groups, tend to give rewards to ingroups rather than outgroups, and are more helpful to ingroups rather than outgroups. However, the psychological factors that create ingroup favoritism may be quite diVerent than those that lead to outgroup derogation. Our analysis of social motivations furthers this important distinction. In sum, evaluations and beliefs about ingroups and outgroups represent part of individuals’ most important social knowledge. Outgroup knowledge and evaluations provide information about the potential approachability and likely outcomes of social interaction with others and thus guide social interactions. Ingroup beliefs provide normative information about appropriate attitudes and behaviors, inform the self‐concept about how the individual is meeting ingroup standards, and help the individual develop ties to other ingroup members. As such, stereotypes and prejudice—rather than being harmful, inaccurate, and something we should try to remove from human thinking—can be conceptualized as a basic and essential aspect of human perception (Lee, Jussim, & McCauley, 1995). Our approach to understanding the acquisition of intergroup knowledge and intergroup attitudes considers them as the outcome of basic social processes. In the first section of this chapter we consider these beliefs as outcomes of the attempts of the individual to meet important social goals. In the second section we consider the processes of social sharing and transmission of information more broadly.

III. Social Motivations and Intergroup Beliefs One approach we have recently taken in our research is to assume that intergroup beliefs have functional significance for individuals. If this is the case, then we may be able to specify the important social motivations of individuals, following up with the prediction that these motivations will help determine our intergroup beliefs. But what are the important social goals? Although diVerent researchers have specified these goals somewhat diVerently, there is nevertheless at least some overlap in the goals they have specified. For instance, in their social psychology textbook, Smith and Mackie (2000) considered the important social goals to be construction of

248

CHARLES STANGOR AND SCOTT P. LEARY

eality, striving for mastery, seeking connectedness, valuing me and mine, and conservatism, whereas Susan Fiske (2003) has argued that the fundamental social motivations are belonging, understanding, controlling, enhancing self, and trusting others. Deci and Ryan (1995) also share a similar approach. We can see that although the specifics diVer, there is still some consistency across the diVerent approaches. In this section of the chapter we will focus on two social motivations that we have found to be important in our prior research—we label them the motivations of Protecting the Self and the Ingroup and Enhancing Social Harmony. These motivations overlap to a great extent with the basic social motivations that have been conceptualized as important by others (viz. ‘‘Valuing Me and Mine’’ and ‘‘Enhancing Self ’’ versus ‘‘Seeking Connectedness’’ and ‘‘Trusting Others’’), but, as we will see, they also incorporate the goal of ‘‘Conservatism.’’

A. PROTECTING AND ENHANCING THE SELF AND THE INGROUP The most basic and primary tendency of all living organisms, and what we will refer to as the protecting and enhancing motivation, is the desire to maintain and enhance one’s own life and the lives of important others. Humans have a basic motivation to find food and water, obtain adequate shelter, and protect themselves from danger. Doing so is necessary from an evolutionary perspective, because we can only pass our genes along to our children if we remain alive to have them and to nurture them. In this we are of course not alone ... organisms of all types love life and strive to maintain it. The idea that people strive to maintain their own lives does not, however, mean that our own life is the only focus of this motivation. There is also a natural tendency for organisms to attempt to enhance the survival of their genes—and this means that we should also be invested in protecting and enhancing our kin—those who are genetically related to us. Therefore, the protecting and enhancing motivation also includes the natural tendency to reproduce and to care for our children and relatives. Furthermore, humans also have the motivation to protect and enhance those who are similar and familiar to us, and with whom we have close social connections, even if these people do not actually share our genes. We can say that the motivation to protect the self and kin spreads out to our immediate social situation—that is, to the members of the ingroup (Brewer, 1999). The perception of similarity (or dissimilarity) between people, the creation of ingroups and outgroups, and the relative preference for ingroup over outgroup represent basic human processes. Both children and adults show

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

249

ingroup favoritism, even when the basis of categorization is minimal and completely arbitrary, for instance wearing diVerent colored tee‐shirts (Bigler, Jones, & Lobliner, 1997), or even random assignment to groups (Locksley, Ortiz, & Hepburn, 1980), and the ingroup‐enhancing tendency has been observed on a wide variety of measures, ranging from point assignment (Tajfel, 1981) to memory (Howard & Rothbart, 1980). The importance of enhancing the self and the ingroup is also part of other theoretical approaches (Allport, 1954; Tajfel & Turner, 2004; Wills, 1981). For instance, the social identity approach is based on the idea that individuals enjoy being part of groups and receive positive regard from their group memberships and that prejudice serves a motivational—that is, an esteem enhancing—function (Hogg & Abrams, 1990). According to social identity theory, individuals prefer to be members of valued ingroups and will inflate perceptions of those groups. In short, the fact that people are motivated to protect themselves and others and to respond positively to ingroup members, coupled with research demonstrating the power of the ingroup‐enhancing motivation, supports the idea that ingroup enhancement is a primary human motivation. Brewer (1999) has proposed that the tendency to categorize into ingroups and outgroups and resulting ingroup favoritism is likely a universal aspect of human beings. She has argued that ingroup liking is psychologically primary, in the sense that familiarity, attachment, and preference for one’s ingroups come prior to development of attitudes toward specific outgroups and have a stronger impact on thought and behavior.

B. AFFILIATION AND SOCIAL HARMONY However, and fortunately, although it is primary, ingroup enhancement is not the only important human social motivation. Rather, a major part of human experience is aYliation more broadly—with other members of ingroups but also with members of outgroups. These ideas have been proposed by others as, for instance, the ‘‘seeking connectedness’’ motivation (Smith & Mackie, 2000) and the ‘‘belonging’’ (Fiske, 2003) motivation (Deci & Ryan, 1995). Humans have a natural inclination to socialize, and this tendency provides substantial benefits (Mackie & Goethals, 1987; Moreland, 1987). We live in families, communities, and cities, work in study groups and work teams, worship together in religious groups, and play together on sports teams and clubs. We aYliate in part because doing so helps us meet the goals of protecting and enhancing the self—we need others to successfully reproduce, and our connections with others also allow us better opportunities for manufacturing, trade, food production, and security. But we also

250

CHARLES STANGOR AND SCOTT P. LEARY

aYliate because we enjoy being with others, being part of social groups, and contributing to social discourse. The important outcome of what we will term the social harmony motivation is the tendency to respect other people—even those from outgroups. In general, people view other people positively, act positively toward them in most cases, help them if they can, and expect others to react positively to them in a similar fashion (Sears, 1983). We are of course particularly likely to act positively toward other members of our ingroups, but aYliation is not limited only to ingroup members. Because people tend to be caring for each other, regardless of who they are or what relationship we have with them, we are able to move through our social worlds without expecting harm from others. In fact, negative behaviors are unusual and unexpected (Hansen & Hansen, 1988). Of course this does not mean that people are always friendly, helpful, and nice to each other—powerful social situations can create extremely negative behaviors, including teasing, bullying, aggression, violence, and even war. The problem is that the ingroup‐enhancement motivation is generally stronger and more universal than the harmony motivation. Thus, in some cases hostility and violence may occur in service of this motivation, even at the expense of social harmony. What the aYliation motivation does mean, however, is that hostility and violence are the exception rather than the rule of human behavior.

C. SUMMARY AND PREDICTIONS In sum, our analysis suggests that there are two primary motivations that determine the acquisition of intergroup beliefs—the overall ingroup‐ enhancing motivation that is likely to be universal and important in many contexts, and the (more secondary) social harmony motivation that is important, but nevertheless secondary. Focusing on these two motivations allows us to make some basic predictions about the acquisition of intergroup beliefs and evaluations. In terms of ingroup enhancement, we should find that individuals hold positive attitudes about the ingroup and particularly beliefs that allow them to create positive self‐esteem and social identity. Thus we can expect that individuals will show ingroup favoritism and that group knowledge that relates to beliefs about ingroups versus outgroups should be learned early in social development. Although the ingroup‐enhancement motivation predicts more favorable attitudes to ingroups, the social harmony motivation suggests that people will not be, by and large, hostile to outgroups. Negative beliefs and evaluations of outgroups are not predicted by either of our two basic

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

251

motivations (or indeed by any of the motivations that we know to be seen as relevant to human behavior). We are arguing that prejudice (defined as negative attitudes or hatred) is not part of the underlying human motivational system because there is no clear evolutionary or functional reason for it. The ingroup‐enhancement motivation does not lead to outgroup derogation and the goal of social harmony actively contradicts such a prediction. We can feel good about our ingroups without hating outgroups, and therefore because there is no general motivation for it, derogation of outgroups should be rare. Furthermore, we predict that these two basic motivations will relate to intergroup beliefs in a particular way. Specifically, variations in the ingroup‐ enhancing motivation should influence evaluations of ingroups. Individuals who are particularly concerned about protecting themselves and others should show more favorable opinions of ingroups. However, because the ingroup‐enhancing motivation is not directly relevant to attitudes about outgroups, variation in this motivation should not necessarily be related to attitudes toward outgroups. On the other hand, individuals for whom the social harmony motivation is particularly important should show more positive evaluations of outgroups, but because this outgroup‐ enhancing motivation is not relevant to attitudes about ingroups, variation in this motivation should not necessarily be related to attitudes toward ingroups.

IV. The Acquisition of Intergroup Beliefs in Childhood Because intergroup beliefs are such an important part of human knowledge, they should be acquired early and they should quickly become an integral part of the self‐concept. Furthermore, because the ingroup‐enhancing motivation is primary, we can predict that it will be evolutionarily powerful. Children should be particularly attentive to knowledge that provides information about the relatively safety and status of the ingroup. But they should also show respect for others. At least some research supports this idea, in the domains of both gender and ethnic attitudes. Young children and even infants have been found to focus on the evaluative aspect of group dimensions—quickly learning the extent to which members of diVerent social categories are good or bad, pleasant or unpleasant, and beneficial or harmful (Alvarez, Ruble, & Bolger, 2001; Heyman & Dweck, 1998). Furthermore, studies have found that 2‐ to 4‐year‐old children distinguish between boys and girls on emotions that relate to positive versus negative beliefs (e.g., cruel and fearful; Ruble & Martin, 1998).

252

CHARLES STANGOR AND SCOTT P. LEARY

Children are also aware of diVerentials along a power dimension, applying high‐power adjectives to boys (e.g., strong, fast, hit) and adjectives related to fear and helplessness to girls (e.g., can not fix bike, needs help). Thus early group knowledge appears to highlight distinctions on the important dimensions of power and valence (Osgood, Suci, & Tannenbaum, 1957), and these data seem to fit with our expectation that children are looking to compare ingroups and outgroups on the basis of evaluative concepts. If the ingroup‐enhancing motivation is primary, then we would expect to find that ingroup favoritism develops quickly in young children. And research supports this contention. Ingroup favoritism, on the basis of both gender and ethnicity, appears by the age of 3 (Aboud, 2003; Aboud & Amato, 2001; Bigler & Liben, 1993; Ruble & Martin, 1998), and increases up to about the age of 6. Young children show greater liking for peers of their own sex and race, and typically play with same‐sex others after the age of 3. Furthermore, it appears that, as predicted by Brewer (1999), ingroup attitudes are learned earlier than outgroup attitudes (Aboud, 2003). However, children are not only ingroup favoring. As expected by the social harmony motivation, gender and ethnic attitudes are also influenced by the goals of tolerance and equality (Huston & Alvarez, 1990), and do not generally show outgroup derogation. Thus, although the ingroup favoritism shown by children has often been equated with prejudice, a recent review (Cameron et al., 2001) has concluded that there is little evidence for outgroup devaluation in children. Furthermore, even ingroup preferences decline with age in children, beginning at about 4 to 5 years and continuing until age 12 and beyond, although the decline is often stronger for girls than for boys (Aboud & Amato, 2001; Bigler & Liben, 1993; Egan & Perry, 2001; Yee & Brown, 1994). Thus by the time children have reached adolescence the ingroup‐favoring motivation is substantially tempered, although not eliminated. At the same age range that the ingroup‐favoring motivation appears to decrease in children, an explicit motivation to value fairness and equality is developing. When we (Theimer, Killen, & Stangor, 2001) asked children if it was OK or Not OK to exclude boys from a ballet class, for instance, even though the stereotype that ‘‘boys can’t do ballet’’ provided a strong incentive and reason for answering ‘‘yes,’’ most children said that it was unfair to exclude someone from an activity because of his or her race or gender. In summary, then, there is evidence consistent with the hypothesis that the two social motivations are important in children, but that ingroup favoritism is primary. Children tend to initially favor the ingroup over the outgroup—this seems to be natural and even automatic. However, they also (and somewhat later in development) begin to show respect for members of outgroups and begin to see equality and fairness as important. And, as

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

253

expected, there is little evidence for outgroup derogation in young children at all.

V. Personality and Intergroup Beliefs Although the self‐enhancement and social harmony motivations are expected to be part of general human nature, there are nevertheless likely to be individual diVerences in the extent to which people hold each of them. If our model is correct and the two underlying motivations relate systematically to group beliefs, then (1) existing measures that are known to predict intergroup attitudes should be related to—indeed subsumed by—the underlying social motivations, and (2) individual diVerences in the underlying social motivations should relate to intergroup evaluations in a predictable manner. Prior research from our lab has begun to investigate both the ways in which existing personality variables that are known to relate to group evaluations fall together into underlying motivational orientations, as well as the extent to which intergroup beliefs can be predicted by these dimensions. Stangor, Leary, and Ottenbreit (2005a) administered a number of individual‐diVerence measures to a sample of 246 college students. We took our list of measures from prior research, which has shown that each of these measures predict intergroup attitudes. In Study 1, the measures administered were Social Dominance Orientation (SDO) (Pratto, Sidanius, Stallworth, & Malle, 1995), Devine’s measure of Internal Desire to Control Prejudice (Plant & Devine, 1998), Humanism and the Protestant Work Ethic (Katz & Hass, 1988), five‐item measure of Conformity developed by the researchers, Right Wing Authoritarianism (RWA) (Altemeyer, 1988), and Need for Structure (Neuberg & Newsom, 1993). We also measured a variety of intergroup attitudes that have been shown to be important in prior research. Specifically, participants expressed their attitudes toward 12 social groups: Americans, Blacks, Whites, Russians, Asians, Homosexuals, Arabs, Hispanics, Christians, Jews, Men, and Women on 9‐point feeling thermometers (1 ¼ unfavorable; 9 ¼ favorable). We then computed individual attitudes toward ingroups and outgroups ideographically, on the basis of each person’s own reported group memberships. The meanscale scores of the personality measures were submitted to a principle components factor analysis, followed by an oblique rotation. Supporting the hypothesis that these varied personality variables actually represent only two underlying motivational orientations, the factor structure produced only two factors. As shown in Table I, SDO (negative loading), Internal Desire to Control Prejudice, and Humanism loaded on the first

254

CHARLES STANGOR AND SCOTT P. LEARY

VARIABLES RELATED

TO INTERGROUP

TABLE I EVALUATIONS PRODUCE TWO UNDERLYING FACTORS Factor 1

SDO Internal desire to control prejudice Humanism Conformity RWA Need for structure Protestant work ethic


.83 .83 .80 – – –

Factor 2 – – .75 .72 .68 .53

Note: Loadings under .40 have been suppressed.

factor, whereas Conformity, Authoritarianism, Need for Structure, and Protestant Work Ethic loaded on the second factor. Furthermore, the factors were relatively independent, r ¼ .10, ns. We interpreted the first of these two factors to represent the underlying motivation of social harmony. The match for this factor is quite good— individuals who desire to control their prejudices, who believe that there should be few diVerences among groups (low SDO), and who endorse a humanistic outlook are those who are expressing a greater valuing of social harmony. However, because the idea of social harmony is a broad social construct that could relate to any type of social interaction, whereas the present factor specifically captures beliefs about the desirability of equality between groups, we labeled this factor Egalitarianism. We expected this factor to be positively related to outgroup evaluations. The match for Factor 2 is less straightforward, because it represents—not the goal of ingroup enhancement per se—but rather a motivational orientation that includes conservatism, conformity, and the desire to maintain the existing state of aVairs (high need for structure). However, exactly this factor has been extensively discussed in prior research. Most notably, Jost, Glaser, Kruglanski, and Sulloway (2003) reviewed literature demonstrating that a dimension of conservatism underlies much social perception in a variety of domains and that it is in large part indexed by a desire to maintain the existing social structure and to conform to group norms. And, as we have seen, conservatism has also been listed by Smith and Mackie (2000) as one of the most important social motivations. This dimension is also an essential component of the human value model of Schwartz (Schwartz, 1992; Schwartz & Bilsky, 1990). We therefore labeled this factor Conservatism.

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE TABLE II HOW UNDERLYING MOTIVATIONS PREDICT INGROUP

AND

255

OUTGROUP EVALUATION

B

t

p

Outgroup evaluations Egalitarianism Conservatism

.16 .06

2.48 .93

.01 .35

Ingroup evaluations Egalitarianism Conservatism

.06 .14

.88 2.15

.38 .03

Note: N ¼ 246.

We then assessed the extent to which each of the two motivational orientations (egalitarianism and conservatism) predicted expressed evaluations of social groups. First, as expected, the mean score for ratings of ingroups (M ¼ 5.75) was significantly greater than the mean score for ratings of outgroups (M ¼ 5.15), p < .001, and both were above the midpoint of the scale (5.00). Thus, across these groups, there was significant ingroup favoritism and yet no evidence for outgroup devaluation. Furthermore, as shown in Table II, when we conducted regression analyses using the factor scores to predict attitudes toward ingroups and outgroups we found that the variables that relate to egalitarianism significantly predicted attitudes toward outgroups (but not ingroups), whereas variables that related to the conservatism orientation significantly predicted attitudes toward ingroups (but not outgroups). We (Stangor et al., 2005a) studied this pattern in two follow‐up studies that used diVerent personality measures. In Study 2 the variables were expanded to include Need for Closure (Webster & Kruglanski, 1994) as well as Need for Structure and also included a measure of Need for Cognition (Cacioppo & Petty, 1982). We also added in this study the ‘‘Big 5’’ personality measure (McCrae & Costa, 1999) and a measure of social identity (Luhtanen & Crocker, 1992) as well as the External Motivation to Control Prejudice measure (Plant & Devine, 1998). We also assessed whether each of the original variables was necessary to create a simple factor structure, deleting those variables that were not. Figure 1 represents the outcomes of the confirmatory factor analysis that provided the best and most concise picture of our two factors on the basis of this study. Again, and despite the inclusion of these extra variables, the measures created two strong factors, one representing egalitarianism and the other representing conservatism. And again, the two factors were uncorrelated. The fit of the model is quite good, and adding the other variables—protestant work ethic, the Big 5 scale, Devine’s external motivation to control prejudice, need for cognition and social identity—does not improve the model fit. That is, these other variables do not

256

CHARLES STANGOR AND SCOTT P. LEARY

Fig. 1. Confirmatory analysis of personality variables.

load with our factors and do not predict group attitudes above and beyond them. Furthermore, RWA was not necessary to accurately predict the conservatism motivation, and so we deleted it from our final solution, and again in this study we found that conservatism predicted only attitudes toward ingroups whereas egalitarianism predicted only attitudes toward outgroups. On the basis of this analysis, in Study 3, a subset of 25 items from the 5 scales shown in Fig. 1 was selected by choosing the items with the highest factor loadings from each of the 5 scales. We administered this measure to a sample of college students and found that this short measure produced the same underlying factor structure and related to ingroup and outgroup attitudes in the same way as did the studies in which we assessed a larger number of variables. This short scale is presented in Table III. Overall then, we have found that all of the individual diVerence measures that we have so far tested are either (1) subsumed by one of the two underlying motivations or (2) not related to group beliefs above and beyond these motivations. Although the specific measures that are used to index the factors have varied from study to study, in each case we find that these personality measures produce two factors—one representing egalitarianism and the other representing conservatism. Furthermore, we consistently find that the former predicts only outgroup evaluations whereas the latter predicts only ingroup evaluations.

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

257

TABLE III THE SHORTENED MOTIVATION MEASUREa 1. It upsets me to go into a situation without knowing what I can expect from it. (STRUC) 2. Sometimes other groups must be kept in their place. (SDO) 3. I find that a well‐ordered life with regular hours makes my life tedious. (STRUC–R) 4. According to my personal values, using stereotypes is OK. (IDCP–R) 5. In dealing with criminals the courts should recognize that many are victims of circumstances. (HUMAN) 6. We would have fewer problems if we treated people equally. (SDO–R) 7. I consider myself to be obedient. (CONF) 8. There should be equality for everyone because we are all human beings. (HUMAN) 9. Acting to protect the rights and interests of other members of the community is a major obligation for all persons. (HUMAN) 10. It would be good if groups could be equal. (SDO-R) 11. Everyone should have an equal chance and an equal say in most things. (HUMAN) 12. No one group should dominate in society. (SDO-R) 13. One should be kind to all people. (HUMAN) 14. I find that a consistent routine enables me to enjoy life more. (STRUC) 15. I enjoy the exhilaration of being in unpredictable situations. (STRUC) 16. I consider myself to be orthodox. (CONF) 17. To get ahead in life, it is sometimes necessary to step on others’ groups. (SDO) 18. Being nonprejudiced is important to my self‐concept. (IDCP) 19. Because of my personal values, I believe that using stereotypes is wrong. (IDCP) 20. I consider myself to be conforming. (CONF) 21. I hate to change my plans at the last minute. (STRUC) 22. I am personally motivated by my beliefs to be nonprejudiced. (IDCP) 23. I consider myself to be unconventional. (CONF) 24. I consider myself to be rebellious. (CONF) 25. I attempt to act in nonprejudiced ways because it is personally important to me. (IDCP) a Stangor et al., 2005a. Note: CONF ¼ conformity; HUMAN ¼ humanism; IDCP ¼ internal desire to control prejudice; SDO ¼ social dominance orientation; STRUC ¼ need for structure; –R ¼ reverse scored item.

VI. Priming Social Motivations If underlying social motivations predict group evaluations, they should do so to a greater extent when these motivations are currently cognitively accessible. Specifically, priming an individual’s social motivations should increase the correlation between the egalitarianism motivation and evaluations of outgroups and at the same time increase the correlation between the conservatism motivation and evaluations of ingroups. Stangor et al. (2005a, Study 4) first measured conservatism and egalitarianism motivations using the shortened measure of motivations shown in Table III. Participants completed this measure at a mass screening held several weeks prior to the experimental session.

258

CHARLES STANGOR AND SCOTT P. LEARY

TABLE IV CORRELATIONS BETWEEN FACTOR SCORES AND GROUP RATINGS CONDITION

AS A

FUNCTION

OF

EXPERIMENTAL

Group rated Prime

Americans

Arabs

US symbols Egalitarianism Conservatism

.14 .45**

.64***
.03

Control images Egalitarianism Conservatism

.18 .25*

.30*
.06

Note: p < .05, **p < .001, ***p < .001.

An analysis of this measure again demonstrated that this measure reliably captured the same two underlying factors—egalitarianism and conservatism—and that the two factors were again uncorrelated. At the experimental session, all participants first completed a priming task in which they were asked to express their liking of seven photographs. These judgments were not used in any data analysis and were made only to assure that the participants attended to the images. In the control condition these photos were innocuous images of nature scenes, whereas in the experimental condition these images were US symbols (the US flag, the White House, and so on). We expected that the latter set of images would prime the relevant social motivations of conservatism and egalitarianism for our participants, as both of these dimensions are relevant to US values. Participants then indicated their attitudes toward Arabs (an outgroup) and Americans (an ingroup) on 7‐point Likert scales. As shown in Table IV, we found that, as expected, the correlations between the measured motivations and group evaluations replicated our prior findings. Conservatism predicted only ingroup evaluations whereas egalitarianism predicted only outgroup evaluations. Furthermore, these correlations were significantly stronger in the priming condition than in the control condition. In sum, this research provides support for several of the ideas that we have proposed. First, across the variety of attitude measures that we have used, we have found that college students do not show negative attitudes toward outgroups—only relatively more positive attitudes to ingroups. Second, evaluations of ingroups and outgroups are predicted by diVerent sets of variables; specifically, ingroup evaluations are predicted by variables that relate to conservatism and conformity, whereas outgroup attitudes are predicted by variables related to egalitarianism. These two dimensions subsume a number

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

259

of existing personality variables known to predict intergroup attitudes and help explain how these many measures relate to each other. And, the correlations between motivations and group evaluations are stronger in social contexts that prime these orientations. These results are very consistent with the idea that people hold beliefs about ingroups and outgroups to the extent that those beliefs help fulfill the basic social motivations.

VII. Group DiVerences in Intergroup Beliefs If social motivations are essential determinants of intergroup attitudes, then diVerences in these motivations may help explain the mean diVerences in intergroup attitudes that are frequently observed across social groups. For instance, prior research has found that girls and women value egalitarianism and equality more than boys and men (Galambos, Petersen, Richards, & Gitelson, 1985; Theimer et al., 2001), whereas boys and men hold more traditional, conservative attitudes (Jackson & Tein, 1998). And there are also gender diVerences in prejudice, such that men show more negative evaluations of outgroups than do women. DiVerences in both social motivations and group evaluations might also be expected across other social categories such as ethnicity and political preferences. Our analysis of social motivations suggests that they may be important in understanding such group diVerences. For instance, if women evaluate outgroups more positively than do men, this might be because they have, on average, higher levels of the egalitarianism motivation. And if Republicans display more ingroup favoritism than do Democrats, this could be because they are more driven by the conservatism orientation. We assessed whether observed diVerences across social groups in group evaluations might be accounted for by diVerences in underlying motivations by returning to the data from Study 1 earlier, in which students completed a variety of personality measures and also indicated their attitudes toward a number of ingroups and outgroups. As shown in Table V, and replicating prior findings, we found that women showed significantly higher levels of the egalitarianism motivation than did men ( p < .01). There was, however, no diVerence between men and women in conservatism. In terms of group evaluations, although we did find that women were more favorable to outgroups than were men ( p < .01), we also found that they were more favorable to ingroups ( p < .001). The gender x group‐rated (ingroup/outgroup) interaction was not significant. Our next question concerned whether the observed group diVerences in the underlying social motivations might mediate the relationship between gender and group evaluations. And this hypothesis was supported. The main

260

CHARLES STANGOR AND SCOTT P. LEARY TABLE V FACTOR SCORES AND GROUP RATINGS AS A FUNCTION

OF

Factor scores

Male Female

PARTICIPANT GENDER Group ratings

N

Egalitarianism

Conservatism

Ingroup

Outgroup

105 136


.19 .16


.09 .08

3.69 4.09

3.36 3.74

eVect of group evaluation (ingroup versus outgroup) was no longer significant when scores on egalitarianism and conservatism were included as covariates. As shown in Table VI, there were also significant diVerences across ethnic groups in the underlying orientations, p < .01, as well as in attitudes toward ingroups versus outgroups, p < .01. More specifically, African‐Americans and Asian‐Americans were significantly higher on the conservatism motivation than were European‐Americans, and African‐Americans and Hispanic‐ Americans held more egalitarian orientations than did Asians. Furthermore, Asian‐Americans and African‐Americans rated ingroups significantly more positively than did European‐Americans. Again, however, when we partialled the factor scores, diVerences ingroup attitudes were reduced to nonsignificance. Although this original dataset did not include information about political orientation, we were able to analyze data from another study that did (Stangor et al., 2005a, Study 5). In this study participants had completed the short version of the motivational orientation measure, along with the group attitudes measure, at a mass screening session. As shown in Table VII, we found that Democrats held more egalitarian values than did Republicans and that Republicans held more conservative values than did Democrats (Independents and No‐orientation participants are included for comparison). Furthermore, political party preference related to attitudes toward ingroups versus outgroups, p < .01, such that Republicans showed more ingroup favoritism than did either Democrats or Independents. Furthermore, indicating that the social motivations can explain these diVerences, when we partialled the factor scores, mean diVerences ingroup evaluations across the four political groups were reduced to nonsignificance. In sum, then, these analyses add evidence that the social motivations of conservatism and egalitarianism are important determinants of intergroup attitudes, because they mediate the relationship between social‐category membership and group evaluations. On an average, diVerent social groups hold diVerent group beliefs. Although our research has not addressed the

261

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE TABLE VI FACTOR SCORES AND GROUP RATINGS AS A FUNCTION

OF

PARTICIPANT ETHNICITY

Factor score

African‐American European‐American Asian Hispanic Other

FACTOR SCORES

Group ratings

N

Egalitarianism

Conservatism

Ingroup

Outgroup

116 489 93 36 44

.08 .00
.13 .15 .06

.17
.08 .26 .16
.20

4.27 3.87 4.04 3.72 3.45

3.73 3.56 3.57 3.49 3.30

AND

TABLE VII GROUP RATINGS AS A FUNCTION

OF

POLITICAL ORIENTATION

Factor score

Democrat Republican Independent None

Group ratings

N

Egalitarianism

Conservatism

Ingroup

Outgroup

435 193 124 154

.16
.31
.03 .00

.00 .23
.23
.02

3.19 3.35 3.17 3.18

2.98 2.88 2.83 2.97

etiology of these diVerences, we have found that these diVerences can be understood in terms of variability in the two underlying social motivations.

VIII. Group Beliefs as Social Norms If intergroup beliefs are as social in nature as we have suggested they are, then they should be part of the everyday social fabric. That is, intergroup beliefs should be shared among people and developed and changed through social discourse. Furthermore, this transmission should be in service of basic social motivational goals.

262

CHARLES STANGOR AND SCOTT P. LEARY

The idea that intergroup beliefs are in essence social norms is of course not a new one. For instance, Pettigrew (1959) proposed that prejudice was in large part a function of individuals conforming to group or societal norms, and Allport (1954) considered group‐norm theory as one of major theories of prejudice (Crandall & Stangor, 2005). As Allport noted, support for this theory of prejudice was found in the relative ineVectiveness of attempts to change individual attitudes—he pointed out that it is easier to change group than individual beliefs (Allport, 1954; Sherif & Sherif, 1953). A host of recent research has assessed the role of norms in prejudice. In this section we review research related to the related hypotheses that: (1) intergroup beliefs are communicated in everyday discourse, (2) that individuals’ perceptions regarding the beliefs of others impact their own beliefs, (3) that group attitudes are shared among peers, and (4) that group attitudes are transmitted from parents to their children. A number of lines of research have focused on the interpersonal communication of group beliefs and thus are relevant to this section (Ruscher, 1998, 2001; Schaller & Conway, 1999). Group beliefs are communicated in everyday interaction in many forms, ranging from the direct expression of negative evaluations and facial expressions, to the use of racial epithets, and the telling of racial and ethnic jokes. Furthermore, this communication is likely to be in the service of the social motivations that we have considered. That is, group beliefs may sometimes be expressed with the goal of helping to protect the ingroup’s status and furthering ingroup enhancement, or alternatively, may be used to communicate the goals of social harmony. Communicating one’s adherence to group beliefs may be used as a method for ingroup members to conform to social norms, thereby helping the individual to gain status by creating positive impressions on other ingroup members. For instance, Lyons and Kashima (2003) found that people communicated more stereotypes when they believed that other ingroup members also held those beliefs. These findings suggest that individuals express beliefs in large part to conform to their perceptions of prevailing social norms and to be accepted by others. Similarly, Ruscher and colleagues have found that, particularly when asked to come to agreement, dyads increasingly focus their conversation around stereotype‐consistent information (Ruscher, 2001; Ruscher & Duval, 1998). These findings are consistent with the idea that the members of the dyads use the communication of stereotypes that they believe are held by others to create a positive group identity and to bolster their status in the group. Schaller, Conway, and Tanchuk (2002) directly tested the hypothesis that group knowledge is communicated to the extent that it is shared by others. They found that the traits that are most likely to be the subject of social discourse (i.e., most communicable) are most likely to persist in ethnic

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

263

stereotypes over time. This result is also consistent both with the idea that intergroup beliefs are communicated and that the motivation for doing so is social conformity. Although research has found that group beliefs are frequently expressed in the service of conformity and social identity, it is also possible that they may be expressed with the goal of communicating egalitarian beliefs and tolerance. We are not aware of any research on this possibility, but we do know that expressing these norms is frequently socially undesirable. People who believe that they or others have been victims of discrimination are frequently unwilling to admit or confront it, because doing so leads to negative perceptions on the part of others (Kaiser & Miller, 2001; Stangor, Swim, Van Allen, & Sechrist, 2001). It is our expectation that it takes more eVort and may, in many cases, be less socially normative to overtly express egalitarian rather than ingroup‐enhancing beliefs. In addition to helping explain how group beliefs are developed and maintained, the fact that they are communicated also has important social implications. For instance, Greenberg and Pyszczynski (1985) found that participants rated a black target more negatively after they had overheard an ethnic slur (‘‘nigger’’). Simply overhearing a derogatory ethnic label can aVect attitudes toward members of an ethnic group, evidently because it sets up a norm in which expressing racist attitudes is acceptable (Kirkland, Greenberg, & Pyszczynski, 1987). In another relevant study, Henderson‐King and Nisbett (1996) had participants overhear another White student talking on a cellular phone about a friend who had been assaulted and robbed. In describing the incident, the student noted that the assailant was either White or Black. Participants then completed several questionnaires, including a measure of perceived antagonism. The authors found that simply overhearing that an African‐American had committed an assault (a stereotype‐consistent act) increased the extent to which Whites perceived Blacks as antagonistic and hostile. There was no eVect on participants’ perceptions of Whites after hearing that a White person had committed the same crime. Taken together, these studies suggest that prejudiced communications, including direct comments and humor containing negative references tend to validate and strengthen negative group evaluations and also reduce perceived sanctions against discrimination (Ford & Ferguson, 2004). We (Sechrist & Stangor, 2001; Stangor, Sechrist, & Jost, 2001a,b) have conducted a number of studies designed to study the extent to which the perceptions of others’ beliefs influences one’s own intergroup evaluations. In these experiments participants learned about the supposed outgroup evaluations of others, both before and after expressing their own beliefs. We found that White participants expressed significantly more positive

264

CHARLES STANGOR AND SCOTT P. LEARY

attitudes toward African‐Americans when they had learned that other people held more favorable stereotypes than they had originally estimated and they became more negative toward African‐Americans when they learned that others held less favorable stereotypes than they had originally assumed. These changes were strong in magnitude and persisted when assessed again up to 2 weeks later (Stangor et al., 2001a). In addition, in a second experiment we found that consensus eVects were stronger for people who were exposed to information about the opinions of ingroup rather than outgroup members, and that this change occurred even when it was assessed in private on a diVerent measure and at an unrelated experimental session. These results again suggest that people hold group beliefs because they perceive them as normative within the ingroup. In addition to creating new intergroup beliefs, we also assessed the hypothesis that perceived consensus would make intergroup beliefs more resistant to change (Stangor et al., 2001b). One week after estimating their own attitudes toward African‐Americans, participants were provided with information that their beliefs were either shared or were not shared by other students at their university. Then, in an attempt to change their beliefs, they were given allegedly ‘‘objective’’ information about the actual traits possessed by African‐ Americans, supposedly as determined by actual research. Results showed that perceptions of agreement with others strengthened racial stereotypes, such that participants who had been given information that others shared their beliefs (high perceived consensus) showed less opinion change as a result of the supposedly ‘‘objective’’ information, on both positive and negative stereotypes, in comparison with participants who were given information that others did not share their beliefs (low perceived consensus). One limitation of the research just described is that it is not always completely clear whether the belief change represented public compliance with social norms or if it was more internalized. However, Blanchard, Crandall, Brigham, and Vaughn (1994) found that the normative features of social situations could influence people’s privately expressed attitudes toward racism. In their research, White participants heard another student from their college either condone or condemn racism before completing five questions regarding how their college should respond to acts of racism. Results showed that hearing another student condemn racism increased participants’ expressed antiracist opinions and hearing someone condone racism reduced antiracist expressions, in comparison to a control condition in which no information about others’ opinions was provided. In addition, these results occurred regardless of whether participants’ responses were spoken publicly in the presence of the person making the comment and the experimenter or written privately on a questionnaire and sealed in an envelope, suggesting that internalization did occur (Blanchard, Lilly, & Vaughn, 1991).

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

265

Sechrist and Stangor (2001) explored the extent to which perceived group beliefs of others could influence implicit attitude and behavioral measures. In Experiment 1, we found that high‐prejudice participants who were provided with information that their (negative) beliefs were shared by other students at their university subsequently sat farther away from an African‐American target than participants who were informed that their beliefs were not shared. On the other hand, low‐prejudice participants who were provided with information that their (favorable) beliefs were shared by others subsequently sat closer to an African‐American target than participants who were informed that their beliefs were not shared. Furthermore, the correlation between expressed attitudes (on the pro‐Black scale; Katz & Hass, 1988) and behavior (seating distance) was significantly greater for participants in conditions where their beliefs were supported by their ingroup. Thus, perceived group beliefs appear to increase attitude–behavior consistency in the domain of racial relations and this occurred even on a nonreactive behavior (the participants did not know that their seating distance was being observed). In a second experiment, we examined whether learning that one’s racial attitudes are consistent with group beliefs would alter the mental representation of those beliefs such that they would become more closely associated with the category label in memory and thus be more quickly activated upon exposure to the relevant category label. This prediction was based on models of stereotypes, which suggest that stereotypes are mentally associated with category labels in memory (Dovidio, Evans, & Tyler, 1986; Kunda & Thagard, 1996; Stangor & Lange, 1994) and thus come to mind when the category is activated. We found that White participants who learned that their stereotypes of African‐Americans were shared with others (consistent with the ingroup norm) were significantly faster at identifying those same stereotypes as words after being primed with a word associating them with African‐Americans (Black) than after neutral primes, but this diVerence did not occur for participants who had just learned that their stereotypes were inconsistent with their ingroup’s beliefs. Thus, as expected, stereotypes that are perceived as shared by the group are highly cognitively accessible, in the sense that they come to mind quickly on exposure to the category label, and thus facilitate their identification as words.

IX. The Sharing and Transmission of Intergroup Beliefs The research reviewed in the prior section indicates that people communicate their group beliefs among each other and that their own group beliefs are influenced by both their perceptions of the beliefs of important others

266

CHARLES STANGOR AND SCOTT P. LEARY

and by the statements made by others. In this section we tackle another question that follows from the basic idea that group beliefs are social constructs—namely that group beliefs should be shared or transmitted among peers and from parents to their children. Although these questions might seem to be at the heart of any true social psychology of intergroup beliefs, they have barely been addressed in the literature.

A. PEER‐TO‐PEER TRANSMISSION If group beliefs are as social as we have claimed they are, then friends should learn from, and share their group attitudes with each other. That is, we should find that peer groups share their beliefs because they communicate them and are influenced by this communication. Surprisingly, there is very little existing research concerning the sharing of group knowledge and evaluations among peers. Indeed the only such study that we are aware of (Ritchey & Fishbein, 2001), which assessed attitudes toward HIV/AIDS, race, homosexual, and fat prejudice in 426 White male and female 9th and 11th graders, found no correlation between the prejudices and stereotypes expressed by friends. Thus the limited literature suggests that evaluations of specific outgroups, at least in high school students, are not shared. Whether group beliefs are shared among other populations is currently unknown, because the hypothesis has not to our knowledge been tested. There is, on the other hand, at least some research that supports the idea that some broader‐level attitudes and orientations are shared among peers. Hamm (2000), for example, found that academic value orientations were similar across friends, although this finding was qualified by race, with White and Asian friends being more similar than Black friends. Similarly, materialism values have been found to be similar among friends, especially when parental involvement with children is low (Flouri, 2004). In an attempt to address this question more fully, we (Stangor, Leary, Freidus, & Ottenbreit, 2005b) studied the extent to which college student friends shared the underlying motivational orientations of conservatism and egalitarianism, as well as the extent to which they shared attitudes toward ingroups and outgroups. Participants were 94 undergraduates at the University of Maryland. Forty‐seven participants were recruited through the university subject pool. When signing up for the study, each participant was asked to bring a close friend to the testing session. Participants consisted of various racial and ethnic groups representative of a typical urban university. Participants were tested in groups of two. The researcher double checked to make sure each pair consisted of a participant from the subject pool and

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

267

his or her close friend. Both participants then proceeded to complete the surveys in separate rooms. As in our prior research, participants completed a battery of individual diVerences measures, this time including SDO, Devine’s measure of Internal Desire to Control Prejudice, Humanism and Protestant Work Ethic, Need for Structure, RWA, and a five‐item measure of conformity developed by the experimenters. Because our measures share a great deal of conceptual similarity with the dimensions proposed in the human values model of Shalom Schwartz (Schwartz & Bilsky, 1990), and because we were interested in the relationship between those variables and the ones we had collected in the past, we also had our participants complete the measures from his value survey. Participants also completed a 5‐item patriotism measure (from Federico, Golec, & Dial, 2005), the pro‐Black scale (Katz & Hass, 1988), the Modern Homophobia scale (Raja & Stokes, 1998), a 5‐item attitudes towards Arabs scale, and a 16‐item attitudes toward religious groups measure. We again found that the variables that we included, this time including items from the Schwartz Value Survey, created two factors. The variables that had been used in our prior studies loaded as expected. And two variables from the Schwartz Value Survey—the values of security and conformity—also loaded on the conservatism factor (the other Schwartz values— power, universalism, benevolence, self‐determination, hedonism, achievement, and stimulation did not load consistently with these factors and so we did not include them in our analysis). The between‐peer correlations for each of the scales, as well as the between‐peer correlations for the factor scores that we derived from the individual scales listed in the table, are shown in Table VIII. We found a very consistent pattern of results. Conservatism motivations, as assessed by each of the individual scales and by the factor scores, correlated highly between friends. On the other hand, egalitarian motivations were not significantly shared. And attitudes toward both religious and ethnic outgroups were significantly correlated, but there was no evidence that friends shared their attitudes toward specific ingroups.

B. PARENT‐TO‐CHILD TRANSMISSION Although the idea seems straightforward, we have not been able to locate a great deal of research considering the transmission of intergroup beliefs from parents to children. Indeed, most reviewers have concluded that the existing evidence is not highly supportive of the notion that parents pass their beliefs about social groups to their children (O’Bryan, Fishbein, & Ritchey, 2004).

268

CHARLES STANGOR AND SCOTT P. LEARY TABLE VIII PEER‐TO‐PEER CORRELATIONS Conservatism motivation Need for structure RWA Conformity Schwartz conformity Schwartz security Patriotism Factor score Egalitarian motivation Humanism Desire to control prejudice SDO Factor score Ingroup attitudes Religious groups Ethnic groups Overall Outgroup attitudes Religious groups Ethnic groups Anti‐Arab prejudice Anti‐Arab prejudice Anti‐gay prejudice Overall

.17 .24* .20þ .26* .19 .35** .35** .14
.02 .22* .16 .01 .03 .06 .26*
.05 .28* .25* .22
.00

Note: þp < .10, *p < .05, **p < .001.

Aboud and Doyle (p. 371, 1996) claimed that ‘‘there is little support for the widespread assumption that children acquire their racial attitudes from parents and friends’’. However, the link may be stronger in cases where parents talk about their intergroup beliefs with their children and may be stronger in minority families (Aboud & Amato, 2001). Perhaps it is not surprising that there is little relationship between the outgroup beliefs of parents and children, given that the particular outgroups that are most socially relevant may vary across cohort. That is, children live in diVerent social environments, with diVerent social norms, than do their parents, and thus—as we would expect if norms are important their beliefs may vary. However, children share the same ingroup status with their parents, and thus it is possible that ingroup beliefs are more likely to be similar. Another possibility is that specific group beliefs are not transmitted from parents to children, but that general goals and motivations—for instance those of ingroup enhancement and egalitarianism—may nevertheless be (O’Bryan et al., 2004; Rohan & Zanna, 1996).

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

269

Indeed, in contrast to the lack of correlations observed on specific group beliefs, the literature suggests that broad values may be transmitted from parents to children (Boehnke, 2001; Flor & Knapp, 2001; Nieuwbeerta & Wittebrood 1995; Rohan & Zanna, 1996; Sidanius & Ekehammar, 1979). For instance, Flor and Knapp (2001) found that parental religious behavior and a parent’s desire for their child to be religious predicted religious beliefs and religious behaviors of their oVspring. Furthermore, suggesting that this reflects social rather than genetic transmission, parental discussions of faith with their children were related to religiosity in children. The research on the transmission of political values also provides at least some evidence for political value transmission from parents to oVspring. Children tend to endorse the same political ideology as their parents (Nieuwbeerta & Wittebrood, 1995; Sidanius & Ekehammar, 1979), but it has been concluded that this relationship is in part genetically determined (Tesser, 1993). O’Bryan et al. (2004) collected several measures of prejudice against people with HIV/AIDS, gays and lesbians, and Blacks, and found that, at least for these groups, the group evaluations of parents was correlated with those of their children, although the eVect size was small and not always significant. However, correlations were generally higher when group ratings were aggregated. Similarly, Acock and Bengtson (1978) found that when they examined multiple items together (such as traditional sexual norms, religious behavior, and legitimacy of Black demands) they found a higher correlation between parent and child compared to when just one value was measured. This approach again suggests that higher level constructs may be more likely to be transmitted than are lower level constructs. We have conducted one preliminary study to assess this issue (stangor et al., 2006b, Study 2). The participants were the same 47 participants who had participated in the peer‐to‐peer study described earlier. Participants were asked to complete a survey in exchange for extra credit. Participants were informed that in order to receive full credit, one parent or guardian would also need to complete the survey. Participants were then given a flyer with a web address where they could complete this survey. Parents/guardians were then able to complete the surveys from any location with Internet access. Data was collected over 2 weeks, at which point data from students were matched with their respective parents or guardians. As shown in Table IX, we found a clear pattern, and one that was very similar to our findings in the peer‐to‐peer study described earlier. Conservatism motivations, as assessed by each of the individual scales and by the factor scores, correlated highly between parents and their children. On the other hand, egalitarian motivations were more weakly shared or transmitted—the correlations for these variables were less significant. Similarly, attitudes toward both religious and ethnic outgroups were significantly correlated

270

CHARLES STANGOR AND SCOTT P. LEARY TABLE IX PARENT‐CHILD CORRELATIONS Conservatism motivation: Need for structure RWA Conformity Schwartz conformity Schwartz security Patriotism Factor score

.24* .49** .38** .29** .14* .49** .40**

Egalitarian motivation: Humanism Desire to control prejudice SDO Factor score

.33* .29* .10 .26*

Ingroup attitudes Religious groups Ethnic groups Overall Outgroup attitudes Religious groups Ethnic groups Anti‐Black prejudice Anti‐Arab prejudice Anti‐gay prejudice Overall

.31*
.21þ .00 .45* .07 .29* .25* .44** .26*

Note: þp < .10, *p < .05, **p < .001.

between parents and oVspring, but there was less evidence that friends shared their attitudes toward specific ingroups.

C. SUMMARY In summary, although there is little existing literature concerning the direct sharing and transmission of group beliefs among peers, and between parents and oVspring, and although we have only two studies currently from our lab, there are perhaps some conclusions we can draw. Although the existing research has generally found little or no correlations in this regard, we have surmised that perhaps some attitudes and beliefs are more likely to be transmitted than others. We have found in our research that conservatism motivations are shared, both between peers, and between parents and their

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

271

oVspring, but egalitarian motivations are not. And to the extent that attitudes toward specific groups are shared, these relationships are found more on outgroup than on ingroup evaluations. The idea that people are more likely to share higher level orientations than specific beliefs fits well with our overall conceptualization. That is, even if parents do not transmit their specific group beliefs to their children, and even if friends do not communicate their specific beliefs, they may nevertheless communicate and transmit the broader social motivations of ingroup enhancement and social harmony. Although not expected, the finding that only conservatism but not egalitarianism motivations are shared fits with our assumption that the motivation of protecting the ingroup is primary. Perhaps ingroup evaluations are communicated more often because they are somehow more evolutionarily important, and they may be genetically transmitted from parents to children. As Rohan and Zanna (1996) noted, the degree of similarity in values between parent and child (and presumably among peers) will be greater to the extent that value is important to them. And, as we have proposed earlier, communicating ingroup attitudes may be more normatively desirable.

X. Conclusions Taken together, our recent research program has focused, broadly, on group beliefs as social variables. In one line of research we have attempted to determine how group beliefs are influenced by the basic social motivations of individuals. In a second line we have focused on the social sharing and transmission of group beliefs. These are lines of research that we feel are vitally important to the social psychology of stereotyping and prejudice. They are less well developed, overall, than more individual social‐cognitive level approaches, and yet seem to be worth mining more fully. The lack of recent interest in these issues in the area of intergroup attitudes is perhaps surprising, given the historical importance of social norms as determinants of social behavior and the avid use of social norm theories in attempts to change real world behavior in other domains—for instance campus alcohol abuse (Prentice & Miller, 2003). Our approach not only shares similarities with historical and contemporary approaches but also has some unique characteristics. In our work on social motivations we have followed the ‘‘motivated social cognition’’ approach proposed by others (Jost et al., 2003), basing our analysis on the idea that there are two fundamental motivations relevant to intergroup attitudes—ingroup enhancement and social harmony. Furthermore, we have

272

CHARLES STANGOR AND SCOTT P. LEARY

argued that these motivations can be indexed by measures of conservatism and egalitarianism, respectively, and that these variables then independently predict ingroup and outgroup evaluations. Thus we have found that the social goals proposed by prominent writers to be important overall are indeed influential within the domain of group attitudes. Prior research within the group relations literature has found that a variety of individual diVerence measures predict group attitudes. But there has not heretofore been an attempt to determine how these measures relate to each other and to determine how, and if, they form higher level factors. Furthermore, the prior work in this area has generally focused on outgroups, rather than ingroup, attitudes. We have consistently found, across a number of studies, that the measures that have been used to predict intergroup attitudes in prior research can be conceptualized as two orthogonal dimensions of conservatism and egalitarianism. This result fits with prior models that have hypothesized similar dimensions (Kerlinger, 1984; Tomkins, 1964). Our results are also at least in part consistent with the claim, proposed by Jost et al. (2003), that two important dimensions of social perception and behavior are resistance to change and acceptance of inequality. These authors claim, however, that these dimensions are normally correlated and conceptualize them both within the conservatism dimension. Our data on the other hand find that—at least within the domain of intergroup attitudes—these two dimensions seem to be independent. Our findings also relate, perhaps most directly, to a model recently developed by John Duckitt (2002). Duckitt proposed that intergroup evaluations were essentially a function of two individual motivational goals: (1) a competitively driven dominance‐power–superiority motivation, and (2) a threat‐driven social control and group‐defense motivation. Duckitt indexed these two motivations using the measures of RWA and SDO, respectively. Our model shares some characteristics with that proposed by Duckitt, but also diVers from it. For one, the models are in agreement that there are two basic underlying motivations that determine intergroup beliefs. And, in both models SDO loads on one of the motivations and RWA loads on the other. An important diVerence, however, is that we conceptualize RWA and SDO not as the important constructs in their own right, but rather as indices of underlying motivations. For instance, Authoritarianism in our model is one potential measure of conservatism, and SDO is one measure of Egalitarianism. We have not relied on one measure to assess our conceptual variables, but rather have argued that there are a number of measures that relate to these dimensions. Second, our model diVerentiates ingroup and outgroup evaluations in a way that Duckitt’s model does not. Our research finds that the conservatism motivation, as assessed in part by Authoritarianism, better predicts attitudes toward ingroups than attitudes toward outgroups, whereas Egalitarianism,

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

273

assessed in part by SDO, predicts attitudes toward outgroups better than attitudes toward ingroups. These findings are in some sense inconsistent with the results reported by Duckitt (2002), who finds that both orientations predict outgroup attitudes. Why these diVerences are observed is not completely clear, and this issue still needs to be addressed. One aspect of Duckitt’s model that we have not yet addressed concerns interactions between situational variables and personality. He argues that the motivational goals in his model are aroused by intergroup threat and inequalities in, or competition over, power and dominance. Our research has found that situations that prime the underlying motivations make them more predictive of group attitudes, but we have not yet determined whether there are particular situational contexts in which one or the other is most predictive of group evaluation. We have also used our approach to help explain observed diVerences ingroup beliefs across social groups. We have found that women—who are known to express more empathy and egalitarianism goals—also tend to show more favorable attitudes to both ingroups and outgroups. And we also found that racial and political groups diVered in their group evaluations. Furthermore, our approach provides a potential explanation for those diVerences. Specifically, if the members of a social group, on an average, weight one of the two motivations as more important than the other, this can explain why the group also tends to have, on an average, diVerent group evaluations. It is also possible that cultural diVerences (for instance collectivism and individualism) may be accounted for in part by diVerences in these underlying social motivations. It has been noted that values are most predictive of behavior when they are in conflict (Tetlock, 1986). And the two motivations that we have specified as critical frequently have this characteristic. Consider, for instance the responses of US citizens to the 9/11 terrorist attacks. One approach— perhaps favored by Republicans—involves security and ingroup protection. The borders are sealed and outgroup members who seem likely to be threatening are questioned and perhaps detained. On the other hand, others—perhaps the preferred response of Democrats—involves reaching out to allies and even to the source of the attack itself. However, these two approaches are in many ways mutually exclusive. It is diYcult to reach out to an enemy at the same time that one is protecting the ingroup—choices need to be made. In these situations the individual’s preferred motivational orientation may be particularly important in determining their opinions and behaviors. We have also argued, as have others, that in most cases the ingroup‐ enhancing motivation will be functionally more important than social harmony (Brewer, 1999). We have found that although both motivations are observed in children, ingroup favoritism seems to occur earlier, only later

274

CHARLES STANGOR AND SCOTT P. LEARY

being tempered by more egalitarian beliefs. And we have found that the conservatism goal is more likely to be shared among peers, and between parents and oVspring. These results fit with the assumption that this motivation is primary. It is therefore not surprising that, particularly when ingroup threat is high, it becomes diYcult to act on the secondary social harmony goal and the ingroup‐enhancing goal prevails. How does Terror Management Theory (TMT) fit with our conceptualization? TMT proposes that the evolutionarily determined desire for continued life with awareness of the inevitability of death creates the potential for paralyzing terror. Thus, from our perspective, terror appears to activate the ingroup‐enhancing goal. Consistent with this expectation, research has shown that mortality salience increases the tendency to conform to the standards of one’s cultural worldview, makes it harder to violate ingroup norms, and increases estimates of social consensus for culturally relevant attitudes among those holding minority positions (Greenberg, Solomon, & Pyszczynski, 1997). Also consistent with our analysis are the results of Greenberg, Simon, Pyszczynski, and Solomon (1992), who found that although conservatives rated outgroup members more negatively after having been reminded of their mortality, liberals rated outgroups more positively. Although not exactly what we might expect, these results seem consistent with our basic approach. Although we have not tested this in our research, it is not impossible that ingroup‐enhancing and social harmony motivations are accompanied by diVerent emotional responses (Skitka, Bauman, & Mullen, 2004; Smith, 1993). For instance, Skitka et al. (2004) found that fear was more associated with bolstering the ingroup than with derogation of outgroup members, and thus seems more associated with the ingroup‐enhancing variable. On the other hand, anger was more associated with moral outrage and outgroup derogation and thus seems to relate to evaluations of outgroups. Our analysis of the role of social norms ingroup beliefs also follows from general social motivations, although it takes a diVerent overall approach. This research finds that group norms are shared through conversation, that this sharing influences group beliefs, and that people’s beliefs are strongly influenced by the perceptions of the beliefs of their peers—and particularly members of the ingroup. Taken together, we are proposing that group beliefs are in large part normative and that these processes occur in part to allow individuals to feel that they are valued members of social groups. One issue that we have clearly not addressed concerns the origins of intergroup prejudice evaluations and behaviors that move from mere dislike of an outgroup to the perpetration of racial and ethnic discrimination, hate crimes, ethnic cleansing, and genocide (Staub, 1989). Although we propose that such beliefs and behaviors will not be common, they clearly do occur.

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

275

Our approach also leads to the suggestion that when such events occur, they are caused more by social than by individual characteristics. Fear and threat are easily communicated and create powerful social norms that can be used to justify bigotry, discrimination, and racism. Children who grow up in a context in which there is great negativity toward certain groups—Israelis toward Arabs and vice versa, for instance—will almost certainly develop the negative beliefs of their community. Again it is the social motivations and goals that prevail, both for good and evil. It is fortunate that these situations are less common than we might expect, but unfortunate that they continue to exist.

Acknowledgments This research was supported by grant BCS0240877 from the National Science Foundation to Charles Stangor. We thank Diane Ruble for comments on an earlier draft.

References Aboud, F. E. (2003). The formation of ingroup favoritism and outgroup prejudice in young children: Are they distinct attitudes? Developmental Psychology, 39, 48–60. Aboud, F. E., & Amato, M. (2001). Developmental and socialization influences on intergroup bias. In R. Brown and S. Gaertner (Eds.), Blackwell handbook in social psychology (Vol. 4, Intergroup Processes, pp. 65–85). New York and Oxford: Blackwell. Aboud, F. E., & Doyle, A.‐B. (1996). Parental and peer influences on children’s racial attitudes. International Journal of Intercultural Relations, 20, 371–383. Acock, A. C., & Bengtson, V. L. (1978). On the relative influence of mothers and fathers: A covariance analysis of political and religious socialization. Journal of Marriage and the Family, 40, 519–530. Adorno, T. W., Frenkel‐Brunswik, E., Levinson, D. J., & Sanford, R. N. (1950). The authoritarian personality. New York: Harper. Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison‐Wesley. Altemeyer, B. (1988). Enemies of freedom: Understanding right‐wing authoritarianism. San Francisco: Jossey‐Bass. Alvarez, J. M., Ruble, D. N., & Bolger, N. (2001). Trait understanding or evaluative reasoning? An analysis of children’s behavioral predictions. Child Development, 72, 1409–1425. Bigler, R., & Liben, L. (1993). A cognitive‐developmental approach to racial sterotyping and reconstructive memory in Euro‐American children. Child Development, 64, 1507–1518. Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorization and the formation of intergroup attitudes in children. Child Development, 68, 530–543. Blanchard, F. A., Crandall, C. S., Brigham, J. C., & Vaughn, L. A. (1994). Condemning and condoning racism: A social context approach to interracial settings. Journal of Applied Psychology, 79, 993–997.

276

CHARLES STANGOR AND SCOTT P. LEARY

Blanchard, F. A., Lilly, T., & Vaughn, L. A. (1991). Reducing the expression of racial prejudice. Psychological Science, 2, 101–105. Bodenhausen, G. V., Kramer, G. P., & Sasser, K. (1994). Happiness and stereotypic thinking in social judgment. Journal of Personality & Social Psychology, 66, 621–632. Boehnke, K. (2001). Parent‐oVspring value transmission in a societal context: Suggestions for a utopian research design—with empirical underpinnings. Journal of Cross‐Cultural Psychology, 32, 241–255. Brewer, M. B. (1999). The psychology of prejudice: Ingroup love or outgroup hate? Journal of Social Issues, 55, 429–444. Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116–131. Cameron, J., Alvarez, J., Ruble, D., & Fuligni, A. (2001). Children’s lay theories about ingroups and outgroups: Reconceptualizing research on ‘‘prejudice.’’ Personality and Social Psychology Review, 5, 118–128. Clark, K., & Clark, M. (1947). Racial identification and preference in Negro children. In E. Maccoby, T. Newcomb, and E. Hartley (Eds.), Readings in social psychology (pp. 602–611). New York: Holt, Rinehart & Winston. Crandall, C. S., & Stangor, C. (2005). Conformity and prejudice. In J. Dovidio, P. Glick, and L. A. Rudman (Eds.), Reflecting on the nature of prejudice: 50 years after Allport. Malden, MA: Blackwell. Deci, E. L., & Ryan, R. M. (1995). Human autonomy: The basis for true self‐esteem. In M. Kernis (Ed.), EYcacy, agency, and self‐esteem (pp. 31–49). New York: Plenum Press. Diehl, M., & Jonas, K. (1991). Measures of national stereotypes as predictors of the latencies of inductive versus deductive stereotypic judgments. European Journal of Social Psychology, 21, 317–330. Dovidio, J. F., Brigham, J. C., Johnson, B. T., & Gaertner, S. L. (1996). Stereotyping, prejudice, and discrimination: Another look. In C. N. Macrae, C. Stangor, and M. Hewstone (Eds.), Stereotypes and stereotyping (pp. 276–322). New York: The Guilford Press. Dovidio, J., Evans, N., & Tyler, R. (1986). Racial stereotypes: The contents of their cognitive representations. Journal of Experimental Social Psychology, 22, 22–37. Doyle, A., & Aboud, F. (1995). A longitudinal study of white children’s racial prejudice as a social cognitive development. Merrill‐Palmer Quarterly, 41, 210–229. Duckitt, J. (2002). A dual‐process cognitive‐motivational theory of ideology and prejudice. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 33, pp. 41–113). New York: Academic Press. Egan, S. K., & Perry, D. G. (2001). Gender identity: A multidimensional analysis with implications for psychosocial adjustment. Developmental Psychology, 37, 451–463. Federico, C. M., Golec, A., & Dial, J. L. (2005). The relationship between the need for closure and support for military action against Iraq: Moderating eVects of national attachment. Personality and Social Psychology Bulletin, 31(5), 621–632. Fiske, S. T. (2003). Social beings. Hoboken, NJ: John Wiley. Flor, D. L., & Knapp, N. F. (2001). Transmission and transaction: Predicting adolescents’ internalization of parental religious values. Journal of Family Psychology, 15, 627–645. Flouri, E. (2004). Exploring the relationship between mothers’ and fathers’ parenting practices and children’s materialist values. Journal of Economic Psychology, 25, 743–752. Ford, T. E., & Ferguson, M. A. (2004). Social consequences of disparagement humor: A prejudiced norm theory. Personality and Social Psychology Review, 8, 79–94. Ford, T. E., & Stangor, C. (1992). The role of diagnosticity in stereotype formation: Perceiving group means and variances. Journal of Personality and Social Psychology, 63, 356–367.

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

277

Gaertner, S. L., Dovidio, J. F., Anastasio, P. A., Bachman, B. A., & Rust, M. C. (1993). The common ingroup identity model: Recategorization and the reduction of intergroup bias. In W. Stroebe and M. Hewstone (Eds.), European review of social psychology (Vol. 4, pp. 1–26). Chichester, UK: Wiley. Galambos, N. L., Petersen, A. C., Richards, M., & Gitelson, I. B. (1985). The attitudes toward women scale for adolescents (AWSA): A study of reliability and validity. Sex Roles, 13, 343–356. Greenberg, J., & Pyszczynski, T. (1985). The eVect of an overheard ethnic slur on evaluations of the target: How to spread a social disease. Journal of Experimental Social Psychology, 21, 61–72. Greenberg, J., Simon, L., Pyszczynski, T., & Solomon, S. (1992). Terror management and tolerance: Does mortality salience always intensify negative reactions to others who threaten one’s worldview? Journal of Personality and Social Psychology, 63, 212–220. Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory of self‐esteem and cultural worldviews: Empirical assessments and conceptual refinements. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 29, pp. 61–139). New York: Academic Press. Hamilton, D. L. (1981). Stereotyping and intergroup behavior: Some thoughts on the cognitive approach. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 37–81). Hillsdale, NJ: Erlbaum. Hamilton, D. L., & Sherman, J. W. (1994). Stereotypes. In R. S. Wyer and T. K. Srull (Eds.), Handbook of social cognition (Vol. 2, pp. 1–68). Hillsdale, NJ: Erlbaum. Hamm, J. V. (2000). Do birds of a feather flock together? The variable bases for African American, Asian American, and European American adolescents’ selection of similar friends. Developmental Psychology, 36, 209–219. Hansen, C. H., & Hansen, R. D. (1988). Finding the face in the crowd: An anger superiority eVect. Journal of Personality and Social Psychology, 54, 917–924. Harding, J., Proshansky, H., Kutner, B., & Chein, I. (1969). Prejudice and ethnic relations. In J. Harding, H. Proshansky, B. Kutner, and I. Chein (Eds.), Handbook of social psychology (Vol. 5, pp. 1–76). Reading, MA: Addison‐Wesley. Henderson‐King, E. I., & Nisbett, R. E. (1996). Anti‐Black prejudice as a function of exposure to the negative behavior of a single Black person. Journal of Personality and Social Psychology, 71, 654–664. Hewstone, M. (1996). Contact and categorization: Social psychological interventions to change intergroup relations. In C. N. Macrae, C. Stangor, and M. Hewstone (Eds.), Stereotypes and stereotyping (pp. 323–368). New York: The Guilford Press. Heyman, G. D., & Dweck, C. S. (1998). Children’s thinking about traits: Implications for judgments of the self and others. Child Development, 69, 391–403. Hogg, M. A. (1992). The social psychology of group cohesiveness: From attraction to social identity. New York: New York University Press. Hogg, M. A., & Abrams, D. (1990). Social motivation, self‐esteem and social identity. In D. Abrams and M. A. Hogg (Eds.), Social identity theory: Constructive and critical advances (pp. 28–47). New York: Springer‐Verlag. Hogg, M. A., Hardie, E. A., & Reynolds, K. J. (1995). Prototypical similarity, self‐categorization, and depersonalized attraction: A perspective on group cohesiveness. European Journal of Social Psychology, 25, 159–177. Howard, J. W., & Rothbart, M. (1980). Social categorization and memory for ingroup and outgroup behavior. Journal of Personality and Social Psychology, 38, 301–310. Huston, A. C., & Alvarez, M. M. (1990). The socialization context of gender role development in early adolescence. In From childhood to adolescence: A transitional period? (pp. 156–179). Thousand Oaks, CA: Sage Publications Inc.

278

CHARLES STANGOR AND SCOTT P. LEARY

Jackson, D. W., & Tein, J.‐Y. (1998). Adolescents’ conceptualization of adult roles: Relationships with age, gender, work goal, and maternal employment. Sex Roles, 38, 987–1008. Jost, J. T., & Banaji, M. R. (1994). The role of stereotyping in system‐justification and the production of false consciousness. British Journal of Social Psychology, 33, 1–27. Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political conservatism as motivated social cognition. Psychological Bulletin, 129, 339–375. Kaiser, C. R., & Miller, C. T. (2001). Stop complaining! The social costs of making attributions to discrimination. Personality and Social Psychology Bulletin, 27, 254–263. Katz, I., & Hass, R. G. (1988). Racial ambivalence and American value conflict: Correlational and priming studies of dual cognitive structures. Journal of Personality and Social Psychology, 55, 893–905. Kerlinger, F. (1984). Liberalism and conservatism: The nature and structure of social attitudes. Hillsdale, NJ: Erlbaum. Kirkland, S. L., Greenberg, J., & Pyszczynski, T. (1987). Further evidence of the deleterious eVects of overheard derogatory ethnic labels: Derogation beyond the target. Personality and Social Psychology Bulletin, 13, 216–227. Kunda, Z., & Thagard, P. (1996). Forming impressions from stereotypes, traits and behaviors: A parallel constraint satisfaction theory. Psychological Review, 103, 284–308. Lee, Y. T., Jussim, L. J., & McCauley, C. R. (1995). Stereotype accuracy: Toward appreciating group diVerences. Washington, DC: American Psychological Association. Locksley, A., Ortiz, V., & Hepburn, C. (1980). Social categorization and discriminatory behavior: Extinguishing the minimal intergroup discrimination eVect. Journal of Personality and Social Psychology, 39, 773–783. Luhtanen, R., & Crocker, J. (1992). A collective self‐esteem scale: Self‐evaluation of one’s social identity. Personality and Social Psychology Bulletin, 18, 302–318. Lyons, A., & Kashima, Y. (2003). How are stereotypes maintained through communication? The influence of stereotype sharedness. Journal of Personality and Social Psychology, 85, 989–1005. Mackie, D. M., & Goethals, G. R. (1987). Individual and group goals. In C. Hendrick (Eds.), Group processes. Review of personality and social psychology (Vol. 8, pp. 144–166). Newbury Park, CA: Sage Pubications. McCrae, R. R., & Costa, P. T., Jr. (1999). A five‐factor theory of personality. In Handbook of personality: Theory and research (2nd ed., pp. 139–153). New York: Guilford Press. Moreland, R. L. (1987). The formation of small groups. In C. Hendrick (Ed.), Group processes. Review of personality and social psychology (Vol. 8, pp. 80–110). Newbury Park, CA: Sage Publications, Inc. Moskowitz, G. B., Wasel, W., Schaal, B., & Gollwitzer, P. M. (1999). Preconscious control of stereotype activation through chronic egalitarian goals. Journal of Personality and Social Psychology, 77, 167–184. Neuberg, S. L., & Newsom, J. T. (1993). Personal need for structure: Individual diVerences in the desire for simple structure. Journal of Personality and Social Psychology, 65, 113–131. Nieuwbeerta, P., & Wittebrood, K. (1995). Intergenerational transmission of political party preference in the Netherlands. Social Science Research, 24, 243–261. Oakes, P. J., Turner, J. C., & Haslam, S. A. (1991). Perceiving people as group members: The role of fit in the salience of social categorizations. British Journal of Social Psychology, 30, 125–144. O’Bryan, M., Fishbein, H. D., & Ritchey, P. N. (2004). Intergenerational transmission of prejudice, sex role stereotyping, and intolerance. Adolescence, 39, 407–426. Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. Urbana, IL: University of Illinois Press.

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

279

Pettigrew, T. F. (1959). Regional diVerences in anti‐Negro prejudice. Journal of Abnormal and Social Psychology, 59, 28–36. Plant, E. A., & Devine, P. G. (1998). Internal and external motivation to respond without prejudice. Journal of Personality and Social Psychology, 75, 811–832. Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1995). Social dominance orientation: A personality variable predicting social and political attitudes. Journal of Personality and Social Psychology, 67, 741–763. Prentice, D. A., & Miller, D. T. (2003). Pluralistic ignorance and alcohol use on campus: Some consequences of misperceiving the social norm. In Social psychology of health (pp. 183–198). New York: Psychology Press. Raja, S., & Stokes, J. P. (1998). Assessing attitudes toward lesbians and gay men: The Modern Homophobia Scale. Journal of Gay, Lesbian and Bisexual Identity, 3, 113–134. Ritchey, P. N., & Fishbein, H. D. (2001). The lack of an association between adolescent friends’ prejudices and stereotypes. Merrill‐Palmer Quarterly, 47(Special issue), 188–206. Rohan, M. J., & Zanna, M. P. (1996). Value transmission in families. In C. Seligman, J. M. Olson, and M. P. Zanna (Eds.), The psychology of values: The Ontario symposium (Vol. 8, pp. 253–276). Mahwah, NJ: Lawrence Erlbaum Associates. Ruble, D., & Martin, C. (1998). Gender development. In W. Damon (Ed.), Handbook of child psychology (5th ed., pp. 933–1016). New York: Wiley & Sons. Rudman, L. A., Greenwald, A. G., Mellott, D. S., & Schwartz, J. L. K. (1999). Measuring the automatic components of prejudice: Flexibility and generality of the Implicit Association Test. Social Cognition, 17, 437–465. Ruscher, J. B. (1998). Prejudice and stereotyping in everyday communication. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 30, pp. 241–307). San Diego, CA: Academic Press. Ruscher, J. B. (2001). Prejudiced communication: A social psychological perspective. New York: Guilford. Ruscher, J. B., & Duval, L. L. (1998). Multiple communicators with unique target information transmit less stereotypical impressions. Journal of Personality and Social Psychology, 74, 329–344. Schaller, M., & Conway, G. (1999). Influence of impression‐management goals on the emerging content of group stereotypes: Support for a social‐evolutionary perspective. Personality and Social Psychology Bulletin, 25, 819–833. Schaller, M., Conway, L. G., & Tanchuk, T. L. (2002). Selective pressures on the once and future contents of ethnic stereotypes: EVects of the communicability of traits. Journal of Personality and Social Psychology, 82, 861–877. Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 25, pp. 1–65). San Diego, CA: Academic Press. Schwartz, S. H., & Bilsky, W. (1990). Toward a theory of the universal content and structure of values: Extensions and cross‐cultural replications. Journal of Personality and Social Psychology, 58, 878–891. Sears, D. O. (1983). The person‐positivity bias. Journal of Personality and Social Psychology, 44, 233–250. Sechrist, G., & Stangor, C. (2001). Perceived consensus influences intergroup behavior and stereotype accessibility. Journal of Personality and Social Psychology, 80, 645–654. Sherif, M., & Sherif, C. W. (1953). Groups in harmony and tension. New York: Harper & Row. Sidanius, J., & Ekehammar, B. (1979). Political socialization: A multivariate analysis of Swedish political attitude and preference data. European Journal of Social Psychology, 9, 265–279.

280

CHARLES STANGOR AND SCOTT P. LEARY

Sinclair, L., & Kunda, Z. (2000). Motivated stereotyping of women: She’s fine if she praised me but incompetent if she criticized me. Personality and Social Psychology Bulletin, 26, 1329–1342. Skitka, L. J., Bauman, C. W., & Mullen, E. (2004). Political tolerance and coming to psychological closure following the September 11, 2001, terrorist attacks: An integrative approach. Personality and Social Psychology Bulletin, 30, 743–756. Smith, E. R. (1993). Social identity and social emotions: Toward new conceptualizations of prejudice. In AVect, cognition, and stereotyping: Interactive processes in group perception (pp. 297–315). San Diego, CA: Academic Press Inc. Smith, E. R., & Mackie, D. M. (2000). Social psychology (2nd ed.). New York: Psychology Press. Snyder, M., & Miene, P. (1994). On the function of stereotypes and prejudice. In M. Zanna and J. Olson (Eds.), The psychology of prejudice: The Ontario Symposium (Vol. 7, pp. 33–54). Hillsdale, NJ: Erlbaum. Stangor, C., & Lange, J. (1994). Mental representations of social groups: Advances in conceptualizing stereotypes and stereotyping. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 26, pp. 357–416). New York: Academic Press. Stangor, C., Leary, S., Freidus, R., & Ottenbreit, A. (2005a). Social values and the prejudiced personality. Manuscript in Preparation, University of Maryland. Stangor, C., Leary, S., Freidus, R., & Ottenbreit, A. (2005b). The transmission of intergroup beliefs among parents and peers. University of Maryland. Manuscript in Preparation. Stangor, C., & McMillan, D. (1992). Memory for expectancy—congruent and expectancy— incongruent information: A review of the social and social developmental literatures. Psychological Bulletin, 111, 42–61. Stangor, C., & Schaller, M. (1996). Stereotypes as individual and collective representations. In C. N. Macrae, Stangor, and M. Hewstone (Eds.), Stereotypes and stereotyping (pp. 3–40). New York: The Guilford Press. Stangor, C., & Sechrist, G. B. (1998). Conceptualizing the determinants of academic choice and task performance across social groups. In J. K. Swim and Stangor (Eds.), Prejudice: The target’s perspective. Santa Barbara, CA: Academic Press. Stangor, C., Sechrist, G. B., & Jost, J. T. (2001a). Social influence and intergroup beliefs: The role of perceived social consensus. In J. P. Forgas and K. D. Williams (Eds.), Social Influence (pp. 235–252). Philadelphia, PA: Psychology Press. Stangor, C., Sechrist, G., & Jost, J. (2001b). Changing racial beliefs by providing consensus information. Personality and Social Psychology Bulletin, 27, 484–494. Stangor, C., Swim, J. K., Sechrist, G. B., deCoster, J., Van Allen, K., & Ottenbreit, A. (2003). Ask, answer and announce: Three stages in perceiving and responding to discrimination. European Review of Social Psychology, 14, 277–311. Stangor, C., Swim, J. K., Van Allen, K., & Sechrist, G. (2001). Reporting discrimination in public and private contexts. Journal of Personality and Social Psychology, 82, 69–74. Staub, E. (1989). The roots of evil: The origins of genocide and other group violence. New York, NY: Cambridge University Press. Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual performance of African Americans. Journal of Personality and Social Psychology, 69, 797–811. Tajfel, H. (1981). Human groups and social categories: Studies in social psychology. Cambridge: Cambridge University Press. Tajfel, H., & Turner, J. C. (2004). The social identity theory of intergroup behavior. In Political psychology: Key readings (pp. 276–293). New York: Psychology Press. Tesser, A. (1993). The importance of heritability in psychological research: The case of attitudes. Psychological Review, 100, 129–142.

INTERGROUP BELIEFS: INVESTIGATIONS FROM THE SOCIAL SIDE

281

Tetlock, P. E. (1986). A value pluralism model of ideological reasoning. Journal of Personality and Social Psychology, 50, 819–827. Theimer, C., Killen, M., & Stangor, C. (2001). Young children’s evaluations of exclusion in gender‐stereotypic peer contexts. Developmental Psychology, 37, 1–10. Tomkins, S. (1964). The polarity scale. New York: Springer. Turner, J. C. (1975). Social comparison and social identity; some prospects for intergroup behavior. European Journal of Social Psychology, 5, 5–34. Webster, D. M., & Kruglanski, A. W. (1994). Individual diVerences in need for cognitive closure. Journal of Personality and Social Psychology, 67, 1049–1062. Wheeler, S. C., & Petty, R. E. (2001). The eVects of stereotype activation on behavior: A review of possible mechanisms. Psychological Bulletin, 127, 797–826. Williams, J., Best, D., Boswell, D., Mattson, L., & Graves, D. (1975). Preschool racial attitude measure II. Educational and Psychological Measurement, 35, 3–18. Wills, T. A. (1981). Downward social comparison principles in social psychology. Psychological Bulletin, 90, 245–271. Yee, M., & Brown, R. (1994). The development of gender diVerentiation in young children. British Journal of Social Psychology, 33, 183–196.

A MULTICOMPONENT CONCEPTUALIZATION OF AUTHENTICITY: THEORY AND RESEARCH

Michael H. Kernis Brian M. Goldman

And if by chance I wake at night and I ask you who I am, oh take me to the slaughterhouse I will wait there with the lamb. —Leonard Cohen Whatever satisfies the soul is truth. —Walt Whitman I prefer to be true to myself, even at the hazard of incurring the ridicule of others, rather than to be false, and to incur my own abhorrence. —Frederick Douglass

In this chapter, we present research and theory pertaining to our multicomponent perspective on authentic functioning. We begin with a historical account of various philosophical perspectives on authentic functioning and briefly review several past and contemporary psychological perspectives on authenticity. We then define and discuss our multicomponent conceptualization of authenticity and describe each of its components and their relationships to other constructs in the psychology literature. Next, we present an individual diVerences measure we have developed to assess dispositional authenticity and each of its components, and we report findings attesting to the adequacy of its psychometric properties. In addition, we present findings from a variety of studies we have conducted to examine how authenticity relates to diverse aspects of healthy psychological and interpersonal functioning. These studies pertain to a wide range of phenomena, including the following: verbal defensiveness, mindfulness, coping styles, self‐concept structure, social‐role 283 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 38 DOI: 10.1016/S0065-2601(06)38006-9

‘‘Authenticity Inventory’’ Copyright ß 2006 Brain M. Goldman and Michael H. Kernis 0065-2601/06 $35.00

284

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

functioning, goal pursuits, general well‐being, romantic relationships, parenting styles, and self‐esteem. Following this, we discuss potential downsides or costs for authentic functioning and describe some future directions for research on authenticity.

I. A Historical Overview of Authenticity Poets, painters, clergy, scholars, philosophers, and scientists have long sought to define who one ‘‘really’’ is. Descriptions of authentic functioning are found among a variety of works and disciplines across the arts and sciences. However, these descriptions are often vague, relegated to peripheral segments of larger works, and lack continuity in their lineage or origin. At times, descriptions of authenticity seem to be at the ‘‘limits of language,’’ being loosely described in such diverse topics as ethics, well‐being, consciousness, subjectivity, self‐processes, and social or relational contexts, or characterized in terms of its opposite (i.e., inauthenticity), with references to inauthentic living, false‐self behaviors, or self‐deception. Despite such limitations, contemporary psychological views of authenticity owe a great debt to the works of philosophy. Within the field of philosophy, authenticity is loosely set within topics, such as metaphysics or ontology, firmly entrenched in particular movements, such as existentialism or phenomenology, and localized to specific authors like Sartre or Heidegger. In the following section, we identify and discuss some of the historical ideas and perspectives within philosophy that contribute to the development of the concept of psychological authenticity. This historical summary points to a portrayal of authenticity as involving a variety of themes. Most notably, authentic functioning is characterized in terms of people’s (1) self‐understanding, (2) openness to objectively recognizing their ontological realities (e.g., evaluating their desirable and undesirable self‐aspects), (3) actions, and (4) orientation towards interpersonal relationships. Portrayals of authentic functioning date back to the Ancient Greek philosophers. Perhaps, the earliest account dates back to Socrates’ stance that the ‘‘unexamined’’ life is not worth living. While self‐inquiry is paramount for Socrates, in his work Nicomachean Ethics Aristotle emphasized the importance of actions. Aristotle viewed ethics in terms of people’s pursuit of the ‘‘higher good.’’ Specifically, he proposed that the highest good is ‘‘activity of the soul in accordance with the best and most complete virtue in a complete life’’ (Hutchinson, 1995). Such pursuits are intimately tied with people’s well‐ being (Waterman, 1993). From this view, well‐being (i.e., ‘‘eudaimonia’’) is attained through self‐realization, that is, by performing activities that reflect

AUTHENTICITY

285

one’s true calling. Such activities do not have happiness or pleasure as their desired end; instead, pleasure is a consequence of a life in which one successfully manages to perform these activities well. This view seems akin to existential philosophers like Nietzsche and Kierkegaard (May, 1960, p. 22), who described man as ‘‘the organism who makes certain values—prestige, power, tenderness, love—more important than pleasure and even more important than survival itself. ’’ The similarity among these perspectives, and many of the subsequent perspectives discussed in this section, is the portrayal of people in a manner that transcends measuring success primarily via hedonic qualities (e.g., happiness), or even basic evolutionary success (e.g., survival). What emerges in its place is a broad depiction of people as being rich in complexity, actively and intentionally pursuing a life in accord with their deepest potentials. Aristotle also discussed how people’s pursuit of the higher good involves diVerent virtues (e.g., continence, pleasure, friendship, and theoretical wisdom). Whereas the highest good refers to the end that people pursue for its sake only, every other good is pursued for the sake of the highest good (Hutchinson, 1995). As such, the described relationship between pursuit of the good and highest good seems to underscore a sense of unity or integration among people’s pursuits—a perspective that contemporary self‐theorists would suggest reflects self‐organization (e.g., Donahue, Robins, Roberts, & John, 1993; Showers & Ziegler‐Hill, 2003), integrated self‐regulation (e.g., Deci & Ryan, 2000), or self‐concordance (Sheldon & Elliot, 1999). Aristotle’s contribution to conceptualizing authenticity is in having paved a connection between people’s self‐knowledge and behavioral self‐regulation. In his view, knowledge of the highest good significantly aVects peoples’ lives because it allows them to organize their lives well ‘‘like an archer with a target to aim at’’ (Irwin, 2003). Thus, from this perspective, authentic functioning is the result of sustained activity in concert with a deeply informed sense of purpose. Renee Descartes’ Meditations oVers a variety of concepts and insights relevant to conceptualizing authenticity. Descartes’ perspective demonstrated a radical departure from his predecessors. According to Groscholz (2003), prior to Descartes, philosophers asked: What must the world be like for it to be intelligible? Following Descartes, they asked: What must the mind be like for the world to be intelligible to it? This shift in focus demonstrates the centrality of cognitive processes in directing and interpreting experience. While such a view clearly advances the role of psychological functioning in experience, perhaps Descartes’ greatest contribution to conceptualizing authenticity lies in his emphasis on subjectivity in mental processes. Descartes’ proclamation ‘‘I think, therefore I am’’ suggests that what ‘‘I am’’ is a thing that thinks; a thing that doubts, understands, aYrms, denies, is willing, unwilling, imagines, and has sensory perceptions. In contrast to

286

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

the epistemological precedent established by Aristotle, Descartes rejected the notion that all knowledge originates in sense perception and sense perception is our conduit to external things (Grosholz, 2003). In lieu of Aristotelian epistemology, Descartes proposed that if one can know objects, one must firstly learn to think them, or reason upon them. Subsequently, with mistrust, one may rely on sensory perceptions, abstracting from them and correcting them, in light of the constructions of reason (Grosholz, 2003). Descartes demonstrates the importance of subjectivity in the case of a piece of wax just taken from a honeycomb. Presumably, the wax may be conceived of as an object of sense perception—retaining some of the scent of the flowers from which it was gathered. However, the piece of wax is not merely an object sensed by sensory perceptions, but rather it is also always thought, that is, submitted to being understood or reasoned upon (Grosholz, 2003). Whereas the qualities of the wax delivered by sense perception depend on sensory information, what one can think about the piece of wax, as it melts, or diminishes in its smell, is what remains constant under all transformations. Thus, individuals may doubt sensory perceptions about objects (since sense perception is just a modality of awareness), but they cannot doubt that they are aware of their perceptions of the objects (Grosholz, 2003). From this perspective, what validates the ontological reality of the object (e.g., what really constitutes the wax) is the quantifiable mental scrutiny of it. That is, the certainty that individuals’ place on known objects is not caused by the objects’ objective reality; rather, certainty of the object results from a formal subjective process of consciousness, constructed by reason. Knowledge of an object is not a function of the contents of an object, but of the contents of our consciousness and mental activities regarding the object. As such, by relying on the formal process of mentally scrutinizing their consciousness, people may attain clarity and distinctiveness in their idea of things, and thereby grasp their very essence (Grosholz, 2003). What then, if the object of one’s attention is one’s ‘‘self ’’? Philosophers like Descartes, Kant, and Dewey struggled with the role of self‐consciousness in people’s emotion, will, and thinking (Hoyle, Kernis, Leary, & Baldwin, 1999). However, conscious attention that regarded the falseness of others’ behaviors seems to have emerged within a particular cultural context. For instance, the cultural historian Burckhardt (Winter & Barenbaum, 1999) concluded that people of the middle ages were conscious of themselves only as a member of a general category, for example, race, party, family, or corporation. Subsequent to the Renaissance, people construed themselves as individuals with personal attributes. Such societal changes appear to have corresponded with people’s specific concerns in perceiving others’ authentic

AUTHENTICITY

287

functioning. Harter (1999) describes this as the historical emergence of interest in false‐self behavior. According to Baumeister (1987), people of the 16th century became interested in distinguishing between others’ private concealment from that which was observable in them. Similarly, Trilling (1971) discusses themes of deception and pretense found among English politics, philosophy, and literature (e.g., Shakespeare). Concerns about self‐concealment were, in Baumeister’s (1987) view, initially limited to perceptions of others—were people hiding their true‐selves from others? Baumeister (1987) notes that with the arrival of Puritanism so too emerged concerns over whether individuals were deceiving themselves. Determining whether one’s own actions were true or false depended on a consideration of one’s standing on characteristics deemed necessary for one to possess in order to enter into heaven (i.e., piety, faith, and virtue). Thus, authentic functioning from this perspective (i.e., being one’s true‐self ) involves regulating one’s actions to be in accord with religious dictates. Historical perspectives on false‐self behaviors demonstrate the vital role that cultural contexts play in people’s perceptions of their own and others’ authenticity. In many respects, false‐self behaviors represent the lower end of an authenticity continuum (i.e., the relative absence of authentic action or experience). Contemporary interest in false‐self behaviors is evident in such varied topics as self‐monitoring (Snyder, 1987), impression management and strategic self‐presentations (GoVman, 1959; Leary, 1995; Schlenker, 1980), and voice (Gilligan, 1982; Harter, Waters, & Whitesell, 1997). In terms of conceptualizing authenticity, the notion of false‐self behaviors reflects the continual tension between the person and the social structure—the interface of personal inclinations and social obligations that form the stage on which authenticity is portrayed. Many of the works from middle‐age philosophers were consistent with the Puritanical interpretation of authenticity by equating falsehood with nonconformity to religious prescriptions. In contrast, philosophy from the Enlightenment and onward often challenged the premise that authentic functioning occurs through acting in accordance with prescribed religious doctrines, or any learned social conventions. For instance, philosophers like Hobbes and Hume discussed morality and the structure of social contexts as central features of ontological concerns. Hume asserted that the concept of oneself is one that people derive through their social interactions with others—a position championed by symbolic interactionists (Cooley, 1902; Meade, 1934) and advocated by current psychological theorists who emphasize the reflected self (e.g., Tice & Wallace, 2003). Thus, Hume asserted that morality and authenticity are

288

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

best understood through the relationships that connect individuals to others. Specifically, Hume described morality in terms of how people judge ‘‘virtues’’—behavior that produces pleasure or reduces pain for the actor or for others (Wilson, 2003). Whereas ‘‘artificial’’ virtues are those that depend on social conventions, and that people evaluate for its social prudence, ‘‘natural’’ virtues reflect behaviors people would perform even if there were no need for social conventions to regulate their occurrence (Wilson, 2003). Thus, in contrast to the implicit conformity found in artificial virtue pursuits, natural virtues, similar to Aristotle’s notion of the pursuit of the higher good, are actions taken for their own sake. Furthermore, Hume describes such actions as emanating from relational concerns that promote social well‐being, and thus, by extension promote individuals’ personal well‐ being. Thus, the distinction between artificial and natural virtues provides an important basis for further diVerentiating authentic functioning in terms of people’s motives, as opposed to merely actions taken in tandem with prescribed social norms. Moreover, Hume’s views provide an important historical basis for considering interpersonal concerns as central to authentic functioning. With the onset of developments in existential philosophy around the 19th century, metaphysical critiques often equated conformity to religious conventions with inauthentic functioning. As a precursor to the Existential movement, Kierkegaard asserted that authentic functioning reflects subjectivity in choices that involve people’s ‘‘essential knowing’’—knowledge that concerns the deepest meanings of their existence. Objective certainty of essential knowledge is neither final nor complete, and thus, its truth is always an approximation (Westphal, 2003). Kierkegaard also observed that cultural institutions tend to produce pseudo‐individuals (i.e., stereotyped members of ‘‘the crowd’’). Whereas ‘‘the crowd is untruth,’’ Kierkegaard states, ‘‘truth is subjectivity’’ (Kierkegaard, 2004). In response to objective uncertainty and institutionalized identity production, individuals must take responsibility for their existential choices (e.g., their choices regarding who they will be) and become who they are beyond culturally imposed identities (McDonald, 2005). In becoming their self, individual’s existential anxiety is aroused. That is, people experience ambivalence regarding how to be, experiencing joy and excitement for their freedom, yet dread for self‐repudiation and the responsibility for choosing how to be. Existential anxiety reflects a form of self‐ alienation or as Kierkegaard (2004, p. 26) put it, ‘‘there is an interiority that is incommensurable with exteriority.’’ Through a process of becoming their own self, individuals pass through the stage of self‐alienation, and subsequently rely on their subjective faith to energize and organize their chosen actions toward their absolute end/goal (e.g., their essential purpose).

AUTHENTICITY

289

For Kierkegaard, faith is not rooted in the clarity of people’s knowledge (e.g., ‘‘How certain is my knowledge?’’), but rather in the embracing of their paradoxes in spite of their absurdities (e.g., ‘‘How deep is my commitment to what I fallibly take to be true?’’). Faith signifies a particular cognitive stance (e.g., recognition of the absurd) that involves a radical transformation in one’s life (Westphal, 2003). The challenge of this transformation occurs in a ‘‘process of highest inwardness’’—whereby people accept their faith as normative and orient their actions toward becoming their ‘‘innermost’’ selves. From this perspective, authentic functioning is not attainable via learning and conforming to norms derived from external dogmatic beliefs (be they religious, or otherwise). Rather, authentic functioning occurs when individuals choose to be in accordance with their absolute end/goal. Personally, for Kierkegaard, existence emerges as a philosophical problem to embrace the paradoxical presence of God, by smuggling ‘‘Christianity out of the system of Christendom.’’ A generation later, Neitzsche’s ‘‘philosophy of the future’’ sought to deconstruct the interpretations and evaluations implicit in cultural authorities (including the prior teachings of philosophy itself). According to Neitzsche, absolutisms in social categories, such as ‘‘good and evil,’’ needed to be reinterpreted and revalued (e.g., ‘‘beyond good and evil’’). By abandoning any, and all culturally constructed absolutes, nihilism emerges—the recognition that life has no intrinsic meaning. In light of this recognition, Nietzsche proposed that some people would fall victim to despair. Alternatively, Halling and Carroll (1999, p.97) note Nietzsche’s proposal of the emergence of a new person—the Ubermensch (‘‘Overman’’), a ‘‘creator of authentic values.’’ The Ubermensch represents a particular mode of existence, found in a person who goes beyond a mere nihilistic devaluation of all prevailing values, to make possible a ‘‘revaluation of values’’ (Schacht, 2003). Thus, by people ‘‘naturalizing’’ their self understanding to fit within a reinterpreted sensible context of their constitution, resources, and circumstances, people realize their life‐aYrming potential. Schacht (2003, p. 412) describes this state as ‘‘a fundamental expression reflecting how one is or how one has come to be constituted,’’ noting that ‘‘it signals no abandonment of commitment to truthfulness, but rather the ascent to a further, highest humanly possible form of it.’’ Thus, for Nietzsche and Kierkegaard alike, the essence of people’s being is unfounded in objective inventories designed to measure what they are, but rather, people’s essence is understood in terms of their way of being. This idea that no general or uniform account of what it means to be human can be put forth, because the meaning of being is decided in and through existence itself, is captured in Sartre’s infamous existential slogan: ‘‘existence precedes essence’’ (Crowell, 2005). Thus, Sartre (2004, p. 344) puts forth the view that subjectivity must be the starting point on which people’s essence is

290

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

predicated. Whereas entities are defined in terms of their essential properties (e.g., what type or kind of thing they are), the essence of people is not fixed by their type, but rather by what they make of themselves. Existential psychologist Rollo May (1960, p. 17) amplifies this view through his assertion ‘‘that only as we aYrm our existence do we have any essence at all.’’ Within the realm of existential philosophy, the studies of Martin Heidegger and Jean‐Paul Sartre are generally regarded as prototypes for characterizing authenticity. While both philosophers employed a phenomenological methodology, Heidegger is credited with having united existential concerns with the phenomenological method posed by the phenomenologist Edmund Husserl . Husserl proposed that through the psychological process epoche´, people clear away their preconceptions about experience and return ‘‘to the things themselves’’ (Halling & Carroll, 1999). This process relies on people’s intentionality—the interaction between the subjective and objective components of consciousness. Rather than just passively registering an object’s existence, people ‘‘cocreate’’ phenomena through intentionality (Halling & Carroll, 1999). Heidegger (1968) implemented the phenomenological method in seeking to understand the question ‘‘What is meant by being?’’ Heidegger framed existence (Dasein or Being‐there) with respect to both its historical and temporal aspects. For Heidegger, authentic possibility exists in relation to Geworfenheit (i.e., ‘‘thrown‐ness’’). Thrown‐ness refers to the idea that people are born into a world that they did not construct, live amid conditions over which they have little control, and are insuYciently equipped to determine solutions to existential questions such as ‘‘Who am I?’’ Consequently, the totality of people’s behaviors is at first a function of the behavioral prescriptions derived from the social environment. In light of the constraints of their ‘‘thrown‐ness,’’ and the inevitability of their finitude (e.g., death), people can embrace their individuality and freedom to live authentically (Halling & Carroll, 1999). By counteracting their thrown‐ness and imminent finitude, the whole of Dasein’s activity—people’s ‘‘Being‐in‐the‐world’’—gains significance from the purpose or aim to which they understand themselves as existing (Heidegger, 1968). Authentic possibility occurs in the condition of self‐making, when having been confronted with the ‘‘nothingness’’ of their existence (e.g., acting solely in accord with social norms), individuals transform their mode of being to reflect a sense of care (i.e., assumed responsibility) toward others and their being themselves. ‘‘Being‐in‐the‐world’’ does not constitute the self as an independent isolate of the world, but rather it reflects an existential modification of how one exists with others (Heidegger, 1968). When such a transformation occurs, the activity of Dasein is governed by the project of existential possibility in which people ‘‘make themselves.’’ Accordingly,

AUTHENTICITY

291

authenticity in German, Eigentlichkeit refers to the attitude through which individuals engage their projects as their own (Crowell, 2005). From Heidegger’s perspective, authentic functioning reflects people resolutely choosing to act with care those projects that permit their Being‐ in‐the‐World. Moreover, authentic functioning is marked by a sense of unity among the temporal and historical aspects of existence. For instance, Crowell (2005) refers to existential temporality in which the future (the possibility aimed at by one’s projects) recollects the past (what no longer needs to be done or completed) in order to give meaning to the present (the things that take on significance in light of what currently needs doing). These facets of existential temporality resemble various cognitive‐motivational terms used by contemporary psychologists to describe people’s purposive behavior (e.g., Cantor & Zirkel, 1990) and seem relevant to the notion of personal narratives or self‐stories (e.g., Gergen & Gergen, 1988; McAdams, 1995, 1999). In particular, existential temporality complements the fundamental concerns of Hermans’ Valuation Theory (1987) in which people’s personal construction of meaning is examined with respect to specific spatio‐temporal instances ascribed to their life stories (Hermans, Rijks, & Kempken, 1993). Thus, an authentic existence is one in which people understand their choices and commit themselves to enact those projects that give shape to their existence. For Sartre, people’s way of ‘‘being’’ is inextricably linked to their choices. Similarly, contemporary psychological theories of motivation (Deci & Ryan, 2000) and psychological well‐being (RyV, 1989) place a premium on people’s autonomy. In Sartre’s view, we are our choices: ‘‘to be’’ is to choose; ‘‘to cease to choose’’ is to cease to be (Flynn, 2003). While Sartre’s basic message attests to people’s conscious decisions and their responsibility for their actions (or inaction), such choices are noted to occur within situations themselves. More specifically, Sartre describes situations in terms of a synthesis of a person’s ‘‘facticity’’ (e.g., life’s givens, such as a person’s past experience, psychological properties, and broader sociocultural milieu) and one’s ‘‘transcendence’’ (e.g., the willful agent capable of going beyond, or surpassing the situations’ facticity). Actions governed by facticity reflect a particular form of determinism, a predilection toward what practically ‘‘is’’ in the situation. Alternatively, actions governed by transcendence reflect a predilection toward what can be. By recognizing that they are radically free to ‘‘choose’’ otherwise, to be other than the way they ‘‘are’’ (e.g., beyond their facticity alone), people exhibit a form of self‐ negation expressed as existential angst. Thus, the kind of being one is, reflects the choices and decisions one makes amid the facts and the possibilities of the situation. In this respect, Sartre frames authentic functioning as a particular instance of peoples’ behavioral self‐regulation. That is, authentic

292

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

actions reflect the intrapsychic resolve that emerges from the choices found among the operative self‐schemas governing individuals’ situated freedom (cf., involving the actual self and the possible self, Markus & Nurius, 1986). For Sartre, ‘‘Bad faith’’ emerges when individuals lie or deceive themselves about their ontological duality. Such deceptions occur when people either dissolve the possibilities of transcendence in the throws of ‘‘facticity,’’ or conversely when they act with only sheer ‘‘transcendent’’ will, and ignore the facts of the situation. Thus, authentic functioning from this perspective emerges when individuals openly embrace the ontological duality of their situated freedom when deciding on how they will behave.

II. Taking Stock of These Various Perspectives: Towards a Psychological View of Authenticity This brief and necessarily selective historical account of philosophical perspectives on authenticity demonstrates the construct’s richness and complexity. These perspectives depict various themes and help to illuminate the development of the construct. First, authenticity reflects self‐understanding. Whereas Socrates equated self‐examination with the very value of a person’s existence, other philosophers emphasized the importance of self‐understanding in organizing one’s actions. Thus, a second aspect of authenticity involves behaviors that are rooted in self‐knowledge, as in Aristotle’s ‘‘pursuit of the highest good,’’ Heidegger’s notion of ‘‘project,’’ Kierkegaard’s essential knowledge and subjective truth, and Husserl’s intentionality. Moreover, authentic behavior reflects particular actions, actions expressive of people’s values (e.g., Hume, Nietzsche), and that are freely chosen with a sense of agency (e.g., Sartre, Kierkegaard, & Heidegger). Third, authentic functioning reflects people’s willingness and capacity for objectively acknowledging and accepting their core self‐aspects. That is, authenticity reflects the relative absence of self‐deception and the relative presence of unbiased recognition of self‐relevant information, including ontological realities (e.g., consider the discussion on false‐self behaviors, or Sartre’s discussion of facticity and transcendence). Fourth, authentic functioning involves a particular orientation towards others (e.g., Heidegger’s notion of Being‐in‐the‐World). Taken as a whole, authentic functioning also reflects a set of processes. The notion of authenticity reflecting a set of processes is essential to the perspectives discussed from Kierkegaard through Sartre. Collectively, the existential philosophy perspective couches authenticity as occurring when people freely choose to commit themselves to engage their activities

AUTHENTICITY

293

with agency, in a process of self‐authoring their way of being. In this respect, the existential view of authenticity is consistent with Trilling’s (1971) description of the Greek ancestry of the word authentic, authenteo, meaning ‘‘to have full power.’’ That is, authentic functioning is reflected in an individual being ‘‘the master of his or her own domain.’’ Taken as a whole, this historical overview of authenticity documents a variety of mental and behavioral processes that account for how individuals discover, develop, and construct a core sense of self and, furthermore, how this core self is maintained over time and situation. While various historical accounts emphasize that authenticity involves a union between thought and action, they often place a premium on whether these actions originate within the self or without by societal expectations, norms, or pressures. We will see many of these same themes in psychological perspectives on authenticity. In the following section, we briefly discuss authenticity from the perspective of several humanistically oriented psychological frameworks and describe how these frameworks informed our own conceptualization of authenticity, to which we then turn.

III. Psychological Perspectives on Authenticity Self‐determination theory (SDT) (Deci, 1980; Deci & Ryan, 1985; Ryan & Deci, 2000, 2002) holds that people are authentic when their actions reflect their true‐ or core‐self, that is, when they are autonomous and self‐determining. Our multicomponent framework of authentic functioning owes a great deal to this conceptualization. Hodgins and Knee (2002) capture many aspects of this convergence in their description of autonomously functioning individuals. For example, they suggest that autonomously functioning individuals ‘‘will meet the continually changing stream of consciousness experience with openness. By ‘openness’ we mean a readiness to perceive ongoing experience accurately, without distorting or attempting to avoid the experience, and a willingness to assimilate novel experiences into self‐structures’’ (p. 88). They further suggest that autonomously functioning individuals ‘‘grow toward greater unity in understanding and functioning’’ (p. 88), ‘‘have a high tolerance for encountering experience without being threatened or defending against it’’ (p. 88–89), ‘‘feel choiceful and endorsing of their behavior’’ (p. 90), and exhibit ‘‘greater honesty in interactions of all types’’ (p. 90). According to SDT, self‐ determination is one of three basic psychological needs (the others being competence and relatedness), the satisfaction of which is critical for optimal psychological health and well‐being. Considerable research supports this claim (Deci & Ryan, 2000).

294

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

Our conceptualization of authenticity also owes a great deal to Rogers’ (1961) conceptualization of a self‐actualizing or fully functioning individual (Maslow, 1968), who possesses the following characteristics (Cloninger, 1993). First, the fully functioning individual is open to experience, both objective and subjective, that life has to oVer. Accompanying this openness is a tolerance for ambiguity and the tendency to perceive events accurately, rather than defensively distorting or censoring them from awareness. Second, fully functioning individuals can live fully in the moment, they are adaptable and flexible, and they experience the self as a fluid process rather than a static entity. Third, they inherently trust their inner experiences to guide their behaviors. Fourth, a fully functioning person experiences freedom. This freedom may be reflected in the attitudes one adopts toward experiences—even if the environment is immovable, one still has a choice about how to respond and feel about it. Fifth, the fully functioning individual is creative in his or her approach to living, rather than falling back on well‐established modes of behavior that become unnecessarily restrictive. This creativity is fueled by a strong trust in one’s inner experiences and a willingness to adapt to ever‐changing circumstances.

IV. A Multicomponent Conceptualization of Authenticity We have seen that most perspectives on authenticity stress the extent to which one’s thoughts, feelings, and behaviors reflect one’s true‐ or core‐self. Moreover, most perspectives emphasize a nondefensive stance toward evaluative information, openness toward, and trust in, internal experiences, and fulfilling interpersonal relationships. In line with these perspectives, we (Goldman & Kernis, 2002; Kernis, 2003; Kernis & Goldman, 2005a,b) define authenticity as the unobstructed operation of one’s true‐ or core‐self in one’s daily enterprise. However, instead of viewing authenticity as a single unitary process, we suggest that authenticity can be broken down into four separate, but interrelated, components. We refer to these components as awareness, unbiased processing, behavior, and relational orientation. Each of these components focuses on an aspect of authenticity that, while related to each of the others, is distinct. We turn now to a description of each component. A. AWARENESS The awareness component refers to possessing, and being motivated to increase, knowledge of and trust in one’s motives, feelings, desires, and self‐relevant cognitions. It includes, for example, knowing what type of food one likes and dislikes, how motivated one is to lose weight, whether one is

AUTHENTICITY

295

feeling anxious or depressed, in what circumstances one is most likely to be talkative, whether one desires to attend graduate or professional school, and so forth. Moreover, it involves being motivated to learn about such things as one’s strengths and weaknesses, goals and aspirations, dispositional characteristics, and emotional states. Having knowledge about one’s propensities and characteristics (i.e., of one’s true‐self ) promotes the integration of one’s inherent polarities into a coherent and multifaceted self‐representation. As Perls and his colleagues (Perls, HeVerline, & Goodman, 1951) and many others have suggested, people are not masculine or feminine, introverted or extroverted, emotional or stoic, and so forth. Instead, while one aspect of these dualities (‘‘figure’’) generally predominates over the other (‘‘ground’’), individuals invariably possess both aspects to some degree. As people function with greater authenticity, they become more aware of the fact that they possess these multifaceted self‐aspects and strive to integrate them into a cohesive self‐ structure. In short, awareness involves knowledge and acceptance of one’s multifaceted and potentially contradictory self‐aspects (i.e., being both introverted and extraverted), as opposed to rigid acknowledgement and acceptance only of those self‐aspects deemed internally consistent with one’s overall self‐concept. As we have noted elsewhere (Kernis & Goldman, 2005a,b), our view diVers from J. Campbell’s conceptualization of self‐concept clarity (Campbell, 1990; Campbell et al., 1996) and is more closely aligned with Sande, Goethals, and RadloV ’s (1988) approach to the multifaceted self‐concept. According to Campbell, endorsing as self‐descriptive both adjectives that reflect endpoints of bipolar trait dimensions (e.g., introversion, extraversion) reflects an internally inconsistent self‐concept. In contrast, for Sande et al. (1988), such an endorsement strategy reflects a multifaceted self‐concept. We believe that this apparent contradiction can be resolved by taking into consideration Paulhus and Martin’s (1988) concept of functional flexibility. Functional flexibility involves having confidence in one’s ability to call into play multiple, perhaps contradictory, self‐aspects in dealing with life situations. An individual high in functional flexibility believes that he or she will experience little anxiety or diYculty in calling forth these multiple selves because they are well‐defined and can be enacted with confidence. These aspects of multiple selves can be thought of as constituting figure– ground aspects of personality because the ‘‘selves’’ under consideration are arranged around the interpersonal circumplex (Wiggins, 1979). In this circumplex model, 16 interpersonal trait characteristics are arrayed around two orthogonal dimensions (dominance and warmth). Examples of trait pairs include ambitious–lazy, warm–cold, dominant–submissive, agreeable–quarrelsome, extroverted–introverted, and arrogant–assuming. For each item constituting the eight pairs, respondents indicate the extent to

296

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

which ‘‘they are capable of being [insert trait] if the situation requires it,’’ ‘‘it is diYcult for them to behave in a [insert trait] manner,’’ ‘‘how anxious they are when they behave in a [insert trait] manner,’’ and ‘‘the extent to which they attempt to avoid situations that require them to behave in a [insert trait] manner.’’ In Paulhus and Martin’s (1988) research, functional flexibility related to a high sense of agency and other indices of adaptive psychological functioning. Kernis, Goldman, Piasecki, and Brunnell (2003) (reported in Kernis & Goldman, 2005b) administered the Functional Flexibility Inventory (Paulhus & Martin, 1988) and the Authenticity Inventory (AI) (Version 2) to a sample of 84 individuals. We created summary indexes of capable, diYculty, anxiety, and avoidance scores by summing responses to the 16 traits (Paulhus & Martin, 1988). Total authenticity scale scores correlated significantly positively with capability, and negatively with diYculty, anxiety, and avoidance (Kernis & Goldman, 2005b). These findings support our contention that authenticity relates to a multifaceted and integrated self that is anchored in strong self‐beliefs, self‐confidence, self‐acceptance, and agency rather than self‐doubt, confusion, and conflict. Later in this chapter, we report additional findings linking authenticity to a ‘‘stronger sense of self ’’ (Kernis, Paradise, Whitaker, Wheatman, & Goldman, 2000). One of the premises underlying our conceptualization is that awareness of self is a component of healthy functioning. Awareness is really just a first step, however. Also important is that this awareness fosters self‐integration and acceptance of self. As integration and acceptance of self‐aspects increase, more information about them will become accessible. An important issue, therefore, is how individuals attain self‐knowledge in ways that foster integration and acceptance of self. A number of techniques are available, some of which stem from the Gestalt therapy framework developed by Fritz Perls and his colleagues (Perls et al., 1951). These techniques emphasize deliberately attending to aspects of self without evaluating their implications. A similar principle underlies the use of techniques or strategies designed to enhance individuals’ mindfulness. Through these exercises, people can become aware of currently ignored or unexamined self‐aspects with which they often are uncomfortable. Other techniques can then be applied to understand and resolve the basis of the uncomfortableness, thereby fostering self‐integration and acceptance.

B. UNBIASED PROCESSING The second component of authenticity involves the unbiased processing of self‐relevant information. This component involves objectivity with respect to one’s positive and negative self‐aspects, emotions, and other internal experiences, information, and private knowledge. In addition, it involves

AUTHENTICITY

297

not denying, distorting, or exaggerating externally based evaluative information. In short, unbiased processing reflects the relative absence of interpretive distortions (e.g., defensiveness and self‐aggrandizement) in the processing of self‐relevant information. To the extent that unbiased processing reflects an aspect of authentic functioning, variables that are theoretically related to authenticity should predict the relative absence of self‐serving biases and illusions. Importantly, highly autonomous and self‐determining individuals do not engage in self‐serving biases following success or failure (Knee & Zuckerman, 1996). Our characterization of the unbiased processing component of authenticity resonates with conceptualizations of ego defense mechanisms that link them to a wide range of important outcomes. For example, whereas adaptive defense mechanism styles that involve minimal reality distortions predict psychological and physical well‐being many years into the future (e.g., Vaillant, 1992), maladaptive or immature defenses that involve considerable reality distortion and/or failure to acknowledge and resolve distressing emotions predict psychological and interpersonal diYculties (e.g., poor marital adjustment) (Ungerer, Waters, Barnett, & Dolby, 1997). Note that our perspective stands in direct contrast to perspectives in which defensive processing is considered to be an adaptive solution to inevitable threats (e.g., Terror Management Theory) (Greenberg, Pyszcynski, & Solomon, 1986; Solomon, Greenberg, & Pyszcynski, 1991). While we agree that people can and do react defensively to threat, we believe that people’s natural inclinations are toward open and nondefensive processing of self‐relevant information (Deci & Ryan, 2000). The major benefit of unbiased processing is that it contributes to an accurate sense of self. This accuracy is highly beneficial for behavioral choices that have either short‐ or long‐term implications. The more important the outcome, the more important is accuracy. Pursuing the right occupation, investing time in developing one’s talents, and even finding a dance partner at a club all benefit from accurate or unbiased processing of evaluative information. Engaging in biased processing may unwittingly limit one’s options because relevant self‐knowledge is ignored or distorted. We believe, as many have before us (e.g., Deci & Ryan, 2000; Rogers, 1961), that people are oriented toward growing, developing, and increasing in complexity. We believe that these processes are inherently geared toward obtaining accurate, not necessarily flattering, information. In essence, we believe that positive self‐illusions generally are less healthy than accurate self‐realities (in contrast to Taylor & Brown, 1988), even though the former may confer short‐term benefits by helping individuals cope with unpleasant emotions (Crocker, 2002). In the end, possessing and portraying accurate

298

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

self‐knowledge is more beneficial than possessing and portraying positive but false self‐knowledge (e.g., Crocker, 2002; Robins & Beer, 2001). Controversy currently exists over whether positive self‐related illusions promote and reflect healthy psychological functioning (Robins & Beer, 2001; Taylor & Brown, 1988). Our view is that often these distortions stem from insecurity rather than strength (Kernis, 2000). In support of this contention, research has shown that people who function autonomously and are self‐ determining do not show such self‐serving distortions (Knee & Zuckerman, 1996). In contrast, people who rely on defense mechanisms that involve major distortions of reality have relatively poor interpersonal and psychological outcomes throughout their lifetimes (Vaillant, 1992). While self‐ illusions may minimize negative aVectivity in the short‐run (Crocker, 2002; Kernis, 2003; Robins & Beer, 2001) and therefore seem to be adaptive, this adaptiveness is itself an illusion, as it does not holdup over time and, in fact, may contribute to poorer outcomes in the end (Robins & Beer, 2001). Other forms of defensive functioning also appear reflective of insecurity and suboptimal functioning and are antithetical to authentic functioning, as we will describe shortly. Individuals high in unbiased processing are motivated to evaluate themselves objectively with respect to both positive and negative self‐aspects. Thus, processing self‐relevant information in an unbiased manner is likely to reflect what NeV (2003) referred to as a sense of self‐compassion (e.g., extending kindness and understanding to oneself rather than harsh self‐ criticism and judgment, and holding one’s painful thoughts and feelings in balanced awareness rather than overidentifying them). Sample items on NeV’s (2003) measure of self‐compassion include ‘‘I try to be understanding and patient towards those aspects of my personality I don’t like’’ and ‘‘I’m disapproving and judgmental about my own flaws and inadequacies (reversed scored).’’ In fact, Goldman, Lakey, and Kernis (2005d) found that higher unbiased processing was associated with greater self‐compassion.

C. BEHAVIOR The third component of authenticity involves behaving in accord with one’s values, preferences, and needs as opposed to acting ‘‘falsely’’ merely to please others or to attain rewards or avoid punishments. In essence, this component reflects the behavioral output of the awareness and unbiased processing components. We acknowledge that instances exist in which the unadulterated expression of one’s true‐self may result in severe social sanctions. In such instances, we expect that, at the very least, authenticity will reflect heightened sensitivity to the fit (or lack thereof ) between one’s true‐self

AUTHENTICITY

299

and the dictates of the environment, and a heightened awareness of the potential implications of one’s behavioral choices. In contrast, blind obedience to environmental forces typically reflects the absence of authenticity (cf., Deci & Ryan, 2000). Authentic behavior can be distinguished from inauthentic behavior by the conscious, motivated intentions that underlie it. Authentic behavior is guided by an honest assessment of one’s self‐aspects via the awareness and unbiased processing components. To the extent that one is conscious of the ‘‘figure’’ and ‘‘ground’’ inherent in one’s self‐aspects, one is aVorded the opportunity to act in a manner that is consistent with these multifaceted self‐aspects. In essence, authentic behavior is choiceful behavior oriented toward a ‘‘solution’’ derived from consciously considering one’s self‐relevant ‘‘problems’’ (e.g., potentially competing self‐motives, beliefs, etc.). In contrast, inauthentic behavior does not reflect a choiceful and conscious regulatory focus designed to eventuate in behavior that resonates with one’s complex, multifaceted self‐aspects. Rather, inauthentic behavior involves being unaware of, ignoring, oversimplifying, and/or distorting self‐aspects relevant to the behavioral context. In essence, whereas authentic behavior reflects the awareness and operation of one’s true‐ or core‐self, inauthentic behavior generally is oriented toward glorification and reverence by self and others (though on occasion it may be oriented toward excessive deprecation by self and others). Authenticity is not reflected in a compulsion to be one’s true‐self, but rather in the free and natural expression of core feelings, motives, and inclinations. When this expression stands at odds with immediate environmental contingencies, we expect that authenticity will be reflected in short‐term conflict. How this conflict is resolved can have considerable implications for one’s felt integrity and authenticity as well as for one’s overall functioning and well‐ being. An important implication of this reasoning is that it is insuYcient to focus exclusively on whether one’s actions per se reflect authenticity. Rather, it is crucial to focus also on the manner in which processes associated with the other authenticity components inform one’s behaviors. For example, Goldman (in press) presents findings indicating that awareness scores negatively correlate with tendencies to engage in social comparison, self‐monitoring, and public self‐consciousness. He argues that such tendencies can undermine one’s behavioral authenticity, because one fails to consider internal self‐knowledge and instead depends primarily on externally derived information (by comparing oneself to others, relying on others’ actions as the norm for one’s own actions, or by habitually focusing on how one publicly appears). More generally, sometimes the needs and values of the self are incompatible with the views of the larger society (e.g., when an artist focuses on a highly controversial subject matter). In these instances,

300

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

authenticity may be reflected in awareness of one’s needs and motives and an unbiased assessment of relevant evaluative information. Sometimes the resulting behavior may also reflect authenticity, but sometimes it may not (as when the aforementioned artist ‘‘sells out’’). Consequently, while the awareness, unbiased processing, and behavior components of authenticity relate to each other, they clearly are separable. We return to this issue shortly.

D. RELATIONAL ORIENTATION The fourth component of authenticity is relational in nature, and bears resemblance to Jourard’s (1971, p. 133) proposition that ‘‘authentic being means being oneself, honestly, in one’s relations with his fellows.’’ In our view, relational authenticity involves valuing and striving for openness, sincerity, and truthfulness in one’s close relationships. In essence, relational authenticity means being genuine rather than fake in one’s relationships with close others. It is characterized by honesty in one’s actions and motives as they pertain to one’s intimates, and to accuracy in beliefs about oneself and one’s intimates. Moreover, it involves endorsing the importance of close others seeing the ‘‘real’’ you and relating to them in ways that facilitate their being able to do so. Furthermore, given that dispositional authenticity involves heightened levels of self‐knowledge and understanding (i.e., awareness), and the capacity to evaluate one’s self objectively (i.e., unbiased processing), higher authenticity levels may enhance self–other perception congruence. Research focused on self‐verification theory suggests that people are motivated by their need for self‐knowledge (Swann, Stein‐Seroussi, & Giesler, 1992) and are drawn toward others who confirm their preexisting self‐conceptions (Swann, 1983). We believe that self‐verification processes in close relationships are especially likely to occur when the other components of authenticity are operative within individuals (e.g., possessing high levels of awareness, unbiased processing, and behavioral authenticity). Conversely, self‐enhancement processes that involve distorted evaluations within close relationships are especially likely to occur among individuals who are low in authenticity (i.e., those who are uncertain who they really are, and who resist accurate self‐evaluation). Stated diVerently, low authenticity may reflect the presence of fragile self‐feelings that motivate self‐enhancement tendencies (Kernis, 2003; Kernis & Goldman, 2002). In such cases, incongruence between individuals’ self‐evaluation and their perceptions of how their intimates evaluate them may stem from motivations that stifle accuracy and consensus in favor of positive self‐views. Consistent with this line of reasoning, Mikulincer, Orbach, and Iavenieli (1998) found that securely

AUTHENTICITY

301

attached people were more accurate in assessing self–other similarity than were insecurely attached people. In short, substantive intimate relationship adjustment is likely to involve feeling understood or ‘‘known’’ to intimates, and accuracy in such appraisals is likely to occur when authenticity is operative. Authentic relationships involve a reciprocal process of self‐disclosure and of mutual intimacy and trust (Reis & Patrick, 1996). Thus, relational authenticity involves developing and achieving secure attachments with intimates that further promotes the genuine expression of core self‐aspects without threat of reprisal or criticism. In support of this contention, Kernis and Goldman reported that higher relational orientation related to higher secure attachment styles and lower preoccupied and fearful attachment styles (2005a), as well lower rejection sensitivity (2005b). In short, we expect that people high in relational authenticity will be involved in healthier, more satisfying, and fully functioning relationships than people low in relational authenticity. Later in the chapter, we report additional data relevant to examining these claims. In other research, Harter, Waters, Pettit, Whitesell, Kofkin, and Jordan (1997) found that relationship partners who each viewed themselves as ‘‘mutual’’ (e.g., exhibiting a balance between one’s personal needs and one’s partner’s needs) reported the highest levels of validation and authentic behaviors, whereas ‘‘self‐focused autonomy’’ partners were perceived as least validating. In terms of well‐being, Harter et al. (1997) found evidence for a process model. Specifically, the relationship between individuals’ perceived validation from their partners and their own well‐being (i.e., self‐esteem and cheerfulness) depended on the extent to which they exhibited authentic self‐ behavior within their romantic relationship. Taken as a whole, Harter et al.’s (1997) findings demonstrate that behavioral authenticity within one’s intimate relationships involves adopting a relationship orientation that fosters mutuality. Furthermore, their findings suggest that how a person’s intimate relationships influence his or her well‐being is aVected by the extent to which one acts in accord with one’s true‐self within those relationships.

E. MORE ON THE SEPARATENESS OF THESE COMPONENTS We view these multiple components of authenticity as related to, but separable from, each other (Table I). For instance, situations invariably exist in which environmental pressures may inhibit the expression of one’s true‐self (e.g., a person may not express his true opinion to a close friend who is highly depressed). Although behavioral (and perhaps relational) authenticity may be thwarted in such instances, authenticity at the levels of awareness and unbiased processing may be operative. Specifically, awareness may

302

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN TABLE I AUTHENTICITY COMPONENTS











Awareness Awareness and knowledge of, and trust in, one’s motives, feelings, desires, and self‐relevant cognitions Includes awareness of one’s strengths and weaknesses, dominant–recessive aspects of personality, powerful emotions, and their roles in behavior Unbiased Processing Minimal, if any, denial, distortion, exaggeration, or ignoring of private knowledge, internal experiences, and externally based self‐evaluative information Objectivity and acceptance with respect to one’s strengths and weaknesses Behavior Acting in ways congruent with one’s values, preferences, and needs Rather than acting merely to please others or to attain rewards or avoid punishments

Relational Orientation Value and make eVorts to achieve openness and truthfulness in close relationships
Important for close others to see the real you, those deep, dark, or potentially shadowy self‐ aspects that are not routinely discussed
Relational authenticity means being genuine and not ‘‘fake’’ in one’s relationships with others


involve active attempts to resolve conflicting motives and desires involved in knowing one’s true opinion and the implications expressing it may have for one’s friendship and the well‐being of one’s friend. In many respects, the awareness component of authenticity is the most fundamental. Self‐knowledge is at the heart of both behavioral and relational authenticity. Although we can envision instances in which behavioral and relational authenticity emerge spontaneously with little or no conscious deliberation, ultimately the self‐aspects that are involved will be available and accessible with the growth of one’s self‐knowledge. Unbiased processing may involve acknowledgment of the fragile underpinnings of one’s attitude. In contrast, inauthenticity may involve actively ignoring or denying one’s opinion or emphasizing the superiority of one’s judgmental abilities. In short, it is possible for a person to be operating authentically at some levels but not at others. Therefore, it is important to examine the processes associated with each component of authenticity (Kernis, 2003).

F. CONNECTIONS WITH OTHER CONSTRUCTS Each of these aspects of authenticity has received some attention in the past, although not usually with explicit reference to the construct of authenticity. For example, researchers have examined aspects of the awareness

AUTHENTICITY

303

component in research on public and private self‐consciousness (e.g., Fenigstein, Scheier, & Buss, 1975). Some implications of biased processing of self‐relevant information have been examined in research on self‐serving biases (e.g., Blaine & Crocker, 1993). Aspects of behavioral authenticity have been examined in research on personality–behavior and attitude– behavior consistency (Koestner, Bernieri, & Zuckerman, 1992; Snyder, 1987). Finally, aspects of relational authenticity have been studied in research on attachment processes and self‐disclosure (Mikiluncer & Shaver, 2005). Readers of this chapter undoubtedly will recognize aspects of our theory in this prior work. However, our theory has the capacity to integrate these various strands of research to explicate the processes associated with the construct of authenticity in a way not done before. For research to be conducted, however, an empirically based measure of authentic functioning is needed. We turn now to our eVorts to develop such a measure.

V. Measuring Individual DiVerences in Dispositional Authenticity: The Authenticity Inventory We started with a large pool of items that we believed would tap into these four components, and we administered them to several samples of male and female college students. We eliminated items based on interitem correlations and exploratory factor analyses. In the research reported in this chapter, we used three successive versions of the scale. The final scale (AI‐3, Goldman & Kernis, 2004) consists of 45 items (Awareness—12 items, Unbiased Processing—10 items, Behavior—11 items, and Relational Orientation—12 items). We include the scale items, along with instructions for its administration and scoring, in the Appendix. CoeYcient alphas for the scale as a whole (.90) and for each of the subscales (Awareness ¼ .79, Unbiased Processing ¼ .64, Behavior ¼ .80, and Relational Orientation ¼ .78) are acceptable. Test retest reliabilities (over approximately 4 weeks, N ¼ 120) were high (Total ¼ .87, Awareness ¼ .80, Unbiased Processing ¼ .69, Behavior ¼ .73, and Relational Orientation ¼ .80). How do these proposed authenticity components relate to the construct of authenticity? One possibility is that these four components of authenticity reflect conceptually distinct but interrelated aspects of authenticity. Stated diVerently, authenticity may be a multifaceted construct that consists of four distinct components. In factor analytic terminology, this would equate to a four‐factor model. A second possibility is that authenticity is a unidimensional rather than multidimensional construct. That is, while the components we introduced may be conceptually distinct, empirically they may be so highly interrelated

304

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

that they are not distinguishable and, therefore, represent a single broad authenticity construct. In factor analytic terminology, each component may load very highly on a single factor. A third possibility combines aspects of the two previous possibilities. That is, on the one hand, authenticity may reflect four conceptually distinct facets as in the first possibility. Nevertheless, it is unrealistic to suppose that these four aspects are going to be completely unrelated to one another. However, it is also unrealistic to expect that they would be completely redundant with one another. Consequently, there may be value in conceiving of a broad authenticity construct at a higher level of abstraction that subsumes each of the four facets of authenticity. In this instance, while the four components are distinct, they may also measure a single latent construct of authentic functioning. Thus, parsimony exists, but at a higher level of abstraction than with a single‐factor model. In other words, a hierarchical structure exists in which overall authenticity exerts its eVects through the four separable, but interrelated, components of awareness, unbiased processing, behavior, and relational orientation. We anticipated finding the greatest support for the third possibility—a second‐order factor model in which interrelations among the authenticity components are not so high that they are redundant with one another, but are high enough so that they are summarized adequately with a single second‐order authenticity factor. If supported, this model would provide evidence for a broad latent construct of authenticity, while simultaneously providing support for treating the components as valid indicators of distinct, but interrelated, aspects of authentic functioning. We used confirmatory factor analysis (CFA) to test these alternative conceptions of authenticity. When developing measurement models for theoretical constructs, one faces a number of options for operationalizing them, ranging from (a) a total disaggregation model in which individual elements (e.g., questionnaire items) are used as manifest indicators of the latent constructs, to (b) some intermediate level of aggregation, such as creating item parcels (‘‘testlets’’), for use as manifest indicators in latent variable models, to (c) a total aggregation model in which a single composite indicator is used to represent the latent construct (Bagozzi & Phillips, 1991; Edwards, 2000). We opted for a meso‐level of aggregation by creating multiple‐item parcels for each authenticity dimension. Use of item parcels poses several advantages over use of individual items as manifest indicators. Specifically, as compared to individual items, item parcels are more reliable, have smaller ratios of unique to common variance, are less likely to violate distributional assumptions, are more parsimonious, are less likely to have unmeasured correlated disturbances, are less subject to sampling fluctuations, and usually result in less biased CFA solutions (Bandalos, 2002; Little,

AUTHENTICITY

Fig. 1.

305

Proposed theoretical models of dispositional authenticity.

Cunningham, Shahar, & Widaman, 2002). There are a number of approaches to forming item parcels (Hagtvet & Nasser, 2004; Hall, Snell, & Faust, 1999; Landis, Beal, & Tesluk, 2000), but random assignment is a generally eVective approach. As such, we randomly assigned items to three item parcels each for the awareness (AW), unbiased processing (UP), behavior (BE), and relational orientation (RO) subscales. The three models we tested are shown in Fig. 1. Figure 1A shows a unidimensional Authenticity model in which all item parcels (shown in rectangles) are presumed to reflect a single authenticity factor. The second model, shown in Fig. 1B, is a four‐factor model, which diVers from the unidimensional model in that it

306

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

proposes that authenticity is comprised of four distinct, yet possibly correlated, components discussed earlier. The final model we tested, shown in Fig. 1C, was a hierarchical model, which proposes that any interrelationships among the four facets of authenticity proposed by the model shown in Fig. 1B can be explained parsimoniously on the basis of their common dependency on a more general, second‐order authenticity factor. Goodness‐of‐fit indices for these three models are shown in Table II. The
2 statistic was significant for each model indicating that all three models should be rejected statistically, but this is a common finding in CFA research. Consequently, we shifted attention to alternative overall model fit indices and comparisons between these rival models. The unidimensional model provided a poor fit to the data by all conventional standards for acceptable model fit (Marsh, Balla, & McDonald, 1988). By comparison, the four‐factor model provided a much better fit to the data [

2 (6) ¼ 242.64, p < .01], and its goodness‐of‐fit indices satisfied (or approached) even more stringent criteria suggested by Hu and Bentler (1998, 1999) (SRMSR  .08, RMSEA  .06, CFI and TLI  .95). This indicates that authenticity is best regarded as a multidimensional construct and supports the discriminant validity of the factors specified in the four‐factor model. The remaining question, however, is whether a general, higher‐order authenticity factor can explain any relationships that exist between the four first‐order authenticity factors. To test this idea, we compared the fit of the four‐factor model to that of the hierarchical model and found that their fit to the data was not significantly diVerent from one another (

2 ¼ 1.89, ns). Alternative goodness‐of‐fit indices were practically identical as well, indicating that the more parsimonious hierarchical model should be preferred as a plausible explanation of the interrelationship among the first‐order authenticity factors. Results

TABLE II CFA MODEL GOODNESS‐OF‐FIT INDICES Model

df


2

SRMSR

RMSEA

CFI

TLI

1. Unidimensional Model 2. Four‐factor Model 1 versus 2 3. Hierarchical Model 2 versus 3

54 48 6 50 2

402.03* 159.39* 242.64* 161.28* 1.89

.073 .048 —– .049 —–

.130 .075 —– .074 —–

.82 .94 — .94 —

.79 .92 — .92 —

*p < .01. Note: df ¼ degrees of freedom,
2 ¼ model chi‐squared statistic, SRMSR ¼ standardized root mean squared error, RMSEA ¼ root mean squared error or approximation, CFI ¼ Bentler’s (1990) comparative fit index, TLI ¼ the Tucker‐Lewis index.

AUTHENTICITY

Fig. 2.

307

Confirmatory factor analysis findings for hierarchical model of authenticity.

(LISREL’s completely standardized factor loadings) for the hierarchical model are shown in Fig. 2. All parameters were statistically significant ( p < .01) and, with the exception of the first item parcel for the Unbiased Processing factor, were uniformly large. We turn now to a research in which we used the AI‐3 (unless otherwise noted) to examine various aspects of psychological and interpersonal functioning and well‐being. First, we focus on aspects of healthy psychological functioning, including verbal defensiveness, mindfulness, coping styles, self‐ esteem, and self‐concept structure. We then turn to examining the implications of authenticity for social role functioning, goal pursuits, well‐being, and close relationships.

VI. Authenticity and Healthy Psychological Functioning A. AUTHENTICITY AND ABSENCE OF VERBAL DEFENSIVENESS Emotions, thoughts, behaviors, or information that are discrepant with one’s consciously held self‐image often are threatening, producing decreases in self‐esteem and/or increases in negative aVect. To ward oV these threats, people may utilize a wide range of defense mechanisms. ‘‘Defense mechanisms can be thought of as motivated cognitive‐behavioral strategies that protect the self from perceived threat, maintain or augment self‐esteem,

308

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

reduce negative aVect, and maintain positive representations of attachment figures (Feldman Barrett et al., 1996)’’ (Feldman Barrett, Cleveland, Conner, & Williams, 2000). That is, defense mechanisms reduce the perception of threat by altering how people represent these events in conscious thought. When people perceive a self‐esteem threat, for example, they may attempt to deal with surfacing unpleasant aVect by controlling whether the threat enters consciousness (awareness) or by controlling the specific content of the thoughts or feelings that enter consciousness (distortion) (Feldman Barrett et al., 2000). The result is that people distance themselves from the threat and their emotional experience to some extent, and they avoid thoughts and feelings that threaten their consciously held self‐image or self‐feelings. The framework presented here suggests that people low in dispositional authenticity will be especially likely to utilize defensive strategies to ward oV potentially threatening events or experiences. We theorize that people high in dispositional authenticity are motivated to understand themselves, to experience aVect as it is felt, and to not distort evaluative information. Thus, they should have the strength and personal resources to acknowledge information that is potentially threatening without being overly defensive. A number of verbal markers of defensiveness exist (Feldman Barrett et al., 2000) that provide clues to the nature of people’s motivational strategies for protecting the self against threat. Do they rationalize by blaming others? Do they deny awareness of conflicting emotions, choosing only to identify positive aVect? Examining the nature of these motivational strategies has the potential to provide significant insight into diVerences in the ways those individuals who are low or high in authenticity deal with threatening events. Feldman Barrett, Williams, and Fong (2002) reported a structured interview technique (and sophisticated coding scheme) for eliciting threatening experiences and defensive processing. Specifically, individuals engage in a taped 40–60‐minute stressful interview about their experiences. Respondents first answer five nonstressful items to acclimate them to the interview context. They then respond to 15 mild to moderately stressful items (e.g., ‘‘Tell me about a time when you felt that your parents were really disappointed in you,’’ ‘‘Tell me about a time when you’ve broken the rules,’’ ‘‘Tell me about a time when you have done something unethical on an assignment,’’ ‘‘Describe a time when someone has come to you for help and you didn’t want to help them,’’ ‘‘Tell me about a time when you have disappointed someone.’’) The interview concludes with five items designed to gradually restore a nonthreatened self‐view. Two highly trained coders rated responses to each of the 15 stressful items, which we summed to form an overall verbal defensiveness score. Raters incorporated two aspects of defensiveness into their ratings: awareness and

AUTHENTICITY

309

distortion. Awareness is defined as the conscious understanding and acceptance of one’s cognitions, emotions, and behaviors in the face of threat. Distortion is characterized as the reinterpretation of events through rationalization or justification to fit one’s preexisting self‐concept (Feldman Barrett et al., 2002). As such, individuals can respond in a way that is nondefensive (high awareness and acceptance and low distortion), mildly defensive (moderate awareness with mild distortion), moderately defensive (limited awareness and moderate distortion), or highly defensive (highly unaware and high distortion of information). The training manual graciously provided to us by Lisa Feldman Barrett contains extensive coding information and numerous examples to facilitate the training of event coders. This measure is well‐grounded in research and theory that have focused on defensiveness and defense mechanisms (e. g., Cramer, 1990; Sackeim & Gur, 1979; Shedler, Mayman, & Manis, 1993; Vaillant, 1992; Weinberger, 2003). The defensive verbal behavior assessment (DVBA) is ‘‘. . .a method for detecting traces left by defensive processes in the content and structure of speech’’ (Feldman Barrett et al., 2002, p. 777). Although individuals may use diVerent defense mechanisms, the DVBA focuses on the shared consequences of using these mechanisms. The DVBA provides a unique opportunity to assess the validity of our authenticity measure. Specifically, some skeptics have argued that people who are highly defensive will falsely answer items on our authenticity inventory so that they appear to be authentic, especially on the subscale of unbiased processing (‘‘Of course I am authentic—are you trying to say I am a phony?’’). The line of reasoning provided by skeptics suggests that higher authenticity would relate to greater defensiveness. Although we recognize that people are motivated to present themselves in a positive light, we attempted to minimize these considerations in the assessment of authentic functioning (with the AI) by avoiding asking people directly about whether or not they are authentic. Instead, we query individuals about the extent to which their motives, emotions, and behaviors reflect processes and mechanisms theoretically linked to authentic functioning. These processes include the tendency not to distort negative self‐relevant information and to be comfortable with experiencing unpleasant emotions, or motivations reflective of one’s ‘‘dark side.’’ Thus, we predicted that overall, greater authenticity would relate to lower, not higher, defensiveness on the DVBA. Moreover, we anticipated that higher awareness and unbiased processing subscale scores would relate to lower defensiveness because these subscales deal specifically with the extent to which people are aware of, and feel comfortable experiencing, unpleasant self‐relevant thought and aVect. Finally, we anticipated that higher behavioral authenticity would relate to lower verbal defensiveness because one’s behaviors are choiceful and reflective of one’s

310

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

true‐self, and therefore one should be more accepting of their implications and consequences regardless of whether they are positive or negative. To the extent that our findings support these predictions, they would provide important construct validation support for the AI. To test these hypotheses, we (Kernis, Lakey, Heppner, Goldman, & Davis, 2005) had 101 male and female undergraduates participate in individual DVBA interviews with one of three trained interviewers. We then trained two additional raters to code the interviews according to the criteria described in detail in a manual provided to us by Feldman Barrett. Interrater reliability was excellent, exceeding .80. Total authenticity correlated inversely with defensiveness (r ¼ .25, p < .02). In addition, awareness correlated inversely with defensiveness, (r ¼ .21, p < .04), as did behavior (r ¼ .28, p < .01), and unbiased processing, although the latter only marginally (r ¼ .19, p < .062). Finally, relational authenticity was nonsignificantly correlated with defensiveness (r ¼ .10). Other data collected in this study indicated that especially high levels of defensiveness were associated with fragile forms of high self‐esteem, namely unstable and contingent high self‐esteem (Kernis, 2003; Kernis & Paradise, 2002). These findings further corroborate conclusions we can draw from measures of overall subjective and psychological well‐being that we administered. To the extent that defensiveness is adaptive and reflective of optimal functioning, greater tendencies toward defensiveness should correlate positively with these measures of well‐being. However, this clearly was not the case. Total scores on RyV ’s (1989) multicomponent measure of psychological functioning were inversely correlated with defensiveness (r ¼ .25, p < .02), as were scores on the Life Satisfaction Scale (r ¼ .25, p < .02). Taken as a whole, our findings indicate that the higher the individuals’ dispositional authenticity, the more they were able to deal with self‐threatening information in an aware and nondistorting manner, which, as it turns out, related to better overall psychological functioning, secure forms of high self‐ esteem and greater subjective well‐being. Whereas the current study examined how dispositional authenticity related to individuals’ defensive reactions to a specifically threatening context, in the next study we report, we sought to examine individuals’ general tendencies toward actively and openly attending to their experiences in a mindful and nonevaluative manner.

B. AUTHENTICITY AND MINDFULNESS Mindfulness refers to a state of relaxed and nonevaluative awareness of one’s immediate experience (Brown & Ryan, 2003). Research has linked mindfulness with positive immediate experiences (LeBel & Dube´, 2001)

AUTHENTICITY

311

and greater psychological health and well‐being (Brown & Ryan, 2003). Moreover, the capacity for mindfulness is an aspect of being fully functioning, so we expected that it would be associated with greater authenticity. The mindfulness measure we used in our earlier research was the Mindful Attention Awareness Scale (MAAS) (Brown & Ryan, 2003). Greater mindfulness, as assessed by the MAAS, relates to greater psychological well‐being and positive aVect and lower stress (Brown & Ryan, 2003). Sample items, endorsement of which reflects low mindfulness, include: ‘‘I could be experiencing some emotion and not be conscious of it until some time later’’; ‘‘I do jobs or tasks automatically, without being aware of what I am doing’’; ‘‘I find myself listening to someone with one ear, doing something else at the same time.’’ Kernis and Goldman (2005) reported that MAAS mindfulness scores correlated significantly with total authenticity scores, as well as with each subscale score. In more recent research, Lakey, Kernis, Heppner, and Davis (2005) administered both the MAAS and the Kentucky Inventory of Mindfulness Skills (KIMS), which measures the specific mindfulness components of observing (OBSERVE), describing (DESCRIBE), acting with awareness (AWARENESS), and accepting or allowing without judgment (ACCEPTANCE). Observing refers to ‘‘observing, noticing, or attending to a variety of stimuli, including internal phenomena, such as bodily sensations, cognitions, and emotions, and external phenomena, such as sounds and smells’’ (Baer, Smith, & Allen, 2004, p, 193). Sample items include ‘‘I pay attention to whether my muscles are tense or relaxed’’ and ‘‘I notice the smells and aromas of things.’’ Describing refers to ‘‘describing, labeling, or noting of observed phenomena by covertly applying words. . . . This type of describing is done nonjudgmentally and without conceptual analysis’’ (Baer et al., 2004, p. 193). Sample items include ‘‘I’m good at finding the words to describe my feelings’’ and ‘‘Even when I am feeling terribly upset, I can find a way to put it into words.’’ Acting with awareness refers to ‘‘Engaging fully in one’s current activity with undivided attention or focusing with awareness on one thing at a time. . ..’’ (Baer et al., 2004, p. 193). Sample items include ‘‘When I’m doing something, I’m only focused on what I am doing, nothing else’’ and ‘‘I’ll get completely absorbed in what I’m doing, so that all my attention is focused on it.’’ Accepting or allowing without judgment refers to ‘‘. . .accepting, allowing, or being nonjudgmental or nonevaluative about present moment experience . . . to refrain from applying evaluative labels such as good/bad, right/wrong, or worthwhile/worthless . . ..’’ (Baer et al., 2004, p. 194). Sample items include ‘‘I make judgments about whether my thoughts are good or bad (reverse‐scored)’’ and ‘‘I tend to make judgments about how worthwhile or worthless my experiences are (reverse‐scored).’’

312

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN TABLE III CORRELATIONS BETWEEN AUTHENTICITY, MINDFULNESS, FUNCTIONING MEASURES

AND

PSYCHOLOGICAL

Measure

Total

Awareness

Unbiased processing

Behavior

Relational orientation

Self‐actualization Vitality Psychological stress Physical symptomatology Mindfulness (KIMS) KIMS‐OBSERVE KIMS‐DESCRIBE KIMS‐AWARENESS KIMS‐ACCEPTANCE Mindfulness (MAAS)

.61** .23* .30** .21 .64** .26* .62** .36** .28** .49**

.53** .26* .31** .22* .67** .24** .60** .44** .27** .45**

.42** .16 .14 .07 .40** .06 .40** .11 .39** .37**

.41** .02 .25* .14 .43** .12 .50** .25** .18 .40**

.41** .27* .13 .15 .45** .29** .42** .27** .10 .28**

* p < .05, ** p < .01. Note: Higher Psychological Stress scores reflect lower levels of stress. See text for description of KIMS subscales.

As shown in Table III, the findings obtained by Lakey et al. (2005) for the MAAS scale replicated those reported by Kernis and Goldman (2005). Specifically, total authenticity scores, as well as each authenticity subscale score correlated significantly with total MAAS scores. In addition (and new to this study), total authenticity and authenticity subscale scores correlated positively with total KIMS scores as well as its subscales. Specifically, awareness correlated with each KIMS subscale, relational orientation correlated with each KIMS subscale with the exception of KIMS‐Acceptance, unbiased processing correlated significantly with KIMS‐Describe and KIMS‐Acceptance, and behavior correlated significantly with KIMS‐ Describe and KIMS‐Awareness. Most of these relationships were moderate in strength. The relationships that emerged between the subscales of the two measures have many interesting theoretical implications. For example, the awareness authenticity subscale, which reflects a basic awareness of, trust in, and openness toward, self‐knowledge, correlated with each of the KIMS subscales. These relationships suggest that an open and trusting stance toward one’s self‐aspects goes hand‐in‐hand with tendencies to observe internal and external stimuli, competence in describing one’s internal states, ability to focus one’s attention on the task at hand, and a nonjudgmental stance in general. In addition, the significant correlations between our unbiased processing subscale and the KIMS describe and awareness subscales suggests that engaging in biased processing may reflect a more general tendency to engage in evaluative judgments. Finally, the fact that

AUTHENTICITY

313

high behavioral authenticity related to competence in describing observed phenomena and to focusing one’s attention on the task at hand is consistent with research and theory on intrinsic motivation. When intrinsically motivated, people are highly absorbed in activities that match their interests and talents (Deci, 1975). Interestingly although, the relational orientation authenticity subscale is explicitly interpersonal in nature, it is related to many intrapersonal aspects of mindfulness processes. Other findings obtained in our lab and reported in Table III indicate that higher authenticity relates to other aspects of positive psychological functioning. Specifically, higher authenticity relates to greater self‐actualizing tendencies (Jones & Crandall, 1986) and vitality (Ryan & Frederick, 1997) and to lower psychological distress (Cohen, Kamarck, & Mermelstein, 1983) and (marginally) physical symptoms.

C. AUTHENTICITY AND THE USE OF VARIOUS COPING STRATEGIES If our assertion that authentic functioning is associated with greater adaptive functioning is correct, we should find corroborating evidence by examining people’s characteristic ways of coping with stressful events. The adaptive value of coping strategies vary from healthy and helpful to unhealthy and counterproductive (Carver, Scheier, & Weintraub, 1989; Moos & Schaefer, 1993; Vaillant, 2000). For instance, Folkman and Lazarus (1980, 1985) described healthy coping styles with respect to problem‐focused and emotion‐focused strategies. Problem‐focused coping strategies aim toward solving the problem or modifying the source of the threat (Folkman & Lazarus, 1980, 1985). Emotion‐focused coping strategies aim toward managing or reducing the emotional distress associated with the threatening circumstances (Folkman & Lazarus, 1980, 1985). Although this distinction has proven highly useful, Carver et al. (1989) argued that each of these broad categories is comprised of a number of distinct coping strategies. They developed a multidimensional coping inventory (the COPE) to assess the various ways that people cope with stressful events. Active coping: taking active steps to remove the threat or reduce its impact (I concentrate my eVorts on doing something about it). Planning: thinking about how to cope with the threat such as the steps to take to deal with the problem (I think about how I might best handle the problem). Suppression of competing actions: putting other things aside to deal with the problem at hand (I put aside other activities in order to concentrate on this). Instrumental social support: seeking information, help, or advice about how to deal with the stressor (I try to get advice from someone about what to do). An example of

314

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

emotion‐focused coping is emotional social support: seeking sympathy, moral support, and the like (I discuss my feelings with someone). In addition, the COPE assesses a number of potentially maladaptive strategies, as in the following: venting one’s emotions—focusing on and venting one’s distress (I get upset and let my emotions out); behavioral disengagement—withdrawing one’s eVort to either deal with the stressor or achieve the goal hampered by the stressor (I just give up trying to reach my goal); mental disengagement—engaging in alternative activities to distract oneself from the problem at hand (I sleep more than usual); substance use—using alcohol or drugs to take one’s mind oV the problem (I drink alcohol or take drugs, in order to think about it less); and denial—refusing to accept that the stressor is real (I pretend that it hasn’t really happened). (The measure contains several other subscales, but they are not discussed here because they did not relate to our authenticity measure.) To test the hypothesis that greater authenticity would relate to greater reliance on adaptive coping styles and to less reliance on maladaptive coping styles, Goldman and Kernis (2005) administered the AI‐3 and then subsequently administered the COPE measure approximately 4 weeks later. The correlations displayed in Table IV indicate that authentic functioning is related to the (self‐reported) use of more adaptive and less maladaptive coping strategies (Goldman & Kernis, 2005). First, scores on each authenticity dimension, as well as total scores, correlated significantly

TABLE IV CORRELATIONS BETWEEN AUTHENTICITY

Coping scale

Total

Awareness

AND

COPING STYLES

Unbiased processing

Behavior

Relational orientation

Active coping Planning Suppress Instrumental support

.48** .31** .14 .15

Problem‐Focused Coping .37** .27* .32** .10 .11 .05 .03 .01

.49** .37** .21* .11

.26* .09 .12 .32**

Emotional support

.19

Emotion‐Focused Coping .03 .08

.09

.38**

Mental disengage Behavior disengage Emotion venting Denial Substance use

.21a .21a .12 .22* .25*

Suboptimal Coping .13 .23* .22* .11 .25* .15 .23* .10 .22* .09

Note: ap < .10, *p < .05, **p < .01.

.17 .10 .03 .07 .29**

.07 .18a .13 .24* .30**

AUTHENTICITY

315

with scores on the active coping subscale. Thus, greater authentic functioning involves ‘‘taking the bull by the horns’’ and directly tackling the problem at hand. Second, higher awareness and behavioral authenticity, as well as total authenticity, related to greater use of planning. This makes sense, as thinking through stressors and how best to deal with them often involves a thorough assessment of one’s qualities relevant to the situation and a willingness to act on one’s values. Third, greater behavioral authenticity related to greater suppression of competing activities. This finding suggests that behavioral authenticity involves the capability to self‐regulate one’s actions with respect to task relevant demands. Fourth, greater relational authenticity related to greater seeking of emotional and instrumental social support. Thus, the more people value and achieve honesty and sincerity with their intimates, the more they are willing to rely on them in times of stress by seeking their informational and emotional support. Authenticity also related inversely to the use of mostly dysfunctional or maladaptive strategies. For example, substance use related to lower overall authenticity, as well as lower awareness, behavior, and relational orientation scores. These findings indicate that authentic functioning relates to constructive and active eVorts to deal with problems and stressors, rather than shying away from them or simply venting one’s emotions. Interestingly, the fact that awareness subscale scores related to lower emotional venting suggests that the desire to know one’s self does not include becoming fixated on one’s emotional distress in times of stress. Instead, becoming fixated on one’s distress appears to signal a relative lack of self‐knowledge. We would argue that authentic self‐knowledge involves knowledge about one’s sensitivities that interact with stressors to produce certain emotions and that such knowledge reflects a level of mature self‐understanding antithetical to the notion of venting one’s emotions, either to self or to others. Finally, total authenticity scores, as well as awareness and relational orientation scores, related to less denial of a stressor. As we have suggested, these aspects of authenticity involve a desire for accuracy in self‐knowledge and comfortableness with close others, each of which would seem to mitigate the need to deny the existence of a stressor. We conducted additional analyses to examine whether overall authenticity predicted coping styles independent of self‐esteem level. Overall authenticity uniquely predicted a number of coping styles, namely, active coping, planning, emotional support, and substance abuse. These data provide important support for considering dispositional authenticity to be an important construct that cannot be reducible to self‐esteem level. Later in the chapter, we report additional findings regarding the independent predictive utility of dispositional authenticity, and we consider in detail the interrelation between self‐esteem and authenticity.

316

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

D. AUTHENTICITY, SELF‐CONCEPTS, AND ROLE FUNCTIONING In his dissertation, Goldman (2004) examined the relationship and predictive utility of dispositional authenticity with respect to a diverse set of measures assessing (1) aspects of self‐esteem and self‐concepts (self‐esteem level and contingency, self‐concept organization, and self‐theories) and (2) social role functioning (markers reflecting general and authenticity‐related aspects of social role adjustment) across the five commonly enacted social roles of being a student, a romantic partner, a son/daughter, a friend, and an employee. Self‐esteem and self‐concept can be represented with an enormous number of variables. Goldman narrowed the field by making reference to the notion of a stronger sense of self, which Kernis et al. (2000) suggested is comprised of three components: (1) feelings of self‐worth that are well‐anchored and secure, (2) actions that reflect a strong sense of agency and self‐determination, and (3) self‐concept that is clearly and confidently defined so that it contributes to a coherent sense of direction in one’s daily experience. With respect to self‐esteem, Goldman’s study included measures of self‐esteem level (Rosenberg, 1965) and contingent self‐esteem (the Contingent Self‐esteem Scale, Kernis & Paradise, 2004; reported in Kernis & Goldman, in press). Previous research and theory indicate that the higher and less contingent (i.e., less dependent on specific achievements or outcomes) one’s self‐esteem, the healthier it is (Deci & Ryan, 1995; Kernis, 2003). Self‐concept organization reflects aspects of cognitive structures that organize and guide the processing of self‐related information. Implicit in most conceptualizations of self‐organization is a hierarchical organization of self‐knowledge wherein specific contents or domains of one’s self‐concept are subsumed by more global self‐representations (e.g., general evaluations of one’s self ). This organization can reflect varying degrees of consistency, unification, coherence, versus fragmentation, diVerentiation, confusion, and the like. A number of variables capture aspects of this organization. Self‐ concept clarity (Campbell et al., 1996) is defined as the extent to which the contents of the self‐concept are clearly and confidently held, internally consistent, and temporally stable. Identity integration (O’Brien & Epstein, 1988) reflects the extent to which one’s self‐concept is eYcacious in organizing and directing life experiences and in assimilating new information. Stated diVerently, identity integration reflects the overall adequacy of one’s self‐ concept in one’s general functioning. Self‐concept diVerentiation (Donahue et al., 1993) reflects the extent to which individuals see themselves as having diVerent personality characteristics in diVerent social roles. Thus, higher diVerentiation reflects greater fragmentation in one’s self‐concept because

AUTHENTICITY

317

one’s personality is judged to diVer depending on the social role being considered. Implicit theories (Dweck, Chiu, & Hong, 1995) pertain to individuals’ beliefs regarding the extent to which such characteristics as intelligence, morality, and personality traits are fixed and unchangeable (entity theory), or are malleable and subject to change and development (incremental theory). Endorsement of an incremental self‐theory reflects a mastery orientation characterized by personal development and self‐ improvement as opposed to performance displays at any given point in time. Self‐organization also involves how individuals adjust their self‐concepts to assimilate experiences into an identity, as they actively cope with emerging social demands and developmental challenges (Erickson, 1959). Berzonsky (1988) proposed that how individuals engage and negotiate identity‐ relevant issues involves specific social‐cognitive processing orientations that he refers to as identity styles. Three identity styles are proposed: informational, normative, and diVuse/avoidant (Berzonsky, 1988, 1990). Individuals characterized by an informational identity style ‘‘actively seek out, process, and evaluate self‐relevant information before making identity decisions. They are skeptical about their self‐constructs, open to new information and alternatives, and willing to revise and modify their self‐views in response to discrepant feedback’’ (Nurmi, Berzonsky, Tammi, & Kinney, 1997, p. 556). Individuals characterized by a normative identity style conform to standards and expectations held by authority figures and significant others, whereas individuals characterized by a diVuse/avoidant identity style are unwilling to confront directly and to deal with problems and identity issues. We anticipated that higher authenticity would relate to higher self‐esteem level, clarity, identity integration, incremental self‐theories, and informational identity styles and would relate to lower contingent self‐esteem, self‐ concept diVerentiation, and normative and diVuse/avoidant identity styles. Table V displays the correlations. As can be seen, the data strongly supported our expectations. Specifically, higher dispositional authenticity scores related to feelings of self‐worth that were both more favorable (higher self‐ esteem level) and more secure (less contingent feelings of self‐worth). In addition, higher authenticity scores related to aspects of self‐organization that are characterized by possessing a self‐concept that (1) is clearly and confidently defined (high self‐concept clarity) and (2) exhibits less variability or fragmentation across one’s social roles (low self‐concept diVerentiation). Furthermore, with respect to identity styles, higher authenticity related to greater tendencies to actively explore identity relevant information (high informational identity styles) and lower tendencies to avoid acknowledging, deciding, or reconciling their identity (low diVuse identity styles). Finally, higher dispositional authenticity reflected heightened tendencies toward growth motivations reflected by possessing implicit self‐theories characterized

318

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN TABLE V CORRELATIONS BETWEEN AUTHENTICITY AND SELF‐CONCEPT VARIABLES (GOLDMAN, 2004)

Self‐concept clarity Identity integration Identity styles: diVusion Self‐concept diVerentiation Entity Contingent self‐esteem Self‐esteem level

Total authenticity

Awareness

Unbiased processing

Behavior

Relational orientation

.68** .57** .51** .32** .24* .58** .60**

.67** .61** .52** .28* .29* .43** .56**

.47** .33** .32** .28* .14 .51** .44**

.55** .56** .42** .16 .23* .50** .37**

.29* .16 .19a .22a .03 .24* .39**

Note: ap < .10, *p < .05, **p < .01.

in incremental terms (i.e., believing one’s eVorts have meaningful implications for changing outcomes in important self‐aspects). In sum, greater dispositional authenticity reflected components of self‐esteem, self‐ organization, and self‐theories that involved a stronger, as opposed to weaker, sense of self. Dispositional authenticity reflects heightened self‐knowledge and understanding and openness toward knowing one’s self accurately. In contrast to most measures of self‐concept organization that focus on structural or meta‐ knowledge features of self‐concept (e.g., how clearly the self‐concept is defined), dispositional authenticity also assesses one’s prevailing motivational tendencies toward acquiring and processing self‐relevant information (i.e., awareness and unbiased processing component). Consequently, the strong relations that emerged between dispositional authenticity and these structural aspects of self‐concept suggest that dispositional authenticity reflects an interface between self‐concept organization and its motivational properties. For instance, higher authenticity reflected greater self‐concept clarity and identity integration. Likewise, greater authenticity related to greater beliefs that people felt they could change themselves through their eVorts (i.e., incremental self‐theorists), a stance toward the self that is central to philosophical and psychological perspectives that emphasize personal choice and responsibility. Goldman (2004) conducted additional analyses to examine the extent to which dispositional authenticity predicted these aspects of a stronger sense of self independently of self‐esteem level. Both self‐esteem level and dispositional authenticity independently predicted a number of these aspects,

AUTHENTICITY

319

namely, self‐concept clarity, contingent self‐esteem, identity integration, and self‐actualization. In addition, whereas only authenticity predicted diVuse and informational identity styles, only self‐esteem level predicted normative identity styles such that higher levels of self‐esteem related to greater normative identity styles. The pattern of findings for identity styles underscores the importance of diVerentiating authenticity from self‐esteem level. Specifically, whereas dispositional authenticity uniquely predicted greater openness in exploring one’s identity, and less avoidance and confusion in reconciling one’s identity, self‐esteem level predicted heightened tendencies to reconcile one’s identity by integrating social norms. In sum, these analyses indicate that dispositional authenticity predicts important outcomes independent of self‐esteem level.

E. SOCIAL ROLE FUNCTIONING A complete framework of authentic functioning necessitates taking into consideration individuals’ social roles. When individuals respond to open ended questions, such as ‘‘Who am I?,’’ they spontaneously describe themselves with reference to specific social roles, in addition to decontextualized personal attributes (Coˆte´ & Levine, 2002; Gordon, 1968). Thoits (1992) reported that 85% of respondents indicated one or more social roles as self‐descriptors to a five‐item ‘‘Who am I?’’ Social roles involve identifying oneself as a certain kind of person in relation to specific role partners. Some scholars assert that social roles represent an especially important component of self‐conception because ‘‘most daily interactions occurs in role relationships . . . and, many, if not most, social roles imply auxiliary or embedded social characteristics’’ (Thoits & Virshop, 1997, p. 123). Numerous investigators have examined the relationship between aspects of individuals’ functioning in their role‐identities and psychological adjustment. For instance, researchers have examined adjustment as it relates to role‐accumulation (e.g., Thoits, 1992), role‐balance (Marks, 1986), role‐ overload (e.g., Hecht, 2001), role‐strain (e.g., Thoits, 1986), and role‐conflict (e.g., Sheldon, Ryan, Rawsthorne, & Ilardi, 1997). Investigations examining role‐accumulation have found that possessing a greater number of roles buVers against threats to adjustment (e.g., Thoits, 1986). Marks (1986, p. 420) examined role balance, defined as ‘‘the tendency to become fully engaged in the performance of every role in one’s total role system, to approach every typical role and role partner with an attitude of attentiveness and care.’’ Individuals higher in role‐balance reported significantly higher role‐ease and self‐esteem, and lower levels of role‐overload and depression, than did those lower in role‐balance. Thus, research on role‐accumulation

320

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

and role‐balance indicates that to the extent one possesses an adequate number of roles in which one is fully engaged, psychological adjustment may be enhanced. In contrast, role‐overload (when a person is faced with too many expectations), role‐strain (when a person’s roles are overly diYcult to enact), and role‐conflict (when concurrent roles are not compatible) all have been found to be associated with heightened stress levels (Biddle, 1986). Of particular importance in the present context is the extent to which one’s role‐identities are personally chosen and experienced as authentic. When enacting roles that foster feelings of choice and authenticity, perceived stress within such roles should be low, and individuals should more fully integrate these roles into their self‐systems (Ryan & Deci, 2003; Thoits, 1992). Sheldon et al. (1997) examined the relation of self‐integration (authentic role functioning) to well‐being and adjustment in two studies. They assessed self‐ integration in terms of participants’ felt authenticity across the same five social roles examined by Goldman (2004). Authentic role‐functioning was assessed through items such as ‘‘I experience this aspect of myself as an authentic part of who I am’’ and ‘‘I have freely chosen this way of being.’’ A number of interesting findings emerged. First, higher scores in role‐specific authenticity significantly correlated with greater satisfaction in all five roles and with greater preference to spend more time in four of five roles (the friend role was the exception). Second, role‐authenticity ratings negatively correlated with self‐concept diVerentiation (r ¼ .61), indicating greater‐felt authenticity was associated with lower levels of self‐fragmentation across social roles. Third, role authenticity ratings and self‐concept diVerentiation (SCD) scores independently predicted other indices of psychological adjustment. In short, Sheldon et al.’s (1997) findings demonstrate the importance of experiencing authenticity in one’s social roles for fostering healthy psychological adjustment. We anticipated that dispositional authenticity would relate to indices reflecting healthier role functioning. Highly operative authenticity presumably provides individuals with a depth of inner resources that serve to enhance their global interpersonal and psychological adjustment. For instance, by having greater self‐understanding, individuals high in authenticity seemingly are capable of self‐selecting appropriate niches in their interpersonal milieu that sustain and promote their interpersonal and psychological adjustment. Whereas some people may experience social roles with great distress, other people may experience them as opportunities for personal growth or meaning. In addition, by exhibiting greater self‐acceptance in processing self‐relevant information in an unbiased manner, individuals with higher levels of authenticity may perhaps inoculate themselves from the adverse influence of others when evaluating themselves in their social roles. That is, individuals who are oriented toward self‐acceptance will find joy and

321

AUTHENTICITY

happiness enacting social roles even when their objective performance may not meet others’ standards (as in a marathon runner who finishes in over 5 hours). In sum, we presumed that dispositional authenticity reflects a wide range of important psychological characteristics that diVerentiate how individuals experience their social environment and social roles. The correlations are displayed in Table VI in which it can be seen that dispositional authenticity related consistently to the role functioning variables as we anticipated. Higher dispositional authenticity related to healthy role functioning across a range of commonly enacted social roles (i.e., being a son/daughter, a student, a romantic partner, a friend, and an employee). For instance, higher dispositional authenticity related to positive aspects of general role functioning including greater satisfaction and positive aVectivity experienced within one’s social roles, as well as greater ‘‘balance’’ of one’s total role‐ system. In addition, higher dispositional authenticity related to less negative aspects of general role functioning as reflected in less stress within their commonly enacted social roles, and less ‘‘overload’’ in their social roles in general. Finally, heightened levels of dispositional authenticity also reflected authentic aspects of role functioning. Specifically, greater dispositional authenticity relates to role experiences that were: (1) reflective of greater expressiveness of their true beliefs and opinions (role‐voice), (2) more fully involved the enactment of their true‐selves (greater true‐self role enactment), (3) subjectively deemed to be authentic (role authenticity), and (4) regulated

TABLE VI CORRELATIONS BETWEEN AUTHENTICITY AND ROLE FUNCTIONING VARIABLES (GOLDMAN, 2004)

Role Variable Balance Overload Satisfaction Stress Strain Voice Authenticity Net positive aVect True‐self Self‐determination

Total authenticity

Awareness

Unbiased processing

Behavior

Relational orientation

.35** .06 .42** .23a .25* .55** .46** .44** .43** .24*

.29* .03 .44** .16 .16 .51** .40** .45** .36** .20a

.17 .17 .36** .14 .30** .28** .25* .28** .29** .18

.25* .06 .21a .23* .13 .36** .29* .27* .28* .20a

.33** .05 .23* .13 .16 .45** .43** .30** .35** .10

Note: ap < .10, *p < .05, **p < .01.

322

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

by more self‐determined reasons. In sum, higher dispositional authenticity related to diverse aspects of healthy role functioning that also included several markers of authentic role experiences. An important characteristic of this study is that social role variables represent ‘‘middle‐level units’’ of personality functioning (Cantor & Zirkel, 1990) that involve a complex of motivations and precepts that function to guide experiences and behaviors. In a similar vein, individuals’ goal pursuits reflect middle‐level units that organize their day‐to‐day experiences and contribute to their overall sense of well‐being. In the following study, we report how authentic goal pursuits are linked with more general aspects of well‐being.

F. AUTHENTICITY, GOAL PURSUITS, AND PSYCHOLOGICAL ADJUSTMENT Goldman, Kernis, Foster, Herrmann, and Piasecki (2005b) examined the extent to which dispositional and goal‐based indexes of authentic functioning related to each other and to markers of well‐being. The eudaimonic view of well‐being calls upon people to live their lives in accord with their daimon, or true‐self (Ryan & Deci, 2001; Waterman, 1993). From this perspective, psychological health and well‐being (eudaimonia) occur when people’s lives are congruent with their deeply held values and core self, that is, when people are authentic. This theme is prominent in a number of major theories. For example, Rogers (1961, p. 351) emphasized the actualization tendency, described as ‘‘. . . the directional trend which is evident in all organic and human life—the urge to expand, extend, develop, and mature—the tendency to express and activate all the capacities of the organism, or the self. ’’ In a similar vein, SDT (Deci, 1980; Deci & Ryan, 1985; Ryan & Deci, 2000, 2002) holds that people are authentic when their actions reflect their true‐ or core‐selves, that is, when they are autonomous and self‐determining. Considerable research supports this claim (Deci & Ryan, 2000). Building on SDT, Sheldon and Elliot (1999) proposed a model of self‐ concordance that links goal strivings, self‐regulation, and basic need satisfaction. Self‐concordant goals are those that satisfy basic needs and are congruent with the true‐self. When individuals select and strive for goals that satisfy their basic needs, they tend to regulate their behavior in a highly self‐determined manner via intrinsic and identified motivations. Several studies show that highly self‐concordant goal strivings enhance psychological adjustment and well‐being (Sheldon & Elliot, 1999; Sheldon & Kasser, 1995, 1998).

AUTHENTICITY

323

Other researchers substantiate the claim that self‐concordant goals promote health and well‐being. McGregor and Little (1998) examined the meaning that people ascribed to their goal pursuits (i.e., personal projects) and its relationship with psychological well‐being. They reported that the more individuals rated their personal projects as reflecting personal integrity, the higher their psychological well‐being as assessed with RyV ’s (1989) measure. In another study, Sheldon et al. (1997) found that subjective role authenticity (i.e., the extent to which an individual feels authentic in each of five social roles) predicted greater psychological well‐being. Thus, theory and research suggest that authentic functioning (operating from one’s daimon) is critical to need satisfaction and optimal well‐being. The study by Goldman et al. (2005b) builds on prior study in several ways. First, we examined how individual diVerences in authentic functioning relate to psychological health and subjective well‐being. We predicted that the greater one’s authenticity, the healthier and more positive one’s psychological and subjective well‐being. Second, we examined how individual diVerences in authentic functioning relate to undertaking goals that are self‐concordant and fulfill one’s basic psychological needs. We predicted that the greater one’s authenticity, the more self‐concordant and need satisfying his or her goal pursuits. Third, we examined whether dispositional authenticity and undertaking need‐fulfilling goals independently predict well‐being. To the extent that they are independent predictors, this would oVer strong support for the assertion that trait level authenticity and authenticity expressed via goal pursuits both are important to incorporate into conceptualizations of well‐being. One hundred and eleven participants completed the AI (AI‐2). Subsequently, approximately 3 weeks later, they self‐identified various goal pursuits (i.e., personal projects) and rated them on various aspects of authentic goal processes (i.e., authenticity, eYcacy, stress/pressure, and intrinsic motivation). Ratings from these aspects were combined to form a project need‐ fulfillment index that reflected the overall degree to which people’s projects aVorded them authentic, need‐fulfilling experiences. Specifically, higher project need‐fulfillment reflected greater goal‐based authenticity (e.g., ‘‘ To what extent is this project consistent with the values which guide your life?’’ and ‘‘. . . to what extent does this project reflect who you really are?’’), eYcacy (e.g., ‘‘How competent are you to complete this project?’’ and ‘‘How successful do you think you will be at this project?’’), and intrinsic motivation (e.g., ‘‘How much do you enjoy working on this project?’’ and ‘‘To what extent is this project pleasurable, i.e., comfortable, relaxing, self‐indulgent, or hedonistic?’’), and lower goal‐based stress or pressure (e.g. ‘‘How diYcult do you find it to carry out this project?’’ and ‘‘How much do you feel the amount of time available for working on this project is adequate?’’).

324

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

Participants also completed measures assessing subjective well‐being, specifically, life satisfaction (Diener, Emmons, Larsen, & GriYn, 1985) and net positive aVect (Brunstein, 1993), and psychological forms of well‐ being, specifically, RyV ’s (1989) multicomponent measure of psychological well‐being, 3 weeks after completing the AI‐2. Our findings indicated that dispositional authenticity and project need‐ fulfillment positively correlated with each other (r ¼ .29, p < .01) and with all assessed markers of well‐being. In addition, as shown in Table VII, both dispositional authenticity and project need‐fulfillment independently predicted participant’s subsequent ratings of hedonic and eudaimonic well‐ being, as well as the vast majority of specific well‐being components (all in the case of dispositional authenticity). These findings suggest that authentic functioning, exhibited at the personality level and in middle‐level goal representations, meaningfully accounts for broad aspects of well‐being. The fact that dispositional authenticity and project need‐fulfillment positively correlate with one another suggests that we can conceptualize and measure authentic functioning in multiple ways. However, this bivariate relationship does not establish a causal direction with regard to these two indices of authentic functioning. Although it may be tempting to presume that dispositional authenticity may account for the extent to which people experience their goals as authentic and need fulfilling, alternative possibilities exist. For instance, people’s direct experiences in their goal pursuits may aVect their degree of dispositional authenticity. In other words, by satisfying needs for autonomy and competence through enacting need‐fulfilling goals, individuals may further promote their overall level of dispositional authenticity and its individual components. For example, awareness may increase when goal pursuits confer opportunities to discover and identify one’s interests and talents. Similarly, by satisfying needs for autonomy and competence, people may become less prone to use self‐esteem maintenance strategies that bias their processing of self‐relevant information (Goldman, in press; Kernis, 2003; Knee & Zuckerman, 1996). Our findings also indicated that individual diVerences in authentic functioning, as measured by the Authenticity Inventory, related to measures of both psychological and subjective well‐being. In fact, individual diVerences in authenticity predicted each facet of RyV ’s measure (and its composite), net positive aVect, and life satisfaction independently of our project need‐ fulfillment index, which itself was a consistent predictor of well‐being. We had anticipated the possibility that the need‐fulfillment index would mediate relationships that emerged between authenticity and other markers of well‐ being, but our findings did not bear this out. Instead, simultaneous regression analyses indicated either that both authenticity and goal need‐fulfillment

325

AUTHENTICITY TABLE VII PREDICTING WELL‐BEING FROM DISPOSITIONAL AUTHENTICITY PROJECT NEED‐FULFILLMENT Predictor

Beta

t

AND

p<

Model R2

Authenticity Project need‐fulfillment

Life Satisfaction .24 2.60 .29 3.11

.01 .01

.18

Authenticity Project need‐fulfillment

Net Positive AVect .26 3.18 .45 5.44

.01 .01

.34

.01 .19

.29

Autonomy Authenticity Project need‐fulfillment

.50 .11

5.90 1.31

Authenticity Project need‐fulfillment

Environmental Mastery .30 3.49 .32 3.71

.01 .01

.25

Authenticity Project need‐fulfillment

Personal Growth .35 3.85 .21 2.32

.01 .02

.20

Authenticity Project need‐fulfillment

Positive Relationships .32 3.46 .17 1.79

.01 .08

.16

Authenticity Project need‐fulfillment

Purpose in Life .29 3.08 .14 1.43

.01 .16

.13

Authenticity Project need‐fulfillment

Self‐Acceptance .24 2.57 .25 2.71

.01 .01

.15

Authenticity Project need‐fulfillment

Eudaimonic Well‐Being Composite .49 6.28 .29 3.78

.01 .01

.41

Authenticity Project need‐fulfillment

Hedonic Well‐Being Composite .27 3.24 .40 4.75

.01 .01

.39

Note: Eudaimonic composite ¼ sum of autonomy, environmental mastery, personal growth, positive relationships, purpose in life, and self‐acceptance. Hedonic composite ¼ sum of life satisfaction and net positive aVect (with z‐score transformation). All Model R2 significant at p < .01.

predicted well‐being or that only authenticity was an independent predictor; in no case was goal need‐fulfillment the only independent predictor. The fact that goal need‐fulfillment did not mediate the impact of dispositional authenticity on other measures of well‐being raises the question of

326

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

what mechanisms or processes might account for this relationship. One possibility is that a chronic tendency for one’s true‐ or core‐self to be operative confers direct benefits on one’s well‐being. That is, being true to oneself in thought, feeling, and behavior may promote eudaimonia (e.g., a sense of meaning, autonomy, growth, and mastery over one’s environment), as well as positive aVect and life satisfaction. This view of authenticity is similar to Rogers’ conceptualization of the fully functioning individual as one who is open to experience, lives fully in the present moment, trusts inner experience, is creative, and possesses a strong sense of freedom. These qualities are akin to those that RyV emphasizes in her model of eudaimonic well‐being.

VII. Authenticity, Relationship Functioning, and Relationship Satisfaction We believe that authentic functioning involves a number of processes and characteristics that aid in the development and maintenance of close relationships. In this section, we elaborate on such processes and characteristics and we report on a study in which we examined their operation within the context of romantic relationships among college students. A. SELF‐DISCLOSURE, TRUST, INTIMACY Our view of authentic functioning emphasizes that an authentic relationship involves placing a premium on valuing, striving, and achieving openness and honesty. Therefore, we presume that highly authentic individuals will foster healthy growth and development of their intimate relationships by disclosing self‐aspects that reflect who they really are, both good and bad. Thus, when authenticity is high, meaningful and honest self‐disclosure emerges and presumably facilitates intimacy that is rooted in feeling accepted by one’s relationship partner for being who one really is. This line of reasoning is largely consistent with Reis and Shaver’s (1988) proposition that two conditions are necessary for self‐disclosure to create intimacy. Specifically, they propose that intimacy develops in circumstances when (1) individuals engage in self‐disclosure that is emotional, as opposed to merely factual, and (2) their partners are responsive to the self‐disclosure, making them feel understood, validated, and cared for. Moreover, prior studies (e.g., Meeks, Hendrick, & Hendrick, 1998; Sprecher & Hendrich, 2004) demonstrate that self‐disclosure tendencies relate positively to relationship satisfaction. In the present investigation, we assessed emotional and intimate forms of self‐disclosure and people’s perceptions of their partners validation of them. To assess emotional self‐disclosure we administered Snell, Miller, and Belk’s

AUTHENTICITY

327

(1988) Emotional Self‐Disclosure Scale, a 40‐item measure that assesses participants’ willingness to discuss with their partner, those times when they felt various emotions (e.g., depressed, enraged, serene, happy). In addition, we administered the Self‐Disclosure Index (Miller, Berg, & Archer, 1983), a 10‐item measure that assesses the degree to which individuals had discussed various intimate topics with their relationship partner (e.g., ‘‘your deepest feelings’’ and ‘‘what you like and dislike about yourself ’’). Ratings were made on five‐point scales (1 ¼ not at all; 5 ¼ fully and completely). Rempel, Holmes, and Zanna (1985) theorized that interpersonal trust in close relationships is comprised of three components: predictability (the belief that the partner’s behavior is consistent), dependability (the belief that the partner can be counted on to be honest, reliable, and benevolent), and faith (the conviction that the partner will be caring and responsive in the future). Rempel et al. (1985) found that faith and dependability subscale scores significantly correlated with reported feelings on a composite measure of relationship adjustment that included love and liking toward one’s partner, and present levels of relationship happiness, satisfaction, and success. Descutner and Thelen (1991. p. 218) proposed that fear‐of‐intimacy reflects people’s ‘‘inhibited capacity, because of anxiety, to exchange thoughts and feelings of personal significance with another individual who is highly valued.’’ Specifically, fear‐of‐intimacy was theorized to be comprised of three features: (1) content (the communication of personal information), (2) emotional valence (strong feelings about the personal information exchanged), and (3) vulnerability (high regard for the intimate other). In two separate studies, Descutner and Thelen (1991) reported that higher fear‐of‐intimacy scores negatively correlated with participants’ satisfaction with the quality of their dating relationships. In our view, fear‐of‐intimacy reflects a barrier to authentic relationship functioning and consequently relationship satisfaction. We administered Descutner and Thelen’s 35‐item individual diVerence measure of fear‐of‐intimacy and modified the instructions to ask participants to make ratings based on their romantic partner. These ratings were made with respect to such content as how they would feel sharing personal things about the past, entrusting their most private thoughts with another person, their comfort with having open and honest communication, and taking the risk of being hurt in the context of a close relationship.

B. PARTNER VALIDATION Perceptions of partner validation were determined by questionnaire responses to a ‘‘Perceived Relationship Qualities’’ scale (Goldman & Kernis, 2005) we had generated for the purpose of the present study. The partner validation

328

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

subscale was composed of six‐items (e.g., ‘‘I feel my partner values me for who I really am as a person,’’ ‘‘I would like my partner to value me more than he/she does (reverse‐scored),’’ and ‘‘I feel my partner is supportive of my feelings on matters central to who I am’’). Internal consistency ratings indicated that the subscale had reasonably good reliability, alpha ¼ .79.

C. REACTIONS TO PARTNER’S BEHAVIORS AND RELATIONSHIP PROBLEMS We also examined how authenticity and relationship satisfaction relate to reactions to partner’s ambiguously negative reactions. Kernis, Goldman, and Paradise (2004) reasoned that individuals with secure high self‐esteem would interpret and react to ambiguously negative actions by their partner by treating them as innocuous, either by minimizing their negative aspects or by oVering a benign interpretation of them. In contrast, Kernis et al. (2004) reasoned and that individuals with fragile high self‐esteem would imbue these events with negative implications, either by internalizing their negative implications or by resolving to reciprocate in kind to get even with their partner. To test these hypotheses, Kernis et al. (2004) developed the Relationship Reaction Inventory (RRI) to gauge the extent to which people report defensive, highly ego‐involved reactions to ambiguously negative actions by their relationship partner. The RRI consists of nine scenarios that depicted ambiguously negative events in which their partner might engage. Each event has multiple plausible causes and implications for self, partner, and the relationship. Participants rated the likelihood that they would respond in each of four diVerent ways designed to capture this multiplicity of potential causes and implications. Two response options signaled overinvestment of the self and implied that the self somehow was threatened by the event. Of these, one (Personalizing) involved magnifying the event’s negative implications for the self. The other (Reciprocating) involved resolving to ‘‘get even’’ with one’s partner as a way to deal with the self‐esteem threat. The two remaining response options captured reactions or interpretations that did not involve overinvestment of the self. One of these (Benign) involved a transient externally based (usually partner related) explanation and the other (Minimize) involved taking the event at ‘‘face value,’’ that is, not making a big deal of it. An example scenario and response options is as follows: Your partner gives you a nice birthday present, but it is not what you have subtly let him/her know that you really wanted. How likely is it that you would . . .

AUTHENTICITY

329

(a) Think that you must not be important enough to him/her (Personalize) (b) Enjoy the present you got (Minimize) (c) Think that circumstances beyond his/her control must have prevented it (Benign) (d) In the future give him/her a present other than what you know he/she clearly wants (Reciprocating) Consistent with these hypotheses, Kernis et al. (2004) found that whereas unstable high self‐esteem individuals reported being most likely to engage in personalizing and get even reactions, stable high self‐esteem individuals reported being least likely (low self‐esteem individuals fell between). Conversely, whereas stable high self‐esteem individuals reported being most likely to engage in benign and minimizing reactions, unstable high self‐esteem individuals reported being least likely (again, low self‐esteem individuals fell between). These findings are important because they point to the operation of dynamics associated with fragile high self‐esteem that until now have been ascribed to low self‐esteem individuals (Murray, Holmes, MacDonald, & Ellsworth, 1998) or to those highly sensitive to rejection (Downey, Freitas, Michaelis, & Khouri, 1998). Elsewhere (Kernis, 2003; Kernis & Goldman, 2004, 2005), we theorized that complementary processes may be operative in authenticity and secure high self‐esteem development, and thus, we reasoned that high authenticity would similarly be related to low ego‐involved reactions (i.e., greater minimizing and benign reactions and less personalizing, and reciprocating reactions). We anticipated these findings for several reasons. First, high authenticity relates to secure forms of high self‐esteem, which Kernis et al. (2004) showed related to this pattern of reactions. Second, high authenticity (particularly the unbiased processing component) reflects objective, nondefensive processing of evaluative information. Third, correlations between authenticity and mindfulness reported in this chapter suggest that authenticity relates to non‐ego‐involved attentiveness and awareness of internal and external stimuli. Whereas the RRI asks respondents to indicate how they would respond to hypothetical partner transgressions, other work has examined various responses that people have when experiencing relationship problems. Specifically, Rusbult, Verette, Whitney, Slovik, and Lipkus (1991) developed a self‐report measure of accommodation tendencies. Participants respond to four separate stems (e.g., ‘‘When my partner is unintentionally unpleasant or thoughtless’’), each of which is followed by four response options. Specifically, participants indicate the extent to which they engage in constructive reactions: voice (actively and constructively attempting to improve conditions, e.g., ‘‘I talk to him or her about what’s going on, trying to work

330

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

out a solution’’) and loyalty (passively but optimistically waiting for conditions to improve, e.g., ‘‘I give my partner the benefit of the doubt and forget about it’’), and destructive reactions—exit (actively destroying the relationship, abusing one’s partner, or threatening to leave or separate, e.g., ‘‘I feel so angry that I want to walk right out the door’’) and neglect (ignoring or criticizing one’s partner, or avoiding spending time or discussing relevant issues, e.g., ‘‘I sulk and try to stay away from my partner for awhile’’). Across a variety of studies, Rusbult et al. (1991) reported that destructive reactions predicted lower relationship satisfaction, and constructive reactions predicted greater relationship satisfaction. In the present study, we predicted that higher authenticity would relate to greater constructive and lesser destructive conflict reactions.

D. RELATIONSHIP MOTIVES Finally, we asked people about their motives for being in their current relationships, using a variant of the self‐regulatory styles measure developed by Ryan and Connell (1989) in which people are asked to report on the reasons why they engage in a particular behavior. In this instance, respondents were asked to report their reasons for their behaviors as a romantic partner. Reasons reflect varying degrees of self‐determination. Intrinsic (e.g., I do things because of the pleasure and fun of doing them) and identified (e.g., I do things because they tie into my personal values and beliefs) reasons reflect high self‐determination, whereas introjected (e.g., I force myself to do things to avoid feeling guilty or anxious) and external (e.g., I do things because somebody else wants me to or because I will get something from somebody if I do them) reasons reflect low self‐determination. Likewise, participants completed the measure of relationship motivation reported by Rempel et al. (1985) that taps into the extent to which individuals’ relationships are intrinsically satisfying to them. Rempel et al. (1985) found that the more individuals endorsed possessing greater intrinsic motives, whereby relationship rewards involve mutual satisfaction, empathic concern, and value for both partners, the more they reported feeling love and faith in their relationship. In contrast, extrinsic motives that involved obtaining rewards exclusively outside the relationship (e.g., parental approval) were unrelated to the measures of relationship adjustment. Thus, possessing motivations that involve rewards obtained within one’s relationship (as opposed to gains obtained external to the couple) appears to confer benefits to the quality of a couple’s relationship. Sample items tapping into intrinsic motives include: (1) ‘‘We have a rewarding intellectual relationship. We have meaningful discussions which are stimulating and enriching.’’

AUTHENTICITY

331

(2) ‘‘We are close and intimate. We have special ways of demonstrating aVection and letting each other know how we feel.’’ (3) ‘‘He/she lets me be myself. He/she doesn’t tie me down and doesn’t try to change me.’’ We predicted that higher relationship authenticity would relate to more self‐ determined self‐regulation and intrinsic relationship motives. In order to examine the relationship between dispositional authenticity and relationship satisfaction, and the previously described relationship functioning variables, Goldman, Brunell, Kernis, Heppner, and Davis (2005a) administered the AI‐3 in an initial session to 61 heterosexual couples involved in a committed relationship of 3 or more months. Participants subsequently completed the remaining measures (except for relationship satisfaction, to be discussed) in two additional sessions that took place over an 8‐week period. We assessed relationship satisfaction twice, during the first and last sessions, using Rusbult’s five‐item satisfaction measure (Rusbult, 1983; Rusbult, Martz, & Agnew, 1998). Items included were ‘‘I feel satisfied with my relationship,’’ ‘‘My relationship is much better than others’ relationships,’’ ‘‘My relationship is close to ideal,’’ ‘‘Our relationship makes me very happy,’’ and ‘‘Our relationship does a good job of fulfilling my needs for intimacy, companionship, etc.’’ Each item was answered on a nine‐point scale (0 ¼ do not agree at all, 8 ¼ agree completely). We first examined the extent to which dispositional authenticity and participant gender predicted Time 2 satisfaction and change in satisfaction from Time 1 to Time 2. In both cases, significant main eVects emerged for authenticity, such that higher authenticity related to higher subsequent satisfaction and greater positive change in satisfaction. In neither case did gender produce any significant eVects, either as a main eVect or as interacting with authenticity. Consequently, we did not include gender in further analyses. We developed composite measures for a variety of relationship processes in order to examine more broadly based theoretical domains of close relationship functioning. Specifically, these relationship process composites include the following: self‐disclosure and relationship motives. For each composite an overall summary score was created by first calculating z‐scores for each measure and then adding the z‐scores. Self‐Disclosure reflected both emotional, (i.e., Snell et al.’s measure) and intimate (i.e., Miller et al.’s measure) forms. In addition, we included fear‐of‐intimacy scores (reverse scored so that higher summary scores reflected less fear) to reflect people’s tendencies to inhibit their thoughts and feelings from being shared with their partners. Relationship motives reflected the degree to which participants reported self‐determined reasons for relationship behaviors (i.e., self‐ determination index scores) and were motivated by rewards obtained within their relationships (i.e., intrinsic relationship motives). In addition, defensiveness in response to hypothetical partner transgressions involved responses on the RRI. Specifically, the defensive reactions

332

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

composite reflected ego‐involved reactions (i.e., personalizing and reciprocating) and the nondefensive reactions composite reflected non‐ego‐involved reactions (i.e., minimizing and benign explanations). With respect to Rusbult et al.’s (1991) accommodation measure, a destructive resolution tactic composite reflected the sum of exit and neglect subscales and a constructive resolution tactic composite reflected the sum of voice and loyalty subscales. We computed correlations between dispositional authenticity and each of the relationship process variables. In addition, we computed separate regression analyses in which we predicted relationship satisfaction from dispositional authenticity and each of these relationship process variables. Table VIII displays the correlations between overall dispositional authenticity and relational orientation scores with the relationship process

TABLE VIII CORRELATIONS BETWEEN AUTHENTICITY PROCESS VARIABLES

AND

RELATIONSHIP

Variable

Authenticity total

Relational orientation

Satisfaction T2 Partner trust Self‐regulation Intrinsic motives Relationship motives composite Miller‐Berg’s self‐disclosure Emotional self‐disclosure Fear‐of‐intimacy Self‐disclosure composite Minimize Benign Internalize Reciprocate Ego‐involved composite Non‐ego‐involved composite Voice Exit Loyalty Neglect Destructive composite Constructive composite

.29** .41** .41** .34** .48** .34** .32** .40** .43** .29** .30** .28** .32** .33** .31** .14 .31** .08 .42** .42** .04

.31** .41** .36** .45** .51** .38** .30** .44** .44** .25** .33** .15 .32** .26** .31** .31** .30** .09 .42** .41** .23**

Note: Relationship motives ¼ self‐regulatory styles and intrinsic relationship motive; self‐ disclosure composite ¼ Miller‐Berg’s self‐disclosure scale, emotional self‐disclosure scale, and fear‐of‐intimacy (reversed); ego‐involved ¼ internalize and reciprocate; non‐ego‐involved ¼ minimize and benign explanation; destructive ¼ exit and neglect; constructive ¼ voice and loyalty.

333

AUTHENTICITY

variables. Consistent with our view that authentic people value openness and intimacy in their close relationships, we find that higher authenticity relates to more self‐disclosure. In addition, higher authenticity relates to greater self‐determined reasons and intrinsic relationship motives. When responding to hypothetical partner transgressions, higher authenticity relates to greater non‐ego‐involved (benign and minimizing) and less ego‐involved reactions (personalizing and reciprocating). In addition, higher authenticity relates to greater constructive reactions and less destructive reactions to relationship problems, although this was less true for passive (loyalty) than for active (voice) constructive reactions. Tables IX and X, respectively, report the regression analyses with overall dispositional authenticity and each relationship process variable entered simultaneously as predictors of subsequent relationship satisfaction and change in satisfaction, respectively. With respect to subsequent satisfaction, we find that the tendency to self‐disclose, intrinsic relationship motives, and trust, each mediate the relationship between authenticity and satisfaction that emerged. That is, when authenticity and disclosure tendencies,

TABLE IX PREDICTING RELATIONSHIP SATISFACTION FROM AUTHENTICITY PROCESS VARIABLES

AND

RELATIONSHIP

Predictor

Beta

t

p<

Model R2

1. Authenticity 2. Authenticity Self‐disclosure 3. Authenticity Relationship motives 4. Authenticity Ego‐involved reactions 5. Authenticity Non‐ego‐involved reactions 6. Authenticity Destructive reactions to problems 7. Authenticity Trust

.29 .08 .51 .07 .48 .22 .20 .26 .12 .15 .34 .03 .63

3.37 0.91 5.90 0.80 5.46 2.37 2.17 2.78 1.34 1.65 3.77 0.43 8.17

.01 .38 .01 .43 .01 .03 .04 .01 .19 .10 .01 .67 .01

.09 .30 .26 .12 .10 .17 .41

Note: Constructive reactions to problems did not predict relationship satisfaction and therefore were not included in regression models. Relationship motives ¼ self‐regulatory styles and intrinsic relationship motives; self‐disclosure composite ¼ Miller‐Berg’s self‐disclosure scale, emotional self‐disclosure, and fear of intimacy; ego‐involved RRI reactions ¼ personalize plus reciprocate reactions; Non‐ego‐involved reactions ¼ benign plus minimize reactions, Destructive reactions ¼ exit and neglect.

334

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN TABLE X PREDICTING CHANGE IN RELATIONSHIP SATISFACTION FROM AUTHENTICITY AND RELATIONSHIP PROCESS VARIABLES Predictor

Beta

t

p<

1. Authenticity Time 1 satisfaction 2. Authenticity Self‐disclosure Time 1 satisfaction 3. Authenticity Relationship motives Time 1 satisfaction 4. Authenticity Ego‐involved reactions Time 1 satisfaction 5. Authenticity Non‐ego‐involved reactions Time 1 satisfaction 6. Authenticity Destructive reactions Time 1 satisfaction 7. Authenticity Trust Time 1 satisfaction

.16 .55 .05 .32 .43 .05 .29 .45 .12 .12 .53 .14 .07 .54 .09 .19 .50 .02 .46 .35

2.09 7.31 0.65 3.84 5.47 0.61 3.42 5.67 1.51 1.49 6.70 1.74 0.92 7.18 1.14 2.33 6.47 0.25 5.63 4.59

.05 .01 .53 .01 .01 .54 .01 .01 .14 .14 .01 .08 .36 .01 .26 .03 .01 .80 .01 .01

Model R2

.37

.43

.41

.38

.37

.38

.49

Note: Constructive reactions to problems did not predict relationship satisfaction and therefore were not included in regression models. Relationship motives ¼ self‐regulatory styles and intrinsic relationship motives; self‐disclosure composite ¼ Miller‐Berg’s self‐disclosure scale, emotional self‐disclosure, and fear of intimacy; ego‐involved RRI reactions ¼ personalize plus reciprocate reactions; non‐ego‐involved reactions ¼ benign plus minimize reactions, Destructive reactions ¼ exit and neglect.

relationship motives, or trust, are entered simultaneously, only disclosure tendencies, relationship motives, or trust remain predictive of relationship satisfaction. These mediation findings indicate that being able to open up and share intimate information with one’s partner, trusting one’s partner, and possessing intrinsic and self‐determined motives for one’s romantic relationship are important components of the relatively high relationship satisfaction experienced by authentic individuals. In contrast to findings from Kernis et al. (2004), RRI reactions did not mediate the association between authenticity and relationship satisfaction in the current study. Rather, change in satisfaction was marginally predicted by authenticity when entered simultaneously with non‐ego‐involved relationship reactions, whereas no significant predictors emerged with ego‐involved

AUTHENTICITY

335

relationship reactions. When predicting subsequent relationship satisfaction (i.e., no change in satisfaction), authenticity was an independent predictor when either ego‐involved or non‐ego‐involved relationship reactions also were in the equation. However, with respect to the Rusbult’s measure of accommodation, the composite assessing destructive reactions to relationship problems did mediate the association between authenticity and change in relationship satisfaction, suggesting that authenticity may influence changes in satisfaction by buVering against the use of destructive reactions. The findings from this study shed some light on how dispositional authenticity relates to the nature of individuals’ intimate relationships. Individuals high in dispositional authenticity report greater self‐disclosure and intrinsic relationship motives and exhibit less defensive ego‐involved reactions to either hypothetical partner transgressions or problems within their relationships. In turn, a number of these factors appear to account for why higher dispositional authenticity relates to higher relationship satisfaction.

VIII. Authenticity and Retrospective Accounts of Parental Authority Styles Kernis (2003) suggested that the awareness component of authenticity is facilitated by what is called intersubjectivity (Stern, 1985), ‘‘a state of connection and mutual understanding that emerges during interaction with another person . . .. A reasonable degree of match between the child’s experience and the adult’s feedback is necessary in order to establish a state of intersubjectivity; diVerent types of mismatches, such as when the caregiver fails to reflect the same emotional tone or energy level that the infant is feeling, can make the infant quite distressed and may lead to a disrupted sense of self ’’ (Stern, 1985; as cited in Hoyle et al., 1999, pp. 31–32). Kernis (2003) further suggested that the most damaging type of exchange for a child’s developing awareness involves a parent explicitly denying the legitimacy of a child’s inner experience, perhaps even punishing it. Continual punishment or contradiction of a child’s inner experiences may lead the child to ignore or dismiss these experiences in favor of those of the parental figure (cf., Deci & Ryan, 1985; Ryan, 1993). Some years ago, Sullivan called these self‐aspects the ‘‘bad me’’ or ‘‘not me.’’ Elements of the ‘‘bad’’ or ‘‘not’’ me also figure in the display of inauthentic or ‘‘false‐self ’’ behaviors (Harter, 1997). In contrast, familial environments that support the child’s expression of core aspects and that facilitate choicefulness should promote authentic functioning (Grolnick & Beiswenger, in press). Thus, it is important to

336

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

examine parental characteristics that promote or undermine the development of authenticity in children and young adults. Goldman et al. (2005c) examined how authenticity among college students related to their memories of their parents’ childrearing styles. Baumrind (1971, 1991), as cited by Berzonsky (2004) identified three major types (authoritative, authoritarian, and permissive) that . . . vary along a number of dimensions including the extent to which parents: establish firm guidelines and limits; explain and justify demands and expectations; assert power and control; and provide emotional support. Authoritative parents set clear, reasonable guideline and they exercise reliable control in a legitimate and loving fashion. They explain and justify their expectations and actions and they are responsive to feedback. However, authoritative parents will assert power and control when adolescents are too immature or self‐centered to listen to reason. . . . Authoritarian parents set definite limits and make rules that are not open to discussion. They make unilateral demands and use power to reinforce them. Permissive parents are responsive and indulging but make few demands and exercise limited control (Berzonsky, 2004, p. 214).

Authoritative parents are warm and supportive, foster autonomy in their children, and allow them voice. From our perspective, such behaviors on the part of parents will promote children’s developing confidence and trust in their own views, desires, motives, and so forth. We anticipated that higher authenticity would relate to higher ratings of authoritative parenting styles. In contrast, authoritarian parents stifle autonomy and children’s voice, instead coldly attempting to exert power and control with their demands. Finally, permissive parents, while warm and indulgent, do not provide the structure that most children need; instead, they leave children to their own devices. Consequently, we anticipated that authoritarian and permissive parental styles would relate to lower authenticity. We measured parenting styles using Buri’s (1991) parental authority questionnaire. Correlations are displayed in Table XI. The strongest relationships emerged between authenticity and both mothers’ and fathers’ authoritative styles; as anticipated, the higher the parents’ authoritative styles, the higher their adult children’s total authenticity, awareness, and unbiased processing. Although behavioral authenticity was unrelated to reports of parents’ authoritative styles, it did relate (inversely) to parental permissiveness in the manner we expected. Unbiased processing inversely correlated with mothers’ permissiveness, and total authenticity marginally correlated inversely with permissiveness. Surprisingly, no relationships emerged between parental authoritarianism and dispositional authenticity. In future research, it will be important to examine more directly the extent to which parents are autonomy‐supportive versus controlling. As noted by Vansteenkiste, Zhou, Lens, and Soenens (2005):

337

AUTHENTICITY TABLE XI CORRELATIONS BETWEEN AUTHENTICITY AND RETROSPECTIVE ACCOUNTS OF PARENTAL AUTHORITY

Measure

Total

Awareness

Unbiased processing

Behavior

Relational orientation

.10 .03 .07

.01 .08 .05 .10 .13 .14

Mother Father Combined

.04 .11 .05

.05 .14 .06

Authoritarian .00 .14 .09

Mother Father Combined

.22a .20a .24*

.01 .02 .04

Permissiveness .31** .17 .26*

.25* .30** .31**

Authoritative .27* .28* .31**

.20a .04 .13

Mother Father Combined

.38** .29* .36**

.47** .36** .45**

.20a .18 .21a

Note: a p < .10, *p < .05, **p < .01.

Parents promote volitional functioning in their children by being attuned to and empathic toward the child’s needs, by encouraging the child to act on his or her personally valued interests, and by minimizing the use of controlling parenting techniques (Grolnick, Deci, & Ryan, 1997). Conversely, parents will induce a controlled regulation when they use overtly controlling strategies (e.g., rewards, deadlines, punishments; see Deci, Koestner, & Ryan, 1999) or more subtle and implicit pressures (e.g., guilt‐induction, shaming, love withdrawal; Barber, 2002; Vansteenkiste, Simons, Lens, Soenens, & Matos, 2005) that are aimed to push adolescents to think, act, or feel in particular ways. Past research in Western samples has clearly demonstrated the beneficial well‐being and learning eVects of parental autonomy support (Grolnick, 2003). Conversely, consistent evidence has documented the negative developmental outcomes of controlling parenting (Barber, 2002, p. 468).

Although autonomy‐support and control are implied in authoritative and authoritarian styles respectively, these styles capture broader orientations than these dimensions per se. We believe that parental autonomy‐support and control are key elements in developing or undermining children’s and adolescents’ authentic functioning, but demonstration of such processes awaits future research. Similarly, studies that explicitly examine degrees of parental support for the components of dispositional authenticity are likely to play a pivotal role in distinguishing the eVects of parenting on authentic functioning.

338

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

IX. Authenticity and Self‐Esteem Earlier in this chapter, we reported findings indicating that dispositional authenticity predicted aspects of coping and self‐concepts independently of self‐esteem level. However, the issue of how authenticity and self‐esteem relate is complex because high self‐esteem has multiple forms, some more closely related to psychological health and well‐being than are others (Jordan, Logel, Spencer, & Zanna, in press; Kernis, 2003; Kernis & Waschull, 1995; Paradise & Kernis, 2002). Specifically, whereas some forms of high self‐esteem reflect secure high self‐esteem, other forms reflect fragile high self‐esteem (Kernis, 2003). Secure high self‐esteem involves feeling worthwhile and valuable, liking and satisfaction with oneself, accepting one’s weaknesses, having a solid foundation, and NOT requiring continual validation or promotion. In contrast, fragile high self‐esteem involves feeling very proud and superior to others, not liking to see weaknesses in oneself, or for others to see such weaknesses, and exaggerated tendencies to defend against possible threats to self‐worth and to engage in self‐promoting activities. Existing theory and literature provide several ways to distinguish between secure and fragile forms of high self‐esteem. Each of these forms has been discussed extensively elsewhere, along with supporting evidence (Kernis, 2003; Kernis & Paradise, 2002), so we only briefly discuss them here. Unstable (fragile) self‐esteem reflects substantial short‐term fluctuations in contextually based immediate feelings of self‐worth, whereas stable (secure) self‐esteem reflects minimal short‐term fluctuations. Contingent (fragile) self‐esteem is dependent on achieving specific outcomes, meeting expectations, matching standards, etc., whereas true (secure) self‐esteem is secure self‐worth that arises naturally from satisfaction of basic psychological needs, and that is not in need of continual validation. A match between individual’s implicit (nonconscious) and explicit (conscious) positive feelings of self‐worth reflects secure high self‐ esteem. In contrast, a mismatch between individual’s implicit (nonconscious) and explicit (conscious) feelings of self‐worth (one is negative) reflects fragile high self‐esteem. Optimal self‐esteem reflects the sum total of these secure self‐ esteem markers. It arises naturally from: (1) successfully dealing with life challenges; (2) the operation of one’s core, true, authentic self as a source of input to behavioral choices; and (3) relationships in which one is valued for whom one is, and not for what one achieves (Kernis, 2003). We believe that authenticity and each aspect of secure (versus fragile) high self‐esteem likely are reciprocally related to each other. That is, authenticity may provide both the foundation for achieving secure high self‐esteem and the processes through which secure high self‐esteem relates to psychological and interpersonal adjustment (Kernis & Goldman, 2004). When

339

AUTHENTICITY TABLE XII CORRELATIONS BETWEEN AUTHENTICITY AND SELF‐ESTEEM VARIABLES Total authenticity

Awareness

Unbiased processing

Behavior

Relational orientation

.52** .56** .39**

.48** .34** .23a

Spring 2005 Sample (61 Individuals) Self‐esteem Level Contingent Stability

.68** .58** .43**

.63** .58** .37**

.57** .43** .42**

Fall and Spring 2004 Sample (101 Individuals) Self‐esteem Level Contingent Stability

.47** .34** .34**

.48** .35** .32**

.34** .37** .31**

.39** .45** .29**

.24* .25 .15

Note: a < .10 * p < .05, **p < .01.

breakdowns in authenticity occur, they are likely to reverberate through the self‐system and cause decreased or more fragile self‐esteem. Conversely, possessing fragile self‐esteem may undermine or interfere with various processes associated with authenticity. For example, to ease the sting associated with failure, people with fragile self‐esteem may be more likely to engage in biased than unbiased processing, or to modify their behavior merely to please a potential evaluator. To date, we have obtained data linking authenticity to secure forms of high self‐esteem. Specifically, in several samples, we have found that higher dispositional authenticity relates to higher self‐ esteem levels, more stable self‐esteem, and less contingent self‐esteem. These data are displayed in Table XII. An important agenda for future research will be to further examine the interplay of these components of self‐esteem and dispositional authenticity.

X. Potential Downside of Authenticity The findings reviewed in this chapter reveal that dispositional authenticity relates to a diverse set of markers reflecting healthy psychological and interpersonal functioning. Why, then, is not everybody highly authentic? Elsewhere (Kernis & Goldman, 2005), we tackled this issue in depth and suggested that authenticity can have its costs. Here, we briefly review these costs.

340

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

A. AWARENESS Certain forms of self‐knowledge may be painful. Becoming aware of the limitations in one’s social skills may be painful, as may finding out that one is not as athletically talented as one had hoped. Perceived self‐discrepancies between one’s actual qualities and ideal qualities or between qualities one believes one should possess may produce negative emotional consequences (Higgins, 1989). In addition, powerful emotional experiences may be unsettling and even threatening, as they expose one’s vulnerabilities and sensitivities. Furthermore, self‐reflection itself may heighten unpleasant aVect, particularly when it involves attempts to understand one’s role in important negative outcomes or experiences. Finally, incorporating a diverse set of social roles into a multifaceted self‐concept may promote role strain and accompanying distress (Thoits, 1986).

B. UNBIASED PROCESSING The major drawback associated with unbiased processing is that it makes one susceptible to encountering negative information about the self. Although undoubtedly true, distorting or repressing negative self‐information is costly for a number of reasons. As we reported earlier in this chapter, defensively processing stressful information about oneself relates to lower, not higher, psychological functioning. Likewise, positively distorting self‐relevant information leaves one open to the social consequences associated with an arrogant self‐image. In any case, distorting self‐relevant information reflects heightened ego‐investment with its attendant rigidity and overreactivity (Hodgins & Knee, 2002).

C. BEHAVIOR Behaving authentically sometimes takes courage because one’s true inclinations may conflict with those of one’s peers or authority figures who have strong evaluative or controlling tendencies (Deci & Ryan, 1995). Sometimes behaving authentically runs the risk of prompting others’ scorn or ridicule, costs which can be very powerful inhibitors. However, behaving in ways that are at odds with one’s true‐self merely to satisfy controlling pressures can also undermine well‐being (Deci & Ryan, 2000; Neighbors, Larimer, Geisner, & Knee, 2004). When people conform to environmental dictates they are not always behaving in ways that conflict with their true‐selves. At times exhibiting false‐self behaviors can be an expression of role‐experimentation whereby people may facilitate self‐discovery and enrich the depth of their

AUTHENTICITY

341

self‐understanding. Harter (1999) notes that people’s motives for their inauthentic behaviors diVerentiate the consequences these actions have for their well‐being. For instance, inauthentic actions that are motivated by self‐uncertainty, or disdain for oneself, are likely to undermine well‐being more so than inauthentic actions that reflect a process of electing to express preexisting self‐ inclinations (e.g., role‐experimentation). Furthermore, people can, and often do, internalize social contingencies and freely adopt them as self‐guides (Ryan & Connell, 1989). In such cases, conflict is minimal or absent and it is relatively easy for people to behave authentically.

D. RELATIONAL ORIENTATION The potential costs of confrontations over relational authenticity are many, including self and partner defensiveness, overreactivity, and lack of intimacy. In turn, these costs contribute to shallow, unsatisfying relationships that are prone to dissolve over time or be continually fraught with problems and challenges (Kernis, Goldman, & Paradise, 2004). One can minimize these costs by choosing partners who also value relational authenticity.

E. IMPLICATIONS FOR HEDONIC AND EUDAIMONIC WELL‐BEING Thus, it is evident that authenticity can have potential costs. For example, accurate self‐knowledge can be painful, behaving in accord with one’s true‐ self may occasion others’ disfavor, and opening oneself up to an intimate makes one vulnerable to rejection or betrayal. Such adverse consequences potentially associated with authenticity are likely to undermine individuals’ hedonic, or subjective, well‐being. For instance, research on identity statuses demonstrates that optimal well‐being occurs when individuals experience identity achievement by resolving their identity crises through engaging in high levels of identity exploration and then committing to ways of resolving them (Marcia, 1966). Thus, for people to attain optimal well‐being through identity formation, they may have to temporarily endure costs to well‐being in the course of exploring who they can be. Similarly, authenticity may not always be pleasurable. However, we would argue that the benefits of authentic functioning to individuals’ eudaimonic well‐being (i.e., the extent to which they are fully functioning, Ryan & Deci, 2000) are substantial. When functioning authentically, people are likely to think, feel, and behave in ways that promote the fulfillment of their needs and heighten the degree to which they are fully functioning (Deci & Ryan, 2000; Rogers, 1961). Thus, people can be faced with choosing between experiencing pleasure (or avoiding displeasure)

342

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

and maximizing the extent to which they are fully functioning. How they resolve this dilemma has enormous implications, both short‐ and long‐term.

XI. Future Directions The findings reported in this chapter provide initial support for our multicomponent conceptualization of authenticity. The Authenticity Inventory appears adequate psychometrically, and confirmatory factor analyses support our view that the four components, while distinct, reflect a broad latent authenticity factor. Although we have accrued considerable validity data, we also recognize that our work is just beginning and many questions remain for us to address. An important avenue for future research is to obtain data relating our measure to outcomes other than that obtained from self‐reports. We have begun to do that (e.g., our study on verbal defensiveness), but more research needs to be done. We recognize that our self‐report measure of authenticity has the same inherent diYculties as the majority of self‐report measures. Respondents may either deliberately misrepresent themselves or have limited access to the information needed for valid responses. One way to deal with such issues is to obtain validity data that cannot easily be explained in terms of response biases. In this chapter, we presented data showing that scores on two self‐ esteem fragility measures (contingent SE and unstable SE, the latter represented by high response variability across multiple assessments) related to lower scores on the Authenticity Inventory. Likewise, we presented data demonstrating that higher authenticity scores related to less defensiveness while answering provocative questions during an interview. Although these findings are informative and diYcult to explain purely in terms of response biases, it would be beneficial to examine additional outcomes that involve behaviors or reactions to experimental manipulations. Elsewhere (Kernis & Goldman, 2005), we described a number of questions that we felt were interesting and important to examine in future research. We rely on that exposition in the discussion here.

A. AWARENESS One question pertains to whether high awareness relates to lower susceptibility to misattribution of arousal eVects. A second question is whether high awareness relates to individuals’ understanding better their emotions, motives, etc. when they describe a meaningful past experience in detail.

AUTHENTICITY

343

A third question is whether high awareness relates to greater interest and investment in gaining knowledge about one’s strengths and weaknesses. B. UNBIASED PROCESSING It would be interesting to examine whether unbiased processing relates to the relative absence of the self‐serving bias, as it does for autonomously functioning individuals (Knee & Zuckerman, 1996). A second question is whether unbiased processing relates to more mature or adaptive defense mechanisms that involve little reality distortion. Additional questions involve whether unbiased processing relates to less self‐enhancing retrospective memories pertaining to one’s performances or personal qualities and less idealized childhood memories. C. BEHAVIOR A number of questions revolve around whether behavioral authenticity actually relates to behaviors that are more congruent with one’s core‐self. For example, if people are made uncomfortable or anxious, does high behavioral authenticity relate to fewer instances of smiling and laughter (behaviors incongruent with one’s internal state)? Furthermore, does high behavioral authenticity relate to less susceptibility to ‘‘symbolic self‐completion manipulations’’ in which people’s actual goal completion eVorts are thwarted and they instead tend to symbolize completion? In addition, does high behavioral authenticity relate to greater behavioral consistency across audiences and contexts and does it relate to greater attitude–behavior consistency? D. RELATIONAL ORIENTATION We would expect relational orientation to relate to a number of relationship process variables. For example, we would expect that high relational orientation relates to valuing and behaviorally engaging in intimate self‐ disclosures with partners. In addition, we would expect that high relational orientation relates to less game‐playing (ludus) and manipulativeness in close relationships, and less idealization of one’s relationship or relationship partner (e.g., greater accuracy in evaluating aspects of the relationship or relationship partners). To the extent that our future research yields theoretically predicted findings to questions such as those just posed, it presumably will quiet concerns regarding the validity of our authenticity measure. Until that time, we will

344

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

remain vigilant in our sensitivity to its limitations. Although the research and theory reported in this chapter are in their early stages, they support the viability of a multicomponent conceptualization of authenticity. To do so, we had to overcome the extreme diYculty of capturing authenticity within a scientific framework. We are convinced that, despite its elusiveness, authenticity deserves its place alongside other critical aspects of the human condition that define who we are and what we are able to become. Our hope is that our work stimulates other scholars to join in our quest to understand it.

XII. Summary Our goal was to present a comprehensive overview of our multicomponent conceptualization of dispositional authenticity. We define authenticity as the ‘‘unimpeded operation of one’s true‐ or core‐self in one’s daily enterprise.’’ Our framework distinguishes four interrelated components of authentic functioning: awareness, unbiased processing, behavior, and relational orientation. We reported confirmatory factor analyses indicating that the AI (AI‐3) (Goldman & Kernis, 2004) measures these four discriminable components, which comprise a higher‐order latent authenticity factor. Thus, researchers can either use the total score as an index of overall authentic functioning, or each of the subscale scores if they are interested in specific aspects of authentic functioning. We reported research indicating that higher dispositional authenticity relates to many aspects of adaptive functioning, including problem‐focused coping strategies, mindfulness, positive role functioning, healthy aspects of self‐concept structure, hedonic and eudaimonic well‐ being, authentic goal pursuits, and low verbal defensiveness. In addition, higher dispositional authenticity relates to higher couple satisfaction and functioning. We considered the relation between authenticity and self‐ esteem, the potential costs of authenticity, and future research directions. Much more study is needed, but we believe that we have provided a solid foundation on which we and other researchers will be able to build.

Acknowledgments Preparation of this chapter was facilitated by support from the Department of Psychology and the Institute for Behavioral Research at the University of Georgia and NSF Grant 0451029. The authors thank Charles Lance for his assistance conducting and reporting statistical analyses and Pam Riddle for her assistance in preparing the chapter.

345

AUTHENTICITY

Appendix AUT3 The following measure has a series of statements that involve people’s perceptions about themselves. There are not right or wrong responses, so please answer honestly. Respond to each statement by writing the number from the scale below, which you feel most accurately characterizes your response to the statement.

1 Strongly Disagree

2 Disagree

3 Neither Agree Nor Disagree

4 Agree

5 Strongly Agree

1. I am often confused about my feelings. 2. I frequently pretend to enjoy something when in actuality I really don’t. 3. For better or for worse I am aware of who I truly am. 4. I understand why I believe the things I do about myself. 5. I want people with whom I am close to understand my strengths. 6. I actively try to understand which of my self‐aspects fit together to form my core‐ or true‐self. 7. I am very uncomfortable objectively considering my limitations and shortcomings. 8. I’ve often used my silence or head‐nodding to convey agreement with someone else’s statement or position even though I really disagree. 9. I have a very good understanding of why I do the things I do. 10. I am willing to change myself for others if the reward is desirable enough. 11. I find it easy to pretend to be something other than my true‐self. 12. I want people with whom I am close to understand my weaknesses. 13. I find it very diYcult to critically assess myself. 14. I am not in touch with my deepest thoughts and feelings. 15. I make it a point to express to close others how much I truly care for them. 16. I tend to have diYculty accepting my personal faults, so I try to cast them in a more positive way. 17. I tend to idealize close others rather than objectively see them as they truly are. 18. If asked, people I am close to can accurately describe what kind of person I am.

346

1 Strongly Disagree

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

2 Disagree

3 Neither Agree Nor Disagree

4 Agree

5 Strongly Agree

19. I prefer to ignore my darkest thoughts and feelings. 20. I am aware of when I am not being my true‐self. 21. I am able to distinguish those self‐aspects that are important to my core‐or true‐self from those that are unimportant. 22. People close to me would be shocked or surprised if they discovered what I keep inside me. 23. It is important for me to understand my close others’ needs and desires. 24. I want close others to understand the real me rather than just my public persona or ‘‘image.’’ 25. I try to act in a manner that is consistent with my personally held values, even if others criticize or reject me for doing so. 26. If a close other and I are in disagreement I would rather ignore the issue than constructively work it out. 27. I’ve often done things that I don’t want to do merely not to disappoint people. 28. I find that my behavior typically expresses my values. 29. I actively attempt to understand myself as best as possible. 30. I’d rather feel good about myself than objectively assess my personal limitations and shortcomings. 31. I find that my behavior typically expresses my personal needs and desires. 32. I rarely if ever, put on a ‘‘false face’’ for others to see. 33. I spend a lot of energy pursuing goals that are very important to other people even though they are unimportant to me. 34. I frequently am not in touch with what’s important to me. 35. I try to block out any unpleasant feelings I might have about myself. 36. I often question whether I really know what I want to accomplish in my lifetime. 37. I often find that I am overly critical about myself. 38. I am in touch with my motives and desires. 39. I often deny the validity of any compliments that I receive. 40. In general, I place a good deal of importance on people I am close to understanding who I truly am. 41. I find it diYcult to embrace and feel good about the things I have accomplished.

347

AUTHENTICITY

1 Strongly Disagree

2 Disagree

3 Neither Agree Nor Disagree

4 Agree

5 Strongly Agree

42. If someone points out or focuses on one of my shortcomings I quickly try to block it out of my mind and forget it. 43. The people I am close to can count on me being who I am regardless of what setting we are in. 44. My openness and honesty in close relationships are extremely important to me. 45. I am willing to endure negative consequences by expressing my true beliefs about things.

THE AUTHENTICITY INVENTORY (AI‐3) Version 3 Goldman and Kernis, 2004 The preceding measure is conceptually designed to assess the unimpeded operation of one’s true‐ or core‐self in one’s daily enterprise. There are four components to how we conceive of authenticity: awareness, unbiased processing, behavior, and relational orientation. These components can be measured via content domains that were constructed as subscales in the Authenticity Inventory and are described below: 1 Awareness: Awareness of, and trust in, one’s motives, feelings, desires, and self‐relevant cognitions. Conceptually, this includes awareness of one’s strengths and weaknesses, figure–ground personality aspects, emotions, and their roles in behavior. 2 Unbiased Processing: Not denying, distorting, exaggerating, nor ignoring private knowledge, internal experiences, and externally based self‐evaluative information. Conceptually then, this includes objectivity and acceptance of one’s positive and negative aspects. 3 Behavior: Acting in accord with one’s values, preferences, and needs. Conceptually, this contrasts acting merely to please others, or to attain rewards, or avoid punishments even if it means acting ‘‘falsely.’’ 4 Relational Orientation: Valuing and achieving openness and truthfulness in one’s close relationships. Conceptually, the relational component presumes it is important for close others to see the real you, good and bad. Moreover, relational authenticity means being genuine and not ‘‘fake’’ in one’s relationships with others.

348

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

Subscales Awareness: 1R, 3, 4, 6, 9, 14R, 20, 21, 29, 34R, 36R, 38 Alpha ¼ .79 Unbiased Processing: 7R, 13R, 16R, 19R, 30R, 35R, 37R, 39R, 41R, 42R Alpha ¼ .64 Behavioral: 2, 8R, 10R, 11R, 25, 27R, 28, 31, 32, 33R, 45 Alpha ¼ .80 Relational Orientation: 5, 12, 15, 17R, 18, 22R, 23, 24, 26R, 40, 43, 44 Alpha ¼ .78 Composite Scale Alpha ¼ .90 ***NOTE: R ¼ Reverse Scored Item

References Baer, R. A., Smith, G. T., & Allen, K. B. (2004). Assessment of mindfulness by self‐report: The Kentucky Inventory of Mindfulness Skills. Assessment, 3, 191–206. Bagozzi, R. P., & Phillips, L. W. (1991). Assessing construct validity in organizational research. Administrative Science Quarterly, 36, 421–458. Bandalos, D. L. (2002). The eVects of item parceling on goodness‐of‐fit and parameter estimate bias in structural equation modeling. Structural Equation Modeling, 9, 78–102. Barber, B. K. (2002). Reintroducing parental psychological control. In B. K. Barber (Ed.), Intrusive parenting: How psychological control aVects children and adolescents (pp. 3–13). Washington, DC: American Psychological Association. Baumeister, R. F. (1987). How the self became a problem: A psychological review of historical research. Journal of Personality and Social Psychology, 52, 163–176. Baumrind, D. (1971). Current patterns of parental authority. Developmental Psychology Monographs, 4(1, part 2), 1–103. Baumrind, D. (1991). The influence of parenting style on adolescent competence and substance use. Journal of Early Adolescence, 11, 56–95. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. Berzonsky, M. D. (1988). Self‐theorists, identity status, and social cognition. In D. K. Lapsley & F. C. Power (Eds.), Self, ego, and identity: Integrative approaches (pp. 243–262). New York: Springer. Berzonsky, M. D. (1990). Self‐construction over the life‐span: A process perspective on identity formation. In G. J. Neimeyer & R. A. Neimeyer (Eds.), Advances in personal construct psychology (Vol. 1, pp. 155–186). Greenwich, CT: JAI. Berzonsky, M. D. (2004). Identity style, parental authority, and identity commitment. Journal of Youth and Adolescence, 33, 213–220. Biddle, B. J. (1986). Recent developments in role theory. Annual Review of Sociology, 12, 67–92. Blaine, B., & Crocker, J. (1993). Self‐esteem and self‐serving biases in reactions to positive and negative events: An integrative review. In R. F. Baumesiter (Ed.), Self‐Esteem: The puzzle of low self‐regard (pp. 55–85). New York: Plenum Press. Brown, K. W., & Ryan, R. M. (2003). The benefits of being present: Mindfulness and its role in psychological well‐being. Journal of Personality and Social Psychology, 84, 922–948.

AUTHENTICITY

349

Brunstein, J. (1993). Personal goals and subjective well‐being: A longitudinal study. Journal of Personality and Social Psychology, 65, 1061–1070. Buri, J. R. (1991). Parental authority questionnaire. Journal of Personality Assessment, 57, 110–119. Campbell, J. D. (1990). Self‐esteem and clarity of the self‐concept. Journal of Personality and Social Psychology, 59, 538–549. Campbell, J. D., Trapnell, P. D., Heine, S. J., Katz, I. M., Covollee, L. F., & Lehman, D. R. (1996). Self‐concept clarity: Measurement, personality correlates, and cultural boundaries. Journal of Personality and Social Psychology, 70, 141–156. Cantor, N., & Zirkel, S. (1990). Personality, cognition, and purposive behavior. In L. Pervin (Ed.), Handbook of personality: Theory and research (pp. 135–164). New York: Guilford Press. Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies: A theoretically based approach. Journal of Personality and Social Psychology, 56, 267–283. Cloninger, S. C. (1993). Theories of personality: Understanding persons. Englewood CliVs, NJ: Prentice Hall. Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385–396. Cooley, C. H. (1902). Human nature and the social order. New York: Scribner’s. Coˆte´, J. E., & Levine, C. (2002). Identity, agency, and culture. Hillsdale, NJ: Lawrence Erlbaum. Cramer, P. (1990). Defense mechanisms in psychology today: Further processes for adaptation. American Psychologist, 55, 637–646. Crocker, J. (2002). Contingencies of self‐worth: Implications for self‐regulation and psychological vulnerability. Journal of Self and Identity, 1, 143–149. Crowell, S. (2005, September). Existentialism. In E.N., Zalta (Ed.), The stanford encyclopedia of Philosophy (Summer 2005 Edition), retrieved from http://plato.stanford.edu/archives/ sum2005/entries/existentialism/. Deci, E. L. (1975). Intrinsic motivation. New York: Plenum Press. Deci, E. L. (1980). The psychology of self‐determination. Lexington, MA: Heath. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self‐determination in human behavior. New York: Plenum Press. Deci, E. L., & Ryan, R. M. (1995). Human agency: The basis for true self‐esteem. In M. H. Kernis (Ed.), EYcacy, agency, and self‐esteem (pp. 31–50). New York: Plenum. Deci, E. L., & Ryan, R. M. (2000). The ‘‘what’’ and ‘‘why’’ of goal pursuits: Human needs and the self‐determination of behavior. Psychological Inquiry, 11, 227–268. Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A metaanalytic review of experiments examining the eVects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125, 627–668. Descutner, C. J., & Thelen, M. H. (1991). Development and validation of a fear‐of‐intimacy scale. Psychological Assessment, 3, 218–225. Diener, E., Emmons, R., Larsen, R., & GriYn, S. (1985). The satisfaction with life scale. Journal of Personality and Social Psychology, 49, 71–75. Donahue, E. M., Robins, R. W., Roberts, B. W., & John, O. P. (1993). The divided self: Concurrent and longitudinal eVects of psychological adjustment and social roles on self‐ concept diVerentiation. Journal of Personality and Social Psychology, 64, 834–846. Downey, G., Freitas, A. L., Michaelis, B., & Khouri, H. (1998). The self‐fulfilling prophecy in close relationships: Rejection sensitivity and rejection by close partners. Journal of Personality and Social Psychology, 75, 545–560. Dweck, C. S., Chiu, C., & Hong, Y. (1995). Implicit theories and their role in judgments and reactions: A world from two perspectives. Psychological Inquiry, 6, 267–285. Edwards, J. R. (2000). Multidimensional constructs in organizational behavior research: An integrative analytical framework. Organizational Research Methods, 4, 144–192.

350

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

Erickson, E. H. (1959). Identity and the life cycle. New York: International Universities Press. Feldman Barrett, L., Cleveland, J., Conner, T., & Williams, N. L. (2000). Manual for the defensive verbal behavior ratings scale (version 3.0). Unpublished manuscript, Boston College. Feldman Barrett, L., Williams, N. L., Boekman, L. F., & Feeney, K. A. (1996). Manual for the defensive verbal behavior rating scale. Feldman Barrett, L., Williams, N. L., & Fong, G. T. (2002). Defensive verbal behavior assessment. Personality and Social Psychology Bulletin, 28, 776–788. Fenigstein, A., Scheier, M. G., & Buss, A. H. (1975). Public and private self‐consciousness: Assessment and theory. Journal of Consulting and Clinical Psychology, 43, 522–528. Flynn, T. R. (2003). Jean‐Paul Sartre, Being and Nothingness (1943): The prodigious power of the negative. In J. E. Garcia, G. M. Reichberg, & B. N. Schumacher (Eds.), The classics of western philosophy: A reader’s guide (pp. 519–524). Massachusetts: Blackwell Publishing. Folkman, S., & Lazarus, R. S. (1980). An analysis of coping in a middle‐aged community sample. Journal of Health and Social Behavior, 21, 219–239. Folkman, S., & Lazarus, R. S. (1985). If it changes it must be a process: Study of emotion and coping during three stages of a college examination. Journal of Personality and Social Psychology, 48, 150–170. Gergen, K. J., & Gergen, M. M. (1988). Narrative and the self as relationship. Advances in Experimental Social Psychology, 21, 17–56. Gilligan, C. (1982). In a diVerent voice: Psychological theory and women’s development. Cambridge, MA: Harvard University Press. GoVman, E. (1959). The presentation of self in everyday life. Garden City, NY: Doubleday. Goldman, B. M. (2004). Dispositional authenticity as a predictor of self‐esteem, self‐concepts, and social role functioning. Unpublished data. Goldman, B. M. (in press). Making diamonds out of coal: The role of authenticity in healthy (optimal) self‐esteem and psychological functioning. In M. H. Kernis (Ed.), Self‐esteem: Issues and answers. New York: Psychology Press. Goldman, B. M., & Kernis, M. H. (2002). The role of authenticity in healthy psychological functioning and subjective well‐being. Annals of the American Psychotherapy Association, 5 (6), 18–20. Goldman, B. M., & Kernis, M. H. (2004). The development of the authenticity inventory, version 3. Unpublished data. Goldman, B. M., & Kernis, B. M. (2005). Dispositional authenticity as a predictor of coping styles. Unpublished data. Goldman, B. M., Brunnell, A., Kernis, M. H., Heppner, W., & Davis, P. (2005a). Dispositional authenticity as a predictor of romantic relationship functioning. Unpublished data. Goldman, B. M., Kernis, M. H., Foster, J. D., Hermann, A., & Piasecki, R. (2005b). Exercising one’s daimon: Dispositional authenticity and self‐concordant goal pursuits as predictors of well‐being. Manuscript submitted for publication. Goldman, B. M., Kernis, M., Stewart, E., Davis, P., Cascio, E., Lakey, C., & Heppner, W. (2005c). Dispositional authenticity as a predictor of parenting styles and parental support for authenticity. Unpublished data. Goldman, B. N., Lakey, C. E., & Kernis, M. H. (2005d). Authenticity and psychological functioning. Unpublished data. Gordon, C. (1968). Self‐conceptions: Configurations of content. In C. Gordon & K. J. Gergen (Eds.), The self is social interaction (pp. 115–136). New York: Wiley. Greenberg, J., Pyszcynski, T., & Solomon, S. (1986). The causes and consequences of a need for self‐esteem: A terror management theory. In R. F. Baumeister (Ed.), Public self and private self (pp. 189–212). New York: Springer‐Verlag.

AUTHENTICITY

351

Grolnick, W. S. (2003). The psychology of parental control: How wellmeant parenting backfires. Mahwah, NJ: Erlbaum. Grolnick, W. S., & Beiswenger, K. L. (in press). Facilitating children’s self‐esteem: The role of parents and teachers. In M. H. Kernis (Ed.), Self‐esteem issues and answers: A sourcebook of current perspectives. New York: Psychology Press. Grolnick, W. S., Deci, E. L., & Ryan, R. M. (1997). Internalization within the family: The self‐ determination theory perspective. In J. E. Grusec & L. Kuczynski (Eds.), Parenting and children’s internalization of values: A handbook of contemporary theory (pp. 135–161). New York: Wiley. Grosholz, E. R. (2003). Meditations on first philosophy (1641): Thought, existence, and the project of science. In J. E. Garcia, G. M. Reichberg, & B. N. Schumacher (Eds.), The classics of western philosophy: A reader’s guide (pp. 217–233). Massachusetts: Blackwell Publishing. Hagtvet, K. A., & Nasser, F. M. (2004). How well do item parcels represent conceptually dewfined latent constructs? A two‐facet approach. Structural Equation Modeling, 11, 168–193. Hall, R. J., Snell, A. F., & Faust, M. S. (1999). Item parceling strategies in SEM: Investigating the subtle eVects of unmodeled secondary constructs. Organizational Research Methods, 2, 233–256. Halling, S., & Carroll, A. (1999). Existential‐phenomenological psychology. In D. Moss (Ed.), Humanistic and transpersonal psychology. A historical and biographical sourcebook (pp. 95–124). Westport, CT: Greenwood Press/Greenwood Publishing. Harter, S. (1997). The personal self in social context: Barriers to authenticity. In R. D. Ashmore & L. J. Jussim (Eds.), Self and identity. Fundamental Issues (pp. 81–105). New York: Oxford University Press. Harter, S. (1999). The construction of the self. New York: Guilford Press. Harter, S., Waters, P. L., Pettit, L., Whitesell, N. R., Kofkin, J., & Jordan, J. V. (1997). Autonomy and connectedness as dimensions of relationship styles in adult men and women. Journal of Social and Personal Relationships, 14(2), 147–164. Harter, S., Waters, P., & Whitesell, N. R. (1997). False self behavior and lack of voice among adolescent males and females. Educational Psychologist, 32, 153–173. Hecht, L. M. (2001). Role conflict and role overload: DiVerent concepts, diVerent consequences. Sociological Inquiry, 71, 111–121. Heidegger, M. (1968). Existence and being: Introduction and analysis by Werner Brock. (W. Brock, Trans., 6th ed.). Chicago, IL: Henry Regnery Company. (Original American edition published 1949). Hermans, H. J. M. (1987). Self as organized system of valuations: Toward a dialogue with the person. Journal of Counseling Psychology, 34, 10–19. Hermans, H. J. M., Rijks, T. I., & Kempen, H. J. G. (1993). Imaginal dialogues in the self: Theory and method. Journal of Personality, 61, 207–236. Higgins, E. T. (1989). Self‐discrepancy: A theory relating self and aVect. Psychological Review, 94, 319–340. Hodgins, H. S., & Knee, C. R. (2002). The integrating self and conscious experience. In E. L. Deci & R. M. Ryan (Eds.), Handbook of self‐determination research (pp. 87–100). Rochester, NY: University of Rochester Press. Hoyle, R. H., Kernis, M. H., Leary, M. R., & Baldwin, M. W. (1999). Selfhood: Identity, esteem, regulation. Oxford, UK: Westview Press. Hu, L., & Bentler, P. M. (1998). Fit indexes in covariance structure modeling: Sensitivity to underparameterization model misspecification. Psychological Methods, 3, 424–453. Hu, L., & Bentler, P. M. (1999). CutoV criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55.

352

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

Hutchinson, D. S. (1995). Ethics. In J. Barnes (Ed.), The Cambridge companion to Aristotle, (pp. 195–232). Cambridge, UK: The Press Syndicate of the University of Cambridge. Irwin, T. H. (2003). Aristotle, Nicomachean Ethics (367–323 BC): A sort of political science. In J. E. Garcia, G. M. Reichberg, & B. N. Schumacher (Eds.), The classics of western philosophy: A reader’s guide (pp. 56–69). Massachusetts: Blackwell Publishing. Jones, A., & Crandall, R. (1986). Validation of a short index of self‐actualization. Personality and Social Psychology Bulletin, 12, 63–73. Jordan, C. H., Logel, C., Spencer, S. J., & Zanna, M. P. (in press). Nonconscious self‐esteem: Is there something you’re not telling yourself? In M. H. Kernis (Ed.), Self‐esteem issues and answers: A sourcebook of current perspectives. New York: Psychology Press. Jourard, S. M. (1971). The transparent self. New York: D. Van Nostrand Company. Kernis, M. H. (2000). Substitute needs and fragile self‐esteem. Invited commentary of Deci & Ryan’s target article. Psychological Inquiry, 11, 227–268. Kernis, M. H. (2003). Toward a conceptualization of optimal self‐esteem. Psychology Inquiry, 14, 1–26. Kernis, M. H., & Goldman, B. M. (2004). Authenticity, social motivation, and well‐being. In J. P. Forgas, K. D. Williams, & S. M. Laham (Eds.), Social motivation: Conscious and unconscious processes (210–227). Cambridge, United Kingdom: Cambridge University Press. Kernis, M. H., & Goldman, B. M. (2005). From thought and experience to behavior and interpersonal relationships: A multicomponent conceptualization of authenticity. In A. Tesser, J. V. Wood, & D. Stapel (Eds.), On building, defending and regulating the self: A psychological perspective. New York: Psychology Press. Kernis, M. H., & Goldman, B. M. (in press). Measuring stability of self‐esteem. In M. H. Kernis (Ed.), Self‐esteem issues and answers: A sourcebook of current perspectives. New York: Psychology Press. Kernis, M. H., & Goldman, B. M. (2002). Stability and variability in self‐concept and self‐ esteem. In M. Leary & J. Tangney (Eds.), Handbook of self and identity (pp. 106–127). New York: Guilford Press. Kernis, & Paradise, A. W. (2002). Distinguishing between secure and fragile forms of high self‐ esteem. In E. L. Deci & R. M. Ryan (Eds.), Handbook of self‐determination research (pp. 330–360). Rochester, NY: University of Rochester Press. Kernis, M. H., & Paradise, A. W. (2004). Contingent self‐esteem scale development. Unpublished data. Kernis, M. H., & Waschull, S. B. (1995). The interactive roles of stability and level of self‐ esteem: Research and theory. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 27, pp. 93–141). San Diego, CA: Academic Press. Kernis, M. H., Goldman, B. M., Piasecki, R., & Brunnell, A. (2003). Authenticity, functional flexibility, and mindfulness. Unpublished data. Kernis, M., Goldman, B. M., & Paradise, A. N. (2004). Fragile self‐esteem, overreactivity and satisfaction in close relationships. Unpublished data. Kernis, M. H., Lakey, C., Heppner, W., Goldman, B. M., & Davis, P. (2005). Fragile self‐esteem predicts verbal defensiveness. Manuscript in preparation. Kernis, M. H., Paradise, A. W., Whitaker, D., Wheatman, S., & Goldman, B. (2000). Master of one’s psychological domain? Not likely if one’s self‐esteem is unstable. Personality and Social Psychology Bulletin, 26, 1297–1305. Kierkegaard, S. (2004). From fear and trembling problema I. In G. Marino (Ed.), Basic writings of existentialism (pp. 7–24). New York: The Modern Library. (Excerpts From Søren

AUTHENTICITY

353

Kierkegaard, Fear and Trembling, V. Howard & E. H. Hong, Trans., 1983, New Jersey: Princeton University Press). Knee, C. R., & Zuckerman, M. (1996). Causality orientations and the disappearance of the self‐serving bias. Journal of Research in Personality, 30, 76–87. Koestner, R., Bernieri, F., & Zuckerman, M. (1992). Self‐regulation and consistency between attitudes, traits, and behaviors. Personality and Social Psychology Bulletin, 18, 52–59. Lakey, C., Kernis, M. H., Heppner, W., & Davis, P. (2005). Authenticity and mindfulness. Unpublished data. Landis, R. S., Beal, D. J., & Tesluk, P. E. (2000). A comparison of approaches to forming composite measures in structural equation models. Organizational Research Methods, 3, 186–207. Leary, M. R. (1995). Self‐presentation: Impression management and interpersonal behavior. Madison, WI: Brown & Benchmark. LeBel, J. L., & Dube´, L. (2001, June). The impact of sensory knowledge and attentional focus on pleasure and on behavioral responses to hedonic stimuli. Paper presented at the 13th annual American Psychological Society Convention. Toronto, Ontario, Canada. Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002). To parcel or not to parcel: Exploring the question, weighing the merits. Structural Equation Modeling, 9, 151–173. Marcia, J. E. (1966). Development and validation of ego‐identity status. Journal of Personality and Social Psychology, 3, 551–558. Marks, S. (1986). Multiple roles and the self: A theory of role balance. Journal of Marriage and the Family, 58(2), 417–433. Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist, 41, 954–969. Marsh, H. W., Balla, J. R., & McDonald, R. P. (1988). Goodness‐of‐fit indexes in confirmatory factor analysis: The eVect of sample size. Psychological Bulletin, 103, 391–410. Maslow, A. H. (1968). Toward a psychology of being (2nd ed.). New York: D. Van Nostrand. May, R. M. (1960). The emergence of existential psychology. In R. May (Ed.), Existential psychology (pp. 11–51). New York: Random House. McAdams, D. P. (1995). The ‘‘imago’’: A key narrative component of identity. In P. Shaver (Ed.), Review of personality and social psychology, 6, 87–113. McAdams, D. P. (1999). Personal narratives and the life story. In L. A. Pervin & O. P. John (Eds.), Handbook of personality theory and research (2nd ed., pp. 478–500). New York: Guilford Press. McDonald, W. (2005, September). Søren Kierkegaard. In E. N. Zalta (Ed.), The Stanford encyclopedia of philosophy (Summer 2005 edition). retrieved from http://plato.stanford. edu/archives/sum2005/entries/kierkegaard/. McGregor, I., & Little, B. R. (1998). Personal projects, happiness, and meaning: On doing well and being yourself. Journal of Personality and Social Psychology, 74, 494–512. Meade, G. H. (1934). Mind, self, and society. Chicago: University of Chicago Press. Meeks, B. S., Hendrick, S. S., & Hendrick, C. (1998). Communication, love, and relationship satisfaction. Journal of Social and Personal Relationships, 15, 755–773. Mikulincer, M., Orbach, I., & Iavenieli, D. (1998). Adult attachment styles and aVect regulation: Strategic variations in subjective self‐other similarity. Journal of Personality and Social Psychology, 75, 436–448. Mikiluncer, M., & Shaver, P. R. (2005). Mental representations of attachment security: Theoretical foundation for a positive social psychology. In M. W. Baldwin (Ed.), Interpersonal cognition (pp. 233–266). New York: Guilford Press.

354

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

Miller, L.,C, Berg, J.,H, & Archer, R. L. (1983). Openers: Individuals who elicit intimate self‐ disclosure. Journal of Personality and Social Psychology, 44, 1234–1244. Moos, R. H., & Schaefer, J. A. (1993). Coping resources and processes: Current concepts and measures. In L. Goldberger & S. Breznitz (Eds.), Handbook of Stress: Theoretical and clinical aspects (2nd ed.). New York: Free Press. Murray, S. L., Holmes, J. G., MacDonald, G., & Ellsworth, P. C. (1998). Through the looking glass darkly? When self‐doubts turn into relationship insecurities. Journal of Personality and Social Psychology, 75, 1459–1480. NeV, K. D. (2003). The development and validation of a scale to measure self‐compassion. Self and Identity, 2, 223–250. Neighbors, C., Larimer, M. E., Geisner, I. M., & Knee, C. R. (2004). Feeling controlled and drinking motives among college students: Contingent self‐esteem as a mediator. Self and Identity, 3, 207–224. Nurmi, J., Berzonsky, M. D., Tammi, K., & Kinney, A. (1997). Identity processing orientation, cognitive and behavioural strategies and well‐being. International Journal of Behavioral Development, 21, 555–570. O’Brien, E. J., & Epstein, S. (1988). The multidimensional self‐esteem inventory: Professional manual. Odessa, FL: Psychological Assessments Resources. Paulhus, D. L., & Martin, C. L. (1988). Functional flexibility: A new conceptualization of interpersonal flexibility. Journal of Personality and Social Psychology, 55, 88–101. Paradise, A. W., & Kernis, M. H. (2002). Self‐esteem and psychological well‐being: Implications of fragile self‐esteem. Journal of Social and Clinical Psychology, 21, 345–361. Perls, F., HeVerline, R. F., & Goodman, P. (1951). Gestalt therapy. New York: Julian Press (reprinted 1965, Dell Press). Reis, H. T., & Patrick, B. C. (1996). Attachment and intimacy: Component processes. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 523–563). New York: Guilford Press. Reis, H. T., & Shaver, P. (1988). Intimacy as interpersonal process. In S. Duck (Ed.), Handbook of personal relationships: Theory, research and interventions (pp. 367–389). Chichester, UK: Wiley. Rempel, J. K., Holmes, J. G., & Zanna, M. P. (1985). Trust in close relationships. Journal of Personality and Social Psychology, 49, 95–112. Robins, R. W., & Beer, J. S. (2001). Positive illusions about the self: Short‐term benefits and long‐term costs. Journal of Personality and Social Psychology, 80, 340–352. Rogers, C. (1961). On becoming a person: A therapist’s view of psychotherapy. Boston: Houghton MiZin. Rosenberg, M. (1965). Society and the adolescent self‐image. Princeton, NJ: Princeton University Press. Rusbult, C. E. (1983). A longitudinal test of the investment model: The development (and deterioration) of satisfaction and commitment in heterosexual involvements. Journal of Personality and Social Psychology, 45, 101–117. Rusbult, C. E., Martz, J. M., & Agnew, C. R. (1998). The investment model scale: Measuring commitment level, satisfaction level, quality of alternatives, and investment size. Personal Relationships, 5, 357–391. Rusbult, C. E., Verette, J., Whitney, G. A., Slovik, L. F., & Lipkus, I. (1991). Accommodation processes in close relationships: Theory and preliminary empirical evidence. Journal of Personality and Social Psychology, 60, 53–78. Ryan, R. M. (1993). Agency and organization: Intrinsic motivation, autonomy, and the self in psychological development. In J. Jacobs (Ed.), Nebraska Symposium on Motivation: Vol.

AUTHENTICITY

355

40. Developmental perspectives on motivation (pp. 1–56). Lincoln: University of Nebraska Press. Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internalization: Examining reasons for acting in two domains. Journal of Personality and Social Psychology, 57, 749–761. Ryan, R. M., & Deci, E. L. (2000). Self‐determination theory and the facilitation of intrinsic motivation, social development, and well‐being. American Psychologist, 55, 68–78. Ryan, R. M., & Deci, E. L. (2002). An overview of self‐determination theory: An organismic‐ dialectical perspective. In E. L. Deci & R. M. Ryan (Eds.), Handbook of self‐determination research (pp. 3–36). Rochester, NY: University of Rochester Press. Ryan, R. M., & Deci, E. L. (2003). On assimilating identities to the self: A self‐determination theory perspective on internalization and integrity within cultures. In M. Leary & J. Tangey (Eds.), Handbook of self and identity (pp. 253–274). New York: Guilford Press. Ryan, R. M., & Frederick, C. (1997). On energy, personality, and health: Subjective vitality as a dynamic reflection of well‐being. Journal of Personality, 65, 529–565. RyV, C. (1989). Happiness is everything or is it? Explorations on the meaning of psychological well‐being. Journal of Personality and Social Psychology, 57, 1069–1081. Sackeim, H. A., & Gur, R. C. (1979). Self‐deception, other‐deception, and self‐reported psychopathology. Journal of Consulting and Clinical Psychology, 47, 213–215. Sande, G. N., Goethals, G. R., & RadloV, C. E. (1988). Perceiving one’s own traits and others’: The multifaceted self. Journal of Personality & Social Psychology, 54, 13–20. Sartre, J. P. (2004). From being and nothingness. In G. Marino (Ed.), Basic writings of existentialism (pp. 369–409). New York: The Modern Library. (Excerpts from Jean‐Paul Sartre, Being and Nothingness, H. Barnes, Trans., 1956, New York: Philosophical Library). Schacht, R. (2003). Friedrich Nietzsche, beyond good and evil (1886): Prelude to a philosophy of the future. In J. E. Garcia, G. M. Reichberg, & B. N. Schumacher (Eds.), The classics of western philosophy: A reader’s guide (pp. 405–415). Malden, MA: Blackwell Publishing. Schlenker, B. R. (1980). Impression management: The self‐concept, social identity, and interpersonal relations. Monterey, CA: Brooks/Cole. Shedler, J., Mayman, M., & Manis, M. (1993). The illusion of mental health. American Psychologist, 48, 1117–1131. Sheldon, K. M., & Elliot, A. J. (1999). Goal striving, need satisfaction, and longitudinal well‐ being: The self‐concordance model. Journal of Personality and Social Psychology, 76(3), 482–497. Sheldon, K. M., & Kasser, T. (1995). Coherence and congruence: Two aspects of personality and integration. Journal of Personality and Social Psychology, 68, 531–543. Sheldon, K. M., & Kasser, T. (1998). Pursuing personal goals: Skills enable progress but not all progress is beneficial. Personality and Social Psychology Bulletin, 24, 1319–1331. Sheldon, K. M., Ryan, R. M., Rawsthorne, L. J., & Ilardi, B. (1997). Trait self and true self: Cross‐role variation in the Big‐Five personality traits and its relations with psychological authenticity and subjective well‐being. Journal of Personality and Social Psychology, 73, 1380–1393. Showers, C. J., & Zeigler‐Hill, V. (2003). Organization of self‐knowledge: Features, functions, and flexibility. In M. Leary & J. Tangey (Eds.), Handbook of self and identity (pp. 47–67). New York: Guilford Press. Snell, W. E., Miller, R. S., & Belk, S. S. (1988). Development of the emotional self‐disclosure scale. Sex Roles, 18, 59–73.

356

MICHAEL H. KERNIS AND BRIAN M. GOLDMAN

Snyder, M. (1987). Public appearances/private realities: The psychology of self‐monitoring. San Francisco: Freeman. Solomon, S., Greenberg, J., & Pyszcynski, T. (1991). A terror management theory of social behavior: The psychological functions of self‐esteem and cultural worldviews. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, pp. 93–159). New York: Academic Press. Sprecher, S., & Hendrich, S. S. (2004). Self‐disclosure in intimate relationships: Associations with individual and relationship characteristics over time. Journal of Social and Clinical Psychology, 23, 857–877. Stern, D. N. (1985). The interpersonal world of the infant. New York: Basic Books. Swann, W. B., Jr. (1983). Self‐verification: Bringing social reality into harmony with the self. In J. Suls & A. G. Greenwald (Eds.), Psychological perspectives on the self (Vol. 2, pp. 33–66). NJ: Erlbaum: Hillsdale. Swann, W. B., Jr., Stein‐Seroussi, A., & Giesler, B. (1992). Why people self‐verify. Journal of Personality and Social Psychology, 62, 392–401. Taylor, S. E., & Brown, J. D. (1988). Illusion and well‐being: A social psychological perspective on mental health. Psychological Bulletin, 103, 193–210. Tice, D. M., & Wallace, H. M. (2003). The reflected self: Creating yourself as (you think) others see you. In M. R. Leary & J. P. Tangney (Eds.), Handbook of self and identity (pp. 91–105). New York: Guilford Press. Thoits, P. A. (1986). Multiple identities: Examining gender and marital status diVerences in distress. American Sociological Review, 51, 259–272. Thoits, P. A. (1992). Identity structures and psychological well‐being: Gender and marital status comparisons. Social Psychology Quarterly, 55, 236–256. Thoits, P. A., & Virshop, L. K. (1997). Me’s and we’s: Forms and functions of social identities. In R. D. Ashmore & L. J. Jussim (Eds.), Self and identity: Fundamental issues (pp. 106–133). New York: Oxford University Press. Trilling, L. (1971). Sincerity and authenticity. Massachusetts: Harvard University Press. Ungerer, J. A., Waters, B., Barnett, B., & Dolby, R. (1997). Defense style and adjustment in interpersonal relationships. Journal of Research in Personality, 31, 375–385. Vaillant, G. (1992). Ego mechanisms of defense: A guide for clinicians and researchers. Washington, DC: American Psychiatric Press. Vaillant, G. (2000). Adaptive mental mechanisms: Their role in a positive psychology. American Psychologist, 55, 89–98. Vansteenkiste, M., Simons, J., Lens, W., Soenens, B., & Matos, L. (2005). Examining the impact of intrinsic versus extrinsic goal framing and autonomy‐supportive versus internally controlling communication style upon early adolescents’ achievement. Child Development, 76, 483–501. Vansteenkiste, M., Zhou, M., Lens, W., & Soenens, B. (2005). Experiences of autonomy and control among Chinese learners: Vitalizing or immobilizing? Journal of Educational Psychology, 97, 468–483. Waterman, A. S. (1993). Two conceptions of happiness: Contrasts of personal expressiveness (eudaimonia) and hedonic enjoyment. Journal of Personality and Social Psychology, 64, 678–691. Weinberger, J. (2003). Genuine, defensive, unconscious, and authentic: How do they all fit together? Psychological Inquiry, 14, 80–82. Westphal, M. (2003). Kierkegaard, Concluding unscientific postscript to philosophical fragments (1846): Making things diYcult for the system and for Christendom. In J. E. Garcia, G. M. Reichberg, & B. N. Schumacher (Eds.), The classics of western philosophy: A reader’s guide (pp. 389–394). Malden, MA: Blackwell Publishing.

AUTHENTICITY

357

Wiggins, J. S. (1979). A psychological taxonomy of trait‐descriptive terms: The interpersonal domain. Journal of Personality and Social Psychology, 37, 395–412. Wilson, F. (2003). David Hume, Treatise of human nature (1740): A genial skepticism, an ethical naturalism. In J. E. Garcia, G. M. Reichberg, & B. N. Schumacher (Eds.), The classics of western philosophy: A reader’s guide (pp. 291–308). Malden, MA: Blackwell Publishing. Winter, D. G., & Barenbaum, N. B. (1999). History of modern personality theory and research. In L. A. Pervin & O. P. John (Eds.), Handbook of personality theory and research (2nd ed., pp. 3–30). New York: Guilford Press.

CONTRIBUTORS

Numbers in parentheses indicate the pages on which authors’ contributions begins.

GEORGE Y. BIZER (1), Department of Psychology, Eastern Illinois University, Charleston, IL 61920 GEOFFREY L. COHEN (183), Department of Psychology, Yale University, New Haven, CT 6520 BRIAN M. GOLDMAN (283), Department of Psychology, Clayton State University, Morrow, GA 30260 PETER M. GOLLWITZER (69), Fachbereich Psychologie, Universita¨t Konstanz, Fach D 39, 78457 Konstanz, Germany; Department of Psychology, New York University, New York, NY 10003 MICHAEL H. KERNIS (283), Department of Psychology, University of Georgia, Athens, GA 30602 JON A. KROSNICK (1), Department of Communication, Stanford University, Stanford, CA 94305 SCOTT P. LEARY (243), Department of Psychology, University of Maryland, College Park, MD 20742 JENNIFER A. RICHESON (121), Department of Psychology, Northwestern University, Evanston, IL 60208 PASCHAL SHEERAN (69), Department of Psychology, University of SheYeld, SheYeld S10 2TP, United Kingdom J. NICOLE SHELTON (121), Department of Psychology, Princeton University, Princeton, NJ 08544 DAVID K. SHERMAN (183), Department of Psychology, University of California Santa Barbara, Santa Barbara, CA 93106 CHARLES STANGOR (243), Department of Psychology, University of Maryland, College Park, MD 20742 PENNY S. VISSER (1), Department of Psychology, University of Chicago, Chicago, IL 60637 ix

INDEX

A Abortion, legal, 46 –married woman’s right to, 46 –psychological consequences for women, 23–24 Abstainer, inclined, 72–73 Accessibility, attitude, 3, 19, 49 and certainty, 18, 38–40 and importance, 17–18, 32–37 of people’s, 39 Activities, self‐symbolizing, 80 Actors, inclined, 72–73 Adaptation, psychological, 184–186 Aesthetics, intellectual, 56 AYrmations diVerent‐domain, 219 moral, 219 potentially detrimental, 220 Ambivalence, 4–5 Antecedents, common, and consequences, 18 Anxiety, individual’s existential, 288 Aristotelian epistemology, 286 Attention, 77 Attitude(s) activation of, 53 causal antecedents of, importance, 53–54 causes –and consequences of, importance, 21 –of, importance and knowledge, 22 certainty, 3, 31, 38, 40, 45, 51 change, 27, 51, 196 –actual, 24 –assessment of, 23 cognitive and behavioral consequences of attaching importance to an, 28 359

composite indices of strength‐related, factors, 50 conceptualization of, strength, 16, 56 correct, 53 dimensionality of, importance, 52–54 embeddedness, 45 extremity, 55–56 impact of repeated expression of, 34–35 implicit racial, 150 importance, 3, 17, 20, 22, 24–25, 30–32, 34, 37, 43 –and knowledge, 18 increased resistance to, change, 53 index, –of overall, strength, 56 –of, strength, 52 intensity, 5 intergroup, 143, 253, 259 measurement, 43 meta‐attitudinal indices of, strength, 50 moderation, 49 negative, 58, 159 –toward Whites, 157 perceived, change, 24 polarization, 23 policy, impact of, on candidate preferences, 29 positive, 58 racial, 129, 157 –and interracial contact, 147 repeated, expression, 34, 37 resistance to, change, 43, 55 stability, 43 strength, 2, 11–12, 59 –composites, 41 –literature, 14 strength‐related attributes of, 3–5, 7, 12–15, 44

360

INDEX

Attitude(s) (cont. ) –of people’s, 13, 59 target, 44 toward legalized abortion, 21 Attitude accessibility, 3, 19, 49 of people’s, 39 and importance, 17–18 –evidence, 33–34 –hypotheses, 32–33 –impact of personal relevance on, 35–37 –impact of repeated attitude expression on, 34–35 and certainty, 18, 38–40 Attributions, divergent, 128–129 implications, 130–132 and racial attitudes, 129–130 Authentic functioning, 283, 319, 307–326 goal‐based indexes of, 322 historical overview on, 284–291 multicomponent conceptualization of, 294–303 psychological aspects of, 292–294 Authenticity and absence of verbal defensiveness, 307 Aristotle’s contribution to conceptualizing, 285 awareness component of, 335 behavioral, 300, 315 concepts and insights relevant to conceptualizing, 285 conceptualization, 283 –of dispositional, 344 –multicomponent of, 294, 344 development of, in children and young adults, 336 dispositional, 303–308, 310, 317–321, 324–325, 331–332, 337–339 and disclosure tendencies, 333 future directions –awareness, 342–343 –behavior, 343 –relational orientation, 343–344 –unbiased processings, 343 and goal need‐fulfillment, 324 and healthy psychological functioning, 308–326 historical overview, 284–292 inventory, 303–307 latent construct of, 304 and mindfulness, 310–313

partner validation, 327–328 potential downside, 339 –awareness, 340 –behavior, 340–341 –relational orientation, 341–342 –unbiased processing, 340 predictive utility of dispositional, 315–316 psychological –perspectives on, 293 –view of, 292 relational, 300–301 relation between self‐concepts, role functioning, and, 316–319 relationship satisfaction, 331 –and relationship between dispositional, 334 retrospective accounts of parental authority styles, 335–338 role, 321 and self‐esteem, 338–339 social role functioning, 319–322 total, 315 and use of various coping strategies, 313–315 Authoritarianism, 273 Automaticity, 84 Auto‐motive theory, 79 Awareness, 294, 296, 309, 340, 342 lack of, 100 B Behavior(s), 298, 340 aspects of nonverbal, during interracial interactions, 167 attitude‐expressive, 11, 25, 29, 31 authentic, 299 change, 71 deadline for performing, 75 defensive verbal, assessment (DVBA), 309 external, of self and other, 122 false‐self, 341 future, 71–72 goal‐directed, 74 healthy, 58 intimacy building, 152 nonverbal, 144 past, 71–72 positive –health, 201 –for higher‐prejudiced Whites, 152

361

INDEX –toward ethnic minorities, 152 risky health, 201 unhealthy, 58 verbal, 144 Belief(s) acquisition of intergroup, in childhood, 251 intergroup, 261, 264 –and attitudes, 143 Bias defensive, 184 hostile media, 23–24, 49, 55 implication of a self–other, 130 racial, 151, 169 relative absence of the self‐serving, 343 self–other, 131, 133 Biased processing, 191 of health information, 195–197 of self‐relevant information, 303 Black(s), 125, 128, 131, 134 confederates’ friendliness judgments, 154 racial attitudes, 158 stereotype‐inconsistency, 154 with negative attitudes toward Whites, 158 C Capability, reduced self‐regulatory, 81 Certainty, 4, 43 and accessibility, 3, 18 comparisons of meta‐attitudinal measures of importance and, 50 and importance, 3, 17, 28–32 Common ground creation, 161–164 Communication, social and exchange, 144 and perception, 144 Concerns, interpersonal and attitudes of Whites, 173 with prejudice and interracial contact, 135 with prejudice, 161 Confirmatory factor analysis (CFA), 304–307 Consensus, false, eVect, 55 Conservatism, 254, 256, 258, 260, 272 motivation, 267, 273 orientation, 259 Consistency, attitude–behavior, 55, 124 Contact(s) dynamics of interracial, 122, 135 interracial, 129

Control theory, 71 Constructs, strength‐related attitude attributes, 5–6 common factor model, 6–7 –evidence, 7–11 Cooperation, intergroup, 143 Coping and resilience, 230–231 strategies, 313–315 Counter arguments, immediate initiation of, 85 Crime and lawlessness, 47 Cross‐race friendship, 134 D Defensiveness, 307–310, 331 Disengagement, behavioral, 314 Dissonance, cognitive, 201, 217 Distortions, social perceptions, 203–205 Distractability, 107. See also Self‐defensiveness Divergent attributions, 128 implications of, 130–132 and racial attitudes, 129–130 reconverging of, 132–135 Divergent experiences, 159–161 E Egalitarianism, 254–260, 268, 272–273 Ego‐depletion, 80–81, 111 Embeddedness, 11, 45 interattitudinal, 44, 46 Ethnic minorities, 123, 125–127, 141–142, 145–146, 153–155, 165 prejudice concerns of –implications for partner, 145–147 –implications for self, 140–145 racial attitudes of –implications for partner, 158–159 –implications for self, 156–158 Ethnicity, 142–143 Existential angst, 291 Exploratory factor analysis, 12, 42 Expression, repeated attitude, 34–35, 38–40, 49 Extremity, 4, 6 attitude, 55–56 F Fear of rejection, 128–131, 134 Friendship, development, 129

362

INDEX

Functional flexibility, 295–296 Functional magnetic resonance imaging technology (fMRI), 160 G Goal(s) achievement, 70–72 attainment, 69–73, 82, 98, 101, 103–108. See also Goal(s), striving domains, 94 intentions, 70–72 –for implementation intention, 86 problems in completion of, 75–81 self‐concordant, 322–326 striving –initiating and shielding, from unwanted influences, 80 –self‐regulation of, 73–75 –strategy for eVective self‐regulation of, 82–83 H Health information, threatening, 193–194 Hostile media bias, 24 I Identity centrality and salience, 217–218 defense of collective, 208–210 goal, 80 integration, 316 social, 127, 250 If‐then plans, 104–109. See also Implementation intentions Implementation intentions, 69–70, 95, 99–100, 105 action control by, 103 component processes of, 83–109 eVective operation of, 85 eVects, 87, 93–94, 96 –on goal achievement, 88–91, 104 –on goal attainment, 98, 103–104 –on goal completion, 106 features of, 107–108 forming, 82–83, 102 in everyday life, 101–102 and goal achievement, 69, 107

group, 92 participants, 104 psychological processes underlying eVects of, 99–101 and self‐regulatory tasks in goal striving, 98–99 moderators of, eVects, 92–94, 106–107 Implicit association test (IAT), 136, 246 Importance and accessibility, 7, 17–18, 32–38 and certainty, 3, 17, 28–32 and knowledge, 17–20 –consequences of, 23 –correlations between, 21 –origins of, 21–27 of self‐integrity maintenance, 188 Individual(s) nonstigmatized, 165 stigmatized, 165 Ingroup, 246, 274 evaluations, 256, 258 favoritism, 252 protecting and enhancing, 248–249 Intentions, implementation. See Implementation intentions Interconnectedness, 121, 133 Interpersonal relationships, 229–230 Intergroup beliefs, 243–246 acquisition of, childhood, 251–253 group diVerences in 259–261 in social motivation and, 247–251 personality and 253–257 sharing and transmission of, 265–271 social motivations for, 257–259 as social norms, 261–265 Intergroup meta‐perceptions, 125–127 Interracial contact, 122–123, 126. See also Interracial interactions dynamics of, 135 Interracial encounters, 133, 138, 173 Interracial interactions conceptual approaches, 121–122 –avoiding, 127–135 –classic revisted, 124–125 –paradigm shift, 122–124 future directions –boundary conditions, 172–173 –reciprocal dynamics, 173–174 –temporal factors, 174 Interventions, 211, 224

363

INDEX Intrasubjectivity, 122, 335 Involvement, value‐relevant, 53 J Judgments, biased, 206 K KIMS scores, 311–312 Knowledge, 3 accumulation, 20 attitude relevant, 17–19, 21, 25–26, 39, 49 importance and, 17–28 operative measures of, 54–55 M MAAS. See Mindful attention awareness scale Mental disengagement, 314 Meta‐attitudinal aggregate, 47 Meta‐perceptions individuals, 132 intergroup, 125–127 Metaphysics, 284 Mindful attention awareness scale (MAAS), 311–312 Minorities, racial, 130. See also Ethnic minorities Moderate variables, 213–214, 219–221 culture, 214–215 identity centrality, 217–218 self‐esteem, 215–217 Mortality salience, 228 Motivation to protect self‐integrity, 188 psychological theories of, 291 and self‐regulation, 69 social, 248, 259 of social harmony, 254 N Negative eVect, 25–26 News media, 21 O Ontology, 284 Orientation, relational, 300–301, 313, 315, 341–344

Outcome relevance, 53 Outgroup evaluations, 256 P Parenting, eVects of, on authentic functioning, 337 Parent(s) authoritarian, 336–337 authoritative, 336–337 to child transmission, 267–270 permissive, 336–337 Partner validation, 327–328 Peer‐to‐peer transmission, 266–267 Personal relevance, impact of, on accessibility and importance, 35–37 Personality and intergroup beliefs, 253–257 traits, 2 Philosophy from the Enlightenment, 287 Planning, 313 Prejudice, 245–247 concerns, 137 –of White participants: implications for self, 136–139 –of White participants: implications for partner, 139–140 –of Ethnic minorities: implications for self, 140–145 –of Ethnic minorities: implications for partner, 145–147 Problems, self‐regulatory, 87, 95–97 Procrastination, 226 R Racial attitudes, Whites’ implication for partner, 153–156 implication for self, 148–152 Racism, perceptions of, 212–213 Reactions, nondefensive, 332 Rejection, race‐based, sensitivity to, 131–132 Relational orientation, 300–301, 313, 315, 341–342, 343–344 Relationship between gender and group evaluations, 259–261 implications for interpersonal, and coping, 229 motives, 330–335

364

INDEX

Relevance, outcome, 53 Reliability, alpha, 42 Response variability, 342 Role‐balance, 319–320 Role‐conflict, 320 Role‐ease, 319 Role‐strain, 320, 340 S Schwartz Value Survey, 267 Selective information gathering, 25 Self‐aYrmation, 183–185 theory, 185–187 –basic tenets of, 187–189 and threats to individual self, 189 –motivation inferences and biased assimilation, 190–193 –threatening health information, 193–197 –stress, 197–201 –cognitive dissonance, 201–202 –motivated distortions in social perceptions, 202–205 and responses to collective threats, 205 –group‐serving judgements, 206–208 –defense of collective identity, 208–210 –stereotype threat and performance, 210–212 –perceptions of racism, 212–213 Self‐concealment, concerns about, 287 Self‐concept, 316–319 diVerentiation, 316–318, 320 organization, 316, 318 Self‐concordance, 285 Self‐control, 80–81 Self‐defensiveness, 107 Self‐determination, 330 theory, 293 Self‐disclosure, 142, 326–327 emotional, 327 index, 327 Self‐discrepancies, perceived, 340 Self‐enhancement, 230–231, 253, 300 Self‐evaluative process, interchangeability of, 222 Self‐esteem, 217, 223, 227 and cognitive dissonance, 216 maintenance processes, 222

people –with defensive high, 217 –with high, 216 positive, 250 scale, 215, sociometer theory of, 228 Self‐integration, 296. See also Self integrity and acceptance, 296 Self integrity, 203, 226, 228. See also Self‐ integration global perceptions of, 189 individual’s sense of, 184 maintenance of, 227 restoration of, and adaptive behavior change, 185 Self‐knowledge, 302, 340 Self‐negation, 291 Self‐organization, 285, 316–317 Self‐reflection, 340 Self‐representations, individuals, 316 Self‐symbolizing activities, 80 Self‐threat, 183, 203, 205, 228–229 Self‐verification theory, 125, 300 Self‐worth, 184–185, 197, 202–203 Social cognitive theory, 71 Social harmony motivation, 250–253, 274 Social loafing, 78 Social motivations aYliation, 249–250 protecting and enhancing motivation, 248–249 and intergroup beliefs, 247–248 priming, 257–259 Social norms, perceptions of prevailing, 261–265, 272 Social perception, 129 motivated distortions in, 202–205 Social support emotional, 314 instrumental, 313, 315 Stereotyping, 245–247 impact of positive mood on, target persons, 79–80 Stigma and interactions, 165 consciousness, 132, 146 –scale‐race, 142 management of, in social interactions, 124 Strength‐related attitude attributes accessibility, 3, 32–38

365

INDEX certainty, 3–4, 38–41 elaboration, 5 extremity, 4 importance, 3, 32–38 intensity, 5 knowledge, 3 structural consistency, 4 Stroop task, 137, 149, 158, 163 Structural consistency, 4–5 Substance abuse, 315 Symbolic interaction theory, 125 T Terror management theory (TMT), 274 Theory of planned behavior, 71 of self‐aYrmation, 185 symbolic self‐completion, 80 terror management, 227 Threat, stereotype, and performance, 210–212 Total authenticity scale scores, 296 Trier social stress task, 198 U Unbiased processing, 296–310, 312, 340, 343

V Value relevance, 21–22, 57 studies, 53 Verbal defensiveness, 307–310 Virtues, artificial, 288 natural, 288

W Wason selection task, 26 Well‐being, psychological, 291, 310–311, 323–324 White House Conference on Global Climate Change, 22 Whites, 123–125, 127–128, 131, 134, 145 high‐bias, 161 high‐prejudice, 130, 155 individuals, 151 lower and higher prejudiced, 153 low‐prejudice, 130, 155 nonverbal friendliness, 154 prejudice, 135 –concerns, 139 racial attitudes, 148–156 targets racial bias, 167 unfair treatment from, 127

CONTENTS OF OTHER VOLUMES

Volume 1 Cultural Influences upon Cognitive Processes Harry C. Triendis The Interaction of Cognitive and Physiological Determinants of Emotional State Stanley Schachter Experimental Studies of Coalirion Formation William A. Gamson Communication Networks Marvin E. Shaw A Contingency Model of Leadership Effectiveness Fred E. Fiedler Inducing Resistance to Persuasion: Some Contemporary Approaches William J. McGuire Social Motivation, Dependency, and Susceptibility to Social Influence Richard H. Walters and Ross D. Purke Sociability and Social Organization in Monkeys and Apes William A. Mason Author Index—Subject Index Volume 2 Vicarious Processes: A Case of No-Trial Learning Albert Bandura Selective Exposure Jonathan L. Freedman and David O. Sears Group Problem Solving L. Richard Hoffman Situational Factors in Conformity Vernon L. Allen Social Power John Schopler

From Acts to Dispositions: The Attribution Process in Person Perception Edward E. Jones and Keith E. Davis Inequality in Social Exchange J. Stacy Adams The Concept of Aggressive Drive: Some Additional Considerations Leonard Berkowitz Author Index—Subject Index Volume 3 Mathematical Models in Social Psychology Robert P. Abelson The Experimental Analysis of Social Performance Michael Argyle and Adam Kendon A Structural Balance Approach to the Analysis of Communication Effects N. T. Feather Effects of Fear Arousal on Attitude Change: Recent Developments in Theory and Experimental Research Irving L. Janis Communication Processes and the Properties of Language Serge Moscovici The Congruity Principle Revisited: Studies in the Reduction, Induction, and Generalization of Persuasion Percy H. Tannenbaum Author Index—Subject Index Volume 4 The Theory of Cognitive Dissonance: A Current Perspective Elliot Aronson

367

368

CONTENTS OF OTHER VOLUMES

Attitudes and Attraction Donn Byrne Sociolinguistics Susan M. Ervin-Tripp Recognition of Emotion Nico H. Frijda Studies of Status Congruence Edward E. Sampson Exploratory Investigations of Empathy Ezra Stotland The Personal Reference Scale: An Approach to Social Judgment Harry S. Upshaw Author Index—Subject Index Volume 5 Media Violence and Aggressive Behavior: A Review of Experimental Research Richard E. Goranson Studies in Leader Legitimacy, Influence, and Innovation Edwin P. Hollander and James W. Julian Experimental Studies of Negro-White Relationships Irwin Katz Findings and Theory in the Study of Fear Communications Howard Leventhal Perceived Freedom Ivan D. Steiner Experimental Studies of Families Nancy E. Waxler and Elliot G. Mishler Why Do Groups Make Riskier Decisions than Individuals? Kenneth L. Dion, Robert S. Baron, and Norman Miller Author Index—Subject Index Volume 6 Self-Perception Theory Daryl J. Bem Social Norms, Feelings, and Other Factors Affecting Helping and Altruism Leonard Berkowitz The Power of Liking: Consequence of Interpersonal Attitudes Derived from a Liberalized View of Secondary Reinforcement Albert J. Lott and Bernice E. Lott

Social Influence, Conformity Bias, and the Study of Active Minorities Serge Moscovici and Claude Faucheux A Critical Analysis of Research Utilizing the Prisoner’s Dilemma Paradigm for the Study of Bargaining Charlan Nemeth Structural Representations of Implicit Personality Theory Seymour Rosenberg and Andrea Sedlak Author Index—Subject Index Volume 7 Cognitive Algebra: Integration Theory Applied to Social Attribution Norman A. Anderson On Conflicts and Bargaining Erika Apfelbaum Physical Attractiveness Ellen Bersheid and Elaine Walster Compliance, Justification, and Cognitive Change Harold B. Gerard, Edward S. Connolley, and Roland A. Wilhelmy Processes in Delay of Gratification Walter Mischel Helping a Distressed Person: Social, Personality, and Stimulus Determinants Ervin Staub Author Index—Subject Index Volume 8 Social Support for Nonconformity Vernon L. Allen Group Tasks, Group Interaction Process, and Group Performance Effectiveness: A Review and Proposed Integration J. Richard Hackman and Charles G. Morris The Human Subject in the Psychology Experiment: Fact and Artifact Arie W. Kruglanski Emotional Arousal in the Facilitation of Aggression Through Communication Percy H. Tannenbaum and Dolf Zillman The Reluctance to Transmit Bad News Abraham Tesser and Sidney Rosen Objective Self-Awareness Robert A. Wicklund

CONTENTS OF OTHER VOLUMES Responses to Uncontrollable Outcomes: An Integration of Reactance Theory and the Learned Helplessness Model Camille B. Wortman and Jack W. Brehm Subject Index Volume 9 New Directions in Equity Research Elaine Walster, Ellen Berscheid, and G. William Walster Equity Theory Revisited: Comments and Annotated Bibliography J. Stacy Adams and Sara Freedman The Distribution of Rewards and Resources in Groups and Organizations Gerald S. Leventhal Deserving and the Emergence of Forms of Justice Melvin J. Lerner, Dale T. Miller, and John G. Holmes Equity and the Law: The Effect of a Harmdoer’s ‘‘Suffering in the Act’’ on Liking and Assigned Punishment William Austin, Elaine Walster, and Mary Kristine Utne Incremental Exchange Theory: A Formal Model for Progression in Dyadic Social Interaction L. Lowell Huesmann and George Levinger Commentary George C. Homans Subject Index Volume 10 The Catharsis of Aggression: An Evaluation of a Hypothesis Russell G. Geen and Michael B. Quanty Mere Exposure Albert A. Harrison Moral Internalization: Current Theory and Research Martin L. Hoffman Some Effects of Violent and Nonviolent Movies on the Behavior of Juvenile Delinquents Ross D. Parke, Leonard Berkowitz, Jacques P. Leyens, Stephen G. West, and Richard Sebastian

369

The Intuitive Psychologist and His Shortcomings: Distortions in the Attribution Process Less Ross Normative Influences on Altruism Shalom H. Schwartz A Discussion of the Domain and Methods of Social Psychology: Two Papers by Ron Harre and Barry R. Schlenker Leonard Berkowitz The Ethogenic Approach: Theory and Practice R. Harre On the Ethogenic Approach: Etiquette and Revolution Barry R. Schlenker Automatisms and Autonomies: In Reply to Professor Schlenker R. Harre Subject Index Volume 11 The Persistence of Experimentally Induced Attitude Change Thomas D. Cook and Brian F. Flay The Contingency Model and the Dynamics of the Leadership Process Fred E. Fiedler An Attributional Theory of Choice Andy Kukla Group-Induced Polarization of Attitudes and Behavior Helmut Lamm and David G. Myers Crowding: Determinants and Effects Janet E. Stockdale Salience: Attention, and Attribution: Top of the Head Phenomena Shelley E. Taylor and Susan T. Fiske Self-Generated Attitude Change Abraham Tesser Subject Index Volume 12 Part I. Studies in Social Cognition Prototypes in Person Perception Nancy Cantor and Walter Mischel A Cognitive-Attributional Analysis of Stereotyping David L. Hamilton

370

CONTENTS OF OTHER VOLUMES

Self-Monitoring Processes Mark Snyder Part II. Social Influences and Social Interaction Architectural Mediation of Residential Density and Control: Crowding and the Regulation of Social Contact Andrew Baum and Stuart Valins A Cultural Ecology of Social Behavior J. W. Berry Experiments on Deviance with Special Reference to Dishonesty David P. Farrington From the Early Window to the Late Night Show: International Trends in the Study of Television’s Impact on Children and Adults John P. Murray and Susan Kippax Effects of Prosocial Television and Film Material on the Behavior of Viewers J. Phillipe Rushton Subject Index

Cognitive, Social, and Personality Processes in the Physiological Detection of Deception William M. Waid and Martin T. Orne Dialectic Conceptions in Social Psychology: An Application to Social Penetration and Privacy Regulation Irwin Altman, Anne Vinsel, and Barbara B. Brown Direct Experience and Attitude–Behavior Consistency Russell H. Fazio and Mark P. Zanna Predictability and Human Stress: Toward a Clarification of Evidence and Theory Suzanne M. Miller Perceptual and Judgmental Processes in Social Contexts Arnold Upmeyer Jury Trials: Psychology and Law Charlan Jeanne Nemeth Index Volume 15

Volume 13 People’s Analyses of the Causes of AbilityLinked Performances John M. Darley and George R. Goethals The Empirical Exploration of Intrinsic Motivational Processes Edward I. Deci and Richard M. Ryan Attribution of Responsibility: From Man the Scientist to Man as Lawyer Frank D. Fincham and Joseph M. Jaspars Toward a Comprehensive Theory of Emotion Howard Leventhal Toward a Theory of Conversion Behavior Serge Moscovici The Role of Information Retrieval and Conditional Inference Processes in Belief Formation and Change Robert S. Wyer, Jr. and Jon Hartwick Index

Balance, Agreement, and Positivity in the Cognition of Small Social Structures Walter H. Crockett Episode Cognition: Internal Representations of Interaction Routines Joseph P. Forgas The Effects of Aggressive-Pornographic Mass Media Stimuli Neil M. Malamuth and Ed Donnerstein Socialization in Small Groups: Temporal Changes in Individual–Group Relations Richard L. Moreland and John M. Levine Translating Actions into Attitudes: An Identity-Analytic Approach to the Explanation of Social Conduct Barry R. Schlenker Aversive Conditions as Stimuli to Aggression Leonard Berkowitz Index

Volume 14

Volume 16

Verbal and Nonverbal Communication of Deception Miron Zuckerman, Bella M. DePaulo, and Robert Rosenthal

A Contextualist Theory of Knowledge: Its Implications for Innovation and Reform in Psychological Research William J. McGuire

CONTENTS OF OTHER VOLUMES Social Cognition: Some Historical and Theoretical Perspectives Janet Landman and Melvin Manis Paradigmatic Behaviorism: Unified Theory for Social-Personality Psychology Arthur W. Staats Social PsychologyfromtheStandpointofa StructuralSymbolicInteractionism: Toward an InterdisciplinarySocialPsychology Sheldon Stryker Toward an Interdisciplinary Social Psychology Carl W. Backman Index Volume 17 Mental Representations of Self John F. Kihlstrom and Nancy Cantor Theory of the Self: Impasse and Evolution Kenneth J. Gergen A Perceptual-Motor Theory of Emotion Howard Leventhal Equity and Social Change in Human Relationships Charles G. McClintock, Roderick M. Kramer, and Linda J. Keil A New Look at Dissonance Theory Joel Cooper and Russell H. Fazio Cognitive Theories of Persuasion Alice H. Eagly and Shelly Chaiken Helping Behavior and Altruism: An Empirical and Conceptual Overview John F. Dovidio Index Volume 18 A Typological Approach to Marital Interaction: Recent Theory and Research Mary Anne Fitzpatrick Groups in Exotic Environments Albert A. Harrison and Mary M. Connors Balance Theory, the Jordan Paradigm, and the Wiest Tetrahedon Chester A. Insko The Social Relations Model David A. Kenny and Lawrence La Voie Coalition Bargaining S. S. Komorita

371

When Belief Creates Reality Mark Snyder Index Volume 19 Distraction–Conflict Theory: Progress and Problems Robert S. Baron Recent Research on Selective Exposure to Information Dieter Frey The Role of Threat to Self-Esteem and Perceived Control in Recipient Reaction to Help: Theory Development and Empirical Validation Arie Nadler and Jeffrey D. Fisher The Elaboration Likelihood Model of Persuasion Richard E. Petty and John T. Cacioppo Natural Experiments on the Effects of Mass Media Violence on Fatal Aggression: Strengths and Weaknesses of a New Approach David P. Phillips Paradigms and Groups Ivan D. Steiner Social Categorization: Implications for Creation and Reduction of Intergroup Bias David A. Wilder Index Volume 20 Attitudes, Traits, and Actions: Dispositional Prediction of Behavior in Personality and Social Psychology Icek Ajzen Prosocial Motivation: Is It Ever Truly Altruistic? C. Daniel Batson Dimensions of Group Process: Amount and Structure of Vocal Interaction James M. Dabbs, Jr. and R. Barry Ruback The Dynamics of Opinion Formation Harold B. Gerard and Ruben Orive Positive Affect, Cognitive Processes, and Social Behavior Alice M. Isen

372

CONTENTS OF OTHER VOLUMES

Between Hope and Fear: The Psychology of Risk Lola L. Lopes Toward an Integration of Cognitive and Motivational Perspectives on Social Inference: A Biased Hypothesis-Testing Model Tom Pyszczynski and Jeff Greenberg Index Volume 21 Introduction Leonard Berkowitz Part I. The Self as Known Narrative and the Self as Relationship Kenneth J. Gergen and Mary M. Gergen Self and Others: Studies in Social Personality and Autobiography Seymour Rosenberg Content and Process in the Experience of Self William J. McGuire and Claire V. McGuire Information Processing and the Study of the Self John F. Kihlstrom, Nancy Cantor, Jeanne Sumi Albright, Beverly R. Chew, Stanley B. Klein, and Paula M. Niedenthal Part II. Self-Motives Toward a Self-Evaluation Maintenance Model of Social Behavior Abraham Tesser The Self: A Dialectical Approach Carl W. Backman The Psychology of Self-Affirmation: Sustaining the Integrity of the Self Claude M. Steele A Model of Behavioral Self-Regulation: Translating Intention into Action Michael F. Scheier and Charles S. Carver Index Volume 22 On the Construction of the Anger Experience: Aversive Events and Negative Priming in the Formation of Feelings Leonard Berkowitz and Karen Heimer Social Psychophysiology: A New Look John T. Cacioppo, Richard E. Petty, and Louis G. Tassinary

Self-Discrepancy Theory: What Patterns of Self-Beliefs Cause People to Suffer? E. Tory Higgins Minding Matters: The Consequences of Mindlessness-Mindfulness Ellen J. Langer The Tradeoffs of Social Control and Innovation in Groups and Organizations Charlan Jeanne Nemeth and Barry M. Staw Confession, Inhibition, and Disease James W. Pennebaker A Sociocognitive Model of Attitude Structure and Function Anthony R. Pratkanis and Anthony G. Greenwald Introspection, Attitude Change, and Attitude– Behavior Consistency: The Disruptive Effects of Explaining Why We Feel the Way We Do Timothy D. Wilson, Dana S. Dunn, Dolores Kraft, and Douglas J. Lisle Index Volume 23 A Continuum of Impression Formation, from Category-Based to Individuating Processes: Influences of Information and Motivation on Attention and Interpretation Susan T. Fiske and Steven L. Neuberg Multiple Processes by Which Attitudes Guide Behavior: The MODE Model as an Integrative Framework Russell H. Fazio PEAT: An Integrative Model of Attribution Processes John W. Medcof Reading People’s Minds: A Transformation Rule Model for Predicting Others’ Thoughts and Feelings Rachel Karniol Self-Attention and Behavior: A Review and Theoretical Update Frederick X. Gibbons Counterfactual Thinking and Social Perception: Thinking about What Might Have Been Dale T. Miller, William Turnbull, and Cathy McFarland Index

CONTENTS OF OTHER VOLUMES Volume 24 The Role of Self-Interest in Social and Political Attitudes David O. Sears and Carolyn L. Funk A Terror Management Theory of Social Behavior: The Psychological Functions of Self-Esteem and Cultural Worldviews Sheldon Solomon, Jeff Greenberg, and Tom Pyszczynski Mood and Persuasion: Affective States Influence the Processing of Persuasive Communications Norbert Schwarz, Herbert Bless, and Gerd Bohner A Focus Theory of Normative Conduct: A Theoretical Refinement and Reevaluation of the Role of Norms in Human Behavior Robert B. Cialdini, Carl A. Kallgren, and Raymond R. Reno The Effects of Interaction Goals on Person Perception James L. Hilton and John M. Darley Studying Social Interaction with the Rochester Interaction Record Harry T. Reis and Ladd Wheeler Subjective Construal, Social Inference, and Human Misunderstanding Dale W. Griffin and Lee Ross Index Volume 25 Universals in the Content and Structure of Values: Theoretical Advances and Empirical Tests in 20 Countries Shalom H. Schwartz Motivational Foundations of Behavioral Confirmation Mark Snyder A Relational Model of Authority in Groups Tom R. Tyler and E. Allan Lind You Can’t Always Think What You Want: Problems in the Suppression of Unwanted Thoughts Daniel M. Wegner Affect in Social Judgments and Decisions: A Multiprocess Model Joseph Paul Forgas

373

The Social Psychology of Stanley Milgram Thomas Blass The Impact of Accountability on Judgment and Choice: Toward a Social Contingency Model Philip E. Tetlock Index Volume 26 Attitudes Toward High Achievers and Reactions to Their Fall: Theory and Research Concerning Tall Poppies N. T. Feather Evolutionary Social Psychology: From Sexual Selection to Social Cognition Douglas T. Kenrick Judgment in a Social Context: Biases, Short-comings, and the Logic of Conversation Norbert Schwarz A Phase Model of Transitions: Cognitive and Motivational Consequences Diane N. Ruble Multiple-Audience Problems, Tactical Communication, and Social Interaction: A Relational-Regulation Perspective John H. Fleming From Social Inequality to Personal Entitlement: The Role of Social Comparisons, Legitimacy Appraisals, and Group Membership Brenda Major Mental Representations of Social Groups: Advances in Understanding Stereotypes and Stereotyping Charles Stangor and James E. Lange Index Volume 27 Inferences of Responsibility and Social Motivation Bernard Weiner Information Processing in Social Contexts: Implications for Social Memory and Judgment Robert S. Wyer, Jr. and Deborah H. Gruenfeld

374

CONTENTS OF OTHER VOLUMES

The Interactive Roles of Stability and Level of Self-Esteem: Research and Theory Michael H. Kernis and Stephanie B. Waschull Gender Differences in Perceiving Internal State: Toward a His-and-Hers Model of Perceptual Cue Use Tomi-Ann Roberts and James W. Pennebaker On the Role of Encoding Processes in Stereotype Maintenance William von Hippel, Denise Sekaquaptewa, and Patrick Vargas Psychological Barriers to Dispute Resolution Lee Ross and Andrew Ward Index Volume 28 The Biopsychosocial Model of Arousal Regulation Jim Blascovich and Joe Tomaka Outcome Biases in Social Perception: Implications for Dispositional Inference, Attitude Change, Stereotyping, and Social Behavior Scott T. Allison, Diane M. Mackie, and David M. Messick Principles of Judging Valence: What Makes Events Positive or Negative? C. Miguel Brendl and E. Tory Higgins Pluralistic Ignorance and the Perpetuation of Social Norms by Unwitting Actors Deborah A. Prentice and Dale T. Miller People as Flexible Interpreters: Evidence and Issues from Spontaneous Trait Inference James S. Uleman, Leonard S. Newman, and Gordon B. Moskowitz Social Perception, Social Stereotypes, and Teacher Expectations: Accuracy and the Quest for the Powerful Self-Fulfilling Prophecy Lee Jussim Jacquelynne Eccles, and Stephanie Madon Nonverbal Behavior and Nonverbal Communication: What do Conversational Hand Gestures Tell Us? Robert M. Krauss, Yihsiu Chen, and Purnima Chawla Index

Volume 29 Counterfactual Thinking: The Intersection of Affect and Function Neal J. Roese and James M. Olson Terror Management Theory of Self-Esteem and Cultural Worldviews: Empirical Assessments and Conceptual Refinements Jeff Greenberg, Sheldon Solomon, and Tom Pyszczynski The Flexible Correction Model: The Role of Naı¨ve Theories of Bias in Bias Correction Duane T. Wegener and Richard E. Petty Self-Evaluation: To Thine Own Self Be Good, to Thine Own Self Be Sure, to Thine Own Self Be True, and to Thine Own Self Be Better Constantine Sedikides and Michael J. Strube Toward a Hierarchical Model of Intrinsic and Extrinsic Motivation Robert J. Vallerand Volume 30 Promotion and Prevention: Regulatory Focus as a Motivational Principle E. Tory Higgins The Other ‘‘Authoritarian Personality’’ Bob Altemeyer Person Preception Comes of Age: The Salience and Significance of Age in Social Adjustments Joann M. Montepare and Leslie A. Zebrowitz On the Perception of Social Consensus Joachim Krueger Prejudice and Stereotyping in Everyday Communication Janet B. Ruscher Situated Optimism: Specific Outcome Expectancies and Self-Regulation David A. Armor and Shelley E. Taylor Volume 31 Affect and Information Processing Robert S. Wyer, Jr., Gerald L. Clore, and Linda M. Isbell Linguistic Intergroup Bias: Stereotype Perpetuation through Language Anne Maass

CONTENTS OF OTHER VOLUMES Relationships from the Past in the Present: Significant-Other Representations and Transference in Interpersonal Life Serena Chen and Susan M. Anderson The Puzzle of Continuing Group Inequality: Piecing Together Psychological, Social, and Cultural Forces in Social Dominance Theory Felicia Prano Attitude Representation Theory Charles G. Lord and Mark R. Lepper Discontinuity Theory: Cognitive and Social Searches for Rationality and Normality— May Lead to Madness Philip G. Zimbardo

375

Ambivalent Sexism Peter Glick and Susan T. Fiske Videotaped Confessions: Is Guilt in the Eye of the Camera? G. Daniel Lassiter, Andrew L. Geers, Patrick J. Munhall, Ian M. Handley, and Melissa J. Beers Effort Determination of Cardiovascular Response: An Integrative Analysis with Applications in Social Psychology Rex A. Wright and Leslie D. Kirby Index Volume 34

Volume 32 The Nature and Function of Self-Esteem: Sociometer Theory Mark R. Leary and Roy F. Baumeister Temperature and Aggression Craig A. Anderson, Kathryn B. Anderson, Nancy Dorr, Kristina M. DeNeve, and Mindy Flanagan The Importance of Being Selective: Weighing the Role of Attribute Importance in Attitudinal Judgment Joop van der Pligt, Nanne K. de Vries, Antony S. R. Manstead, and Frank van Harreveld Toward a History of Social Behavior: Judgmental Accuracy from Thin Slices of the Behavioral Stream Nalini Amabady, Frank J. Bernieri, and Jennifer A. Richeson Attractiveness, Attraction, and Sexual Selection: Evolutionary Perspectives on the Form and Function of Physical Attractiveness Dianne S. Berry Index Volume 33 The Perception–Behavior Expressway: Automatic Effects of Social Perception on Social Behavior Ap Dijksterhuis and John A. Bargh A Dual-Process Cognitive-Motivational Theory of Ideology and Prejudice John Duckitt

Uncertainty Management by Means of Fairness Judgments Kees van den Bos and E. Allan Lind Cognition in Persuasion: An Analysis of Information Processing in Response to Persuasive Communications Dolores Albarracin Narrative-Based Representations of Social Knowledge: Their Construction and Use in Comprehension, Memory, and Judgment Robert S. Wyer, Jr., Rashmi Adaval and Stanley J. Colcombe Reflexion and Reflection: A Social Cognitive Neuroscience Approach to Attributional Inference Matthew D. Lieberman, Ruth Gaunt, Daniel T. Gilbert, and Yaacov Trope Antecedents and Consequences of Attributions to Discrimination: Theoretical and Empirical Advances Brenda Major, Wendy J. Quinton, and Shannon K. McCoy A Theory of Goal Systems Arie W. Kruglanski, James Y. Shah, Ayelet Fishbach, Ron Friedman, Woo Young Chun, and David Sleeth-Keppler Contending with Group Image: The Psychology of Stereotype and Social Identity Threat Claude M. Steele, Steven J. Spencer, and Joshua Aronson Index

376

CONTENTS OF OTHER VOLUMES

Volume 35 Social Identity and Leadership Processes in Groups Michael A. Hogg and Daan van Knippenberg The Attachment Behavioral System in Adulthood: Activation, Psychodynamics, and Interpersonal Processes Mario Mikulincer and Phillip R. Shaver Stereotypes and Behavioral Confirmation: From Interpersonal to Intergroup Perspectives Olivier Klein and Mark Snyder Motivational Bases of Information Processing and Strategy in Conflict and Negotiation Carsten K. W. De Dreu and Peter J. Carnevale Regulatory Mode: Locomotion and Assessment as Distinct Orientations E. Tory Higgins, Arie W. Kruglanski, and Antonio Pierro Affective Forecasting Timothy D. Wilson and Daniel T. Gilbert Index Volume 36 Aversive Racism John F. Dovidio and Samuel L. Gaertner Socially Situated Cognition: Cognition in its Social Context Eliot R. Smith and Gu¨n R. Semin Social Axioms: A Model for Social Beliefs in Multicultural Perspective Kwok Leung and Michael Harris Bond Violent Video Games: Specific Effects of Violent Content on Aggressive Thoughts and Behavior

Craig A. Anderson, Nicholas L. Carnagey, Mindy Flanagan, Arlin J. Benjamin, Jr., Janie Eubanks, and Jeffery C. Valentine Survival and Change in Judgments: A Model of Activation and Comparison Dolores Albarracı´n, Harry M. Wallace, and Laura R. Glasman The Implicit Volition Model: On the Preconscious Regulation of Temporarily Adopted Goals Gordon B. Moskowitz, Peizhong Li, and Elizabeth R. Kirk Index Volume 37 Accuracy in Social Perception: Criticisms, Controversies, Criteria, Components, and Cognitive Processes Lee Jussim Over Thirty Years Later: A Contemporary Look at Symbolic Racism David O. Sears and P. J. Henry Managing Group Behavior: The Interplay Between Procedural Justice, Sense of Self, and Cooperation David De Cremer and Tom R. Tyler So Right it’s Wrong: Groupthink and the Ubiquitous Nature of Polarized Group Decision Making Robert S. Baron An Integrative Theory of Intergroup Contact Rupert Brown and Miles Hewstone Says Who?: Epistemic Authority Effects in Social Judgment Arie W. Kruglanski, Amiram Raviv, Daniel Bar-Tal, Alona Raviv, Keren Sharvit, Shmuel Ellis, Ruth Bar, Antonio Pierro, and Lucia Mannetti Index

CONTENTS Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

Exploring the Latent Structure of Strength‐Related Attitude Attributes Penny S. Visser, George Y. Bizer, and Jon A. Krosnick I. II. III. IV. V. VI. VII. VIII. IX.

Introduction.. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Defining Strength‐Related Attitude Attributes .. . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Underlying Constructs? . .. . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Attitude Importance and Knowledge. . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Importance and Certainty . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Importance and Accessibility . . . . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Accessibility and Certainty . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Attitude Strength Composites and Their Constituents . . .. . . . . . .. . .. . . .. . .. . . .. . General Discussion . . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . References . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. .

1 3 5 18 28 32 38 41 48 60

Implementation Intentions and Goal Achievement: A Meta‐Analysis of EVects and Processes Peter M. Gollwitzer and Paschal Sheeran I. II. III. IV. V.

Introduction.. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Goal Intention Strength and Goal Achievement . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Self‐Regulation of Goal Striving . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Present Review. . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Conclusions .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . References . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. .

69 70 73 86 109 111

Interracial Interactions: A Relational Approach J. Nicole Shelton and Jennifer A. Richeson I. II.

Conceptual Approaches to Study Interracial Interactions. . . . . . .. . .. . . .. . .. . . .. . Intergroup Meta‐Perceptions . . . . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . v

121 125

vi

CONTENTS

III. IV. V. VI. VII. VIII.

Avoiding Interracial Interactions . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Dynamics of Interracial Contact: A View from Both Sides. .. . . . . . .. . .. . . .. . . .. . Interpersonal Concerns with Prejudice and Interracial Contact . . . .. . .. . . .. . . .. . Racial Attitudes and Interracial Contact . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Methodological Considerations .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Suggestions for Future Directions . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . References. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. .

127 135 135 147 164 172 175

The Psychology of Self‐Defense: Self‐AYrmation Theory David K. Sherman and GeoVrey L. Cohen I. II. III. IV. V. VI. VII.

Introduction . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Self‐AYrmation and Threats to the Individual Self . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . Self‐AYrmation and Responses to Collective Threats.. . . . . . .. . .. . . .. . .. . . .. . . .. . Moderator Variables and Qualifying Conditions . . . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Underlying Processes and Superordinate Functions . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . Implications for Interpersonal Relationships and Coping . . .. . . . . . .. . .. . . .. . . .. . Conclusions . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . References. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. .

183 189 205 213 221 229 231 232

Intergroup Beliefs: Investigations from the Social Side Charles Stangor and Scott P. Leary I. II. III. IV. V. VI. VII. VIII. IX. X.

Introduction . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . What are Stereotypes and Prejudice?. .. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Social Motivations and Intergroup Beliefs . .. . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . The Acquisition of Intergroup Beliefs in Childhood . .. . . . . . .. . .. . . .. . .. . . .. . . .. . Personality and Intergroup Beliefs . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Priming Social Motivations . . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Group DiVerences in Intergroup Beliefs . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Group Beliefs as Social Norms .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . The Sharing and Transmission of Intergroup Beliefs. .. . . . . . .. . .. . . .. . .. . . .. . . .. . Conclusions . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . References. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. .

243 245 247 251 253 257 259 261 265 271 275

A Multicomponent Conceptualization of Authenticity: Theory and Research Michael H. Kernis and Brian M. Goldman I. II.

A Historical Overview of Authenticity. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Taking Stock of These Various Perspectives: Towards a Psychological View of Authenticity . .. . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. .

284 292

CONTENTS

vii

III. Psychological Perspectives on Authenticity . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . IV. A Multicomponent Conceptualization of Authenticity . . .. . . . . . .. . .. . . .. . .. . . .. . V. Measuring Individual DiVerences in Dispositional Authenticity: The Authenticity Inventory . . .. . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . VI. Authenticity and Healthy Psychological Functioning. . . . .. . .. . . .. . .. . . .. . .. . . .. . VII. Authenticity, Relationship Functioning, and Relationship Satisfaction. . . . . .. . VIII. Authenticity and Retrospective Accounts of Parental Authority Styles. . . . . .. . IX. Authenticity and Self‐Esteem. . . . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . X. Potential Downside of Authenticity . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . XI. Future Directions. .. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . XII. Summary . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . References . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. .

293 294 303 307 326 335 338 339 342 344 348

Index ........................................................................................... Contents of Other Volumes ...............................................................

359 367