Personal Coping
This page intentionally left blank
PERSONAL COPING Theory, Research, and Application EDITED BY
Bru...
191 downloads
1775 Views
15MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Personal Coping
This page intentionally left blank
PERSONAL COPING Theory, Research, and Application EDITED BY
Bruce N. Carpenter
PRAEGER
Westport, Connecticut London
Library of Congress Cataloging-in-Publication Data Personal coping : theory, research, and application / edited by Bruce N. Carpenter. p. cm. Includes bibliographical references and indexes. ISBN 0-275-93012-2 (alk. paper) 1. Adjustment (Psychology) 2. Stress (Psychology) I. Carpenter, Bruce N. BF335.P48 1992 155.2'4-
The paper used in this book complies with the Permanent Paper Standard issued by the National Information Standards Organization (Z39.48-1984). P In order to keep this title in print and available to the academic community, this edition was produced using digital reprint technology in a relatively short print run. This would not have been attainable using traditional methods. Although the cover has been changed from its original appearance, the text remains the same and all materials and methods used still conform to the highest book-making standards.
Contents
PREFACE
1. Issues and Advances in Coping Research Bruce N. Carpenter 2.
3.
6.
1.
1
Conceptual and Methodological Issues in Current Coping Assessments Arthur A. Stone, Eileen Kennedy-Moore, Michelle G. Newman, Melanie Greenberg, and John M. Neale
15
Making the Case for Coping Susan Folkman
31
4. Coping with Psychosocial Stress: A Developmental Perspective Bruce E. Compas, Vanessa L. Malcarne, and Gerard A. Banez 5.
Vll
47
Situational Determinants of Coping Robert R. McCrae
65
Individual Differences in the Coping Process: What to Know and When to Know It Suzanne M. Miller
11
Interpersonal Aspects of Coping Bruce N. Carpenter and Susan M. Scott
93
vi 8.
9.
10. 11. 12. 13.
Contents Perceived Control, Personal Effectiveness, and Emotional States Herbert M. Lefcourt
111
Temporal Factors in Stress and Coping: Intervention Implications Stephen M. Auerbach
133
Life Crises and Personal Growth Jeanne A. Schaefer and Rudolf H. Moos
149
Outcome Expectancies and Psychosomatic Consequences Holger Ursin and Karsten Hytten
171
Religious Beliefs and Practices and the Coping Process Crystal Park and Lawrence H. Cohen
185
A Theory of Family Competence and Coping Luciano V Abate
199
BIBLIOGRAPHY
219
NAME INDEX
253
SUBJECT INDEX
263
ABOUT THE CONTRIBUTORS
267
Preface
A natural outgrowth of attention to stress issues over the past two decades has been a tremendous increase in the research on coping. Although mental health practitioners have necessarily dealt with coping for many years, the development of coping models is mostly a recent phenomenon. The recognition that coping is not simply a reflection of pathology has led researchers to study the area apart from the traditional models of psychopathology, yielding clearer and more powerful models. Thus, many persons are interested in updating their knowledge of this area to reflect this shift. Also, the large volume of research now being conducted in this area is scattered across many topical areas, making it difficult for the average reader to find the common themes and perceive the structure. It is hoped that the essays contained in this volume will serve these needs. Because coping necessarily implies a problem, much of the writing in the area has been integrated with concepts of stress. A large proportion of the research on coping, then, describes coping with a given life problem or stressor. This has been a useful approach and has clarified some of the dynamics of those particular situations. However, three relevant points have become increasingly evident: (1) the variety of stressors is extremely large, making a comprehensive examination of stressful situations unwieldly; (2) the obvious and usual ways of distinguishing stressors (e.g., marriage, divorce, and death of spouse) are often not along the dimensions that have implications for coping; and (3) a number of coping process models appear to have relevance across many situations. It seems appropriate, therefore, to focus on features of coping which have broad relevance, applying to many situations. As will become evident, many contributors to the coping literature have adopted the process model proposed by Lazarus and Folkman in their landmark 1984 book, States, Appraisal, and Coping. However, as characterized by the
vm
Preface
chapters in this volume, not all find it necessary to refer to such a model; some openly disagree with elements of that approach, and problems in studying coping from such a perspective are emerging. Several aspects of this controversy are highlighted in the following chapters. Such treatments will be useful to those conducting research on coping or designing intervention programs. For the most part, each chapter focuses on some limited aspect of coping, examining the dynamics of that component and its relevance to the whole process. Topics are wide-ranging, from how coping with extreme stressors can contribute positively to personal growth, to the role of personal control, to situational influences on coping. In total the chapters provide an overview of major principles of coping and description of important, emerging trends.
Personal Coping
This page intentionally left blank
1
Issues and Advances in Coping Research Bruce N. Carpenter
The recent past has seen a tremendous expansion in our attention to and knowledge about stress. Both mental health professionals and the public have focused on stress and its consequences, at least in part because research has demonstrated its importance and begun to divulge its mechanisms. A natural outgrowth of this attention to stress is a parallel increase in the research on and understanding of coping. Many of the questions, and perhaps many of the insights, regarding coping are not new. Much of psychology and related disciplines has sought to understand how we interact with our environment, especially when its demands are harsh. Mental health practitioners have necessarily dealt with coping for many years, although perhaps traditionally emphasizing variables quite different from those to which stress research points. Developmental psychologists focus on principles similar to coping, but with greater attention to adaptation—how the individual changes in response to normal environmental pressures. Similarly, groups as diverse as sociologists studying the impact of poverty to endocrinologists seeking to understand the role of social variables on diabetic control explore processes akin to coping. In spite of this, general coping models are mostly a recent phenomenon. The ability of stress concepts to tie together diverse kinds of environmental demands and human reactions appears to provide the needed unifying focus. Coping, then, is necessarily connected to stress, and many of the emerging principles have become evident from our understanding of stress. Both lay and scientific perspectives on coping typically highlight efforts of the individual or system to better respond to stress. The recognition that coping is not simply a reflection of pathology or specific to a particular medical condition has led researchers to study the area apart from the traditional models of psychopathology and across
2
Personal Coping
narrow conditions, yielding clearer and more powerful models. However, even though stress has provided a unifying focus, the boundaries and mechanisms of this relationship are still unclear.
BACKGROUND There are a number of constructs and principles relevant to coping which are important throughout this book. Although there is still considerable discussion about ideal definitions, much agreement about central features and basic concepts does exist. Therefore, a brief overview of several is given here to help readers not already familiar with these concepts. Stress The term "stress" has been used in a variety of ways. Most often it carries one of two meanings. First, it is viewed as the complex constellation of reactions one has when demands are perceived as exceeding readily available resources (e.g., Selye, 1982). From this view, then, stress is how one feels and reacts to heavy demands. This response might include physiological, cognitive, affective, and behavioral components, and is usually perceived by the individual as noxious. In contrast, some prefer to view stress as the situation of high demands and limited resources (e.g., Cox, 1978). To distinguish between the two views the first might be considered the stress response, and the second the stressful situation, although both are clearly important for understanding stress. Alternatively, it has been proposed that stress is best used as a general term of the total process linking demands to reactions and other outcomes (e.g., Lazarus & Folkman, 1984). Stressors Most researchers now distinguish stressors from stress, using the former to refer only to the demands themselves that may lead to a stress reaction. Others use the term more generally, including events and situations that are potentially stressful because they make demands or because they lead to stress for many people. Distinguishing stress from stressors makes it easier to account for wide differences in individual reactions to seemingly similar environmental demands. Resources Whereas some models posit that stress results directly from threat or the perception that demands can lead to negative outcomes, many models explicitly consider coping potential, or resources, in the development of a stress response. For example, these latter models may view stress as resulting from the condition in which stressor demands exceed readily available resources or use up those
Issues and Advances
3
resources faster than they are typically replenished. Resources might include anything that one can use to help deal with either the environmental demand or the resulting stress. These resources are categorized in different ways: for example, they might be thought of as including personal resources (e.g., abilities and personality) and environmental resources (e.g., social support and tangible resources like money). Appraisal Cognitive models of stress and coping typically emphasize that a judgmental process occurs, comparing demands to resources. Appraisal, because it is subjective and depends on one's evaluation of a situation, helps account for individual differences in reactions to similar situations. Also, appraisal is viewed as ongoing, allowing for new information and shifts in one's beliefs about a situation, making it easier to explain how the stress response may change over time. Lazarus and Folkman (1984) describe two main types of appraisal, which they label primary and secondary. Primary appraisal refers to evaluation of a situation to determine if demands are likely to be stressful, whereas secondary appraisal refers to judgments regarding availability and suitability of resources and likely outcomes. COPING DEFINITIONS Among the basic questions facing those wishing to understand the process of stress and coping is that of definition. There is still no general agreement about what ought to be considered coping. Because we are at an early level of understanding, it is quite appropriate to have several competing perspectives. However, efforts to compare research findings and to apply principles must take definitional differences into consideration, thus making integration and application of results more difficult and more prone to error. The idea of coping is quite general, encompassing a wide variety of behaviors, as well as several traditionally distinct lines of research. Thus, when various researchers and clinicians invoke the term "coping," they often mean quite different things. Frequently, a definition is not even given, perhaps inferring reliance on lay conceptions of coping or forcing the careful reader to form some operational definition based on the measures used. Given the lack of agreement on the best definition for the construct, such failure by an author to specify what is meant by the term "coping" can be a serious problem. The relevance of such research remains vague, and results are more difficult to integrate with other findings. Perhaps more importantly, those engaging in such research and those using research findings to design interventions without clear specification of the construct are often themselves unclear about what processes are important or are being examined.
4
Personal Coping Coping Activities versus Coping Outcomes
From the better articulated perspectives of coping several definitional issues emerge which seem to make a difference. One clear trend has been to distinguish coping from its consequences. This is in contrast to common lay usage of the term, which frequently intermingles coping efforts with coping outcomes. (For example, to say that someone "is coping with the death of the spouse" means, for most persons, that this person is doing well under the circumstances.) As scientists, however, it becomes quite useful to separate efforts from outcomes. To suggest that only the former is coping may not be satisfactory to some, who may prefer to think of coping as a process that ties behavior to outcomes. Thus, it becomes reasonable to talk about coping efforts, activities, or behaviors, and to distinguish them from coping outcomes. In fact, to suggest that coping is only the activity portion seems to create other problems when defining coping. In particular, as we seek to distinguish which of our many behaviors qualify for coping it may be advantageous to define them as those for which we have particular outcome expectancies. This approach still allows outcomes to be separated from coping activities, but suggests how those outcomes impact the system. Chapter 11, by Ursin and Hytten, discusses this concern at length. Similarly, without consideration of additional variables, such as outcome expectancies, it becomes difficult to distinguish many coping behaviors from stress itself and from coping outcomes. For example, when a person experiencing stress withdraws from others it may not be clear if the withdrawal is a coping behavior (e.g., reduce demands, deny problems) or part of a stressinduced depression, unless we consider expectancies which might motivate the person to withdraw. Separating coping activities from coping outcomes also helps remove a priori judgments about the value of particular coping behaviors (see Folkman, Chapter 3). This helps us account for findings that coping behaviors traditionally thought of as primitive, regressive, or negative can be quite helpful at times, and that many coping activities can have both positive and negative consequences. This neutral acceptance of behaviors without regard to consequence might be problematic for some. For example, some may not want to consider the person who abuses his spouse as a response to stress to be coping, even if the individual feels better afterwards. Coping as an Individual Difference Another clear trend in recent definitions is to see coping as "what one does" rather than "what one is." Historically, efforts to understand coping as personality attributes, or even as styles, have not been very successful. The problems inherent in such an approach are carefully discussed by Folkman (Chapter 3). In spite of this, most models suggest that individual differences ought to be influential in the process of selecting and using coping activities and in perceiving
Issues and Advances
5
coping outcomes. To the extent that this is true, then, the impact of stable personality characteristics ought to lead to coping activities which operate like individual difference variables. The relationship of personality to coping activities is seemingly one instance of the debate over consistency of behavior and personality (e.g., Epstein, 1979; Mischel, 1973). Thus, at the specific behavior level we may well observe little consistency, but as coping behaviors are organized into meaningful classes, stability may emerge (see the discussion of coping measurement below). Several of the chapters in this volume address these issues, with far more positive results than were suggested from past research. In particular, Miller (Chapter 6) describes evidence for her theory of blunting and monitoring, which suggests that people's use of these coping activities approaches the concept of a coping style. The work of McCrae (Chapter 5) gives strong evidence regarding the stability of coping behavior. His work is especially interesting because he examines the role of individual differences simultaneously with that of situational determinants of coping. Lefcourt (Chapter 8), and Carpenter & Scott (Chapter 7) also suggest ways in which personality factors influence coping. Coping as a Response to Stress Because coping behaviors are, in large measure, much like other behaviors, some defining feature is needed. In the case of coping this feature is stress. To remove consideration of stress from definitions of coping would open these definitions to an unacceptably wide array of behaviors. But the union has created a variety of complexities, largely because of wide individual variability in response to similar stressful conditions. If we rely on situational parameters, such as life events scales, to define stress, and hence coping, we cannot account for these differences. Similarly, the differences in the experience of stress appear to be at least in part due to difference in coping, creating a definitional nightmare. At present it seems most models view coping behavior as a response to a stress reaction, thereby accounting for individual variability by making a stress response a necessary precursor. Coping occurs only when one is experiencing stress. Although apparently useful, at least two problems emerge from this approach. First, the definition relies heavily on another definition, that of the stress response, about which lack of agreement also exists. Second, this approach makes it difficult to accept stress avoidance or stress prevention behaviors as coping. Rigid application of the principle would specify that before a stress response starts (and perhaps after one ends) any behavior would not be considered coping. From this perspective, the individual who anticipates stress and works to avoid the stress, even though perhaps doing exactly what someone already stressed by the same situation is doing, is not coping. At present we have little data with which to evaluate the utility of separating stress avoidance and prevention from other forms of coping. We may well find that such behaviors do indeed operate like coping behaviors performed in re-
6
Personal Coping
sponse to a current stress reaction, arguing against the distinction. But until we have such evidence, researchers should, at the least, be clear about how the behaviors they study are related to stress reactions, even if they choose to view avoidance and prevention behaviors as coping. Coping as a Process In an effort to study coping it is often true that we attempt to take a "snapshot" to more carefully examine what is going on at a particular moment. In keeping with the trend away from viewing coping as an attribute of the person, Folkman and Lazarus (1985) emphasize that coping is a process, with complex feedback loops and constant updating. Thus, as coping progresses over time, one's behavior might change. To the extent that coping is viewed as behaviors in which one engages rather than as characteristics of oneself, there seems to be relatively little disagreement on this point. As researchers study the phenomenon of coping they should, therefore, remember the process nature of the construct is not well represented by their relatively static measures. A number of the models articulated in this book specifically allow for updating of the model variables over time. However, little empirical work has been done clarifying how coping progresses over time and what variables are most salient. Chapter 9, by Auerbach, examines a portion of this process in detail. Hopefully, future work will better attend to temporal features, especially clarifying the interplay between the process of coping and the relatively stable, personality based features which are emerging. Effort and Intentionality Lazarus and Folkman (1984) limited coping behaviors to effortful responses, specifically ruling out activities that are automatic. Thus, they take a subset of behaviors, adaptive responses, and divide them further so that only effortful responses are viewed as coping. It appears that their primary reason for doing so is that effort can serve in part as a substitute for effectiveness ("trying" as opposed to actually "succeeding") when defining coping. The issue is a bit confusing, perhaps in part because automatic behavior is probably not really the same as effortless behavior. For example, repression or denial may for some operate quite automatically but take considerable effort, whereas getting extra sleep may require essentially no effort but is usually not automatic. Some theoreticians might further object because other behaviors, such as smoking, might appear both automatic and effortless, but can still function much like other coping behaviors. Lazarus and Folkman's distinction may yet prove to be a useful one, although there is probably not yet sufficient evidence with which to evaluate the issue. For example, we may find that relatively automatic behavior does not fit as well into models of coping as does more effortful behavior. However, if the intention
Issues and Advances
7
is to separate coping from adaptation and from outcomes, we may find the use of expectancies and intentionality to provide a more useful way. Unfortunately, intentionality has similar problems, suggesting awareness, and thereby making it difficult to account for nonconscious coping activities.
MEASUREMENT AND STRUCTURE OF COPING Even after selecting a definition of coping, the measurement of coping remains a difficult exercise. Efforts to empirically examine the structure of coping activities has, of necessity, relied on available coping measures and yielded muddled results as a consequence. These two issues, measurement and structure, are briefly overviewed here. Measurement In keeping with older definitions of coping, several trait measures, such as those of ego strength and repression, have served as measures of coping. Although such measures have been related to outcomes, efforts to relate them to behavioral definitions of coping have yielded discouraging results. More recently, the measurement of coping has focused on coping activities. With this approach the usual procedure has been to ask people what they have done when faced with a particular stressor. The coping activity approach, then, has tended to be rather behavior-specific, in terms of both the eliciting stimuli with which one copes and the reactions of the individual to those stimuli. Even when open-ended approaches are used ("tell how you coped w i t h . . . " ) , people quite naturally tell what they did or thought, rather than describe what kind of person they are. Similarly, structured approaches usually list a number of potential coping behaviors and ask the respondent to report the frequency with which each was used. Stone and his colleagues (Chapter 2) discuss in detail the most widely used of the objective, self-report measures, the Ways of Coping Scale (Folkman & Lazarus, 1988), highlighting features of this procedure which are relevant to most measurement approaches. Several difficulties in measuring coping seem especially evident. First, although a focus on coping behaviors may make sense from a definitional and theoretical perspective, measurement of specific behaviors may not be very revealing. It has become evident, for example, that specific coping behaviors vary widely among individuals, even when they experience the same stressor (e.g., Scott & Carpenter, 1990). They also vary within individuals across stressors. Consequently, emphasis at the level of specific behaviors highlights the idiosyncratic nature of coping and does not seem to lead to meaningful patterns and laws. It would be useful, then, to organize coping behaviors into broader categories. The typical empirical procedures for structuring data and forming classes of behavior, such as factor analysis, do not work in this case, however. For example,
Personal Coping
8
the Ways of Coping Scale has tended to result in rather different factor structures across studies, suggesting that the underlying structure is quite unstable (Stone et al., Chapter 2). (An exception seems to be the problem-focused versus emotion-focused distinction of coping activities, which does seem to be a somewhat reliable finding.) The reason for this lack of replication may result from the differences in populations and stressors across studies. In particular, it seems that many of the items on coping scales are so specific that they are irrelevant for particular stressors. Thus, in one study a large chunk of items is endorsed by essentially no one, whereas in another study, using different stressors, a different set of items are irrelevant and remain unused. Unendorsed items will not load on any factor, and those endorsed by only a few might yield spurious loadings. Thus, the applicability of the items to the stressor or group at hand seems to drive the analysis more than any real underlying patterns. A related issue is the multifaceted nature of even rather specific behaviors. From the perspective of coping, one might engage in a behavior for one or more reasons. For example, a scale item like "talked to a friend" might be endorsed if someone was seeking advice, looking for emotional support, asking for tangible help, or simply distracting oneself. Scott and Carpenter (1990) found, for example, that even relatively specific coping behaviors served multiple coping functions for many subjects. One must question, then, whether measurement of specific behaviors is the correct approach. In contrast, to ask directly about specific coping functions assumes that respondents have sufficient insight to recognize their real motivations for engaging in an activity. Perhaps some combination of inquiry about both coping behaviors and coping functions will be most suitable. Finally, it is not clear what about coping is most important to measure. Some measures ask only if a behavior was performed, whereas more complete measures ask how frequently behaviors were used. It might be important to go beyond frequency and ask extent of use or proportion of time spent in each coping activity. Similarly, the amount of effort expended on each activity might also clarify the utility of coping behaviors. Further, because studies reveal that people engage in multiple coping activities to achieve multiple functions, it may not be sufficient to assess outcomes separately from coping; it may be necessary to determine the relationship of each activity or function to outcomes. For example, one might simultaneously engage in two coping activities, relaxation and problem solving, the former to control arousal created by the latter's focus on solving the stressful problem. Teasing that process out is very difficult without measuring the separate outcomes relevant to each coping activity. Structure In spite of these difficulties a structure of coping seems to be emerging. Perhaps the clearest and most useful distinction is that of problem-focused versus emotionfocused coping (Folkman & Lazarus, 1980). The distinction suggests that even
Issues and Advances
9
if coping occurs within the context of stress and is intended to reduce stress, this can be accomplished in ways other than by directly reducing the stress response. Thus, functions that remove or prepare one to remove stress by impacting stressors or appraisal processes can qualify as coping. In addition to allowing for indirect means for reducing stress, the distinction clarifies the relationship between components of the stress-coping process and has empirical support. We might expect that stress dimensions should have implications for the structure of coping. This appears to be true. A stress response seems more likely, for example, when there is low perceived control, lack of predictability, and long duration of stressor exposure. Paralleling this is the distinction in coping of avoidant and nonavoidant coping strategies (e.g., Suls & Fletcher, 1985). Research on these two classes of coping behaviors suggests that avoidant behaviors work best on short-term stressors, for which nonavoidant strategies may be too costly in terms of effort and arousal, given that the stressor will end shortly anyway. In contrast, avoidance of potentially long-term stressors eventually yields more negative consequences than do nonavoidant strategies because with the former, problems are neither solved nor reduced. Nonavoidant strategies, however, do not seem superior when one has little control over the stressor; in this case avoidant strategies seem superior regardless of temporal factors. In this regard, Miller's work on blunting/monitoring (Chapter 6) is quite relevant. Lefcourt's discussion of control (Chapter 8) is also quite helpful for understanding the impact of control and personal characteristics affecting that process. Carpenter and Scott (Chapter 7) briefly present data on the structure of stressors suggesting that a relational component may be significant, and then discuss coping resources that are organized around this distinction.
COPING OUTCOMES Distinguishing Outcomes from Stress Much of the interest in stress research has been fueled by the recognition that stress leads to negative outcomes. Quite naturally, these traditional stress outcomes have become a primary focus when applying coping principles or when researching coping outcomes. But recent efforts to personalize stress to account for variability, as mentioned above, lead us toward defining stress in terms of a stress response (e.g., Cohen, Kamarck, & Merrnelstein, 1983). Although this is not a problem for separating coping activities from coping outcomes, it presents a real challenge for deciding if an outcome is best thought of as a stress response or a negative consequence of stress. Some studied outcomes are clearly longterm and cumulative, such as heart disease, and are readily distinguishable from a limited stress response. But others, such as affective states like depression and anxiety, or more transient physical conditions like headaches, can be part of
10
Personal Coping
both a stress response and stress outcomes. If we feel forced to make a distinction, no clear point at which to draw the line is apparent. To get around this difficulty, we can instead emphasize the process nature of stress, coping, and outcomes, viewing outcomes as part of the stress response. This might be especially justified for short-term outcomes that are readily confused with the stress response. But even outcomes that are long-term and manifest only after chronic or high levels of stress over a period of time might be considered part of stress itself; the distinction between these and other reactions typically thought of as stress becomes one of realizing different features of the response at different times in the process. Alternatively, we might reject the idea of stress as one's reaction, instead reserving the construct for "stressful situations." This approach lumps the stress response with outcomes, rather than the other way around. One manifestation of such a blurring between traditional outcomes and the stress response is the trend toward emphasizing the stress experience itself as negative, regardless of other outcomes like poor health. Many stress reduction programs take this approach—the goal is to stop "feeling stressed." Similarly, the research literature includes as targets of coping both the stress response and more peripheral outcomes. In spite of the problems, it is probably premature to scrap the distinction between the stress response and other outcomes. Most likely we will continue making a somewhat arbitrary distinction between the stress response and stress outcomes. For example, the emotion-focused problem-focused approach to organizing coping activities matches this distinction. The stress response, as a negative emotional state, seems to be directly influenced by emotion-focused coping; other outcomes are only indirectly affected by coping as the stress response is controlled. Similarly, problem-focused coping, with its emphasis on the stressor, mostly affects the stress response indirectly. Impacting Outcomes Recognizing the ways in which coping outcomes relate to stress outcomes can also help us postulate possible mechanisms by which coping can contribute to positive results. For example, coping can act (1) as a buffer of the effects of stress; (2) to counteract the effect of stress by directly leading to improved outcomes; or (3) to remove stress and thereby reduce its impact. Similarly, the stress-coping paradigm, including related features such as appraisal and resources, readily implies four ways in which coping can impact the system: (1) minimize the stress response; (2) remove or reduce demands; (3) increase resources; or (4) alter appraisal. Outcomes as Feedback Both short- and long-term outcomes of coping apparently serve as feedback regarding stress and coping efficacy. However, although intuitively obvious,
Issues and Advances
11
little research has been done on this process. Most likely, traditional learning principles are in effect reinforcing and shaping various beliefs about how demanding situations are and how coping activities impact them. It is probably true that from our perceptions of outcomes come most beliefs that affect our primary and secondary appraisal, thereby influencing our stress reactions and our choices for coping activities. The case of more extreme or traumatic stress may be different, although it is perhaps simply more dramatic. When faced with such stress one's traditional coping may prove woefully inadequate, and beliefs that drive appraisal can be strongly challenged. Schaefer and Moos (Chapter 10) discuss in some detail the changes that can occur in persons as a consequence of extreme conditions and the exploration of alternative coping processes and beliefs that may follow. In spite of the negative nature of the stress itself, Schaefer and Moos draw attention to the positive change that can occur as one forms new beliefs and develops new skills. These more extreme cases remind us of the essential ties between coping and outcomes; outcomes seem to set the system, altering the beliefs that are responsible for appraisal and choice of coping response. Clearly, people are able to largely ignore mild challenges to this belief system, as evidenced by persistent use of ineffective and self-damaging coping. In contrast, dramatic challenges cannot be ignored and reveal the process. At present our understanding of the natural systems for monitoring progress and success of coping is poorly understood. ADDITIONAL ISSUES Development of Coping The learning of effective life skills and the development of positive characteristics are the focus of much of psychology. Even so, we are just beginning to understand how those skills and characteristics impact one's response to stress. For example, Terman's (Oden, 1968) landmark study of very bright persons demonstrated the link between intelligence and life achievement. It seems likely that intelligence has similar linkage to coping choice and effectiveness, but such links remain largely untested. Similarly, it is well known that most persons exposed to severe trauma do not develop post-traumatic stress disorder. Important factors beyond the trauma itself appear critical, yet efforts to distinguish resilient persons from those manifesting the disorder have yielded largely negative results. An important parallel to studying skills and characteristics impacting coping in adulthood is the direct examination of stress and coping in childhood and adolescence. This line of research clarifies changes in the stress response, appraisal, and coping choice and ability. It may also provide clues about critical stages in the development of effective coping and about how some persons become especially vulnerable to particular stressors. Compas, Malcarne, and Banez (Chapter 4) examine coping from a developmental perspective.
12
Personal Coping Context of Coping and Level of Analysis
The emphasis of this book, as stated by the title, is on individual coping. From this perspective, environmental influences are usually viewed as either stressors or resources. However, it is of course true that the coping individual operates within a system, and coping activities both affect and are affected by that system. For example, the coping of those with chronic health problems is typically quite dependent on caregiver behavior, especially that of family. We now know that demands on these resource people often lead to caregiver strain and burnout (e.g., Brody, 1985), which in turn affects the coping of the ill person. Because the contexts in which most coping occurs are organized and somewhat stable, analysis at the level of the system could yield new insights. The interaction between coping individuals and their social environments suggests that system models may hold promise, especially as we focus more on coping as a process rather than a static event. Further, we might extend the principles of coping to the systems themselves, studying how families, work groups, and neighborhoods cope with problems affecting the system. L'Abate's family model of coping (Chapter 13), perhaps the most complete of its kind, is included as an example of the systems approach.
Cognitions, Beliefs, and Values The concept of appraisal introduces cognitions into the stress-coping process by suggesting that persons make judgments about stressors, resources, and their coping efforts. In fact, a number of thoughts and beliefs, not readily obvious from the construct of appraisal, might help account for differences in stress response, coping activities, and coping outcomes. At least two types of beliefs are in need of research. First, in appraisal one need not have an actual stressor before judgments might be made about demands and one's ability to meet them— imagined stressors appear to also activate this process. Imagined stressors could include worry over possibilities that are not realized, anticipation of real stressors (threat), and pure fabrication. Further, actual stressors might be greatly distorted as well. With physical stressors it is likely that coping activities and effectiveness change as one moves from simply anticipating the stressor to actually experiencing the demand. In contrast, the impact of psychological stressors on coping may not be distinguishable from that of imagined stressors because no direct physical demand is placed on the body. Second, it seems likely that more global beliefs and values affect the appraisal and coping process, but the nature and mechanisms of that impact are not very clear. In Chapter 12, Park and Cohen examine the impact of religiosity, one such variable, on the coping process. It should not surprise us that global beliefs help determine one's stressors, available resources, appraisal, and choice of coping options.
Issues and Advances
13
Application Given the negative nature of stress and long-term consequences on health and well-being, it is natural to apply our growing understanding of the coping process to help people cope more effectively. Further, the skills developed through effective coping may generalize to other life endeavors, and stress reduction can free up valuable resources for other uses. In fact, stress reduction programs have become very popular, although this author is concerned that many have unnecessarily limited effectiveness because principles of coping are not well understood or utilized. We now understand that the stress-coping process is highly individualized. Intervention efforts will be most effective when similarly individualized. Careful examination of the process for a given person can reveal key points of difficulty; in contrast, a presumption that everyone can cope in the same way is likely wrong. For example, many intervention programs attend primarily to emotionfocused coping, perhaps by teaching relaxation. Too ready use of a limited set of techniques leaves other avenues unexplored, such as problem-focused coping, cognitive reappraisal, resource building and utilization, and demand reduction. Failure to consider multiple ways of impacting the stress-coping process may be even more important in group interventions, such as with stress reduction classes, groups for those with serious illness, and the like. Because these interventions are not tailored to the individual, teaching a relatively broad spectrum of coping alternatives may be best. This should not be taken to mean that a given problem is best attacked with many techniques; using a large number of coping techniques may diffuse one's efforts, and it is only reasonable to assume that a person will only become proficient at a few. But without targeting the particular needs of a person, teaching several options and the principles for selecting among them may be best for group interventions. As we come to better understand the role of enduring individual differences, we may find particular interventions better suited to particular types of people. This suggests that our assessment will need to extend beyond the usual evaluation for amount and type of stressor. Even the somewhat common evaluation for the "stress-prone personality" may not tap into characteristics of the person relevant to coping. Finally, the links between various coping techniques and optimal outcomes need to be better articulated. We have much to learn about which coping techniques are best suited to various situations. Those with broad application may be most valuable, even if more difficult to learn.
This page intentionally left blank
2
Conceptual and Methodological Issues in Current Coping Assessments Arthur A. Stone, Eileen Kennedy-Moore, Michelle G. Newman, Melanie Greenberg, and John M. Neale
The concept of coping has been in the literature for decades. In recent years it has become a research focus of many seeking to understand the role of stress in illness. Early attempts to examine the stress-illness relationship focused on univariate models and examined the association between life event checklists and illness measures. However, these studies met with limited success, with correlations rarely exceeding .30 (Rabkin & Struening, 1976). In an attempt to improve the prediction of illness, multivariate models were proposed and coping was introduced into the stress-illness equation as a moderator of the effects of stress. Early models of coping were based on psychoanalytic ego psychology. Menninger (1963), Haan (1969, 1977), and Vaillant (1977) each developed hierarchical systems of coping styles to describe the conscious and unconscious processes people use in dealing with their environment. All of these models specified a priori that some processes were more adaptive or functional than others and classified individuals according to their pervasive use of particular coping processes (Lazarus, 1984). Determination of coping style was, however, based mainly on clinical interview assessments which often had a complex intrapsychic theoretical basis. This limited the work of these researchers to studies with small samples, and it limited the acceptance of their work to those scientists subscribing to psycho-dynamic theory. Drawing upon these ego psychology models of coping styles, other researchers developed trait measures of coping. Coping traits were regarded as dispositional characteristics which predispose people to react in certain ways in certain situations (Lazarus, 1984). Unlike the earlier models, coping traits were usually assessed with questionnaires rather than interviews; thus, they were considerably easier to assess. Yet these measures assessed coping along a single dimension, such as repression-sensitization (e.g., Shipley et al., 1978, 1979) or coping-
16
Personal Coping
avoiding (Goldstein, 1959, 1973), and thus probably underestimated the complexity and variability of actual coping attempts (c.f., Lazarus & Folkman, 1984). Like the earlier models, the coping trait approach assumed cross-situational consistency in coping efforts, designated certain types of coping as more adaptive, and were strongly tied to dynamic theory. In the late 1970s, researchers including Lazarus, Moos, and Pearlin began quantifying coping thoughts and actions using direct self-report assessments. These new coping researchers were not tied to dynamic theory. They were willing to explore behaviors and cognitions available to an individual's awareness instead of focusing solely on inferred cognitive structures. Menaghan (1983a) noted optimistically in her review article that "instead of relying on a clinician to reveal the underlying conflict and interpret the defensive behavior, investigators can objectively identify the stressor, and the individual can directly report on what he or she does to cope with it. It is this possibility for objective research that distinguishes the work of the past few years" (p. 919). Just as Holmes and Rahe's (1967) scale quantified life stress and opened the way for stress research with the general population, development of nonpsychodynamic self-report questionnaires to quantify coping led to an enormous increase in research on coping. By the end of the 1970s, each of the new coping researchers mentioned above had published self-report coping questionnaires (Billings & Moos, 1981; Folkman & Lazarus, 1980; Pearlin & Schooler, 1978). These questionnaires can be completed by untrained subjects in under thirty minutes and thus have widespread appeal. The most popular of these questionnaires is Folkman and Lazarus's (1980, 1988a) Ways of Coping Inventory (WCI) and literally hundreds of scientific reports have used it. In this chapter we will describe the theoretical basis of the WCI, how it was developed, and how it is typically used. We will then raise a number of conceptual issues concerning the assumptions that the WCI makes about coping and discuss what effect these assumptions might have on the field of coping research. Although the comments will be focused on the WCI, they apply to all current selfreport coping assessments (namely, those developed by Billings & Moos, 1981, and Pearlin & Schooler, 1978).
DESCRIPTION OF THE WCI Theoretical Basis The WCI was designed to assess individuals' coping efforts with respect to a particular stressful episode, based on Lazarus' transactional theory of psychological stress (e.g., Lazarus, 1966; Folkman, Schaefer, & Lazarus, 1979; Lazarus, Kanner, & Folkman, 1980; Lazarus & Launier, 1978). According to this theory, stress is defined as "a relationship between the person and the environment that is appraised by the person as taxing or exceeding his or her resources
Conceptual and Methodological Issues
17
and as endangering well-being." Lazarus suggests stress is mediated by two processes: appraisal and coping. Appraisal is the method through which a person cognitively evaluates, for a particular stressful encounter, what is at stake (primary appraisal) and what can be done to remedy the situation (secondary appraisal). If the demands of the situation exceed the individual's available coping resources, the situation will be appraised as stressful and will be perceived to involve harm-loss, threat, or challenge. Laboratory studies have demonstrated that the way an event is appraised indeed influences the subsequent coping strategy that is chosen (e.g., Averill, O'Brien, & DeWitt, 1977; Folkins, 1970; Monat, Averill, & Lazarus, 1972). Coping is defined as an individual's cognitive and behavioral efforts to manage stress. The term coping process refers to what the individual actually thinks and does in a specific encounter and the way in which these efforts change over the course of the stressful encounter. Appraisal and coping continuously influence each other: appraisal of a situation as stressful motivates coping efforts, which may change the stressful situation, leading to reappraisal. Thus, the transactional theory describes the individual and the environment in a continuous, bidirectional relationship. Coping serves two major functions: regulating distressing emotions (emotion-focused coping) and changing the problem that is causing the discomfort (problem-focused coping). Problem- and emotion-focused functions of coping have demonstrated theoretical importance in relatively independent major research literatures, for instance, sociology (Mechanic, 1962), industrial-social psychology (Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964), and developmental psychology (Murphy, 1974). The WCI reflects the transactional theory in that: (1) it is designed to be used in reference to a specific stressful encounter, rather than stressful encounters in general, and therefore does not assume cross-situational consistency in coping efforts; (2) it refers to what the individual actually thinks or does, rather than what they usually do or what they think they should do; (3) it is designed to assess all coping strategies (of which the person is aware) that are used, and not just those that were somehow effective, in an effort not to confound process and adaptational outcome; (4) it contains items pertaining to both emotion-focused and problem-focused coping; (5) it assesses coping as a multidimensional, "complex amalgamation of thoughts and behaviors," rather than a unidimensional trait; (6) it is designed to include general coping strategies which may be applicable in a variety of specific stressful encounters or in various stages of a single stressful encounter. Development of the Original Ways of Coping Checklist Decisions made during the development of the WCI reflect a number of assumptions and were obviously important to the final product. Because the decisions are central to our discussion of conceptual issues concerning the WCI,
18
Personal Coping
we will describe, in detail, the methods Lazarus and his colleagues used in developing the WCI and a revised version of the WCI (R-WCI). Items for the original WCI covered a broad range of cognitive and behavioral coping strategies. They were both rationally derived and taken from suggestions offered in previous literature (Mechanic, 1962; Sidle, Moos, Adams, & Cady, 1969; Weissman & Worden, 1976-1977). Examples of some of these items are "Got the person responsible to change his or her mind," "Stood your ground and fought for what you wanted," "Looked for a * silver lining' so to speak; tried to look on the bright side of things," and "Tried to forget the whole thing." In the early stages of development, the theoretical distinction between problem-focused (PF) and emotion-focused (EF) coping directed much of the work. Subjects' responses were summarized into scores on two rationally derived subscales. The problemfocused scale (P-scale) contained 24 items concerned with changing the problem itself and the emotion-focused scale (E-scale) contained 40 items concerned with managing emotions related to the stressful event. Folkman and Lazarus (1980) subsequently modified the original WCI using the following data sources to identify weak items: correlations between individual items and their respective scales, interitem correlations, item-factor correlations, frequency of item use, and rater evaluation of items. In order to be a candidate for deletion an item had to be identified as weak by several of the above sources. On the basis of these criteria, four items (two from each subscale) were dropped from the checklist, and one item was moved from the E-scale to the P-scale, yielding a total of 64 items. Folkman and Lazarus state, "Revisions of the scales were made very conservatively because of the danger of artificially creating relationships by modifying scales on the basis of data to which they were to be applied" (p. 225). The two rationally derived subscales of the original WCI were tested for internal consistency in several ways. Personnel from Lazarus' laboratory classified items into PF and EF scales, undergraduates classified each item as representing PF or EF coping, a factor analysis of the coping responses was run, and interitem agreement analyses were performed. It should be expected that laboratory personnel who were familiar with the scales would classify items into their scales rather well, and they did (91 percent agreement). On the other hand, only 78 percent of undergraduates' item classifications were correctly assigned to the PF or EF scales designated by Lazarus' group. Five percent of the items were given an assignment opposite to that of Lazarus' assignment, but this was explained as being due to the "situational context," because the undergraduates classified WCI items according to the items' function in vignettes depicting hypothetical stressful situations. The next test of internal consistency, a factor analysis calling for two orthogonally rotated factors, was performed on WCI data from a community sample. Of the 27 items rationally classified as PF, 21 (78 percent) correlated more strongly with thefirstthan the second factor, and of the 41 rationally classified EF items, 28 (61 percent) correlated more strongly with the second rather than
Conceptual and Methodological Issues
19
the first factor. Internal consistency analysis of two randomly selected WCI administrations treated independently yielded alpha coefficients of .81 for the P-scale and .80 for the EF scale. Overall, Lazarus and his group interpreted these results as "clear support for a problem-focused and emotion-focused scale." Finally, for three randomly selected administrations (excluding those used in the scale revision) the correlations between the PF and EF scales were .35, .52, and .44. On this basis Folkman and Lazarus concluded that the PF and EF scales have enough unshared variance to warrant their independent use. Aldwin et al. (1980) also used a principle components analysis with varimax rotation on the community sample data to "empirically elucidate coping strategies." This resulted in seven scales: one problem-focused coping scale, five emotion-focused scales, and one mixed scale. Essentially the same methods were used by the developers of the other major self-report coping questionnaires mentioned above. Although PF and EF scales were not found in those studies, their results are consistent, in general terms, with analysis of the WCI that have yielded many factors. Development of the Revised Ways of Coping Questionnaire Folkman and Lazarus (1988a) revised the WCI using a sample of undergraduate students studied before, during, and after midterm examinations. Even before data were collected from the students, redundant and unclear items in the original WCI were deleted or reworded and three items were added based on the suggestions of subjects from the community sample, yielding a 66-item checklist. Further modifications of the WCI-R were made based on the data collected from the students. Nine items were removed because their distribution was too skewed or because they had restricted variability. The remaining 57 items were then factor analyzed, using only the 108 subjects who had completed the questionnaire for all three time periods (prior to exam, immediately after the exam, after receiving grades for the exam). Subjects who did not complete all three assessments were eliminated from these analyses in order to avoid overrepresentation of Time 1. Observations were pooled across the three time periods because Folkman and Lazarus "wanted a common metric with which to compare coping on the three occasions based on as large a sample size as possible... even though this involved dependence in the data" (p. 157). They state, "A six-factor solution, using common factor analysis with oblique rotation yielded the most conceptually interpretable set of factors" (p. 157). Fifteen items were then deleted because they did not load clearly on any one factor. One of the six factors was divided into three groups of emotion-focused items, "to provide greater theoretical integrity" (p. 157). The final product was eight coping scales: one problem-focused, six emotionfocused, and one mixed. Internal consistency coefficients on the eight scales ranged from .56 to .85. The mean intercorrelations of the eight coping scales
20
Personal Coping
averaged across the three time periods, ranged from .64 for problem-focused coping and seeking social support to .13 for distancing and self-blame. Of the 28 possible intercorrelations, 22 were .31 or greater. In particular, self-blame and tension-reduction were highly correlated with all other scales. Folkman et al. (1986) used the Revised Ways of Coping checklist in a study involving eighty five married couples interviewed once a month for six months. The most stressful event during the preceding seven days was the targeted problem in this study. Four principle factor analyses with oblique rotation were used, all of which yielded similar results. Five of the eight scales generated were comparable to those found by Aldwin et al. (1980) and Folkman and Lazarus (1985). The other three scales were self-control, confrontive coping, and distancing. Marshall and Dunkel-Schetter (1987) tallied a total of eleven studies which involved factor analysis of some version of the Ways of Coping (including those by Folkman and Lazarus, described above), and the resulting factor structures are consistent across these studies. They suggest six core dimensions of coping based on these studies: focus on problem (also labeled problem-solving, information seeking, and instrumental action); seek support (also called support mobilization); focus on positive (also labelled positive reappraisal, growth, cognitive restructuring, and seek meaning); threat minimization (also called distancing, or detachment); escape-avoidance (also called wishful thinking); and accept responsibility (also labelled self-blame). To the credit of the WCI, all of these studies found essentially the same factor structures across different populations of subjects. Scoring and Typical Use The original Ways of Coping checklist (Folkman & Lazarus, 1980) was a yes/ no checklist. Subjects were asked to indicate whether or not they had used particular coping strategies in coping with a specific stressful event. The number of items marked "yes" were tallied in order to render separate P-scale and Escale scores. The R-WCI (Folkman & Lazarus, 1988a) has a four-point Likert scale response format. Subjects are asked to indicate "to what extent" they used a particular strategy in dealing with a specific stressful event. Response options range from "0 = Does not apply and/or not used" to "3 = Used a great deal." Typically subjects are asked tofillout the WCI by thinking of the most stressful event experienced during the past month, giving a written description of it, and indicating those strategies that were used in each episode. However, the WCI does not appear to be used in one standardized manner. Subjects have been asked about the most stressful situation during the previous week (e.g., Folkman, Lazarus, Gruen, & DeLongis, 1986), the past two months (e.g., Scheier, Weintraub, & Carver, 1986), any time during the past year and a half (e.g., McCrae, 1984), the most stressful time ever (e.g., Felton, Revenson, & Hinrichsen, 1984), and how they generally behave during a stressful situation (e.g., Vingerhoets & Flohr, 1984). Although most studies ask about the event after it is over, some
Conceptual and Methodological Issues
21
researchers ask about upcoming stressors (e.g., Folkman & Lazarus, 1985) as well as ongoing stressors (e.g., Felton & Revenson, 1984). Just prior to describing how they coped with an event, subjects are required to respond to items pertaining to appraisal. In various studies, Folkman and Lazarus have assessed global appraisal (Folkman & Lazarus, 1985), primary appraisal (Folkman & Lazarus, 1986; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986), and secondary appraisal (Folkman & Lazarus, 1980, 1986; Folkman et al., 1986). Global appraisal is assessed by asking subjects to indicate to what extent they experienced various emotions reflecting threat, challenge, harm, or benefit. The emotions corresponding to these global appraisals were derived from the transactional theory and pilot study data (Folkman & Lazarus, 1985). The mean alphas for the four global appraisal scales ranged from .59 to .84. Primary appraisal is assessed using a thirteen-item scale describing various things a subject might have "at stake" in a particular stressful encounter. The items on this scale were derived from previous literature and subjects' responses to open-ended questions (cf. Folkman & Lazarus, 1980). Responses on this scale are summarized into two subscales, and four individual items. The two subscales, "threats to self-esteem" and "threats to a loved one's well-being," were derived from a principle factor analysis with oblique rotation, and have alphas of .78 and .76, respectively (Folkman et al., 1986). Secondary appraisals are measured by a four-item scale reflecting coping options. Subjects are asked to indicate to what extent their stressful encounter (1) was changeable, (2) required acceptance, (3) required that they obtain further information before acting, and (4) required that they refrain from acting. Contribution of the WCI It is difficult to overstate the contribution of the WCI to the field of coping research. Not only has it been widely used, but it represents a significant improvement upon previous measures. Unlike previous measures, the WCI enables efficient assessment of a wide variety of coping responses, without making a priori assumptions about which strategies were most efficient. Both in terms of its conceptualization of coping as cognitive and behavioral efforts to manage stress and its use of factor analytic methods of scale construction, at the time of its development it represented the forefront of coping theory and research (cf. Billings & Moos, 1981; Pearlin & Schooler, 1978). However, perhaps because of the pressing need for an efficient tool for the assessment of coping, a number of conceptual issues concerning the WCI have not been addressed. CONCEPTUAL ISSUES CONCERNING THE WCI The Transactional Theory Folkman and Lazarus intended the WCI to reflect their transactional theory of stress and coping. In some ways it does, as described earlier. However, it is
22
Personal Coping
not clear that the WCI measures coping and appraisal as two distinct processes which reciprocally influence each other, nor is it clear that the WCI captures changes in coping over time. First, does it distinguish between coping and appraisal? According to the transactional theory, appraisal and coping are separate processes, which reciprocally influence each other over time. However, the WCI typically measures coping and appraisal at the same time and in relation to one stressful event which occurred during the past month. This method of measurement gives no indication of how these processes mutually affect each other. In fact, at times the assessment of coping and appraisal seems redundant: secondary appraisal is measured by the subject's choice of one of four alternatives, each of which is a reworded version of the coping items in the WCI. For example, the appraisal item that suggests that the situation was one "in which you had to hold yourself back from doing what you wanted to do" is similar to the coping strategy "tried not to act too hastily or follow my own hunch." Appraisal measured in this manner may only be an indication that subjects are consistently choosing the same coping strategy. The meaning of appraisalcoping correlations is, therefore, unclear and future research should examine the distinction between the concepts carefully. Second, does the WCI measure changes in coping over time? Folkman and Lazarus emphasize that coping is a process which changes over time. "To speak of a coping process means speaking of change in coping thoughts and acts as a stressful encounter unfolds" (Folkman & Lazarus, 1984, p. 142). Changing patterns in coping have been demonstrated in research on coping with polio and bereavement (Visotsky et al,, 1961). Yet Folkman and Lazarus' retrospective assessment of coping over the past month means that information about the patterning of coping is lost. This method does not reveal when during the month and in what order particular strategies were used. Since the factoring routines used in the development of the WCI could not utilize information about patterns of coping, scales resulting from factoring may have confounded pattern information with "pure" coping concepts. Perhaps this is why a mixed factor was observed in some analyses of WCI responses. Only one study by Folkman and Lazarus (1985) measured subjects' coping at different times during the same stressor (i.e., before, during, and after an exam). However, in this as well as their 1980 study, they treated each coping assessment period independently in their analyses and made no attempt to examine changes in the coping process over time. As Menaghan (1983a, p. 187) states, "Thus far however, analyses of these data [i.e., Folkman & Lazarus, 1980] have involved a resolute disregard for the repeated measures aspect of their design. Instead, each individual's 4-18 coping reports have been treated as independent data, and analyses have discussed some 1500 'coping episodes' as if there were no dependency among them." If the transactional processis to be studied, multiple prospective assessments at relatively brief time intervals will be required in future research.
Conceptual and Methodological Issues
23
Coping Responses At face value, the WCI appears to tap a number of reasonable concepts that pertain to coping. However, in contrast to Pearlin and Schooler (1978) who developed the items for their coping scale based on open-ended, exploratory interviews concerning four different role areas, with 2,300 people representative of a large census-defined area, and including a broad range of demographic characteristics, the development of the item pool for the WCI (the original set of items prior to items being eliminated due to low frequency, poor factor loadings, etc.) was based primarily on previous speculations and clinical observations about coping processes. Furthermore, Folkman and Lazarus revised the WCI using a specific population (college students) coping with a specific stressor (an exam). It is not likely that this strategy would adequately represent the coping efforts of other populations dealing with other problems. In order to obtain a comprehensive item pool, researchers would have to sample empirically across a range of factors thought to related to coping. There already exists theoretical and empirical support for the existence of several such factors. Several authors (Goldstein, 1973; Krohne, 1978; Pearlin et al., 1981) have found evidence that characteristics of people reporting coping may impact on the type of coping reported. In addition, there is theoretical and empirical support for the idea that the characteristics of various situations affect the type and amount of coping that is observed (McCrae, 1984). Menaghan's (1983) review echoes this sentiment: "Since different kinds of coping are appropriate for different kinds of stressors, no single form of coping may be the most useful as a mediator of stressful outcomes. Future research might profit from dividing stressors into losses, threats, and challenges and examining coping mechanisms that are used particularly with each group" (Menaghan, 1983a, p. 928). Thus, sampling across the range of factors thought to relate to coping is critical to the content of the final item pool. Future studies should examine the domain of coping derived from different subject samples. The result may be several coping instruments specifically tailored to particular groups of subjects. Measurement of Coping The WCI's dichotomous response format assumes that use of all the coping items is equivalent and that the number of scale items used represents the extent of a person's coping. Thus the scores reflect only the diversity of coping efforts. For example, a person who uses three PF coping items briefly would have a higher score than someone who used a single PF item for an extended period of time. The differential time spent on the activities is entirely lost. It is apparent that the dichotomous scheme found in the WCI not only eliminates information about the coping process, but it may also distort our understanding of how a person is coping with a problem. The R-WCI improves upon the WCI's response format in that a four-point
24
Personal Coping
"extent of usage" key is employed. However, neither the dimension nor the standard by which subjects are to judge "extent of use" is clear. In other words, it is not clear whether "extent" refers to frequency, duration, or intensity of use of a particular coping strategy. It is also not apparent what subjects are to use as a basis of comparison in judging "extent" for a particular strategy. Possibilities include: how much they used this strategy prior to the stressor, how much they used other strategies during the stressor, how much they used this strategy during previous stressors, and how much they think other people would use this strategy under these circumstances. As with the original Ways of Coping Checklist, subjects' scores are strongly influenced by the variety of coping strategies which they endorse. For a particular scale, a subject who uses four strategies "somewhat" would have a higher score than a subject who uses one strategy "a great deal." Future research is needed to determine the basis of responses to the R-WCI and whether the assumptions of the scaling process are valid. Opportunities to Cope Another issue is raised by the fact that the WCI is typically used to ask subjects about the problems occurring during the past month. This means that the problem to be coped with could have occurred from one to thirty-one days prior to reporting. Certainly, someone whose problem occurred two days ago would have less opportunity to employ various coping strategies than someone whose problem occurred thirty days earlier. This suggests that some form of adjustment for coping "amounts" be computed based on the time that a person has had to cope with the problem. The WCI and R-WCI do not have such an adjustment, a condition that is likely to cause distorted coping reports and to bias coping's relationships with other measures (such as appraisals, emotions, and health outcomes). The solution to this problem appears relatively straightforward—assess the opportunity to cope—and future research should consider this issue. Validity of Monthly Reports An important assumption about the retrospective report of coping as usually measured by the WCI is that it accurately portrays the day-to-day coping with the reported problem. People do not always remember events and their actions veridically, as has been well demonstrated in the life events and hassles literature (Jenkins et al., 1979; Uhlenhuth et al., 1977). Daily reports of coping may be a more accurate means of measuring coping since they are completed while the event and coping are fresh in the subject's mind and because they are less subject to memory distortions such as selective memory and failures of memory (Stone & Neale, 1984). Future research is needed to examine the correspondence between daily reports of the coping process and the summarized versions of the month's coping as is typically measured by the WCI.
Conceptual and Methodological Issues
25
Coping Remembrance Lazarus (1984) specifically discriminates between coping functions and coping outcomes. He states: "If progress is to be made in understanding the relationship between coping and outcome, that is, what helps or hurts the person and in what ways, coping must be viewed as efforts to manage stressful demands regardless of outcome" (p. 140). Yet, with Folkman and Lazarus' usual methodology, subjects' reports of their coping efforts are likely to be confounded with their knowledge of encounter outcomes. The WCI is typically used to ask subjects to report on how they coped with an event which occurred over the past month. With these instructions, at least some subjects will report on events that have already been concluded and whose final resolution is known. Knowledge of how successfully a stressful event was resolved could profoundly bias both how subjects label or describe their coping efforts as well as which efforts they recall ("effort after meaning;" Brown & Harris, 1978). Truly perspective designs eliminate this bias and are needed in future coping research. Appropriateness of the Use of Factor Analytic Methodologies for Scale Construction in the Assessment of Coping In the development of the WCI and the R-WCI, factor analytic techniques were used to construct scales from the domain of coping items that had been established in previous research. The rationale for using factor analytic techniques to construct scales was largely developed in the personality and attitude assessment areas, and these methods have yielded many instruments with good psychometric characteristics. Thus, this appeared to be a natural direction for coping researchers to follow. However, there are several characteristics of factor-based scales that may render them undesirable for use in the coping area. First, factor-based scales use factor loading as an inclusion-exclusion criterion. For the purposes of coping assessment, the practice of eliminating items which do not load highly on any one factor or which load on multiple factors may be problematic. In attitude assessment, items are generated to represent particular opinions and as many items as are needed can be generated by rephrasing the same basic statement (as in the construction of parallel forms). There is no defined set of items that represent the opinion's domain, and including five items on a short version of the test versus twenty items on a long version presents no problem other than one of decreased reliability for the shorter version. The item domain is basically different for coping responses; eliminating half of the items representing a coping scale may affect not only reliability but also scale validity, because the domain of coping responses suffers. An analogous situation occurs in the domain of certain diagnostic entities, for example, depression. Several symptoms (crying, dysphoria, hopelessness, etc.) define the domain of depressive symptomatology and are, in some inventories (the Beck Depression Inventory or the Hamilton Rating Scale), represented by single items. It would be unwise
26
Personal Coping
to eliminate half of the items on the scale because the content of the scale would be compromised. Yet this occurs when factor analytic methods are applied to the WCI items. Items are eliminated when they load poorly on factors or load on multiple factors. Second, factor-based scales assume interitem covariation, yet this assumption has been questioned in the literature by the developers of one of the coping checklists. "Researchers should recognize that psychometric procedures such as internal consistency and factor-analytic techniques may have only limited usefulness in evaluating the adequacy of measures of coping. These techniques assume positive interitem covariation on similar coping responses, even though the successful use of one response in a domain may effectively reduce stress and thus lessen the utilization of other responses in that domain" (Billings & Moos, 1981, p. 225). We agree with this assessment and wonder what effects the reliance on factor analysis in the development of coping scales has had on the validity of the resulting inventories. Third, factor-based scales emphasize items loading on a single factor. This does not take into account either multiple functions or situationally specific functions of coping strategies. Although the WCI divides coping categorically into problem-focused and emotion-focused functions, in practice the distinction is often less clear. In research from our laboratory (Stone & Neale, 1984), we found that subjects viewed particular items as representing more than one coping scale. For example, we found that an item such as "Went out shopping" could represent the concepts of handling the problem directly (e.g., if the problem to be coped with was the need for a gift) or distracting oneself from the emotional distress caused by the problem, and that the particular view of the item could depend upon the context of the problem. Using factor-based scales, if an item happens to serve multiple functions, it will load on more than one scale (say, direct action and distraction). Then, it may be eliminated by the investigator as not being "pure," implying that the shopping did not constitute any form of coping. Furthermore, McCrae (1984) notes that combining widely different coping mechanisms into summary scores such as PF and EF might obscure that "five percent of the i t e m s . . . apparently reflected a sensitivity to the situational context of the vignette" (p. 225). Folkman and Lazarus (1980) themselves, in explaining why undergraduates' classification of WCI items did not always agree with the items' scale assignment, suggest that whether a specific item is problem- or emotion-focused is dependent upon the situation. Given this statement, one would expect that Folkman and Lazarus would define the function of each item only in terms of the particular situation under consideration or, alternatively, would ask individuals to appraise the function of the coping strategy themselves. However, their studies typically contain an amalgam of different stressful events, and the ways in which people coped with these events are divided on a purely factor-analytic basis, without reference to the particular situation.
Conceptual and Methodological Issues
27
Establishment of the Construct Validity Attempts to establish the construct validity of the WCI have focused on nomological validity (Sechrest, 1984). That is, various studies have attempted to examine the relationship between coping as assessed by the WCI and other variables which theoretically should be associated with coping. These studies have taken two approaches, corresponding to two theoretical assumptions about coping: (1) People who differ on some criterion variable will use different coping strategies; (2) People who use different coping strategies will differ on some criterion variable. If these theoretical differences are evident when coping is assessed by the WCI, it is suggestive evidence that the WCI is measuring some aspect of coping. Strategy One: Group differences and coping. Several correlational studies have attempted to show that groups of subjects characterized by different amounts of current psychological symptoms also differ in the type of coping that they use, as assessed by the WCI. Differences in WCI scores for a recent or current stressful episode have been found for depressed versus nondepressed subjects (Folkman & Lazarus, 1986), medical students in group therapy versus those not in group therapy (Vitaliano et al., 1985), and patients with panic disorder versus those with simple panic or no panic (Vitaliano et al., 1987). However, one study (Coyne et al., 1980) failed to find differences in coping strategies between depressed and nondepressed subjects. Strategy Two: Coping differences and outcome. The relationship between coping as assessed by the WCI and various outcome variables has been examined in a number of studies. Outcome variables have included emotions, psychological adjustment or symptomatology, and recovery from surgery. Two studies have found a relationship between WCI scores and changes in emotions during the course of a stressful encounter. Folkman and Lazarus (1985) found that, controlling for variance due to grade received, appraisal and coping scores on the R-WCI accounted for 28 percent of the variance in students' positive and negative emotions after they received their exam grades. In another study, involving both a younger and an older community resident sample, Folkman and Lazarus (1988b) found that several coping strategies, as assessed by the R-WCI, were associated with changes in emotion (reported retrospectively) over the course of a stressful encounter. The significant coping strategies were planful problemsolving, positive reappraisal, confrontive coping, and distancing. Three studies have found a relationship between WCI scores and psychological symptomatology or adjustment. Folkman et al. (1986) found that the R-WCI scores of community residents describing the most stressful event of the past week were significantly related to their ratings of current psychological symptoms as well as their ratings of satisfaction with encounter outcomes. Using three separate samples consisting of medical students, spouses of Alzheimer's patients, and psychiatric outpatients, Vitaliano et al. (1985) found that WCI scores per-
28
Personal Coping
taining to a current serious stressor were significantly related to depressive symptomatology. One intervention study assessed the relationship of scores on a modified version of the WCI to subsequent surgery recovery. Martelli et al. (1987) attempted to manipulate the coping of patients about to undergo preprosthetic surgery by randomly assigning them to either a mixed-focused stress management, problemfocused, or emotion-focused intervention. Differences in problem-focused coping on the modified WCI, in conjunction with treatment, were the best predictors of pain response to surgery. Evaluation of the construct validity of the WCI. The best evidence offered thus far for the construct validity of the WCI comes from correlational studies in which higher levels of use of certain types of coping strategies are associated with higher or lower levels of psychological symptoms. However, because all of these studies are correlational, they do not establish that coping in a certain way inexorably leads to particular outcomes. The alternative explanations that anxiety and depression cause use of coping strategy or that stressfulness of the encounter determines both coping strategy and symptomatic status have not been conclusively ruled out in studies using the WCI. Furthermore, the most basic question of construct validity has not been addressed: Do people who indicate on the WCI that they do a lot of coping strategy X actually do a lot of coping strategy X? In other words, are we measuring what we think we're measuring? Surprisingly, although this question is standard from a psychometric viewpoint, it has received little attention in the coping literature. Research is needed to examine concurrent validity, both convergent and divergent, as well as predictive validity. Studies examining convergent validity should compare WCI self-reported responses to other types of measures of coping, such as spouse reports or objective behavioral indices (e.g., time spent watching TV, number of trips to the library). Studies examining divergent validity are necessary to insure that the WCI is not tapping some personality dimension, psychological adjustment, or social desirability. The potential influence of social desirability considerations on subjects' responses to coping questionnaires is suggested by Wortman's (1983) research on victimization. There is "a great deal of evidence to suggest that when they interact with others, victims may conceal their emotional distress" (p. 216) and "it is important to determine whether [statements about their cognitive coping strategies] are really believed by the victims who make them, or whether victims come to learn that optimistic statements will make them more attractive to others" (p. 208). CONCLUSIONS Research on how people cope with problematic situations has progressed rapidly during the last two decades. Several new paper and pencil self-report questionnaires have been developed and presented in the literature, and they have renewed interest in the effects of coping. These assessments have enabled
Conceptual and Methodological Issues
29
researchers to incorporate the coping concept into a wide variety of studies and literally hundreds of reports have now been published using these questionnaires. Whereas some of the issues raised here may be remedied rather simply, many of the issues present serious threats to the validity of the information gathered with these methods. We questioned the assumptions of their development, means for summarizing the data collected with instruments, and the quality of the raw information collected. Although we cannot be sure about the seriousness of any particular criticism, we believe that given the number of potential problems with the questionnaires, at least some of the issues have a high probability of invalidating the assessments. Certainly, additional research is called for to confirm or disconfirm these issues. Rather than simply continuing to assess coping with currently available measures, we suggest that it is time to take stock of the methods themselves. As a moderator of the stress-health link, coping may turn out to be an extremely important construct. But much hard work lies ahead if this potential is to be realized. Until the answers to some of the questions are in, considerable caution is needed in interpreting the results from these coping assessments.
This page intentionally left blank
3
Making the Case for Coping Susan Folkman
Despite the impression conveyed by the media that stress is a killer, not everyone becomes sick, depressed, or dysfunctional when faced with stress. In fact, most people manage to maintain reasonable health and functioning. Researchers since the 1970s have tried to explain this observation by examining personality dispositions (e.g., Kobasa, 1979; Scheier & Carver, 1985), biological vulnerabilities (e.g., Murphy & Moriarity, 1976), environmental conditions (e.g., Cohen, Evans, Stokols, & Krantz, 1986), and historical phenomena (e.g., Elder, 1974). The literature indicates that all these factors probably matter to one degree or another. The observation that not everyone succumbs to stress has also been responsible for the burgeoning interest in coping. During the 1960s and 1970s a number of books and articles appeared that examined coping from theoretical and conceptual perspectives (e.g., Coelho et al., 1974; Haan, 1977; Lazarus, 1966; Lazarus & Launier, 1978; Mechanic, 1962; Menninger, 1963; Miller, 1980b; Pearlin & Schooler, 1978; Vaillant, 1977). These writings set the stage for the development of new assessment techniques (e.g., Carver, Scheier, & Weintraub, 1989; Folkman & Lazarus, 1980; McCrae, 1982; Miller, 1987; Moos, 1974; Pearlin & Schooler, 1978; Stone & Neale, 1984; Vaillant, 1977), which in turn led to empirical studies in behavioral medicine, health psychology, life-span psychology, clinical psychology, developmental psychology, anthropology, sociology, social welfare, and nursing.
Portions of this chapter appear in S. Folkman (1991), "Coping across the life span: Theoretical issues," in E. M. Cummings, A. L. Greene, and K. H. Karraker (eds.), Life-span developmental psychology: Perspectives on stress and coping, pp. 3-19. Hillsdale, New Jersey: Lawrence Erlbaum Associates. Reprinted by permission.
32
Personal Coping
Despite the steady increase in coping research, we still know remarkably little about the extent to which coping accounts for individual differences in response to stress and the mechanisms through which coping mitigates the harmful physical and psychological effects of stress. Three issues need clarification if we are to advance our understanding of the role coping plays in stress processes: the theoretical models that underlie current assessment procedures and the ways they shape the data used in empirical studies, the evaluation of coping, and social aspects of coping. ASSESSMENT MODELS Most coping assessments are based on one of three theoretical models: the ego-psychology model; the trait/dispositional model; and the contextual model. The assumptions of these models differ in important ways and lead to different portrayals of the coping process. Further, each of these models presents obstacles for measurement that can constrain understanding of the mechanisms through which coping affects outcomes (see also Lazarus & Folkman, 1984). The Ego-Psychology Model The ego-psychology model of coping is based on the concept of defenses, that is, unconscious adaptive mechanisms that are a major means of managing instinct and affect. Vaillant's (1977) assessment scheme provides one of the best illustrations of this approach. Vaillant arranges defense mechanisms hierarchically according to their maturity. The highest or most mature of these mechanisms consist of adaptive processes such as sublimation, altruism, suppression, and humor. The next lower group consists of neurotic mechanisms, including intellectualization, repression, reaction formation, displacement, and dissociation. Next are the immature mechanisms, including fantasy, projection, hypochondriasis, passive-aggressive behavior, and acting out. And finally, the least mature are the psychotic mechanisms, including denial of external reality, distortion, and delusional projection. Ego-psychology models have also been proposed by Haan (1977) and Menninger (1963). The term "coping," when used in the context of these models, refers to the most advanced or mature of the ego processes. The ego-psychology model has three limitations. First, the model does not easily take into account the nature of the stressful circumstances with which the person is dealing, or the timing of the coping/defense process within a given stressful event. Instead, ego processes are usually judged on an a priori basis as inherently adaptive or maladaptive, mature or immature. Yet an immature strategy, such as denial, can at times be highly adaptive (Lazarus, 1983; Meyerowitz, Heinrich, & Schag, 1983), which Vaillant notes, and a mature strategy, such as humor, can be maladaptive if used inappropriately. Second, coping defined in terms of defense processes emphasizes tension reduction and the restoration
Making the Case for Coping
33
of equilibrium and gives only minor attention to the problem-solving functions of coping. Maintaining emotional equilibrium is an important function of coping, but so is problem solving. Third, the rating of an ego mechanism is made in the context of a stressful transition or stressful event by a trained clinical interviewer. These ratings depend on inference based on incomplete information, and satisfactory interrater reliability is often difficult to achieve (Morrissey, 1977). The ratings are also likely to be confounded with the outcome in that they are frequently made on the basis of the information about how well the person functions or the person's general lifestyle (e.g., Vaillant, 1977). The Trait/Dispositional Model Within the trait/dispositional model, coping is defined as a personality trait such as repression-sensitization (Byrne, 1964), coping-avoidance (Goldstein, 1973), or monitoring-blunting (Miller, 1987). The traits are usually assessed through self-report questionnaires. The assumption underlying this model is that coping, as a personality variable, influences behavior in a wide range of situations. Thus, it is assumed that an individual's coping behavior in diverse circumstances can be predicted from the individual's score on a measure of a coping trait or disposition. Unfortunately, measures of coping traits and dispositions are generally not predictive of how a person copes in an actual, naturally-occurring, stressful event (Cohen & Lazarus, 1973; Kaloupek, White, & Wong, 1984). An exception is reported by Miller (1988), who found that the Miller Behavioral Style Scale, a measure of monitoring and blunting, predicted information seeking and avoidance in experimental conditions. Further, trait measures of coping tend to be unidimensional, usually along an approach-avoidance continuum. Coping, however, has been shown to be a multidimensional phenomenon (Aldwin & Revenson, 1987; Billings & Moos, 1984; Carver et al., 1989; Felton & Revenson, 1984; Folkman, Lazarus, Dunkel-Schetter, et al., 1986; McCrae, 1982; Stone & Neale, 1984; Vitaliano, Russo, Carr, Maiuro, & Becker, 1985). This multidimensional ity cannot be described with unidimensional measures. Trait measures, therefore, are generally inadequate with respect to describing the richness and complexity of actual coping processes. The Contextual Model Within the contextual model, coping is assessed in relation to specific stressful conditions (e.g., Pearlin & Schooler, 1978) or specific stressful situations (e.g., Aldwin & Revenson, 1987; Billings & Moos, 1981; Felton & Revenson, 1984; Folkman & Lazarus, 1980, 1985; Folkman, Lazarus, Dunkel-Schetter et al., 1986; Forsythe & Compas, 1987; Stone & Neale, 1984). The assumption underlying the contextual model is that coping thoughts and acts are influenced by
34
Personal Coping
the relationship between the person and the environment in a given stressful encounter. The version of the contextual model with which I work is based on a cognitive model of stress (Lazarus, 1966; Lazarus & Folkman, 1984). Within this model, coping is defined as the changing thoughts and acts that an individual uses to manage the external and/or internal demands of a specific person-environment transaction that is appraised as stressful (Lazarus & Folkman, 1984). Three features distinguish this definition of coping from the definitions offered by the ego-psychology and trait/dispositional models. First, coping is defined as what the person actually thinks and does to manage the demands of a particular encounter; no attempts are made to infer unconscious processes. Second, coping processes are not hierarchically arranged on the basis of maturity or efficacy. No coping strategy is evaluated on an a priori basis, but rather on contextual criteria, which I shall describe later. Third, the definition points to coping as a changing process, rather than a stable feature of personality. Since this model serves as the basis of my discussion, I shall summarize it here. The coping process begins with a person's cognitive appraisal of a personenvironment relationship. The appraisal includes an evaluation of the personal significance of the encounter (primary appraisal) and an evaluation of the options for coping (secondary appraisal). In primary appraisal the person asks "What do I have at stake in this encounter?" and in secondary appraisal the question is "What can I do?" (Lazarus, 1966; Lazarus & Folkman, 1984). Together, primary appraisal and secondary appraisal shape emotion quality and intensity and influence the coping response. Coping processes continuously change as a function of ongoing appraisals and reappraisals of the shifting person-environment relationship. Shifts may be due to coping efforts directed at changing the environment or the meaning or understanding of the event. Shifts may also be the result of changes in the environment that are independent of the person. Any shift leads to a reappraisal of the situation, which in turn influences subsequent coping efforts. Thus, coping changes as an encounter unfolds and from encounter to encounter. Within the contextual model, coping is viewed as having two major functions: the management of the problem (problem-focused coping) and the regulation of emotion (emotion-focused coping). These two functions have been recognized by a number of coping researchers including Kahn, Wolfe, Quinn, Snoek, and Rosenthal (1964), Mechanic (1962), and Billings and Moos (1981). Several broad conceptual categories of coping have been suggested in addition to the problem-focused and emotion-focused distinction. Pearlin and Schooler (1978) speak of responses that change the situation out of which the stressful experience arises, responses that control the meaning of the stressful experience after it occurs but before the emergence of distress emotion, and responses that control distress emotion. Folkman and Lazarus (1988a) refer to cognitive activity that influences the deployment of attention, cognitive activity that alters the subjective meaning or significance of the encounter for well-
Making the Case for Coping
35
being, and actions that alter the actual terms of the person-environment relationship. Additional distinctions have been made at the empirical level. Pearlin and Schooler (1978), for example, defined seventeen types of coping processes, some of which are limited to specific role areas such as family, marriage, parenting, and work. McCrae (1982) defined twenty-eight types of coping, and Stone and Neale (1984) defined eight types, including distraction, situation redefinition, direct action, catharsis, acceptance, seeking social support, relaxation, and religion. Folkman, Lazarus, DeLongis, et al. (1986) also defined eight types including intrapersonal processes that are directed at problem solving and the regulation of emotion (distancing, self-control, positive reappraisal, and escapeavoidance) and interpersonal problem-focused processes (confrontation) and emotion-focused processes (the seeking of informational and emotional support). Carver et al. (1989) defined thirteen scales, five measuring distinct aspects of problem-focused coping, five measuring aspects of emotion-focused coping, and three that the authors say "arguably are less useful." The contextual model is not without its problems. To have an effect on outcomes of importance such as psychological well-being, physical health, and social functioning, there has to be some stability in coping processes over occasions and time. Contextual assessments of coping have shown that people vary their coping from context to context, depending, for example, on whether the event is a harm, loss, or threat (McCrae, 1984), the social role that is involved (Menaghan, 1982), environmental and social factors (Parkes, 1986), and what is at stake and what the options for coping are (Folkman & Lazarus, 1980, 1985; Folkman, Lazarus, Gruen, & DeLongis, 1986). The challenge is to identify stable aspects of the coping process, which can be done by repeatedly assessing coping across contexts and over time. This approach has led to findings that indicate some stability in at least some types of coping. For example, in their study of community residents, Folkman, Lazarus, Gruen, and DeLongis (1986) found that average intraindividual autocorrelations on eight types of coping over five encounters ranged from .17 (for seeking social support) to .47 (for positive reappraisal). An autocorrelation of .47 represents a relatively high degree of stability. Dolan and White (1988) assessed stability in terms of the rank order of use of eight different types of coping. They asked twenty-five professional and academic females and twenty-eight undergraduate males to report how they coped with daily stressful events each day for about two weeks. Intraindividual Kendall's coefficients of concordance among the females, based on an average of twenty-five events, ranged from .06 to .52, with an average of .27. Among the males the coefficients of concordance ranged from .07 to .63, with an average of .36, based on an average of twenty-eight events. Using a different approach, Carver et al. (1989) examined the correlations between scores from a situational version and a dispositional version of the Coping Orientations to Problems Experienced (COPE). Of the fourteen scales that they assessed, the correlations
36
Personal Coping
from four scales (the emotional social support scale, the focus venting emotion scale, the mental disengagement scale, and the alcohol/drug use scale) were greater than .35. It is important when doing research on the relationship between coping and various outcomes that the sample of coping be reliable. Often, studies of this question use estimates of coping that are based on one or two samples of a person's coping processes in specific contexts, which do not provide a reliable estimate of the individual's stable patterns of coping across contexts, let alone over time. To expect a small, unreliable sample of coping to predict an outcome such as depression or anxiety is analogous (though in the opposite direction) to expecting a broad personality disposition to predict how a person would behave in an isolated instance. A second set of problems has to do with the multidimensional nature of coping that is revealed through the contextual model. It is difficult to deal empirically and conceptually with multiple types of coping. To date most researchers have treated the various kinds of coping separately, without regard to their patterning at a given time. A few investigators have made limited attempts to address this problem. Vitaliano, Maiuro, Russo, Katon, DeWolfe, and Hall (1990) used profile analysis to graph scores for five coping scales. Although this approach is conceptually appealing, the analysis, which requires that the profiles be compared through adjacent paired contrasts, is cumbersome to report and interpret. Billings and Moos (1981) and Folkman and Lazarus (1980) looked at the relative amounts of problem-focused and emotion-focused coping in specific stressful events. But reducing coping to these two categories tends to mask the rich variety of coping strategies that they subsume. A challenge is to find the critical number of types of strategies to evaluate. Two are too few, but eight may be too many (cf. Stone et al., 1988) because they create an overwhelming number of combinations of coping. The problem of how to summarize coping patterns awaits a creative solution. The sequencing of coping over time must also be examined. In a metaanalysis, Suls and Fletcher (1985) evaluated the effects of two sequences: approach followed by avoidance, and avoidance followed by approach. They found that when studies were classified according to the temporal interval between stressor onset and measurement of adaptation, for the most part avoidance fared better than attention. However, as the interval between stressor onset and measurement of adaptation grew, attentional strategies tended to be associated with more positive adaptation. Suls and Fletcher's findings point up the importance of looking at the sequence of coping over time (see also Horowitz, 1976, 1983; Lazarus, 1983), but their conceptualization of coping along approach-avoidance dimensions (see also Roth & Cohen, 1986) leads to the same problem of oversimplification that exists when we reduce coping to problem-focused and emotion-focused dimensions. In the studies my colleagues and I have reported, two types of avoidant coping, escapeavoidance and distancing, are differentially related to adaptive outcomes, as are
Making the Case for Coping
37
two forms of approach coping, planful problem-solving and confrontation (Folkman, Lazarus, Dunkel-Schetter, et al., 1986; Folkman, Lazarus, Gruen, et al., 1986). These important distinctions within approach and avoidant types of coping are lost if all coping is reduced to the approach-avoidance dimension. Nevertheless, Suls and Fletcher's work illustrates why we need to understand the temporal ordering of coping. Despite the limitations posed by the contextual model, it appears to be the most useful of the three models for fully describing coping processes and for addressing basic questions about how coping influences the relationship between stress and outcomes of importance such as mental and physical health, and social functioning. The contextual model has, in fact, become dominant in contemporary coping research. EVALUATING COPING It is not enough simply to describe coping. We must also confront the issue of determining effective versus ineffective coping. The determination of effective and ineffective coping presents important choices for the researcher. Two general models for evaluating coping have evolved. In the first, which I call the outcome model, coping is judged by its effects on outcomes of importance. In the second, which I call the goodness-of-fit model, attention is given to the quality of coping, regardless of its outcome. The Outcome Model The basic assumption of the outcome model is that the quality of a coping process is evaluated according to its effect on an outcome of importance. Put simply, a coping strategy that promotes a favorable outcome such as good job performance, positive morale, or good health, is considered an effective coping strategy, whereas a coping strategy that has the opposite effect is considered an ineffective coping strategy. The ability of this model to estimate the effects of coping on outcomes depends a great deal on the choice of outcome. The selection of an inappropriate outcome can lead either to an underestimation or overestimation of coping effects. Unfortunately, outcomes are often chosen almost by rote, without much thought as to their appropriateness. Two characteristics in particular need to be considered: the proximity of the outcome and its relevance. Proximity. In general, the more proximate an outcome is in relation to a specific coping process, the greater will be our confidence that coping is causally related to that outcome. The more distal an outcome is, the more difficult it is to rule out the causal contributions of cofactors. This issue can be illustrated with emotions, which are a proximate outcome to a stressful encounter, and mood, which can be thought of as a more distal outcome. The ways a person copes with an immediate problem, such as a parent-child relationship or a pressing
38
Personal Coping
deadline on the job, should have a direct effect on his or her emotions at that time. For example, coping accounted for changes in positive and negative emotions during stressful encounters that were reported by community residents (Folkman & Lazarus, 1988a). When participants in this study used problemfocused forms of coping their emotional state improved, whereas when they used confrontive coping their emotional state deteriorated. The most parsimonious explanation for these changes, and the explanation that made most conceptual sense, was that coping, and not cofactors such as aging, the environment, or physical health, accounted for the changes in emotion. Unfortunately, though the use of proximate outcomes may yield information about cause-effect relationships in which we have some confidence, these relationships are not the ones that interest most researchers, because they do not tell us much about the person's mental or physical health. Anxiety or sadness, for example, that are experienced with great intensity in the context of a specific encounter may be entirely appropriate to the encounter. Even when inappropriate, such expressions do not necessarily signify that the person is an anxious or distressed person. (See Thoits, 1984, for an excellent discussion of the meaning of emotion in coping processes.) The case for coping is more difficult to make with mood, which is less immediate to a specific encounter than is emotion and thus is a less proximate and more distal outcome. Mood is a background affective state that is determined by multiple factors such as what went on during the previous week, other events of the day, or the anticipation of upcoming events. Thus, the way a person copes with a specific stressful encounter is but one determinant among many of the person's mood, which means that causal statements about coping in a given stressful encounter and mood cannot be made with great confidence. Stone, Helder, and Schneider (1988) present a thoughtful discussion of this problem. They followed 79 couples for an average of 100 days. Subjects completed questionnaires on a number of variables including how the subject coped with the day's most bothersome problem and daily mood. The investigators chose next-day mood as an outcome. However, they point out that what is defined as next-day mood is simultaneously the same-day mood for the next day and is thus likely to be influenced both by previous-day coping and present-day coping. Establishing causality with outcomes that are more distal than mood, such as depression and anxiety, is even more difficult. For example, several researchers have found a relationship between escape-avoidance and psychological distress (Billings & Moos, 1984; Coyne, Aldwin, & Lazarus, 1981; Aldwin & Revenson, (Billings & Moos, 1984; Coyne, Aldwin, & Laz 1987; Folkman, Lazarus, Gruen, et al., 1986; Manne & Sandier, 1984; Silver, Auerbach, Vishniavsky, & Kaplowitz, 1986; Vitaliano, Katon, Russo, Maiuro, Anderson, & Jones, 1987). What cannot be determined from these findings is whether escape-avoidance is a symptom of psychological distress, causes psychological distress, or is in a recursive relationship with distress. Further, the more distal an outcome is, the more likely it is to be affected by variables such
Making the Case for Coping
39
as environmental events, genetic predispositions, or aging, and the more difficult it is to track the effects of coping. Another problem with distal outcomes concerns their tendency to be stable over long periods. General health status variables, for example, tend to be quite stable in the general population. The probability that such variables might change during a period that is arbitrarily chosen for study is not great. Kasl (1983) illustrates this point: "taking a sample of blue collar workers (35 years and older) and following them for 5 years in order to discern the mental health impact of a boring, monotonous job may miss the phenomenon altogether: the casualties of inadequate adaptation may have disappeared from observation and the remainder have adapted 'successfully' (e.g., giving up on expecting work to be a meaningful human activity), but the costs of such 'successful' adaptation can no longer be reconstructed through the belated follow-up" (p. 90). To manage the problems of distal outcomes, care must be taken to select outcomes that have the potential for changing during the period of study. And predictor variables and outcome variables need to be repeatedly assessed to help detect their true causal sequence (cf. Kasl, 1983) and increase their reliability (cf. Epstein, 1979). The more distal the outcome, the more important these strategies become. Relevance. One of the key conditions for evaluating the effects of coping on an outcome is that the coping variable be independent of the outcome variable (Folkman, 1991). Absence of independence results in confounding between coping and its outcome. An outcome must also be relevant. By relevant, it has to be an outcome that coping can theoretically be expected to affect. We would not, for example, expect the way a person copes with stressful events to directly affect attitudes towards political candidates. Relevance applies to all variables, from the most proximal to the most distal. To achieve relevance, some investigators call for specifying the adaptive tasks that are associated with a particular stressful context. For example, in a threat situation where one has little control over the outcome, such as awaiting news of a biopsy, the regulation of anxiety is a relevant outcome. A relevant outcome in a situation of harm or loss is the restoration of morale or the formation of new goals or commitments. In a highly charged decision-making situation, a relevant outcome is high quality decision making, which includes the management of distress emotions that might cause premature closure on the deliberative process. Menaghan (1983a) notes the comments of several investigators in this regard. She cites Hackman (1970), who states that we must develop a classificatory system for describing task demands; Turk (1979), who suggests that we specify the adaptive tasks posed by an illness and the range of coping options that are available for accomplishing those tasks; and Hirsch (1981), who suggests that we specify particular coping objectives, adaptive tasks, and useful strategies according to criteria of social adaptation for particular subgroups. This strategy has several drawbacks, each of which can be addressed although
40
Personal Coping
probably not eliminated. At a practical level, it is difficult to identify adaptive tasks for diverse contexts. This task can be made less daunting by broadening the definition of adaptive tasks such that they apply to a category of contexts. An example of this approach is provided by Cohen and Lazarus (1979), who identify five adaptive tasks of illness: "(1) to reduce harmful environmental conditions and enhance prospects of recovery, (2) to tolerate or adjust to negative events and realities, (3) to maintain a positive self-image, (4) to maintain emotional equilibrium, and (5) to continue satisfying relationships with others" (p. 232). Similarly, Janis and Mann (1977) identified the conditions, or adaptive tasks, that are required for making good decisions under stressful circumstances. At a conceptual level, the strategy is based on an assumption that a given stressful event will have a normative set of adaptive tasks for a broad range of individuals. However, differences in individual values, commitments, and goals can affect a person's appraisal of the adaptive tasks in a given encounter. Thus, the adaptive tasks involved in coping with being passed over for a promotion may differ for a person who places a high value on status in the work place and a person who places a high value on the relationships he or she has with coworkers in the present job. The possibility that adaptive tasks are not necessarily normative can be explored by asking the individual to specify his or her own goals for that context. Using multiple outcomes. We might assume that successful coping would affect outcomes in similar ways. Effective coping in a specific encounter, for example, ideally should contribute to a desirable problem outcome and reduced distress, and coping over many occasions should contribute to sustained good health, morale, and social functioning. In reality, we cannot assume that coping will have a consistent effect on diverse outcomes. Outcomes from different domains, such as health and morale, that are assessed cross-sectionally may even show contradictory results. For example, consider a person who is advised by his doctor to modify his competitive, hurried behavior in order to help reduce blood pressure. The person may be successful in modifying his behavior and even in reducing his blood pressure, but at the same time he may become demoralized because the modified behavior pattern is inconsistent with, and even contradicts, strongly held values. Or consider the junior high school student who chooses to use drugs in order to gain acceptance with her peers; acceptance may improve morale, but the behavior is likely to cause problems in other areas of functioning including school performance and physical health. There is also the possibility of long-term costs of coping strategies that may appear adaptive in the short-term, which Cohen, Evans, Stokols, and Krantz (1986) refer to as secondary effects of coping. These effects include fatigue and overgeneralization of the strategy to domains where its use is inappropriate. To deal with the problem that coping may effect some outcomes positively and others negatively, researchers may have to rank order the multiple outcomes in a study according to criteria such as the goals of their research, cost-benefit
Making the Case for Coping
41
analyses, or value judgments. Contradictory results can also be used to advantage to rule out alternative hypotheses. For example, consider a psychoneuroimmunological study of stress in which it is hypothesized that increases in stress have an immunosuppressive effect that, in turn, increases host susceptibility to infection. But what if certain coping processes trigger immune responses that increase rather than decrease host resistance? This effect on the immune response is a possible outcome of coping strategies, such as humor, positive reappraisal, or meditation, that induce positive emotion states (e.g., Scheier, Matthews, Owens, Magovern, Lefebvre, Abbott, & Carver, 1989). In such a case, there could be an absence of illness in the presence of psychological stress and distress. Without information about the immune response, the absence of illness would be difficult to interpret. Is it that the psychological stress and/or distress was never severe enough to have an immunosuppressive effect, or is it that the stresstriggered coping process modified distress, increased morale, and enhanced immune functioning? Evidence of the latter possibility would have important implications for preventive interventions. A major potential pitfall of the outcome model that the investigator needs to consider is that it can lead the investigator to equate effective coping with coping that solves problems and reduces distress, without regard for what the person is coping with. Many situations of daily life cannot be mastered. Problems are sometimes insoluble and distress is sometimes intense and not easily regulated. To judge coping by its effect on outcomes may do a great disservice to the efforts that people make to cope with difficult, intractable, and unrelenting conditions of life. Indeed, the presence of distress can indicate that adaptive coping processes are taking place, as when a person confronts a loss and tries to come to terms with it. This pitfall can be avoided to a certain extent by carefully considering the context in which coping is being judged, especially the controllability of the demands with which the person is coping. It is clear that outcomes must be selected with care, and separate hypotheses should be made for each one. If the study is longitudinal, hypotheses should also be made about short-term vs. long-term effects. Despite the complexity that is involved, I strongly urge investigators to make the effort to use multiple outcomes, for only by so doing can we gain a comprehensive understanding of the stress process. The Goodness-of-Fit Model An alternative approach for evaluating coping emphasizes the process rather than the outcome. The central idea of this approach is goodness-of-fit, the appropriateness of the coping strategy given the demands and constraints of the situation. Effective coping depends on two fits: the fit between reality and appraisal and the fit between appraisal and coping (Folkman, 1984; Folkman, Schaefer, & Lazarus, 1979; Lazarus & Folkman, 1984; Menaghan, 1983a). The fit between reality and appraisal refers to the match between what is
42
Personal Coping
actually going on in the person-environment transaction and the person's appraisal of the personal significance of that transaction and his or her options for coping. Serious deviation from veridical appraisals in either of these domains can lead to maladaptive coping. For example, failure to appraise a situation as potentially harmful means that necessary anticipatory coping will not take place, and appraising a situation as threatening when in fact it is benign can lead to unnecessary coping that deflects attention and resources from other more pressing tasks. A person may realistically appraise what is happening, but be unrealistic in his or her appraisal of resources for coping with the demands. An overly pessimistic appraisal of coping resources could result in restricted coping efforts, whereas an overly optimistic appraisal of resources could lead to excessive disappointment and self-blame for the unsatisfactory outcome. The fit between appraisal and coping refers to the fit between situational appraisals of controllability (secondary appraisal) and actual coping processes. In general, problem-focused coping is appropriate in encounters that hold the potential for personal control, whether over the outcome of the particular encounter or its recurrence in the future, whereas emotion-focused coping is appropriate in encounters where there is little the individual can do to control the outcome or recurrence. The issue is more complicated than this prescription suggests because of the interplay between problem- and emotion-focused forms of coping that occurs in most stressful situations. For example, problem-focused coping in the form of an information search is often necessary to determine that a situation is essentially beyond the person's control. Conversely, emotionfocused coping is sometimes needed to facilitate problem-focused coping in encounters that have controllable outcomes. For example, a person who is very anxious may need to reduce anxiety in order to concentrate on the problem at hand. Thus, we should expect that patterns of coping will include both major coping functions, but the relative amounts of problem- and emotion-focused coping strategies should differ depending on secondary appraisals of control (cf. Folkman & Lazarus, 1980). Poor fits between appraisals and patterns of coping should decrease the possibility for the management or reduction of distress, and indeed, could even lead to increased distress. By over-reliance on problem-focused coping in situations that are not controllable, for example, a person remains engaged in a frustrating transaction that is likely to result in increased distress as well as fatigue (Cohen et al., 1986). And by failing to use problem-focused coping in situations that are controllable, a person is likely not to take action that is necessary to manage or resolve the problem (e.g., Katz, Weiner, Gallagher, & Hellman, 1970). The advantage of the goodness-of-fit model compared to the outcome model is that it takes the context into account and avoids some of the problems involved in the selection of outcomes. Given the appraised options for coping, we can evaluate the appropriateness of the coping strategy by determining whether the use of problem-focused and emotion-focused forms varies according to the appraised controllability of the outcome (or some aspect of the outcome). On the
Making the Case for Coping
43
other hand determining the veridicality of appraisals poses problems. If the event or condition is one that is highly stressful or experienced by wide numbers of people, it may be possible to gather enough information to determine whether or not appraisals are reasonably veridical. But many events of day-to-day living do not fall into these categories and we may have to rely heavily on subject selfreport for judgments as to what is going on and what the options for coping are. This reliance on self-report is troublesome. However, the reliability of self-report data can be strengthened by using repeated assessments over time and occasions (Epstein, 1979). Homeostatic models of stress, in which stress is seen as a discrepancy between environmental demands and individual capacities or resources, or between environmental resources and individual needs and goals (French, Rodgers, & Cobb, 1974; Hobfoll, 1988; Menaghan, 1983), are conceptually similar to the goodnessof-fit model in that they focus on processes rather than outcomes. Effective coping reduces the discrepancy and restores homeostasis. However, I am not aware of a method for effectively measuring the discrepancy in naturalistic settings. Although the goodness-of-fit model does not involve reference to outcomes, presumably effective coping (as defined in terms of goodness-of-fit) should more often than not have a positive influence on adaptational outcomes. Forsythe and Compas (1987) tested this hypothesis and found that level of psychological symptoms varied as a function of the match between appraisals of control and coping in life events. The use of relatively more problem-focused than emotionfocused coping in events that were appraised as controllable as compared to events that were uncontrollable was associated with an adaptive outcome; conversely the use of relatively more emotion-focused than problem-focused coping was associated with an adaptive outcome for events that were appraised as uncontrollable as compared to controllable. Vitaliano, DeWolfe, Maiuro, Russo, and Katon (1990) also tested the goodness-of-fit hypothesis in a study of the appraised changeability of a stressor, coping, and depression in people with psychiatric, physical health, work, and family problems. They found support for the hypothesis except in those people with psychiatric conditions. Problemfocused coping and depressed mood were negatively related when a stressor was appraised as changeable but were unrelated when a stressor was appraised as unchangeable, while emotion-focused coping was positively related to depression when a stressor was appraised as changeable. Support for the goodness-of-fit models is also found in two field studies that were not originally designed to examine this hypothesis. Collins, Baum, and Singer (1983) evaluated the use of problem-focused and emotion-focused coping in residents of Three Mile Island following the nuclear accident there and found that problem-focused coping in dealing with the aftermath of the event (which was uncontrollable) was positively associated with psychological symptoms. In a study of a black community in the rural South, Dressier (1985) found that active coping buffered the effects of stressors for females, but not
Personal Coping
44
for males. To explain the gender differences in symptoms of distress, Dressier suggested that the stressors the females were coping with were more controllable than those the males were coping with. The coping efforts of the females were thus more likely to meet with success than were the coping efforts of the males. A Multimethod Approach In some cases the evaluation of coping efficacy may be best served by relying on appropriate outcome measures. In other cases, especially those in which the designation of outcomes presents problems, the goodness-of-fit approach may be more appropriate. Until we have had more experience with each of these approaches, it is a good idea to use both whenever possible. The use of two methods should increase our understanding of coping processes and increase our understanding of the strengths and weaknesses of both approaches. SOCIAL ASPECTS OF COPING Coping literature has developed largely within the tradition of psychology and psychiatry. As a result, much of the literature emphasizes the individual and his or her cognitive and emotional processes. However, coping rarely takes place in a social vacuum; most stressful events of daily living involve other persons. Coping must therefore be viewed within a social context and as part of a dynamic social process. Curiously, only recently have the links between coping and social processes begun to be considered. In a seminal paper that links coping and social support, Thoits (1986) discusses social support as coping assistance. She points out that coping and social support have a number of functions in common. These include instrumental functions, which in social support terms include tangible assistance and aid, and in coping terms include problem-focused coping; emotional functions, which in social support terms include emotional support and in coping terms include emotionfocused coping; and perceptional support, which in social support terms includes informational support that alters perceptions of meaningful aspects of stressful situations and in coping terms includes cognitive reappraisal or restructuring. Thoits hypothesizes that the same coping methods utilized by individuals in response to their own stressors are also the methods that they apply to others as assistance (see also Thoits, 1984). Thoits' purpose in linking coping and social support is to explicate how and when support attempts work. Until now, research investigating the buffering effects of social support has focused primarily on establishing that there is a buffering effect. By looking at social support as coping assistance, Thoits provides a model for testing hypotheses about the conditions under which various types of support are used and the conditions under which they may or may not be effective in reducing distress and solving problems.
Making the Case for Coping
45
A second approach, which has come primarily from the coping literature, emphasizes social support as part of contextual coping processes. Pearlin and Schooler (1978), for example, assessed advice seeking as an aspect of the coping process in their community study. The Ways of Coping, which has now been used by a number of investigators, assesses the seeking of emotional and informational support during specific stressful encounters (e.g., Folkman & Lazarus, 1985; 1988b). Viewing social support contextually, as part of the coping process, leads to questions about the personal and environmental conditions under which social support is offered and received (e.g., Dunkel-Schetter et al., 1987), changes in the type of social support that is sought as a stressful encounter unfolds (e.g., Folkman & Lazarus, 1985), and the timing of social support (Jacobson, 1986). Such questions are not likely to arise when social support is assessed as a stable feature of the social environment. One of the most interesting observations that comes from conceptualizing social support as a dynamic process is that people do not necessarily perceive social support as helpful (Davidowitz & Myrick, 1984; Lehman et al., 1986; Wortman & Lehman, 1985). Whether or not social support is helpful depends at least in part on the efficacy of the person who is seeking the support as well as the efficacy of the person who is providing it. With respect to the person seeking social support, the effectiveness of the perceived support involves the support seeker's social skills (e.g., Cohen et al., 1986), other coping strategies that the support seeker uses (Dunkel-Schetter et al., 1987), and personal outlook (Vinokur, Schul, & Caplan, 1987). Overall, the more socially skillful a person is, the more rational the person is, and the more positive his or her outlook, the more likely he or she is to elicit social support. With respect to the person providing support, Lehman et al. (1986) suggest that support is most successfully provided when the provider's anxieties are minimal. Thoits (1986) suggests that sociocultural and situational similarity enhance effective support because they increase the likelihood of perceiving and receiving empathic understanding. Cohen and his colleagues (Cohen & McKay, 1984; Cohen & Syme, 1985b) state that social support is most effective when there is a match between the needs elicited by the stressful event and the functions of support that are perceived to be available. This concept is close to the goodness-of-fit model described above, though I would place less emphasis on the stressful event per se and more emphasis on the needs of the person so that the match, or fit, would be between the type of support that the support seeker needs and the type of support the provider is comfortable offering. I have two reasons for making this slight adjustment. First, a given stressful event might create the need for instrumental support in one person and emotional support in another person, depending on individual experience, resources, plans, and goals. Focusing on the event rather than the person's appraisal of his or her needs vis a vis that event decreases the likelihood of capturing these individual differences in social support needs. Second, certain types of support, especially emotional or appraisal support, may be needed as a consequence of a series of events or a diffuse condition of living
46
Personal Coping
rather than as a consequence of a specific event. These needs are more likely to be uncovered by asking about the person's support needs at a given time than the needs associated with a specific event. The general rule for increasing the probability of receiving effective social support, I suggest, is that the support seeker know the preferred styles of his or her support providers and, whenever possible, seek support from a person whose preferred style matches his or her needs. A mismatch not only will not produce beneficial effects, it might even make things worse. For example, consider a woman who turns to her husband for comfort and understanding following a difficult day at work. It happens that the husband is more comfortable providing advice and solutions to his wife's problems than he is providing comfort and understanding. Upon offering advice and solutions (with the best of intentions), he is surprised tofindthat his wife becomes irritated and impatient. She needed emotional support but received informational support. She would have been better off turning to someone, such as a friend, who may not have been as close to her as her husband, but who might have been much better at providing emotional support. Despite the widespread belief that coping is an important variable in explaining why people differ in their physical and emotional responses to stressful events, the empirical evidence is still relatively scant. The case for coping will ultimately rest on our ability to prove that coping influences the stress-outcome relationship. In this chapter I attempted to clarify issues having to do with coping assessment and evaluation to help the case-building process. I want to note that these aspects cannot be treated independently. An evaluation of coping that is valid and theoretically relevant is dependent on a clear understanding of the theoretical model that underlies the conceptualization and assessment of coping. Coping is a complex phenomenon. Coping research is no less so.
4
Coping with Psychosocial Stress: A Developmental Perspective Bruce E. Compas, Vanessa L. Malcarne, Bruce E. Compas, Vanessa L. Malcarne,
Gerard A. Banez
Facilitating adaptive coping in children, adolescents, and adults depends in part on adequate knowledge of how coping skills develop across the life span. Therefore, an important step in understanding the nature of the coping process is to examine this process from a developmental perspective. When considering the emergence of coping resources, skills, and styles, three questions are of central importance: How do children learn to cope? How does coping change with development? How is continuity in coping reflected throughout development? The purpose of this chapter is to examine these questions and related issues regarding the development of coping. First, our conceptualization of coping is briefly outlined, drawing primarily on work with adult populations. Second, recent studies that have examined coping in children and adolescents are summarized and evaluated. Third, a variety of developmental factors that may be related to coping are described, emphasizing the importance of cognitive-social learning theory as a useful model for understanding the development of coping. Fourth, developmental studies related to two types of coping, social problemsolving and cognitive distraction, are examined in greater detail. Finally, directions for future research are described. CONCEPTUALIZATION OF THE COPING PROCESS Discussions of the coping process have originated from several theoretical perspectives, including animal experimentation (e.g., N. E. Miller, 1980), ego psychology (e.g., Haan, 1977), and cognitive-transactional models (e.g., Lazarus & Folkman, 1984; Moos & Billings, 1982). The few attempts to conceptualize coping during childhood and adolescence have drawn on these broader theoretical models (e.g., Cummings & Cummings, 1988; Maccoby, 1983; Mur-
48
Personal Coping
phy, 1974; Murphy & Moriarity, 1976). Although a complete discussion of the many issues in conceptualizing the coping process is beyond the scope of this chapter, we will highlight the central issues here (see Lazarus & Folkman, 1984, for a more detailed discussion). We have drawn extensively on the conceptual framework of Lazarus and Folkman in our research on stress and coping in children and adolescents (e.g., Compas, 1987a, 1987b). Since their definitions of stress and coping reflect the interdependence of these two constructs, it is important to begin by defining each. Lazarus and Folkman define stress as "a particular relationship between the person and the environment that is appraised by the person as taxing or exceeding his or her resources and endangering his or her well-being" (1984, p. 19). Thus, whether an individual perceives that he or she is faced with a stressful situation depends in part on his or her perceptions of coping resources. They go on to define coping as "constantly changing cognitive and behavioral efforts to manage specific external and/or internal demands that are appraised as taxing or exceeding the resources of the person" (Lazarus & Folkman, 1984, p. 141). In discussing coping from a developmental perspective, it is important to highlight a number of the critical issues and problems that researchers and theorists concerned with adult coping have already confronted. These concerns include: (1) distinguishing coping as an effortful process of adaptation; (2) defining coping as a process as opposed to a trait; (3) distinguishing between coping and psychological outcomes; (4) distinguishing among coping resources, strategies, and styles; and (5) identifying the different functions served by coping. Each of these issues will be discussed here only briefly; readers are referred elsewhere for more extensive discussions (e.g., Lazarus & Folkman, 1984; Menaghan, 1983; Moos & Billings, 1982). Coping as Effortful Responses to Stress When confronted with stressful circumstances, individuals will respond in a variety of ways, some of which reflect purposeful actions and others that are representative of innate reflexes and overlearned, automatic responses. If the concept of coping included all of these adaptive responses, it would be so overly broad that it would become useless in describing human functioning under stressful circumstances (cf. Lazarus & Folkman, 1984). Murphy and her colleagues (Murphy, 1974; Murphy & Moriarity, 1976) addressed this issue in their work on children's coping by placing coping at the middle of a continuum ranging from reflexes that are present at birth to automatized mastery responses that have been learned to the extent that they are no longer under conscious control. A similar distinction between controlled and automatic processing of information has been made in basic learning models (e.g., Shiffrin & Schneider, 1977). In fact, under conditions of high stress some forms of automatic responding may break down, making effortful coping responses all the more important (cf. Nor-
A Developmental Perspective
49
man, 1976). Although distinguishing among these various types of adaptive responses presents a difficult challenge to researchers, it is important to make this distinction in order to avoid a conceptualization of coping that is too broad to be useful. Coping as a Process Stressful encounters in people's lives cannot be adequately reflected by the notion of discrete events. Rather, stressful encounters unfold over periods of time, some lasting only a few moments and others stretching over periods of months or years. Coping is a process that extends over the course of a stressful episode, often beginning in an anticipatory phase prior to the occurrence of the "stressful event" and continuing until some sense of resolution is achieved. Although it is likely that individuals develop preferred styles of coping with stress, recent evidence indicates that individuals change the ways they cope during the course of a single stressful episode and across different stressors (e.g., Compas, Forsythe, & Wagner, 1988; Folkman & Lazarus, 1985). Conceptualizations of coping that focus on enduring personality traits that are consistent over time and/or across situations cannot capture the changing nature of the coping process. Coping and Psychological Outcomes Coping is frequently equated with successful adaptation. However, a given coping strategy will not necessarily be equally effective in a variety of different stressful situations or at different points during the coping process. That is, the effectiveness of a particular coping strategy cannot be determined independently of the context in which it is used. Coping effectiveness depends on both the realities of the stressful situation in which a strategy is used and on the individual's cognitive appraisals of the situation. For example, Forsythe and Compas (1987) and Compas, Malcarne, and Fondacaro (1988) found that the effectiveness of coping strategies intended to change a stressful situation and those intended to regulate distressing emotions during a stressful encounter varied as a function of the perceived controllability of the stressor. Specifically, efforts to change the stressor were associated with fewer psychological symptoms when the stress was perceived as controllable, and attempts to palliate negative emotions were associated with fewer symptoms when the stressor was perceived as uncontrollable. Coping Resources, Strategies, and Styles The coping process can be viewed on at least three levels (Menaghan, 1983a). First, attention must be paid to the personal and social resources an individual has available to cope with stressful circumstances. Personal resources include
50
Personal Coping
perceptions of personal competence and efficacy, problem-solving skills, and a sense of optimism. Social resources include the availability of a supportive and helpful social network that can provide both practical and emotional support during stressful encounters. Second, coping is most often understood in terms of the specific cognitive and behavioral strategies that individuals use in attempting to handle stressful circumstances and their emotional reactions to these circumstances (see below). Finally, coping styles reflect consistencies in the strategies used by an individual over time in a single stressful episode and/or across different stressors. However, coping styles are not the same as personality traits, as styles may reflect preferred ways of coping under certain circumstances rather than a consistent style independent of situational demands. Further, individuals may vary in the degree of consistency they display in coping, with some individuals repeatedly using the same strategies and others varying their coping with the demands of the situation (Compas, Forsythe, & Wagner, 1988). Functions of Coping Coping strategies can serve a wide variety of functions, including changing an aspect of a stressful environment or distracting one's attention from the stressful situation. Researchers have attempted to classify the various functions of coping at both macro- and micro-analytic levels. Macro-level analyses have typically distinguished between what Lazarus and colleagues (e.g., Lazarus & Folkman, 1984; Lazarus & Launier, 1978) have termed problem-focused coping (efforts to act on the stressful situation) and emotion-focused coping (efforts to regulate the emotional states associated with or resulting from the stressor). Micro-level analyses have led to the identification of a variety of subtypes of coping, with little agreement on the nature of these subtypes. For example, Stone, Helder, and Schneider (1988) distinguished among social support, information seeking, religiosity, situation redefinition, avoidance, tension reduction, and problem solving. Alternatively, Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen (1986) identified eight types of coping: confrontive coping, distancing, self-control, seeking social support, accepting responsibility, escapeavoidance, planful problem-solving, and positive reappraisal. Although the finegrained analyses of subtypes of coping will probably be necessary to understand the subtleties of the coping process, there is little agreement about an appropriate set of categories at this time. In all likelihood, the structure of coping subtypes will vary as a function of the characteristics of the sample (e.g., level of development, gender) and types of stress being studied (e.g., chronic vs. acute). STUDIES OF COPING IN CHILDREN AND ADOLESCENTS Although research conducted from a variety of different traditions has been concerned with coping processes in children and adolescents (Compas, 1987a), we will limit our discussion here to the few studies that we are aware of that
A Developmental Perspective
51
have examined the broad categories of problem-focused and emotion-focused coping in response to a wide range of stressors in these age groups (Asarnow, Carlson, & Guthrie, 1987; Band & Weisz, 1988; Compas, Malcarne, & Fondacaro, 1988; Glyshaw, Cohen, & Towbes, 1989; Tolor & Fehon, 1987; Wertlieb, Weigel, & Feldstein, 1987; Wills, 1986). (See Cummings and Cummings, 1988, for a discussion of this model regarding children's coping with a specific type of stress, adults' angry behavior). The findings of these studies generally support the usefulness of Lazarus and Folkman's model in describing coping strategies used by children and adolescents. For example, in three studies, raters were able to achieve adequate reliability in classifying children's open-ended reports of coping as representing either problem-focused or emotion-focused coping (Band & Weisz, 1988; Compas et al., 1988; Wertlieb et al., 1987). There is also some evidence that children and adolescents vary the use of problem- and emotion-focused coping as a function of situational factors and cognitive appraisals. Band and Weisz (1988) found that children used more problem-focused (primary) coping with school failure and more emotion-focused (secondary) coping with medical stressors. The authors hypothesized that this difference occurred because school failure may have been perceived as more controllable than medical stressors. Consistent with this hypothesis, Compas, Malcarne, and Fondacaro (1988) found that children and adolescents generated more problem-focused coping alternatives for dealing with controllable than with uncontrollable stressors. Evidence distinguishing among coping resources, strategies, and styles has been more rare. In an effort to distinguish between coping resources and strategies, Compas, Malcarne, and Fondacaro (1988) asked children and adolescents to report both the different ways that they could have coped (i.e., coping resources) as well the ways they actually coped (i.e., coping strategies) with recent stressors. Resources and strategies were highly related with each other and both were related with mothers' reports and children's self-reports of emotional/behavioral problems. With regard to styles or patterns of coping, this sample of children and adolescents displayed moderate levels of consistency in their reports of coping with two different types (academic and interpersonal). Wills (1986) reported higher consistency in adolescents' reports across five different types of stress on a measure designed to assess their intentions in coping (Stone & Neale, 1984). Thus, evidence regarding the level of cross-situational and temporal stability in children's and adolescents' coping remains unclear. Studies to date have not examined coping processes in children and adolescents. Instead, measures have focused on types of coping aggregated over the course of a single event (Band & Weisz, 1988; Compas, Malcarne, & Fondacaro, 1988; Wertlieb et al., 1987), how children and adolescents typically cope with stress (Glyshaw et al., 1989; Wills, 1986), or how they would cope in response to a set of hypothetical stressful situations (Asarnow et al., 1987; Tolor & Fehon, 1987). Changes over the course of a stressful episode have not been studied. These studies have, however, provided some evidence on the association
52
Personal Coping
between various types of coping strategies and adjustment in children and adolescents. Compas, Malcarne, and Fondacaro (1988) found that problem-focused alternatives and strategies were associated with lower levels of maternal and self-reported emotional/behavioral problems, whereas emotion-focused alternatives and strategies were positively associated with maladjustment. In contrast, Wertlieb et al. (1986) did not find an association between problem- or emotionfocused coping and mothers' reports of children's behavioral problems. Using a micro-analytic system for classifying coping, Glyshaw et al. (1989) and Wills (1986) found that several subtypes of coping were related to adolescents' selfreported depression, anxiety, and substance abuse. For example, Glyshaw et al. (1988) found that decision making, peer support, and parental support were negatively related to depression and anxiety. In spite of the contributions made by these studies, they all share two major limitations. First, they have not examined how coping strategies change with development during childhood and adolescence. Several of the studies have examined changes with chronological age and some consistent findings have emerged. For example, it appears that the use of emotion-focused coping increases during late childhood and early adolescence (Band & Weisz, 1988; Compas et al., 1988; Wertlieb et al., 1986). However, chronological age cannot be equated with developmental level and, to date, no studies have examined basic aspects of development as they relate to coping. A second limitation of prior studies is that they have failed to shed any light on how children learn to cope. Although authors have speculated about the mechanisms involved in the development of coping (e.g., Band & Weisz, 1988), no studies have examined the direct processes by which children learn to cope with different types of stress.
DEVELOPMENTAL FACTORS RELATED TO COPING Understanding the development of coping resources, strategies, and styles presents researchers with the forboding task of integrating the effects of biological, cognitive, social, and emotional development on the coping process. Further, it is important to recognize that just as coping processes are affected by aspects of development, the ways in which a child learns to cope are likely to influence his or her cognitive, social, and emotional development. In this section, we will briefly describe some of the implications of biological, cognitive, social, and emotional development for understanding the issues outlined above, followed by a discussion of cognitive-social learning theory (e.g., Bandura, 1981, 1986b; Mischel, 1981, 1984) as an integrative perspective for studying the development of coping. Biological Development Children bring into the world a number of innate biological characteristics that are likely to have a bearing on how they learn to cope with stress, foremost
A Developmental Perspective
53
among which is temperament. As discussed elsewhere, a child's temperament may influence his or her degree of responsivity to the environment and, thus, the number and types of situations that he or she experiences as stressful (Compas, 1987a). Further, temperament may affect the child's actual coping responses to stress. Temperamental factors lie at one end of a continuum of adaptative responses to stress, representing reflexive response patterns. Recent evidence indicates that early temperamental characteristics are predictive of coping styles in adulthood, indicating a high degree of continuity in the coping styles of at least some individuals (Steinberg, 1985). Biological development is also related to the experience and expression of emotion . For example, Izard (1978) posited that the development of various biological mechanisms, including specific elicitor-receptor connections, the specificity of CNS receptors, and the connections between receptors and states and between states and expressions, influences the process of emotional development. Similarly, Maccoby (1983) hypothesized that maturation of the nervous system contributes to children's increasing ability to inhibit crying and frustration reactions, and to maintain behavioral organization. It will be important to understand how biological factors are related to the ways children learn to cope with their own emotions and the emotional reactions to stress exhibited by others in their environment. The work of Lewis and Michalson (1983) suggests that a full understanding of these processes will require an examination of the interactions between various biological factors and social-environmental forces.
Cognitive Development Lazarus and Folkman's (1984) theoretical model of stress and coping heavily emphasizes the role of cognitive processes in determining what is experienced as stressful and how one copes. In applying this model to children and adolescents, it is important to account for changes in cognitive functioning with development. Researchers must go beyond chronological age and assess children's levels of cognitive functioning, as these two variables cannot be considered equivalent. An example of the effects of early cognitive changes on stress and coping processes can be observed in studies of attachment and separation anxiety in infants (Bowlby, 1969, 1973). Prior to the age of approximately eight months, infants typically do not exhibit signs of distress at the departure of the caregiver. However, during the period of approximately eight to eighteen months infants show significant upset and distress over the absence of the primary caregiver. Researchers have hypothesized that prior to age eight months infants lack the cognitive skills to recognize the caregiver as a separate individual and are unable to cognitively understand the meaning of her or his departure. Our point is that cognitive processes are probably involved in the appraisal of and coping with stress from a very early point in development. Although many adaptive responses
54
Personal Coping
in early development may reflect innate response patterns, volitional coping responses are likely to emerge with very early advances in cognitive functioning. Other changes in cognitive functioning which occur later in development are also likely to affect how children and adolescents cope. For example, the development of metacognitive functioning (e.g., Flavell, 1981) is hypothesized to allow children to think about and analyze cognitive coping strategies. We will discuss this process as it relates to the use of cognitive distraction as a coping strategy in more detail in a later section. Social Development Siblings, friends, parents, and other adults serve as sources of social support for children and adolescents in the process of coping with stress (e.g., Cauce & Srebnik, 1989; Sandier, Miller, Short, & Wolchik, 1989). However, in light of changes in the nature of family and peer relationships with development, it is likely that the nature of social support processes also changes with development. For example, infants' initial distress responses serve the function of eliciting assistance from caretakers. As such, infants are embedded in a network of supportive relationships that are essential in managing the stresses of early life. Parents remain significant sources of support for coping with stress throughout childhood and adolescence and peers emerge as an increasingly important resource for coping during childhood. Whether the utilization of supportive peer relationships in coping with stress increases in a linear fashion throughout childhood and adolescence, or there are qualitative changes in support processes with the onset of adolescence warrants further research. Emotional Development Given that one of the central tasks of coping involves managing one's emotions in stressful situations, it will be important to understand how children's knowledge, experience, and expression of emotions change with development, and how these changes relate to the development of coping. Specifically, given that one of the central features of coping involves emotional regulation, children may be responding to very different experiences of emotions and possess different capacities to express emotions as a part of the coping process at different points in development. For example, Radke-Yarrow (1986) reported that children's reactions to emotional stimuli become increasingly differentiated and more sophisticated with age. Thus, the tasks involved in emotion-focused coping are also expected to become more complex with development. Cognitive-Social Learning Theory: Implications for the Development of Coping In light of the wide variety of factors that may relate to coping, a broad-based theory is needed to integrate these various influences in understanding the de-
A Developmental Perspective
55
velopment of coping. We have selected cognitive-social learning theory (e.g., Bandura, 1981,1986b; Mischel, 1973,1984) for this purpose for several reasons. First, cognitive-social learning theory is generally consistent with the cognitivetransactional framework of Lazarus and Folkman (1984) which has guided adult research on coping. Second, cognitive-social learning theory has been applied to a number of aspects of development (e.g., self-control, perceptions of selfefficacy) that are relevant to the study of coping. Third, this theory is concerned with the learning and acquisition of behaviors and thoughts, and thus, it is relevant to the question of how children learn to cope. A detailed description of cognitive-social learning models is well beyond the scope of this chapter. In this section, we will highlight three specific aspects of these models that have particular relevance to understanding the development of coping: observational learning processes, cognitive processes, and transactions between person and situation variables. Observational Learning. Although many actions and thoughts are learned through direct experience, the process of human development would be extremely inefficient if it were dependent solely on the effects of an individual's own actions. Much or most human behavior is learned through observation of behavior modeled by others (Bandura, 1986) and observational learning is likely to play a central role in the acquisition of cognitive and behavioral coping strategies. We do not, however, mean to imply that children will simply mimic important and salient models in their social environment. Observational learning is a complex process which includes verbal modeling of thought processes, abstract modeling of generative and innovative behavior, and conceptual learning of rules and principles that underlie complex behavior patterns (Bandura, 1986). The types of behaviors and cognitions that a child is capable of learning through observation are affected by her or his level of cognitive, social, emotional, and biological development. Observational learning processes affect the entire stress and coping process, from learning to perceive situations as stressful to the acquisition of coping strategies and styles. The work of Mineka and her colleagues on the acquisition of fears in rhesus monkeys is instructive in this regard (e.g., Mineka, 1985; Mineka & Cook, 1987). These researchers found that adolescent and adult rhesus monkeys rapidly acquired an intense fear of snakes by observing wild-reared adult monkeys behaving fearfully in the presence of snake stimuli (e.g., Mineka, Davidson, Cook, & Keir, 1984). However, a subsequent study showed that prior exposure to a model behaving nonfearfully with snakes effectively immunized monkeys against the subsequent effects of exposure to models behaving fearfully with snakes (Mineka & Cook, 1986). Further, early experiences of control and mastery over the environment have been shown to be related to the later development of adaptive, active coping behaviors (Mineka, Gunnar, & Champoux, 1986). Although findings obtained in animal experimentation cannot be directly generalized to human populations, these results suggest that vicarious learning experiences play a role in determining the types of stimuli and situations that
56
Personal Coping
will be experienced as stressful, and in acquiring adaptive patterns of responding in the face of potentially stressful stimuli. Cognitive processes. Since many important coping strategies involve thoughts rather than overt behaviors, a conceptual model used to explain the development of coping must be able to account for the acquisition of cognitive skills. From a cognitive-social learning perspective, a variety of cognitive competencies and skills are seen as integral parts of adaptive functioning. The competencies emphasized in this model have direct relevance for understanding the development of coping. For example, perceptions of personal efficacy reflect an important resource for the coping process (Bandura, 1981). Self-efficacy beliefs influence the degree to which situations are experienced as threatening or stressful and whether and with what intensity coping behaviors are enacted. A more detailed discussion of a second example of cognitive skills, those that are useful in effectively delaying gratification and in coping with frustration and threat, is provided below. Person-situation transactions. Finally, cognitive-social learning models assume that the learning and acquisition of new behaviors and cognitions are the result of ongoing processes of mutual influence between the person and the environment. This assumption is reflected in Bandura's (1978, 1986b) notion of reciprocal determinism in which "behavior, cognitive and other personal factors, and environmental influences all operate interactively as determinants of each other" (1986b, p. 23). This reciprocal perspective is consistent with the cognitive-transactional model of stress and coping of Lazarus and Folkman (1984), and thus, offers a useful model for elaborating the development of coping. Within the stress and coping paradigm, a number of reciprocal relations among components of the model have been delineated. For example, the relation between emotions and coping is a bidirectional process, beginning with a cognitive appraisal of a stressful situation which leads to an emotion (Folkman & Lazarus, 1988d). The appraisal and initial emotion then lead to a coping response, which alters the stressful relationship between the person and the environment, resulting in a reappraisal and an altered emotional state. Over extended periods of time, stressful events and circumstances in people's lives lead to the development and maintenance of psychological symptoms, which in turn lead to increases in stress (e.g., Compas, Howell, Phares, Williams, & Giunta, 1989; Compas, Wagner, Slavin, & Vannatta, 1986). TWO EXAMPLES OF COPING: SOCIAL PROBLEMSOLVING AND COGNITIVE DISTRACTION Our goals in the preceding sections were to highlight a variety of developmental factors that could be related to coping and to propose that cognitive-social learning theory is a viable model for guiding investigations in this area. In this final section we will discuss research concerned with two specific types of coping, social problem-solving and cognitive distraction, which reflect some of these
A Developmental Perspective
57
developmental principles and can guide investigations of the development of other types of coping. Social Problem-Solving Social problem-solving (SPS) is an example of problem-focused coping with one type of stressful event, interpersonal or social stressors (e.g., Rubin & Krasnor, 1986; Spivack & Shure, 1982; Weissberg, Caplan, & Sivo, 1989). SPS refers to a cognitive-behavioral process of identifying and implementing adaptive means of responding to problems encountered in interpersonal situations. Component SPS skills include: (1) the ability to recognize problems; (2) the capacity to generate alternative solutions; (3) the ability to conceptualize relevant means to a goal; (4) the capacity to anticipate consequences of solutions; (5) the capacity to anticipate consequences of interpersonal acts; (6) the ability to perceive the feelings and perspectives of the people involved; (7) the capacity to implement appropriate solutions with behavioral skill; and (8) the ability to self-monitor behavioral performance and to modify solutions and goals when necessary (Dodge, 1986; Rubin & Krasnor, 1986; Spivack & Shure, 1982; Weissberg, Caplan, & Sivo, 1989). SPS is an effortful process that requires hypotheticalreflective thinking and represents problem-focused coping in the context of social situations that are appraised by the child as taxing or exceeding his or her resources. Development of SPS skills. A major limitation of early SPS research was a lack of attention to the normative development of SPS skills. Although early investigations of the behavioral correlates of SPS skills provided evidence that at least some SPS ability is evident by the early childhood period, studies designed to provide normative data on the development of SPS skills have emerged only recently. For example, Krasnor (1982) and Rubin and Krasnor (1986) found that children as young as four years of age consider situational factors such as the age or sex of the people involved when conceptualizing interpersonal goals and the means to meet those goals. In addition, results reported by Levin and Rubin (1983) and Rubin and Krasnor (1986) suggest that the ability to generate alternative means to solve interpersonal problems which have met with initial failure increases with age. Unfortunately, at present, there are more questions than answers regarding the normative development of SPS skills (Rubin & Krasnor, 1986). The relative significance of different SPS skills for social adjustment at different age levels has also received attention. Spivack, Platt, and Shure (1976) theorized that an SPS skill can only emerge as significant for adaptive interpersonal functioning in an age group that has the cognitive capacity to exhibit the skill. For example, the ability to generate alternative solutions is believed to be a developmental precursor to other SPS skills such as means-ends and consequential thinking, which are presumed to require cognitive skills that tend to be nonexistent in early childhood. In support of this perspective, Spivack et al.
58
Personal Coping
(1976) examined the relation between SPS skill and social adjustment at different age levels. As predicted, the most significant SPS skill in the early childhood period is that of being able to generate alternative solutions to interpersonal problems. In a series of studies comparing samples of well-adjusted preschool children to matched samples of preschoolers exhibiting emotional and behavioral problems, alternative thinking was the only SPS skill found to discriminate the two groups. In the middle childhood years, the ability to generate alternative solutions to problems remains a dominant SPS skill, but it is joined by the theoretically more "advanced" skill of means-ends thinking. By adolescence, means-ends thinking appears to be the best predictor of social adjustment. The generation of alternative solutions remains important, and the spontaneous consideration of consequential pros and cons in problem situations begins to take on significance. Parental factors and children's SPS skills. Numerous parental and familial factors have been theorized to correlate with and to facilitate the development of children's SPS abilities (Elias, Ubriaco, & Gray, 1985; Rubin & Krasnor, 1986; Spivack, Platt, & Shure, 1976). Parents' SPS skills, their values and expectations, and childrearing styles have been the focus of the research to date. Parental SPS skills and ability may have a prominent role in the development of SPS ability in children. Spivack, Platt, and Shure (1976) found that mothers' means-ends thinking skill was significantly related to their daughters' sensitivity to problems and alternative thinking skills. No relation, however, was found between mothers' SPS skills and their sons' SPS ability, leading Spivack et al. to hypothesize that maternal SPS ability differentially affects the ways in which mothers respond to daughters and sons. In addition, they suggested that the development of SPS ability in boys may be more strongly related to fathers' SPS skills. The beliefs, values, and expectations held by parents are also likely to have important roles in the development of SPS ability (Elias & Ubriaco, 1986; Elias, Ubriaco, & Gray, 1985; Pettit, Dodge, & Brown, 1988). Elias and Ubriaco (1986) postulated that parental beliefs about how children learn are related to both parental teaching strategies and children's SPS skills. Specifically, parental beliefs are thought to influence the ways in which parents interact with their children, which, in turn, affect children's SPS ability. Findings from case studies of four families provide partial support for this conceptualization of the role of parental beliefs (Elias, Ubriaco, & Gray, 1985). In addition to their beliefs, the values and expectations of parents concerning social behavior are also likely to contribute to the development of SPS skill. For example, Pettit, Dodge, and Brown (1988) found that deviant maternal values (e.g., endorsing aggression in interpersonal problem situations) and expectations (e.g., making hostile attributions about their child in hypothetical contexts) were related to lowered levels of SPS abilities in children. Pettit et al. (1988) indicated that these deviant parental qualities may be transmitted through verbal means, affecting the way children learn to process social information in peer settings. Finally, parental childrearing style, broadly defined, may have a major influ-
A Developmental Perspective
59
ence on the development of SPS ability. Spivack, Platt, and Shure (1976) identified three specific childrearing qualities believed to enhance the development of SPS ability: (1) extracting from the child his or her own thinking; (2) willingness and ability of the parent to act as a catalyst, model, or guide in the child's attempts at problem solving in social situations; and (3) parental support when the child manifests SPS skills and makes decisions based upon them. Parental verbal instructional styles and broader parenting style variables (e.g., Baumrind, 1971; Hoffman, 1983) are also likely to have a significant role in facilitating the development of SPS skills (Elias, Ubriaco, & Gray, 1985; Rubin & Krasnor, 1986). At present, little data exist to support or contradict the significance of these parental childrearing factors in the development of SPS abilities. In the study noted earlier, Elias, Ubriaco, and Gray (1985) reported that parents most effectively elicited means-end verbalization through direct questioning. They indicated, however, that the more general ways in which parents socialize their children for competence may be more important for coping and adjustment than the specific skill training or reinforcement techniques that they use. In sum, several studies to date have identified parental and familial factors related to SPS skill in children, but no information is available regarding the processes by which these factors facilitate or hinder SPS skill development. The most frequently theorized mechanisms include verbal instruction, modeling, inducing or encouraging the use of SPS skills, and direct reinforcement of SPS performance. Clearly, additional research is needed to determine which parental and familial factors are most predictive of SPS skill in children and to clarify the processes by which these factors facilitate SPS skill development. SPS training programs. A number of training programs have been developed to teach SPS skills to children and adolescents (see reviews by Denham & Almeida, 1987; Kirschenbaum & Ordman, 1984; Pellegrini & Urbain, 1985; Spivack & Shure, 1982; Weissberg & Allen, 1986). SPS programs have used a wide variety of instructional methods (e.g., verbal instruction, modeling, coaching, self-instructional training, role plays, direct reinforcement) in the effort to promote SPS skill improvements which are generalized and maintained (Pellegrini & Urbain, 1985; Spivack & Shure, 1982; Weissberg, Caplan, & Sivo, 1989). The many teaching strategies and techniques used in these programs are consistent with the cognitive-social learning model described earlier. At present, the extent to which the cognitive-social learning model adopted in these training programs approximates the natural process of SPS skill acquisition or development remains unclear. However, the positive findings of programs based on a cognitive-social learning theory suggest that it may have considerable utility in explaining the development of SPS and other psychosocial coping skills. Cognitive Distraction Distraction is one example of the large group of cognitive strategies included under the rubric of emotion-focused coping (Lazarus & Folkman, 1984). By
60
Personal Coping
distracting themselves from various aspects of a stressor, children can potentially reduce their emotional distress in a stressful encounter. Two research literatures, one focusing on children's abilities to delay gratification in the face of frustration and one focusing on children's abilities to cope in the face of threat, speak to the potential importance of distraction as a coping strategy for regulating emotional response to a stressor (Miller & Green, 1985). Distraction and coping with frustration. In a series of studies spanning fifteen years, Mischel and his colleagues (see Mischel, 1983, for review) have studied children's abilities to effectively self-impose delay of gratification. Mischel and his colleagues presented children with reward choices (e.g., two marshmailows versus one) in which the larger reward could only be obtained by delaying gratification for a period of time, while the smaller reward could be obtained immediately. This paradigm creates frustration for children, who must resolve the conflict between the desire for immediate versus preferred but delayed rewards. The conflict and resultant frustration inherent in the delay paradigm appear conceptually similar to those found in some daily stressors. Therefore, a developmental understanding of how children learn and use strategies to cope with the frustration in the delay-of-gratification paradigm should prove useful to understanding children's emotion-focused coping on a broader level as well. The work of Mischel and his colleagues has demonstrated that children's attention to presented awards can affect their ability to self-impose delay of gratification. In an early study, Mischel and Ebbesen (1970) found that delaying gratification was significantly more difficult for children when reward objects were visible. Further, children who delayed most effectively in the presence of the rewards used a variety of techniques to distract themselves from the rewards and the waiting process (e.g., singing, covering their eyes). The researchers concluded that for these children, feelings of frustration were alleviated somewhat through distraction and thus waiting was facilitated. Mischel and his colleagues reasoned that if distracting one's thoughts from desired rewards eases the aversiveness of the situation by decreasing frustration, then giving children distraction techniques to use should help them cope more effectively with frustration and facilitate delay of gratification. In a series of studies (Mischel, Ebbeson, & Zeiss, 1972), children who were given specific instructions on ways to distract themselves from visually present rewards (i.e., by thinking about "fun" or happy things) waited longer than children instructed to think about the rewards. Although this finding did not hold when the rewards were absent from view, Mischel et al. concluded that this suggests that children can more easily generate and apply their own distraction strategies when rewards are visually absent, while specific instructions about distraction strategies appear to be particularly helpful when the reward is visually present. Overall, it appears that children who distracted themselves from the rewards may have experienced less frustration in the waiting process and were thus better able to delay gratification. Further studies by Mischel and his colleagues have shown that distracting
A Developmental Perspective
61
children from attending to certain aspects of rewards facilitates waiting. Mischel (1973) found that when children's attention was distracted from the consummatory aspects of the rewards and onto the reward's nonconsummatory, more abstract qualities, children were able to delay longer and more easily. It appears from these results that distraction from the more arousing aspects of rewards may decrease frustration and thus make delay less challenging. In sum, this work supports the notion that use of distraction can help children cope with frustration, at least as presented in the delay-of-gratification paradigm. The findings raise the question of how children's understanding and use of distraction as a coping strategy develops. In a 1983 study, Mischel and Mischel examined the development of children's metacognitive understanding of strategies for coping with frustration in the delay paradigm. Their results showed that children do not appreciate the benefits of hiding rewards from view until about age six; before that age children either show no preference or actually prefer exposure to the rewards. Interestingly, some children younger than six spontaneously suggest distraction as an effective strategy for coping with delay, but seem to believe, erroneously, that they can successfully distract themselves from visible rewards (see Mischel, 1974, 1983). By age six, children appear to have developed the understanding that it is preferable to distract themselves (i.e., wait with rewards covered rather than exposed) and to avoid consummatory in favor of task-oriented ideation. In children just a little older (about age eight), Mischel and Mischel (1983) found a shift toward suggestion of task-oriented rather than distraction strategies, and hypothesized that there might also be a shift in the efficacy of delay strategies at this time. As yet, however, the differential effectiveness of cognitively available strategies for delay, and specifically of distraction, at various points in development has yet to be established. By the time children reach sixth grade, they appear to recognize that abstract ideation is facilitative of waiting, something younger children do not seem to understand. Some tentative findings (Mischel & Mischel, 1983) suggest that older children may have more alternative strategies available to them, and thus when asked may suggest a more varied approach to the problem, ultimately perhaps using distraction strategies less overall than somewhat younger children, despite their greater understanding and appreciation of the efficacy of such strategies. The research findings on strategies for effective delay of gratification presented by Mischel and his colleagues are made particularly provocative by their recent findings that the ability to effectively delay gratification for the sake of larger rewards is predictive of more general cognitive-social competence ten years later (Mischel, Shoda, & Peake, 1988). Mischel et al. found that the number of seconds preschoolers delayed when they participated in delay-of-gratification studies in the 1960s and 1970s was significantly related not only to concurrent parental ratings of competence (Mischel, 1983), but to parental ratings of competence and coping skills when the preschoolers had reached adolescence ten years later. Preschoolers who were able to more effectively utilize strategies
62
Personal Coping
such as distraction to cope with the frustration of the delay-of-gratification paradigm were reported by their parents as using more effective strategies for coping with the stresses of adolescence. Mischel et al. (1988) suggest that these findings argue for the existence of relatively stable personal competencies and that future research should attempt to identify important determinants and correlates of these competencies. To date, however, little research has examined or attempted to identify factors that influence development of coping strategies such as distraction. Distraction and coping with threat. In their review of children's coping with threat from impending medical and dental procedures, Miller and Green (1985) describe similarities between the attentional strategies children use to cope with the frustration involved in delaying positive outcomes (as in the delay-ofgratification paradigm) and the attention strategies they use to cope with the threat of impending negative outcomes. Distraction has been identified as an important contributor to the effectiveness of filmed peer modeling in the reduction of distress in children about to undergo medical procedures. Further, investigators have demonstrated the efficacy of cognitive-coping interventions that teach children to distract themselves from threat-relevant cues as well as interventions that distract children from threatening cues by providing sensory/experiential information (Miller & Green, 1985). Miller (1992) suggests that there may be important contextual links between the process of coping with threat and coping with frustration during delay of gratification. In both cases, it appears that distraction from the more threatening/ frustrating aspects of uncontrollable stressful experiences reduces emotional distress in children. Recent work by Miller, Greene, and Mischel (1988) showed that six-year-old children who effectively used distraction to cope with the frustration of waiting for deferred rewards were also able to effectively employ distraction in the face of threat. However, little is known about how distraction strategies are learned or how their use develops.
DIRECTIONS FOR FUTURE RESEARCH The study of the development of coping skills across the life span represents a major priority in future stress and coping research with children, adolescents, and adults. We have highlighted a number of issues which are important when considering coping from a developmental perspective. Existing studies of coping in children and adolescents have provided some evidence of age-related changes in the ways in which youngsters cope with stressful events. No research, however, has examined the development of coping skills and styles per se. We have proposed that cognitive-social learning theory presents a viable model for integrating the many developmental factors which are believed to influence how children learn to cope. Future studies of coping need to examine the utility of this theory in understanding at least three major developmental issues. First, future research on the development of coping needs to examine how
A Developmental Perspective
63
coping strategies change with development during childhood, adolescence, and adulthood. Studies of two particular types of coping, social problem-solving and cognitive distraction, have provided evidence that the development of these coping strategies is associated with changes in cognitive development. Future investigations need to examine how changes in cognitive and other aspects of development relate to the development of the broad categories of problemfocused and emotion-focused coping, as well as more specific types of coping, such as situation redefinition and tension reduction. A related issue is the relative significance of both broad and specific categories of coping for emotional/behavioral adjustment at different stages of development. Second, research is needed to clarify the processes whereby children learn how to cope with stressful experiences. From a cognitive-social learning perspective, one factor of particular importance is the salience of different models in the acquisition of coping skills across the life span. Clearly, parents and other family members represent significant coping models throughout childhood and adolescence. As children grow older, however, peers may assume increased salience as models for coping with stress. An understanding of the parental, familial, and peer factors related to the development of coping, as well as the mechanisms through which these factors influence acquisition of coping strategies, will facilitate the prediction of children at risk for the development of maladaptive coping skills and may inform the development of coping skills intervention programs. Finally, future studies need to examine continuities in coping throughout development. The work of Mischel, Steinberg, and others indicates that behaviors and skills evident early in development are related to coping responses in adolescence and adulthood. Further longitudinal studies are necessary to examine how coping skills and styles during childhood are related to coping with stressful events in adolescence and adulthood.
This page intentionally left blank
5
Situational Determinants of Coping Robert R. McCrae
One of the difficulties facing researchers in the field of coping and stress management is the exquisite specificity of behavioral responses to situational demands. Relaxation techniques are inappropriate when the house is burning down, and clipping coupons may be better than seeking social supports in coping with the high costs of food. The specific behaviors needed to deal with particular problems are, in fact, usually not the concern of psychologists. Lawyers, mechanics, doctors, educators—anyone with special expertise—may be a more appropriate resource for dealing with some particular stressor. Coping falls within the province of psychology only at a more abstract level. What general strategies are there for dealing with stressors and the distress they cause? Which are most effective? For whom? Under what circumstances? In the past decade these topics have begun to be intensely researched (Aldwin & Revenson, 1987; Carver, Scheier, & Weintraub, 1989; Costa, Zonderman, & McCrae, 1991; Lazarus & Folkman, 1984; Menaghan, 1983b). Some investigators have focused on characteristics of the stressor as determinants of coping (Felton & Revenson, 1984; McCrae, 1984; Paterson & Neufeld, 1987; Perrez & Reicherts, 1988); the present chapter considers two questions about situational influences on coping. The first section reports data on the use of coping mechanisms as a function of general features of the situation: type of stressor, controllability, chronicity, and severity. The second section discusses the relative importance of person and situation in determining the choice of coping strategy. The data in this chapter, like most in the field, concern the use of coping mechanisms. They tell us which mechanisms are likely to be used in certain conditions. A more important question for practitioners is likely to be, Which coping mechanisms should be used to optimize adaptation? Fortunately, studies of coping efficacy suggest that there is a strong correlation between reported use
66
Personal Coping
and perceived effectiveness (McCrae & Costa, 1986). The coping mechanisms most widely used—in normal samples—appear to be the most effective in solving problems and reducing distress. At least provisionally, it appears that data on coping use may provide a guide to better coping. CHARACTERISTICS OF THE STRESSOR Although the number of different situations that may be stressful is limitless, it may be possible to make meaningful generalizations about the influence of the stressor on the choice of coping mechanisms if stressors can be organized in terms of common features or dimensions. Type of stressor (i.e., loss, threat, or challenge), controllability, chronicity, and severity all appear to be potentially important dimensions of the stressor. Data from two studies in 1980 and one in 1987 provide some evidence on the influence of these characteristics on the choice of coping mechanism. Subjects were participants in the Baltimore Longitudinal Study of Aging (BLSA; Shock et al., 1984)—generally healthy, well-educated, communitydwelling volunteers ranging in age from twenty to ninety-three. Slightly over half the subjects were males. All completed versions of a Coping Questionnaire (CQ) based on the Ways of Coping (Folkman & Lazarus, 1980), with the addition of fifty new items. The CQ was scored for twenty-seven specific coping mechanisms; two broader coping factors were later identified (McCrae & Costa, 1986). (A twenty-eighth mechanism, taking one step at a time, was not measured in some of the studies, and is not discussed here.) The number of items in each scale ranged from one to seven; for scales with two or more items, internal consistency ranged from .35 to .83, with a median of .67 (McCrae, 1984). Longer, more reliable scales would of course be desirable, but the current scales appear to be useful for exploratory research on a variety of ways of coping. The studies differed in the ways in which stressors were elicited and classified. Prior to the 1980 studies, subjects had completed a life events checklist. We classified events on the list a priori as losses (involving harm or disappointment), threats (involving worry and potential danger), or challenges (taxing opportunities), using a conceptual classification suggested by Lazarus and Launier (1978). We asked subjects who had checked one of these events to complete the CQ with regard to that event. Subjects who had checked two or more eligible events were assigned the most stressful. Complete coping data were received from 255 men and women. Analyses of these data show the effects of type of stressor as classified by the investigator. In a parallel study, subjects who had checked none of the classified events were asked to identify three events—a loss, a threat, and a challenge—and complete a shortened version of the CQ with regard to each. Complete data from 151 subjects were received for this study (see McCrae, 1982, 1984, for details). Analyses of these data show the effects of type of stressor as perceived by the subject.
Situational Determinants of Coping
67
In 1987, a follow-up study was designed. The full CQ was readministered to subjects in both the studies conducted in 1980 and to a supplementary sample of individuals who had joined the BLSA since 1980. Subjects who had previously been assigned to report on a loss, threat, or challenge were assigned to report on that type of stressor again; others were randomly assigned to one of the three conditions. Subjects in the loss condition were asked to recall a stressful event they had experienced in the past six months "in which you were hurt, disappointed, or lost something of value to you." Subjects in the threat condition were asked about "some kind of threat or danger, in which you were worried about how things would turn out;'* those in the challenge condition were asked about "some challenge or big opportunity/' All subjects were asked to describe the incident in a paragraph, rate its stressfulness, indicate how they appraised the event, and then decide whether they had ever used each of the 118 coping responses in dealing with the stressor. Although the type of stressor was determined by the investigator, the specific event was selected by the subject, so these data again speak to perceived type of stressor. As a check on the effectiveness of the instructions and as an alternative way of approaching the classification of stressors, a judge, blind to both condition and the coping responses, read the paragraph describing the event and classified it as a loss, threat, or challenge. The same judge also rated the severity, controllability, and chronicity of the stressor. These ratings cannot be considered truly objective, since they are based entirely on the description given by the subject. They do, however, provide a different perspective on the nature of the .stressor. Complete data from both subject and judge were available from 332 men and women in 1987 (McCrae, 1989). Overall agreement between the subject and judge on the classification of the stressor was 47 percent—a figure which, although significantly beyond the 33 percent expected by chance, was considerably less than perfect. Most discrepencies concerned confusions between threats and losses, which is understandable given the association between these two categories. Threats are often potential losses, and a situation that begins as a threat may become a loss. Similarly, some losses (e.g., loss of a job) may pose threats to future well-being. Disagreement in the classification of such events is expectable. Types of Stressor Together, these studies provide four opportunities to examine effects of type of stressor on the use of coping mechanisms. The first 1980 study compares three groups differing in investigator-classified stressors. The second 1980 study employs a repeated measures design to contrast the ways in which the same individuals coped with three different types of events they themselves selected. The 1987 data can be analyzed in terms of the three conditions for eliciting stressors from the subject, and in terms of the judge's classification of the elicited events. These analyses employ very different operationalizations of the classi-
68
Personal Coping
fixation of stressors. Differences between studies are not readily interpretable: They may point out meaningful consequences of different ways of eliciting and classifying stressors, or they may simply indicate unreplicable effects. However, when the same pattern of results is seen across two, three, or four studies, it is evidence of a robust situational effect. An earlier study compared the results of the two 1980 studies (McCrae, 1984). When the comparison is extended to include the 1987 analyses, a similar pattern of results emerges. However, it may be useful to organize results in terms of a factor analysis (McCrae & Costa, 1986) which identified two broad coping factors: neurotic coping and mature coping (see Table 1 for specific mechanisms in each factor). Several coping mechanisms were unassigned to either factor, and are discussed separately. The mechanisms in the neurotic coping factor are generally regarded by both clinicians and the individuals who use them as ineffective, and are consistently related to the personality dimension of neuroticism (McCrae & Costa, 1986). They are not, however, related to type of stressor. With the exception of indecisiveness, which is used more in threatening than in challenging situations in two of the four studies, none of the Neurotic Coping mechanisms shows a replicated effect. By contrast, replications across three or four of the studies provide strong evidence that all five mature coping mechanisms are used most often in facing challenges, and least often in dealing with losses. These patterns are consistent with the factor interpretations: neurotic coping appears to be a reflection of personal vulnerability rather than a rational reaction to the problem; mature coping is clearly responsive to situational demands. The usassigned coping mechanisms show a variety of patterns of situational influence. Seeking help, isolation of affect, distraction, substitution, avoidance, active forgetting, and assessing blame have no replicated effects, and may be equally suited for all three types of stressor. Fatalism, social comparison, wishful thinking, and faith are all more likely to be used in facing a loss or threat than a challenge. Intellectual denial, drawing strength from adversity, and humor are more frequently used in challenging situations than in harmful or threatening situations. Expression of feelings appears to be used more often in dealing with a loss or challenge than with a threat. On the whole, these results are reasonable. It is surely easier to use mechanisms like positive thinking and humor when taking on a difficult but promising challenge than when reacting to harm or danger. The relatively passive mechanisms of fatalism, faith, and wishful thinking may be appropriate (and are certainly understandable) as responses to harm or to dangers over which the individual may have little control. These data show clearly that different types of stressors—whether classified subjectively or objectively—elicit different coping responses, and the findings may be useful in suggesting the ways of coping most appropriate for particular problems. However, even where effects are consistently replicated, they are generally not large in magnitude. Omega-squared values for the 19 significant
Situational Determinants of Coping
69
effects shown in the first 1980 study ranged from .02 to .16, with a median of .08; and proportions of variance accounted for by the effects in the second 1980 study were generally smaller (McCrae, 1984). In part, the modest magnitude of the effects is doubtless due to the imprecision with which events were classified. Type of stressor was operationalized in terms of three mutually exclusive categories, when in fact many events have both threatening and challenging aspects, or begin as a threat and end as a loss. It might be more meaningful to ask subjects (or judges) to rate events on the degree to which they were harmful, threatening, and challenging. Studies in the future may also wish to target a specific period in the course of a stressful event, since the demands placed by the situation often change over time (cf. Folkman & Lazarus, 1985). Controllability and Chronicity Lazarus and Folkman (1984) have argued that the major determinant of coping responses is the individual's appraisal of the stressor. The Ways of Coping questionnaire frequently includes a section that asks for the subject's view of the situation as one "that you could change or do something about," "that must be accepted or gotten used to," "that you needed to know more about before you could act," or "in which you had to hold yourself back from doing what you wanted to do." An analysis of the use of coping mechanisms as a function of this appraisal shows a number of significant associations, most of which attest to the meaningfulness of responses. Use of rational action and perseverence is most frequent when subjects perceive the event as one they can do something about; use of fatalism is most common when the situation must be accepted. While reassuring, it is not clear whether such associations are truly informative. Appraising a situation as one that must be gotten used to is itself an act of fatalism. Stimulus and response are confounded in ways that make interpretation difficult (Paterson & Neufeld, 1987). Lazarus and his colleagues, of course, are aware of this issue, and argue that the ongoing, transactional nature of the coping process is best represented in a model that avoids a rigid and artificial distinction between event and response, but it is not yet clear how such a model should be evaluated. An alternative to subjective appraisal is provided by the judgment of an observer who can assess the nature of the stressor independent of the coping responses it elicited. The same judge who, blind to coping responses, classified events as losses, threats, or challenges also rated the situations described on controllability, using a five-point scale. There was significant but not perfect agreement between subjective appraisals and ratings of controllability: Events that subjects thought "must be accepted or gotten used to" were judged to be less controllable than other events, r = .43, p < .001. The fact that subjects' appraisals of their control over the event are generally congruent with a more or less objective assessment of controllability suggests that the literature on personal control and coping (Folkman, 1984) is relevant to this dimension.
70
Personal Coping
The first column of Table 5.1 gives correlations between judged controllability and the use of coping mechanisms. Controllable situations elicited rational action, restraint, expression of feelings, and humor, but also hostile reactions and selfblame. As Folkman (1984) pointed out, believing that one has control over a situation may heighten stress in some cases. Uncontrollable situations were strongly associated with fatalism, and also with the use of distraction, social comparison, wishful thinking, and faith. Again, coping choices clearly appear to be constrained to some extent by the objective features of the situation. However, controllability is not independent of the type of stressor, particularly when rated by the same individual. Events the judge considered challenges were rated highest in controllability by the same judge; losses were rated lowest. When the effects of event classification were controlled by partialling two dummy variables representing the three stressor categories from the coping response measures, correlations with controllability were substantially reduced, although the use of fatalism and faith were still significantly associated with low controllability, and the use of self-blame was still associated with high controllability. It appears that type of stressor and controllability are overlapping but not identical dimensions. Threats probably vary more widely in controllability than either challenges (which are generally controllable) or losses (which are generally uncontrollable). Studies of the effects of controllability within different stressor types would be particularly illuminating (cf. Felton & Revenson, 1984). Some stressors are chronic, like extended illness or marital problems; some are acute, like a robbery or the arrival of unexpected guests. Like other situational dimensions, chronicity is more complex than it at first appears. Some individuals will be permanently traumatized by an acute event; some individuals adapt to a continuing difficulty so rapidly that it appears to be only a temporary setback. One function of coping is to prevent temporary misfortunes from becoming longlasting disabilities, and it is encouraging to note that, in the long run, the great majority of individuals successfully adapt to even the most stressful events (McCrae & Costa, 1988). Events were judged to be acute (N = 214) or chronic (N = 118). Chronicity was not related to the classification of events as losses, threats, or challenges, and ratings of chronicity showed very small correlations with ratings of controllability and severity, so it appears to be an independent dimension of the stressor. The second column of Table 5.1 suggests that chronicity has a rather limited impact on the nature of the coping mechanisms used. Chronic stressors appear to be somewhat demoralizing, leading to withdrawal, passivity, indecisiveness, social comparison, and wishful thinking; humor is a more characteristic response to acute stressors. In these studies, subjects were simply asked if they had or had not used the ways of coping suggested. In more recent work (Folkman, Lazarus, DunkelSchetter, DeLongis, & Gruen, 1986), Folkman and colleagues have adopted a four-point scale that assesses the frequency of use. Chronic stressors may lead to more frequent use of the same kinds of coping mechanisms.
Table 5.1 Correlations of Stressor Characteristics with Use of Coping Mechanisms Judge-rated Coping mechanism
Controllability Chronicity
Subjective Severity
Severity
Neurotic Coping factor: Hostile reaction
.14**
.01
-.01
.22***
Escapist fantasy
.07
.01
-.02
.13*
Self-blame
.1Q***
.00
-.06
.07
Sedation
.06
.11
-.02.
.20***
Withdrawal
-.01
.14*
-.02
.10
Passivity
-.08
.13*
.17**
.11*
Indecisiveness
-.02
.12*
.07
.18**
Mature Coping factor: Rational action
.25***
-.02
-.13*
Perseverence
.07
-.01
-.04
Positive thinking
.10
-.05
-.01
-.05
Restraint
.14**
.08
-.06
.08
Self-adaptation
.10
.08
.03
.06
Seeking help
.01
.04
.11*
.18**
Isolation of affect
.08
-.04
-.17**
-.38***
.04 .11*
Unassigned:
Fatalism
-.41***
-.02
.25***
.00
.13*
-.08
-.03
.12*
-.12*
.08
.08
.14*
Intellectual denial
.00
.02
-.03
Social comparison
-.13*
.17*
.11
.04
.01
.01
-.06
-.03
Drawing strength from adversity
-.03
.03
.22***
.07
Avoidance
-.07
.05
.11
-.03
Wishful thinking
-.25***
.13*
.11
.34***
-.02
.00
-.09
-.25***
Expression of feelings Distraction
Substitution
Active forgetting
_ 24***
72
Personal Coping
Table 5.1 (continued)
Coping mechanism Humor Assessing blame Faith Note: JV =332. */7<.05 **p<.01
Judge-rated Controllability Chronicity
Severity
Subjective Severity
.18**
-.13*
-.25***
-.29***
.10
-.09
-.14**
.05
-.22
.07
.21***
-.03
***/>< .001
Severity of Stressor Subjective ratings of severity show only a modest agreement with judged severity, r = .27, p < .001. In part, this may reflect the fact that stressfulness is a function of both the event and the individual, and is often idiosyncratic (McCrae, 1990). Death of a pet might seem a minor loss to objective judges, but it would be personally devastating to many pet-lovers. This divergence of perspective is seen in the correlates of objective and subjective severity, shown in the last two columns of Table 1. The two sources agree that greater stressfulness is associated with the increased use of passivity and seeking help, and decreased use of isolation of affect and humor. In other respects, they differ. Objectively assessed severity is associated with the use of several distinct mechanisms: passivity, seeking help, fatalism, drawing strength from adversity, and faith. Low severity is associated with rational action, isolation of affect, humor, and assessing blame. It appears that more serious situations call into play more passive forms of coping. Part of the reason a situation may be deemed severely stressful is that it exceeds the capacity of the individual to deal with it him- or herself. Active problem solving and minimizing strategies are more characteristic of less stressful situations. Several hypotheses are suggested by the correlates of subjective severity (Table 5.1, last column). First, it appears that most forms of coping are affected in some way by subjective severity. Ways of coping may be more salient to individuals who perceive that they are under greater stress. Second, subjective severity is more strongly correlated with neurotic coping mechanisms than with mature coping mechanisms, so the presence of personality and mood confounds may be suspected. Individuals high in neuroticism or in temporary periods of distress (cf. Cohen, Towbes, & Flocco, 1988) may perceive greater stressfulness,
Situational Determinants of Coping
73
and may be more likely to use immature and ineffective ways of coping (McCrae & Costa, 1986). Finally, four coping mechanisms—isolation of affect, intellectual denial, active forgetting, and humor—show significant negative correlations with perceived severity of the stressor. All of these mechanisms serve to minimize the perceived seriousness of the situation, and their associations with severity may be interpreted in two ways. Perhaps ratings of low stressfulness are the result of the effective action of these coping devices, which convinced the individual that the situation was really not serious. However, the direction of causation may be reversed. It is possible that individuals are able to turn to these comforting coping mechanisms only in situations that are in fact relatively benign. Some evidence for this latter interpretation is seen in a comparison of individuals who were and were not at Three Mile Island during the period of the nuclear accident there. Subjects distant from this objectively severe stressor—that is, those not facing a serious threat—were more likely than local residents to use isolation of affect and active forgetting in dealing with the incident (Costa & McCrae, 1989a). CONTRASTING SITUATIONAL AND DISPOSITIONAL INFLUENCES One of the controversial issues in research on coping is the relative importance of situational versus dispositional contributions in the choice of coping mechanisms. Early research, often based on psychodynamic models of defense, typically assumed that coping reflected consistent styles of responding to stress— indeed, Haan (1977) defined personality itself as "the fundamental and persistent organizational strategies [of coping and defending] that people use to regulate various aspects of themselves'* (p. 1). More recently, the pendulum has swung the other way. When Folkman and Lazarus (1980) assessed coping responses to a series of naturally-occurring stressors in the daily lives of their subjects, they concluded that, "on the whole, coping patterns were not greatly determined by person factors" (p. 229). It seems appropriate to consider recent evidence on dispositional influences on coping to set in context the situational findings just discussed. Dispositional influences are seen in consistent behavior of individuals across time and situation. Some data on consistency are offered by the second 1980 study (McCrae, 1984), in which 151 men and women completed a short version of the CQ with regard to three different events. Correlations between the use of each coping mechanism in dealing with loss versus threat, loss versus challenge, and threat versus challenge situations yielded 81 coefficients of cross-situational consistency. These ranged in magnitude from - . 0 1 to .59, with a median of .29; 67 (or 83 percent) of them were statistically significant, p < .05. The three sets of correlations are quite similar, suggesting differential stability for different mechanisms. The most cross-situationally consistent were sedation, faith, escapist fantasy, and hostile reaction, three of which come from the neurotic
74
Personal Coping
coping cluster. The least consistent coping mechanisms were seeking help, assessing blame, isolation of affect, and avoidance. These appear to be more strongly influenced by the specific requirements of the situation. It is possible to assess the percentage of variance attributable to person, situation, and their interaction in this study (McCrae, 1983). Using Golding's (1975) model, persons accounted for an average of 27 percent of the variance, whereas situations accounted for only about 3 percent. Using Endler's (1966) procedure on the 10 mechanisms that had multiple items, persons were found to account for about 13 percent of the variance, situations for about 2 percent, and their interaction for about 11 percent of variance in individual item responses A second approach to consistency examines the temporal stability of coping mechanisms. To the extent that coping responses are dispositionally based, they should be predictable over relatively long periods of time. In a seven year longitudinal follow-up of the 1980 studies, retest correlations were examined for 113 individuals from the first 1980 study and 78 individuals from the second study (using a shortened version of the CQ). In both cases, the type of event was the same at both administrations, although the specific events were of course different. Of the 54 retest correlations, 35 (65 percent) were significant; they ranged from - .10 to .60. Again, the neurotic coping mechanisms were among the most stable. A total neurotic coping score showed a seven-year retest correlation of .51 in thefirststudy and .55 in the second, bothp < .001 (McCrae, 1989). Although these correlations are substantially lower than those typically seen for personality dispositions (Costa & McCrae, 1988), they do suggest a moderate and enduring dispositional basis for coping. Perrez and Reicherts (1988) used pocket computers to record coping behaviors during or immediately after stressful episodes. A sample of sixty Swiss undergraduates were trained in the use of the computer and recorded responses to about forty separate events over a period of four to five weeks. When split into two successive time periods and aggregated over events within each period, highly significant retest correlations were seen for all coping mechanisms, ranging from .52 for passivity/hesitation to .74 for evasion/avoidance. Active influence, using social support, palliation, reevaluation, and blaming self or others showed intermediate levels of stability. These data suggest that there are meaningful patterns of coping that might be designated as coping styles. Carver, Scheier, & Weintraub (1989) created a new coping measure, the Coping Orientations to Problems Experienced (COPE), in both dispositional and situational versions: The former asked how respondents generally reacted to stress; the latter, how they had reacted in a specified situation. Subjects completed both versions, using the most stressful event of the previous two months as the situational stimulus. The COPE is scored for 14 coping strategies; correlations between dispositional scores and corresponding situational scores ranged from .07 for restraint coping to .76 for religion (recall that faith was one of the most consistent coping mechanisms in the 1980 study). All but three of the mechanisms showed significant agreement in the sample of 117 students. It appears that
Situational Determinants of Coping
75
questions about how people generally act give some insight into how they will act in a specific instance. Aggregating across specific instances to form a more reliable criterion would probably increase predictive validity of the dispositional COPE. Finally, there is evidence that at least some of the consistency and stability of coping responses is due to the influence of enduring personality traits. The personality dimensions of neuroticism, extraversion, and openness to experience, as measured by self-reports and spouse and peer ratings on the NEO Personality Inventory (Costa & McCrae, 1985, 1989b) were systematically related to coping mechanisms in both 1980 studies (McCrae & Costa, 1986). Neuroticism was associated with increased use of hostile reaction, escapist fantasy, self-blame, sedation, withdrawal, wishful thinking, passivity, and indecisiveness. Extraversion was correlated with the use of rational action, positive thinking, substitution, and restraint. Open individuals were more likely to use humor in dealing with stress; closed individuals were more likely to use faith. In part, at least, consistency in coping appears to mirror a more general consistency of personality. CONCLUSIONS AND COMMENTS Coping is, by definition, a response to a stressful situation, and although some ways of coping ("Don't panic!") are of nearly universal applicability, the specific behaviors needed to deal with a particular problem are frequently dictated by the nature of the situation. Whether we should dial 911, and if so, whether we should ask for the fire department, ambulance, or police, depend on the crisis at hand. However, as soon as attention shifts to questions about how general features of the situation affect strategies of coping, the links are considerably weakened. In most cases, there are alternative ways of dealing with a stressor, and individuals appear to make their choices on the basis of personal preferences as well as situational requirements. In some cases, in fact, situational factors appear to be almost irrelevant. Individuals high in neuroticism use coping mechanisms such as hostile reaction, sedation, and indecisiveness when confronted by stressors of all kinds; even, in fact, when they are not confronted by stressors at all (Costa & McCrae, 1989a). One interpretation of this phenomenon is to suggest that such "coping responses" are in fact misattributions. Individuals may blame their moodiness, guilt, or escapist tendencies on some perceived external stressor, when in fact they are caused by enduring internal dispositions (Costa & McCrae, 1989a). Alternatively, these forms of coping may be better seen as defense mechanisms: relatively automatic and inflexible responses to the individual's own anxiety rather than the environmental demands. Certainly, it is worthwhile to reexamine older concepts of defense, which have recently been eclipsed by the study of self-reported coping (Costa, Zonderman, & McCrae, 1991). One of the most serious problems in the coping literature is the lack of any consensus on the number and nature of coping mechanisms, and the consequent
76
Personal Coping
lack of standard coping measures. In the exploratory phases of research such a situation is inevitable and perhaps desirable, preventing premature closure. By now, however, sufficient data may have been accumulated to justify attempts to systematize the area (cf. Carver et al., 1989). Findings on the situational determinants of coping can contribute to this effort. Because neurotic coping mechanisms appear to reflect characteristics of the person far more than the situation, it may be advisable to omit them from coping inventories, or reclassify them as stress reactions (or attributions) rather than coping efforts. Because positive thinking is used more in facing challenges and social comparison is used more in facing threats and losses, it would appear to be inappropriate to combine them in a single category of problem-focused coping (Folkman & Lazarus, 1980) or meaning manipulation (Pearlin & Schooler, 1978). To the extent that different coping mechanisms are functionally equivalent with regard to characteristics of the person and situation, it may make sense to combine them; insofar as they differ in situational or dispositional determinants, it is better to maintain their distinction.
6
Individual Differences in the Coping Process: What to Know and When to Know It Suzanne M. Miller
Individual differences in self-regulatory styles—viewed as the tendency to process and respond to threat in a characteristic manner—have long been highlighted as important components of the process of coping with threat. However, the history of research and theorizing in this area has been largely disappointing (see Miller, 1990, for a review). In recent years, investigators have tended to focus on the situational determinants of adaptational outcomes and their transaction with the use of particular coping strategies—viewed as the individual's response in a given situation (e.g., Lazarus, DeLongis, Folkman, & Gruen, 1985). While this approach has been fruitful, it still seems useful to try to identify differences in dispositional coping styles, and to explore their interaction with situational variables and ongoing coping strategies. From the situational perspective, three variables have emerged as playing a critical role in promoting adaptation. These are the amount and type of control, information, and coping interventions made available to the individual. In this chapter, I first review theory and evidence relevant to these situational parameters. I then go on to identify individual difference factors that moderate these effects, focusing on dispositional preferences to seek or to cognitively avoid and transform threatening cues. Finally, I explore aspects of the interaction between situations, dispositions, and ongoing strategies, including the impact of situational demands, person-by-situation matching, and exposure to clinically relevant outcomes.
Preparation of this chapter was supported in part by Temple University Research Incentive Fund and by NIH grant CA 46591.
78
Personal Coping
DEFINITIONAL AND THEORETICAL FRAMEWORK While information, coping, and control have often been construed so broadly as to render them indistinguishable, previous work suggests that it may be theoretically and empirically useful to keep them distinct (see Miller, 1979a, 1980a and b, 1981; Miller & Birnbaum, 1988b). Control is defined as the individual's perception that he or she can execute (or has the potential to execute) some action that changes an aversive stimulus. In contrast, information (predictability) merely implies that the individual knows something about the event, whether or not he or she can do anything to change it. Finally, for the present purposes, coping is defined as the regulation of stressful emotions via attention deployment and the modulation of internal arousal. Relevant techniques include relaxation, distraction, reinterpretation, calming self-talk, and so forth. This definition corresponds to Folkman and Lazarus' (1980) notion of emotion-focused coping, while information and control are viewed as orthogonal components of their notion of problem-focused coping. Perceptions, preferences, and expectations relevant to these constructs have been explored in a variety of naturalistic contexts. Yet, relatively few real-life studies have attempted to directly measure or manipulate these variables, while simultaneously exploring their impact on stress. The majority of this work has focused on the preparation of patients for stressful medical procedures. Taken together, the results have been extremely uneven and suggest a different pattern of effects for information, coping, and control interventions. While information and control can sometimes facilitate patient adjustment, this is not always the case. In contrast, the consequences of coping interventions, such as relaxation, are generally beneficial and superior to those of information. However, information does appear to have value for a subset of individuals, while control may have value when the response is a meaningful one (see Miller, Combs, & Stoddard, 1989, for a review). These discrepant findings can be integrated within two complementary theoretical frameworks: The Minimax Hypothesis (Miller, 1979, 1980; Miller & Birnbaum, 1988b) and the Monitoring and Blunting Hypothesis (Miller, 1981, 1989, 1990; Miller & Birnbaum, 1988a). The Minimax Hypothesis primarily addresses the issue of control. This view predicts that control should be stressreducing, when it enables the individual to limit how bad the situation can become. However, in the contexts under study, effective forms of control are not generally available, thereby undercutting the individual's ability to minimize the aversiveness of the situation (see also Bandura, 1986b). Further, when another person is in a better position to provide an upper limit on the maximum aversiveness of the outcome, then individuals will generally prefer to relinquish control to the identified expert. There also appear to be individual differences in control preferences that determine when or where controlling responses are engaged (e.g., Clark & Miller, 1990; Burger, 1984, 1985). The Monitoring and Blunting Hypothesis mainly addresses the impact of
Individual Differences
79
information and coping interventions. According to this view, coping interventions (such as distraction and relaxation) enable individuals to selectively process aversive events in a less negative way. This helps them to reduce anxiety and manage their emotions in the face of threat. This, in turn, can attenuate the experience of pain and thereby may minimize the total aversiveness of the experience for the individual, in accord with the Minimax Hypothesis. Information, on the other hand, generally focuses the individual on the negative aspects of the threatening event. This interferes with the use of anxiety-reducing cognitive techniques and so arousal remains high. However, for the subset of individuals who cannot or will not use these cognitive techniques, the arousal effects of information can be compensated for by the increased certainty that predictive cues allow (Weiss, 1970). In addition, consistent with the Minimax Hypothesis, information can also be useful to individuals by providing them with access to and contact with external sources of expertise. IDENTIFICATION OF COPING DISPOSITIONS The hypotheses presented above predict that the effects of situational manipulations can, in part, be reconciled by considering dispositional preferences for information, coping, and control. In support of this, we have found that Type As (who overvalue control) are more reluctant to yield or share control when it is appropriate to do so than are Type Bs (Clark & Miller, 1990; Miller, Lack, & Asroff, 1985). Similarly, Burger and his colleagues have found that individual differences in the desire for control moderate the individual's response to a variety of control-related situations (e.g., Burger, 1984, 1985; Burger & Cooper, 1979; Smith, Wallston, Wallston, Forsberg, & King, 1984). Our program of research, focusing on individual differences in the ability and/ or inclination to seek out or to cognitively avoid and transform threat-relevant information, is also illustrative of the utility of such a person-by-situation approach. Most of the available measures of informational styles had been shown to be psychometrically lacking or unvalidated (e.g., Goldstein, 1959) or were very narrowly defined and delimited to a specific context, such as the medical situation (e.g., Krantz, Baum, & Wideman, 1980). Therefore, we devised a new self-report instrument, that was both closely tailored to the kinds of informational choices of interest (see also Krantz et al., 1980), but that also tapped the individual's inclinations in response to a variety of physically and/or psychologically stressful life events (Miller, 1987). The Miller Behavioral Style Scale (MBSS) consists of four hypothetical, uncontrollable, stress-evoking scenes. For example, imagine that you are afraid of flying and have to go somewhere by plane. Each scene is followed by eight statements, which represent different ways of coping with the situation. Half of the statements following each scene are of a monitoring variety (e.g., in the airplane situation: "I would listen carefully to the engines for unusual noises and would watch the crew to see if their behavior was out of the ordinary;" or
80
Personal Coping
"I would read and reread the safety instruction booklet."). The other half of the statements are of a blunting variety (e.g., "I would watch the inflight movie even if I had seen it before"). The subject simply marks all the statements following each scene that might apply to him or her. Three measures are derived from this scale. The monitoring measure is the sum of all the items endorsed on the monitoring subscale. Subjects scoring above the median are high monitors, and those scoring below are low monitors. The blunting measure is the sum of all the items endorsed on the blunting subscale and subjects are divided into high and low blunters. Finally, the monitor/blunter measure is obtained by subtracting the total number of items endorsed on the blunting subscale from the total number of items endorsed on the monitoring subscale. Using this difference score, subjects are divided into monitors and blunters. Recently, given the complex nature of the coping process, there has been an interest in exploring the distinctive effects of the separate monitoring and blunting dimensions (Folkman, 1984). However, the research discussed here has generally tended to rely on the difference score, except where indicated. This scale has been validated in the laboratory, showing that individuals scoring as high monitors (on the monitoring subscale) or low blunters (on the blunting subscale) typically seek out information relevant to physical and psychological threats. In contrast, individuals identified as low monitors or high blunters typically prefer to avoid and distract from threat-relevant information (Miller, 1987). Findings consistent with this have been obtained with medical populations. Miller and Mangan (1983) studied gynecologic patients at risk for cervical cancer who were about to undergo an aversive diagnostic procedure for the first time (colposcopy). They found that monitors generally desired more sensory and procedural information than they had received. Monitors are also more likely to desire risk information relevant to aversive medical procedures than blunters are, and to feel less anxious after receiving such information (see also Watkins, Weaver, & Odegaard, 1986). Similarly, in a study of women about to undergo gynecologic surgery, Steptoe and O'Sullivan (1986) found that monitors wanted more details than did blunters (Vi vs. lA). It is perhaps surprising that a greater number of monitors did not desire more information, but perhaps they had already exhausted what there was to learn. Indeed, monitors were found to have significantly more factual knowledge about the procedure. Among all coping groups, fewer of the subjects who reported a high level of understanding wanted more information. However, blunters' satisfaction with their understanding appears not to be a function of genuine knowledge, but an aspect of coping style. This was explored by comparing factual knowledge—what they actually knew—with reported understanding—what they claimed they knew. Monitors who claimed to have better levels of understanding actually gave more correct answers. In contrast, factual knowledge was generally low and unrelated to reported understanding in blunters. In a study of patients visiting a primary care setting for acute medical problems, Miller, Brody, and Summerton (1988) found that while both high and low
Individual Differences
81
monitors (on the basis of the monitoring subscale) desired information about the diagnosis and seriousness of their medical problem, high monitors were significantly more likely to desire additional information pertaining to the cause of their medical problem, how healthy they were in general, what they could do to prevent future health problems, and possible medication side effects. Thus, high monitors/low blunters appear to desire more voluminous and detailed levels of information and preparation than do low monitors/high blunters. In addition to predictive validity, the scale appears to show good reliability as well as good discriminative and convergent validity. For example, it has been found to be unrelated to demographic variables, such as sex, race, age, educational status, and marital status (Miller, 1987; Miller & Mangan, 1983). It has also been found to be unrelated to trait measures such as repressionsensitization, depression, anxiety, Type A, extra version-introversion, attributional style, and social desirability (e.g., Caspi, 1987; Efran, Chorney, Ascher, & Lukens, 1989; Miller, Brody, & Summerton, 1988; Miller & Mangan, 1983; Steptoe, 1986). Further, test-retest analyses show the MBSS to be highly stable (in the .8 range) over a three-month period (Miller, 1987). Thus the MBSS appears to represent a distinct dimension, which does not overlap with demographic or traditional trait measures. Some evidence bears on the relations between these informational styles and the use of ongoing coping strategies. Carver, Scheier, and Weintraub (1989) found that high and low monitors (defined on the basis of the monitoring subscale) demonstrated divergent patterns of coping under threat. In one of the few studies to measure coping styles and strategies simultaneously, they showed that high monitors coped with stress by focusing on and ventilating their emotions, as would be expected. Interestingly, they also tended to become less behaviorally disengaged (i.e., they did not give up their goals but kept trying to solve the problem). However, rather than engaging in more active coping or planning, they sought out social support for instrumental reasons (i.e., they tried to get advice and find someone who could do something concrete about the problem). They also tended to turn more to religion (i.e., they sought God's help and put their trust in God). Other researchers have found that individuals make use of both problemfocused coping (entailing efforts to act on the stressor by means of informational and controlling behaviors) and emotion-focused coping (entailing efforts to regulate one's accompanying emotional state) during stressful encounters. However, the type of coping strategy engaged depends, in part, on situational factors (Band & Weisz, 1988; Folkman & Lazarus, 1980). The present results also suggest that certain individuals may show a delineated pattern of coping, displaying a combination of emotion-release and indirect (secondary) forms of problemfocused coping (see also Compas, Forsythe, & Wagner, 1988). These findings further indicate that the notion of problem-focused coping may be overly broad. While high monitors and low blunters appear to prefer information, this is not necessarily linked to the performance of controlling actions.
82
Personal Coping
Rather, information seekers appear to be more likely to yield control to another, more competent, individual. This is consistent with evidence showing that while high monitors (and low blunters) in a primary care setting typically preferred to have detailed information and reassurance about their medical condition, they did not appear to seek this information for its instrumental value (Miller, Brody, & Summerton, 1988). Indeed, compared to low monitors (and high blunters), they were more likely to desire to play a passive role in their medical care. If anything, greater knowledge may make individuals more acutely aware of the limits of personal control. These results would be predicted by the Monitoring and Blunting and the Minimax Hypotheses. First, information seekers appear to prefer information in order to reduce uncertainty and thereby put themselves in the presence of safety signals. Second, seeking information enables individuals to gain access to important sources of expertise (e.g., medical specialists) who, in turn, are in a critical position to minimize the maximum aversiveness of the situation for them. Finally, in the face of threatening events, high monitors/low blunters typically demonstrate greater subjective, behavioral, and physiological morbidity than do low monitors/high blunters (Efran et al., 1989; Gard, Edwards, Harris, & McCormick, 1988; Miller, 1979b; Miller, Brody, & Summerton, 1988; Miller, Leinbach, & Brody, 1989; Miller & Mangan, 1983). This is consistent with the Monitoring and Blunting Hypothesis (Miller, Combs, & Stoddard, 1989), which states that information seeking can maintain high levels of stress and arousal, particularly when it has no instrumental value (see also Ward, Leventhal, & Love, 1988). Thus, high monitors and low blunters may represent a vulnerable population in the face of threat and may benefit from training in blunting-type strategies (e.g., Fleischer & Baron, 1988; Miller, 1989). COPING AS A FUNCTION OF SITUATIONAL DEMANDS Do individuals who differ in coping style dispositions always differ? To the extent that coping tendencies are activated by a particular stimulus configuration, such as threat, then the adaptive concomitants of these styles should only be evident in certain situations. More broadly, this bears on the general issue of when one can detect and analyze basic personal dispositions. Mischel (1988) has proposed a conditional approach to dispositions. According to this view, evidence of a set of ptfototypic stable behaviors is only expected to emerge under a specific set of diagnostic conditions. This implies an important, moderating role for situations, as determining when and where particular dispositions can be identified. It argues against a more traditional trait approach, where global individual differences are typically assessed across aggregates of situations. One possibility is that dispositional differences become manifest in situations that are highly stressful for the individual. For example, with respect to monitoring and blunting, differential arousal between the groups should emerge in response to aversive events, but not in response to nonaversive or base-level
Individual Differences
83
conditions. Further, arousal differences during threat should be most evident as the process of coping unfolds. That is, while the groups may respond with equally high levels of initial arousal, low monitors/high blunters should show greater habituation over time, because of the way in which they are selectively processing and filtering out the negative aspects of the event. High monitors/low blunters, who attend to and focus on threatening cues, should show less habituation and more sustained arousal over time. There is evidence showing that both the subjective and physiological concomitants of monitoring and blunting are apparent under high threat but not under low threat or nonthreatening conditions. With respect to subjective arousal, Phipps and Zinn (1986) looked at a group of at-risk pregnant women, about to undergo amniocentesis. Measures of distress (as measured by the Profile of Mood States; McNair, Lorr, & Droppleman, 1971) were obtained after genetic counseling, after amniocentesis, and after obtaining the results (which were good in every case). Monitors showed greater anxiety than blunters, after genetic counseling. This differential arousal was even more heightened immediately after amniocentesis. Similar patterns were obtained for measures of depression and anger, with monitors evidencing far greater distress than blunters, both after genetic counseling and immediately after amniocentesis. The heightened arousal on the part of monitors dissipated after hearing the results and no differences emerged between the different coping style groups at this point. Importantly, no affective differences were observed between monitors and blunters at equivalent points in time, in a comparison group of pregnant women not at risk, who did not undergo amniocentesis. A similar pattern emerges for physiological arousal. Sparks & Spirek (1988) had undergraduate subjects complete the MBSS. Five weeks later, they watched the frightening film Nightmare on Elm Street for the first time. Average skin conductance levels were calculated at five points in time: (1) baseline; (2) a beginning low stress segment; (3) a moderate stress boobytrap segment; (4) a mild stress segment on descending the stairs; and (5) a high stress chase segment. Monitors showed a significantly greater increase in skin conductance from baseline across the entire film segment than did blunters. However, the results also showed a significant interaction between coping style and time periods. Specifically, monitors displayed a moderate increase in skin conductance during the moderate stress segment and a large increase in skin conductance during the high stress segment. Monitors did not increase in skin conductance during the low stress and mild stress segments and did not differ from blunters during these segments. This latter group generally maintained a low level of electrodermal arousal throughout the film. Not only are monitoring and blunting differences sensitive to threat vs. nonthreat conditions, but they also show a different pattern of habituation within a given stressful situation. The various groups do not differ from each other on initial measures of state or trait distress or on initial measures of psychophysiological arousal, in both laboratory and field settings (Miller, 1979b, 1987;
84
Personal Coping
Miller & Mangan, 1983; Phipps & Zinn, 1986; Watkins et al., 1986). Despite this, the different coping modes are sometimes accompanied by distinctive patterns of emotional responsivity over time. In the laboratory, when faced with the threat of electric shock (to the fingers), monitors and blunters do not differ in their self-reports of tension and anxiety after the first trial (Miller, 1987). However, those who attend to information show sustained higher arousal and less habituation across trials than blunters, who eventually decrease in arousal. This differential habituation is also evident using other self-report measures and physiological indices (Miller, 1979a, 1987). Similar findings have been obtained in field settings. For example, Miller and Mangan (1983) studied gynecologic patients at risk for cervical cancer who had to undergo an aversive diagnostic procedure (colposcopy). Results showed that monitors expressed more anxiety and discomfort than did blunters during the five days following the procedure. Specifically, blunters showed a steady decrease in amount of discomfort expressed. Monitors, on the other hand, showed a more gradual decline in pain and discomfort, and did not feel significantly better even at the third day. It would seem that in order to demonstrate different affective consequences of informational dispositions, such as monitoring and blunting styles, it is important to test for such differences under a specific set of circumscribed situations—for example, the occurrence of aversive events—and to explore the pattern of such differences over time. These results are consistent with those found in other contexts. For example, Wright and Mischel (1987) demonstrated that disturbed children with problems of aggression tended to display signs of aggressive behavior in situations that were psychologically taxing for them: situations that demanded cognitive, self-regulatory, and social skills. Conversely, children with withdrawal problems tended to display signs of withdrawal behavior, but again only in these psychologically demanding contexts. In a recent longitudinal study, Mischel, Shoda, and Peake (1988) showed that the delay-of-gratification ability of preschoolers was predictive of parental ratings of various competence and coping abilities during adolescence. The original delay of gratification situation was then separated into children who underwent (1) a psychologically easy task (delaying with the rewards absent); and (2) a psychologically difficult task (delaying with the rewards present; Shoda, Mischel, & Peake, 1990). Only the latter condition proved to be diagnostic of long-term functioning. Specifically, those who delayed longer (as preschoolers) under the more frustrating condition of having to wait with the rewards in view were rated by their parents ten years later as more attentive, planful, better able to resist temptation and delay gratification, and more mature. Even more striking, preschooler delay tolerance with the rewards exposed was significantly related to both verbal and quantitative SAT scores. Differences in delay behavior under nondemanding conditions (rewards not present) were not related to either parental ratings or SAT scores. Overall, the pattern of findings suggests that individual differences—and their
Individual Differences
85
affective, behavioral, and cognitive correlates and sequellae—are best identified under a specific set of diagnostically significant situations. In the realm of affectively positive outcomes, this seems to entail demanding conditions under which the individual must demonstrate effective self-regulatory skills for managing frustration and social interactions. In the realm of affectively negative outcomes, this seems to entail uncontrollable threatening conditions under which the individual must demonstrate effective self-regulatory skills for modulating stress and arousal. Future research should help to further define individuals with at-risk, self-regulatory profiles and to delineate the specific circumstances that elicit them. PERSON-BY-SITUATION MATCHING When individuals encounter stressful or demanding situations that activate their coping tendencies, can the adaptive consequences of these dispositions be modified by the specifics of the stimuli they experience? One possibility is that individuals fare better when aspects of the situation (e.g., the availability of information, control, and coping) are well matched with the individual's coping style. To the extent that situational factors permit individuals to execute welllearned and well-rehearsed coping repertoires, it may enhance their sense of personal efficacy and thereby reduce stress (Bandura, 1986b). In the case of informational styles, for example, those who typically seek out threat-relevant information should show less arousal when such information is made available than when it is withheld. Conversely, individuals who typically avoid or distract from threat-relevant information should show less arousal when unwanted information is not imposed upon them. Some evidence supports the value of person-by-situation matching. In an early study along these lines, Miller and Mangan (1983) found that blunters showed less physiologic, subjective, and behavioral arousal in response to an aversive diagnostic procedure for cervical cancer, when they received a distracting preparatory intervention than when they received voluminous preparatory information. In contrast, monitors receiving voluminous preparatory information showed less physiologic arousal by the end of the procedure than monitors receiving a distracting preparatory intervention. However, subjective and behavioral indices of arousal were not reduced when monitors were provided with voluminous information. In a similar study, Watkins et al. (1986) found that monitors who received high levels of procedural and sensory information were less subjectively and physiologically aroused during an aversive diagnostic procedure (cardiac catheterization) than monitors who received more minimal procedural information. Conversely, blunters who received more minimal procedural information were less aroused than blunters who received a combination of procedural and sensory information. A further study explored differential pain responses among monitors and blun-
86
Personal Coping
ters exposed to the cold pressor task in the laboratory (Efran et al., 1989). Blunters who were induced to distract from the sensations in their hands and to use rational statements (e.g., "One of the good things about this experience is that it might help psychologists learn more about people") showed higher pain thresholds (i.e., they took longer to perceive the stimulus as painful) than blunters who were induced to attend to and focus on the sensations they were experiencing. Monitors in the attention condition did not show higher pain thresholds than those in the distraction condition, although the results were in the predicted direction. Monitors also expressed less confidence in the effectiveness of the techniques they were shown (see also Gattuso, Litt, & Fitzgerald, 1992). Taken together, while the available data show some support for a person-bysituation approach, the results in general are stronger and more consistent for blunters than for monitors. While monitors show some benefits of interventions that provide them with voluminous information or that orient them to ongoing sensations, these effects do not hold for all measures (e.g., behavioral) and tend to emerge over time. Thus, it may be easier to design effective, standardized interventions for blunters than for monitors. Blunters prefer to avoid and distract from details of the threatening event and seem to benefit from a variety of distraction—as well as relaxation and reinterpretation—interventions. The exact nature and timing of the distractor (or transformation technique) may not be very important to them, as long as it is absorbing and engaging. In contrast, information may be most productive when it allows the individual to attend to and interpret bodily sensations in an accurate and nonthreatening manner. This means that monitors may develop distinctive, moment-to-moment agendas, as they actively process incoming sensations and experiences and compare them with expected occurrences (Leventhal, 1988). As such, they may find an initial preparatory communication too diluted, nonspecific, and unresponsive to their unfolding informational needs. This, in turn, may undermine their sense of personal efficacy in the situation (Bandura, 1986b). In line with this reasoning, we have found that high monitors (and low blunters) with acute medical problems are able to delineate quite specifically the kinds of information they desire (Miller, Brody, & Summerton, 1988). Further, they— but not low monitors (or high blunters)—are highly concerned about the nature of their relationship with the health care provider. This may be because they have an ongoing need to recognize and attach appropriate meanings to their bodily symptoms. Therefore, they may benefit from a health-care interaction which enables them to freely ask questions and discuss their experiences. Indeed, if doctor-patient communications were more systematically structured to amplify the meaning of situational cues and to exclude threatening interpretations, they might be more beneficial to blunters as well as to monitors (Leventhal, 1988). COPING WITH CLINICALLY RELEVANT OUTCOMES The majority of work on the coping process has tended to focus on delineating the strategies and, in some cases, the styles that individuals bring to bear on
Individual Differences
87
stressful situations (e.g., Lazarus & Folkman, 1984). Less attention has been devoted to the affective consequences of different strategies and styles. Researchers have theorized that aspects of the situation—such as its predictability and controllability—should impact on the adaptiveness of various modes of coping. For example, there is evidence that the use of coping strategies of a problem-focused nature is beneficial in situations that are perceived to be controllable. In contrast, the use of coping strategies of an emotion-focused nature is beneficial in situations that are perceived to be uncontrollable (Compas, Malcarne, & Fondacaro, 1988; Folkman & Lazarus, 1980). The thrust of our own results on coping styles suggests that individuals who are high monitors and/or low blunters generally suffer more psychological and physical morbidity in response to common stressful life events that are unpredictable and uncontrollable than do individuals who are low monitors and/or high blunters. Increasing the amount of predictability available, but not necessarily the amount of controllability available, tends to offset somewhat these ill effects. Recently, however, we have been interested in exploring whether and when there are adaptive consequences of employing a high monitoring or low blunting mode of coping. In health-related contexts, one possibility is that high monitors/ low blunters are actually healthier than are low monitors/high blunters. Since they are more risk-aversive, they may take more preventive actions and adhere more carefully to self-help regimens (Kahneman & Tversky, 1984). While the evidence shows that monitors are indeed more likely to undertake health behaviors that entail the successful reduction of uncertainty (e.g., undergoing routine pap smears and breast self-examinations), they may be less likely to undertake behaviors that entail the effective modulation of stress (by exercising regularly, moderating alcohol intake, getting enough sleep, etc.; Miller, Brody, & Summerton, 1988; Steptoe & O'Sullivan, 1986). Thus, monitoring in the face of health threats may be a double-edged sword, facilitating greater compliance with screening protocols but perhaps maintaining more unhealthy, stress-driven behaviors. Future research should help to clarify this issue (see also Ward, Leventhal, & Love, 1988). Another possibility, among clinically anxious populations, is that individuals high in information seeking may respond better to techniques for the management and reduction of anxiety. The most effective treatment for anxiety disorders, such as phobias and obsession-compulsions, appears to be exposure to the phobic object or situation. Further, it has been demonstrated that increased attention to feared stimuli during exposure treatment enhances the reduction of subjective and physiological indices of anxiety and promotes better outcomes (Foa & Kozak, 1986). That is, this treatment works best when the individual thoroughly processes the event or object. Extrapolating from our previous work, high monitors and low blunters should be more attentive to threatening cues during exposure treatment. This means that they should evidence greater initial reactivity than do low monitors and high blunters. However, they should also show greater
88
Personal Coping
habituation and anxiety reduction, as they subsequently process and recode incoming information. This, in turn, should promote better therapeutic outcomes. To test this, snake and spider phobic community volunteers were divided into monitors and blunters, based on the difference score between the two subscales (Steketee, Bransfield, Miller, & Foa, 1989). Subjects participated in exposure treatment over two sessions, which is typically sufficient to reduce such phobias. The first forty-five minute treatment session of in vivo exposure was as follows. A caged animal was placed on a trolley four feet away from the subject who was informed that the cage would be moved closer at intervals, unless he or she objected strongly. Subjective ratings of anxiety were obtained at five-minute intervals. For the first fifteen minutes the cage was placed four feet from the subject. For the next fifteen minute period, it was placed two feet from the subject, and for the last period, it was placed next to the subject who was asked to touch the cage. The experimenter modeled touching the cage. The second treatment session was held on the next day. In vivo exposure was then begun with the animal placed four feet away for five minutes, two feet away for the next five minutes, and then next to the subject who was asked to touch the cage. The subject then placed a gloved hand inside the cage while the experimenter modeled touching the animal, for a ten minute period. Thereafter, the subject was asked to touch the animal with a gloved hand for ten minutes, and then with bare hands for the final ten minutes. Results showed that monitors demonstrated greater habituation of subjective and physiological arousal than did blunters. During session one, monitors showed a decline in subjective anxiety from the beginning to the end of the session whereas blunters' anxiety remained unchanged. That is, monitors eventually felt less afraid. Monitors also showed greater habituation of physiological arousal from session one to session two. Specifically, monitors' heart rate declined across sessions whereas blunters' heart rate increased. Indeed, monitors showed a considerable reduction in heart rate (sixteen bpm) from the baseline of session one to the baseline of session two, indicating a reduction of anticipatory arousal. Blunters, on the other hand, showed a mild increase in baseline heart rate from the first to the second session. Thus, the responses of high monitors and low blunters appear to reflect increased attention to phobic cues and greater processing of exposure information. This suggests that they may be ideal candidates for this type of therapy. These results are in accord with other findings, showing that across-session habituation of heart rate is related to outcome of exposure therapy with obsessive-compulsives (Foa & Kozak, 1986). High monitors and low blunters typically scan their environments for threat-relevant cues. Therefore, they appear to react strongly to the anticipation of threat (presentation of the phobic stimulus) and to attend carefully to information embedded in the exposure situation. Since the situation is, in fact, less aversive than anticipated, they are able to relax rapidly and remain relatively calm during a second exposure. Low monitors and high blunters, on the other hand, who typically protect
Individual Differences
89
themselves from arousal, probably do not engage as fully in the exposure situation and therefore are not able to fully process information about the absence of danger. With continued exposure, cognitive avoidance appears to become increasingly difficult. This results in increased processing of the feared object, evidenced by increased arousal. With additional exposure sessions, low monitors and high blunters also would be expected to habituate (see also Leventhal, 1988). Another related, clinical context where there may be advantages to being a high monitor or a low blunter has to do with the development and persistence of post-traumatic stress reactions in the face of exposure to intensely stressful situations. A subset of individuals who encounter a distressing event outside the range of usual human experience show characteristic symptoms which involve reexperiencing the traumatic event, avoidance of stimuli associated with the event or associated numbing of general responsiveness, and increased arousal and hypervigilance (Brett, Spitzer, & Williams, 1988). On the one hand, high monitors and low blunters might be expected to be more vulnerable to the development of such symptoms, since they attend to threat and fail to distract themselves. On the other hand, to the extent that successful adaptation to such events requires emotional processing of intrusive imagery, information seekers may show greater initial arousal but less prolonged disturbance than those who avoid or distract from such information (Horowitz, 1969). Recent preliminary data show that among groups at risk for medical problems, high monitors/low blunters are more likely to development symptoms of post-traumatic stress disorder (Miller, 1992). These results may reverse for the persistence of post-traumatic symptoms, once they have developed. Solomon, Mikulincer, and Arad (1988) explored the adaptiveness of monitoring and blunting styles for the maintenance of combatrelated post-traumatic stress disorder (PTSD). They looked at 348 Israeli soldiers who had suffered a combat stress reaction episode during the 1982 Lebanon War. All soldiers had participated in front-line battles and had been diagnosed on the battlefield. None of them had experienced serious physical injury or other psychiatric complications (e.g., brief reactive psychosis or factitious disorder). Two years later, the soldiers were reassessed for mental health status, including post-traumatic stress disorder, general psychiatric symptomatology (using the Self-report Checklist-90; Derogatis, 1979), problems in social functioning (Solomon & Mikulincer, 1987), and trauma-related intrusion and avoidance tendencies (using the Impact of Event Scale; Horowitz, Wilner, & Alvarez, 1979). Coping style was assessed on the basis of the MBSS. Building on previous work with this measure, analyses examined main effects of high vs. low scores on the monitoring and blunting subscales separately, as well as their interaction. That is, four groups were created, using the two subscales: high monitors/low blunters; high monitors/high blunters; low monitors/low blunters; and low monitors/high blunters. This allowed for an examination of the effects of information seeking vs. avoidance (high vs. low monitoring) as well as of distraction vs. no distraction (high vs. low blunting) and their interaction.
90
Personal Coping
Results showed that high monitor/low blunter subjects endorsed fewer PTSD symptoms, evidenced less severe general psychopathology, and reported fewer problems in social functioning than all other subjects. They also reported less severe intrusion and avoidance tendencies than subjects in the other three groups. That is, the non-use of monitoring strategies (low monitor/low blunter group), the exclusive reliance on blunting strategies (low monitor/high blunter group), and the addition of blunting strategies to monitoring strategies (high monitor/ high blunter group) all appear to have had a negative impact on mental health and emotional reactions to war. Although only preliminary and correlational in nature, these results raise the intriguing possibility that while low monitoring and high blunting are useful techniques for dealing with more "everyday" stressors, especially of a medical nature, they may be ineffective when dealing with a major traumatic event. Exposure to an event beyond the realm of normal human experience may so disrupt the individual's basic sense of safety, meaning, and orderliness that it activates an intense, generalized, and readily accessible fear structure (Foa, Steketee, & Rothbaum, 1989). The maintenance of post-traumatic symptomatology over time in some individuals may be due to an inability to attend to, process, and work through the feared experience, thereby ultimately setting the stage for both the overwhelming intrusion and avoidance of traumatic imagery. Individuals who do not repeatedly expose themselves to trauma-related cognitions and stimuli are unable to habituate and thereby resign themselves to the painful event. Further, they are unable to incorporate new, corrective knowledge into their stored memory structures. This means that information that is incompatible with their original associations (e.g., "There is nothing to look forward to," "I have no future") cannot challenge these memories. As a result, the painful thoughts are not modified and continue to intrude into the individual's mental life (e.g., through nightmares and flashbacks), leading to the arousal and hypervigilance commonly observed in PTSD. This pattern can also lead to affective numbing, since the threatening cognitions are often not connected with appropriate fear responses (see Foa, 1988). To the extent that the above characterization is accurate, it suggests that low monitors and high blunters may represent a vulnerable population in the face of the persistence of stress symptoms following exposure to traumatic events. Therefore, they may benefit from early interventive programs designed to foster emotional processing and successful adaptation. CONCLUSIONS The present chapter has reviewed evidence relevant to the role of dispositional styles in the coping process, focusing on preferences to seek and/or to cognitively avoid and transform threatening cues. The results suggest that it is possible to reliably identify individual differences in informational styles. Further, high monitors/low blunters appear to suffer greater psychologic, behavioral, and phys-
Individual Differences
91
ical morbidity in response to life events than do low monitors/high blunters. These effects are strongest under high threat conditions and tend to emerge as the process of coping unfolds. Adaptational outcomes of high monitors/low blunters can be partially improved by increasing the amount of information available. These individuals also appear to benefit from gaining access to sources of expertise, who are in a position to help mitigate and offset the aversiveness of the experience on their behalf. Finally, in clinically relevant contexts, there appear to be some benefits of adopting a high monitoring/low blunting mode of coping. Specifically, phobics who are high monitors/low blunters show greater habituation during the standard course of exposure treatment. Further, high monitors/low blunters who experience highly traumatic conditions (e.g., war) appear to be less likely to develop persistent post-traumatic symptomatology. The results suggest several interesting lines of research to pursue, relevant to the origins, maintenance, and modification of these dispositional tendencies. For example, how do these styles develop in young children, how consistent are they across different affective contexts, and to what extent are they influenced by the individual's metacognitive understanding of adaptive coping (Miller & Green, 1985)? We have found that six- and seven-year-old children vary in their preferences for monitoring and blunting strategies, that these strategies can be predicted, in part, on the basis of their dispositional coping styles, and that there is some consistency in their choice of informational strategies when dealing with threat and with the frustration of waiting for a preferred but delayed reward (Miller, Savage, & Mischel, 1988). In addition, we have found that children who show impaired self-regulatory skills also show poor comprehension of effective strategies for successful self-regulation (Gallagher, Miller, & Mischel, 1988). In future research, it will be important to continue to delineate these patterns and to explore the extent to which the execution of coping styles can be taught and shaped to fit with situational constraints.
This page intentionally left blank
7
Interpersonal Aspects of Coping Bruce N. Carpenter and Susan M. Scot
The interpersonal domain constitutes a large portion of our daily functioning. Therefore, it is not surprising that stress and coping theorists have included the role of others in their theories. For example, Lazarus and Folkman (1984) cite the social environment as both a source of stress and a provider of resources for support. Similarly, the very large body of literature on social support focuses primarily on when and how others help one cope. Interpersonal factors may even effect coping in ways that appear far removed from one's present coping efforts; for example, adults who are adept at coping with stress frequently enjoyed childhoods characterized by the personal security that results from warm, supportive relationships (Sarason, Sarason, and Johnson, 1985). In spite of this general recognition that social functioning is important in the coping process, there have been relatively few systematic attempts beyond work in social support to link the interpersonal domain to theories of coping. The purpose of this chapter, then, is to present a model of interpersonal dynamics in the coping process. We will first present a variety of reasons for focusing on interpersonal functioning as relevant to coping. We will then describe constructs we consider important for understanding this area and propose a model for conceptualizing interpersonal aspects of coping. Finally, we will present measurement strategies, discuss several findings from the perspective of the model, and propose additional mechanisms and potentially fruitful areas for study. RELEVANCE OF THE INTERPERSONAL DOMAIN So much of what people do involves interaction with others; hence, interpersonal functioning has the potential to be of great importance in the stress-coping
94
Personal Coping
response. We propose that there are at least three primary ways in which social functioning contributes to stress and coping. First, the social environment in which we necessarily operate produces stressors of an interpersonal nature. That is, relationships, no matter how good, are inherently problematic and place demands on us which can be stressful. Typically a good relationship is not viewed as one that places no demands on either partner, but rather as one in which the benefits to both outweigh the demands. Positive social functioning is often the best way to minimize relational stress. Similarly, coping with the relational demands and interpersonal problems that do arise most often includes adjustments in our social functioning. Second, many stressors and coping activities, which are not themselves interpersonal, take place in a social context. Our social situations strongly influence us and our environments in ways that impact our choice of coping activities and the effectiveness of our coping efforts when faced with nearly any demand, interpersonal or not. Thus, social functioning contributes indirectly to a wide variety of stressors and coping activities. Third, social variables constitute an important set of coping resources. These resources may be personal, including attributes of the individual which contribute to interpersonal functioning, or they may be environmental, including the actions of others and the social milieu in which the individual behaves. These three contributions to stress and coping are described below. Interpersonal Stressors We agree with the contention of Hobfoll (1986, 1988) that effective coping with stress depends on a good match between the demands of the stressor and the resources utilized. Such a perspective underscores the emphasis on relevant resources, rather than on amount or on a single favored class of resources, and encourages the development of a framework for the stress-coping process. If interpersonal stressors are prominent, then we are encouraged to determine which resources are most relevant for coping with such stressors. Our examination of findings leads us to conclude that interpersonal stressors are especially important and social resources are especially relevant to coping with them. The problematic nature of relationships. That social relationships themselves can be a major source of stress is relatively obvious, but until recently has not been a significant focus of the stress literature. In fact, cursory examination of the stress literature might lead one to believe that relationships are not an important source of stress, primarily because discussion of relationships within that literature has mostly focused on social support and its beneficial effects. However, considerable evidence from other research areas, as well as some recent work in social support, has emphasized the problematic nature of relationships. Marriages, parent/child relationships, business relationships, and friendships all have the potential for providing positive support, but at the same time harbor the very real potential for stressful interaction. In much of the
Interpersonal Aspects of Coping
95
previous research, it has often been assumed that social ties with family and friends lent positive support to the individual. In contrast, Rook (1984), in a study of older women, found that 38 percent of those who created problematic situations for respondents were friends, and another 36 percent were family. Further, negative social relationships were found to have more potent effects on well-being than did positive relationships. Hobfoll (1986) has gone so far as to point out that much of what could fit under some definitions of social support is not supportive and might even be negative. Thus, not only are many of the very people on whom one relies for support likely to be a major source of stress, but because those relationships are ongoing, the potential for continued stress is high. In an interesting recasting of the role of social support, a number of writers have suggested that the correlation between support and emotional outcomes arises, at least in part, because a lack of social support is stressful (e.g., Berkman, 1985). That is, the absence or loss of positive relationships places additional demands, probably of an emotional nature, on the individual. For example, from this perspective, isolation is stressful because social bonds are lacking, and relocation is stressful because such bonds are lost. Rather than simply being viewed as ancillary to the process, relationships themselves become the focus as either producing stress when they are lacking or promoting well-being when they are present. A variety of literatures are consistent with this hypothesis. For example, the loneliness literature (cf. Hansson, Jones, & Carpenter, 1984; Peplau & Perlman, 1982) describes the consequences and correlates of loneliness in a way that is similar to those of stress. The widowhood and bereavement literatures emphasize the difficulties which result from the loss of significant relationships. In a review of interpersonal dynamics in the elderly, Hansson and Carpenter (1990) concluded that the act of providing support to others can often be stressful for the provider, producing strains on the relationship that can ultimately affect the support recipient. Thus, it is not surprising that those with major chronic problems experience a decline in support over time as their supporters experience caregiver strain. Horowitz and Vitkus (1986) provide our final example of stress outcomes from poor interpersonal functioning. Calling attention to the frequent occurrence of interpersonal problems in psychiatric populations, they proposed that psychiatric symptoms are related to relational difficulties in important ways. Utilizing a systems approach they point out that clinical symptoms viewed as syndromes typically include interpersonal components which result in social interactions which maintain the system. From their perspective, dysfunctional interpersonal behavior is critical to the development and continuation of the psychiatric disturbance. Stress dimensions. That interpersonal aspects of stress are important is evident from a content analysis of stress measures. Authors of the widely used measures chose to include a number of problems of a relational nature. Because the items chosen for stress measures usually focus on naturally occurring stressors, rela-
96
Personal Coping
tively few are strictly interpersonal; but for many items the interpersonal features are significant. Thus, prominent among most life events scales are death of loved ones, divorce, and marriage. Similarly, the Daily Hassles Scale (Kanner, Coyne, Schaefer, & Lazarus, 1981) includes items inquiring about problems with family members, difficulties with coworkers, troublesome neighbors, and so forth. A number of writers have proposed various types or dimensions of stress (e.g., Hobfoll & London, 1986; Kanner, Coyne, Schaefer, & Lazarus, 1981; Pearlin, Lieberman, Menaghan, & Mullan, 1981), often focusing on features like duration, controllability, and intensity. In contrast, formulations structured around the content of stressors, including social content, are rare. Thus, even though interpersonal stressors are often prominent in descriptions of stress, an interpersonal dimension is not usually proposed. Some of our own work suggests, however, that interpersonal stressors form an important dimension and yield a distinct pattern of coping activities. One of our recent projects (Runge and Carpenter, 1989) specifically examines dimensions of stress. In this study we analyzed judges' similarity ratings of major life events and of daily hassles. Both sets of ratings, when grouped using both multidimensional scaling and clustering techniques, revealed that judges see interpersonal stressors as quite similar to one another. In contrast, these interpersonal stressors tended to be rated as relatively distinct from other types of life events or hassles, depending upon how purely interpersonal the stressor was. The fact that judges tend to use the interpersonal component of stress situations as a grouping characteristic suggests that stressors with such a component tend to be viewed as similar and may lead to similar emotional and coping reactions. Social functioning and other forms of stress. In addition to the stressful potential of relationships themselves, interpersonal functioning appears to contribute to nonsocial forms of stress as well. Several examples might illustrate how the interpersonal domain may add to or ameliorate stress. First, the negative outcomes which result from poor social functioning can themselves act as stressors. We recognize that stressors, coping, and stress outcomes are not distinct entities; rather, they are ways of characterizing a process at a given time. Thus, what is viewed from one perspective as a stress outcome can be viewed from another perspective as a stressor. For example, the depression that often results from poor social functioning is an outcome of ineffective coping; but it also acts as a new stressor, placing additional demands upon the individual. Similarly, failure to resolve marital conflicts can lead to the financial difficulties many people experience following divorce. Second, those with better interpersonal functioning are more likely to develop relationships with people who themselves have high functioning. The resulting environment may often be relatively low in many forms of stress. For example, marriage to a high functioning spouse can contribute to a more orderly home life or to relative freedom from financial difficulties. Getting along well with one's boss can result in better promotions and favored work assignments. A positive parent-child relationship can contribute to development of better esteem,
Interpersonal Aspects of Coping
97
skills, and sense of responsibility in the child, thereby minimizing family conflicts and encouraging a sharing of workload in the home. Finally, some interpersonal skills may lead one to minimize stressful demands which are often placed on us by others. For example, an assertive individual may turn down requests by others, such as coworkers, which could be stressful. Some individuals may accept demands from others because they believe that such willingness and the subsequent efforts in behalf of others enhance relationships with those others. But the presence of other skills for enhancing relationships can make one feel less dependent upon playing a burdensome and overwhelming instrumental role in an effort to maintain social ties. The Social Context of Stress and Coping Although we do not always view it as such, the environment in which we normally operate depends heavily on an interpersonal component. Regardless of the specifics, we are usually obliged to deal with stress and its concomitant, coping, within the context of our interactions with other people. As a result, how successful we are at arriving at appropriate and adequate coping strategies can depend upon the social context in which we find ourselves. Certain demographic characteristics, such as marital status, socioeconomic level, or gender, have been cited as correlated with vulnerability to stressors (Pearlin, 1985a). Raising a handicapped child, for example, can present very different problems and be variably stressful depending upon the social environment in which the family lives. Being poor, uneducated, and unmarried may present a much more stressful scenario than would being married, employed, andfinanciallycomfortable when faced with the demands of a handicapped child. It is also clear that, given identical situations, people will respond differently depending upon the social context in which the stressor occurs. Retirement, for example, can result in depression in the individual who is suddenly cut off from activities and people he or she enjoyed. Retirement can also be the source of joy for the individual who disliked his job and feels a sense of freedom and relief. It is not the specific event, necessarily, which determines the degree of stress, but the consequences of the event given the social context in which it occurs (Pearlin, 1985a). Social Resources Coping and coping resources. When the presence of a stressor is sufficient to tax one's readily available resources, a coping response results. In keeping with Lazarus and Folkman (1984), coping is defined as whatever one does when a stress response occurs. The coping response, therefore, often entails drawing upon additional resources. Our focus in this chapter is upon two sets of social resources, social support and relational competence. Social support is most often seen as an environmental resource—what others do for us—although some have
Personal Coping
98
highlighted the fact that it operates much like an individual difference variable, characterizing the person as much as the social environment (e.g., Sarason, Sarason, & Shearin, 1986). In contrast, relational competence is characterized here as a personal resource, representing the interpersonal skills of the individual (Hansson etal., 1984). SOCIAL SUPPORT AND RELATIONAL COMPETENCE In the preceding section we attempted to show that many important coping situations are interpersonal in nature or take place in an interpersonal context. We believe that two classes of social variables mediate such difficulties. First, there are characteristics of the relationships themselves. Although there is considerable disagreement on the perspective that best characterizes a relationship, for our purposes here we will limit ourselves to what others do to us and for us. Second, there are individual difference variables, highlighting differences in the skills and preferences people possess for a variety of social tasks related to coping. These tasks include, at the least, solving or soothing interpersonal difficulties, functioning well in a variety of social situations so that coping options are increased, and effectively entering into and utilizing relationships in time of need. The model proposed below characterizes these two classes of variables as two sets of coping resources. We believe that the first set is well represented by the construct of social support. Although not synonymous with the definition offered above, social support has much to do with what others do to us and for us, acting primarily as an environmental resource. The second, individual differences in interpersonal skill, is well represented by our construct of relational competence, defined as those personal skills which contribute to effective interpersonal functioning. Such characteristics are best thought of as personal resources. We briefly highlight below some of the main features of social support and then define in some detail our construct of relational competence. Social Support It is not the intent of this chapter to discuss the construct of social support in any detail. Biegel, McCardle, and Mendelson (1983), for example, recently abstracted over 1300 references on the topic, highlighting the complexity and breadth of research in the area. For more complete definitions and reviews of social support, readers are referred to Cohen and Syme (1985a), Gottlieb (1981), and Sarason and Sarason (1985). Overview. In spite of considerable research, our understanding of social support has moved forward rather slowly, in part because we lacked clear definitions of the construct (e.g., Thoits, 1982). Easily thought of as a rnultifaceted construct, numerous definitions have emerged. Most of the competing definitions can fall under a broad characterization of social support as aspects of the social environment that facilitate our survival and well-being. However, in research it
Interpersonal Aspects of Coping
99
is necessary to narrow the concept so that a measurement strategy emerges; and from a theoretical perspective it is highly desirable to differentiate the components of any broad definition and ascertain their respective contributions and the interplay between them. To help organize various definitions, Lin (1986) has formulated a three-dimensional model. This model distinguishes between perceived (subjective) and actual (objective) support, between the layers of relationships (community, network, and partner) from which support derives, and between two provisions (instrumental and expressive) of support. Most definitions of social support, having made similar distinctions or having chosen to emphasize only a subset of these components, can be examined from the context of such a model. For example, Gottlieb (1985) argued for a focus on supportive behaviors, carving off perceptions as related, but separate; whereas Sarason, Levine, Basham, and Sarason (1983) highlighted the role of the recipient's perceptions. Similarly, Barrera (1981) focused on network size, in contrast to Caplan (1974), who underscored the need for loving, caring relationships as found primarily in close partnership. Much of the social support literature presents studies which examine the covariation between support and outcomes. Although the relationships are usually modest, support has been shown to be related to health (e.g., Cohen & Syme, 1985a), mood states (e.g., Lin, Dean, & Ensel, 1986), and psychiatric disorder (e.g., Monroe & Steiner, 1986). The mechanisms of action are not yet understood (e.g., Pearlin, 1985b), although studies appear to support two beneficial effects which might underlie the correlations between social support and positive outcomes. First, some forms of social support provide tangible benefits. Supportive others can provide money or other goods, they can give us needed information, or they can perform some task which benefits us. From the perspective of coping, then, such support is of use primarily because others either absorb some of the environmental demands which would otherwise fall upon the coping individual, or increase tangible resources available to the person. Although mostly untested, we might hypothesize that such support would best help one cope when the stressors are tangible or when the coping response is problem focused and utilizes tangible resources like information or money. The second mechanism by which support appears to benefit the individual is that social support contributes to the image one has of self and the world. Thus, good support might lead to a sense of well-being and mastery or esteem, and might cause one to feel that problems are surmountable and solutions are within one's grasp. Such emotional support might, during the act of coping, encourage a positive emotional state which counteracts that brought on by stress and leads to more positive appraisal. Similarly, it might over time impact personal resources through lasting changes in self-image. Relational Competence Definition. In our work over the past few years we have come to believe that individual difference variables often mediate the supportive and problematic effects
100
Personal Coping
of relationships. In this context one of us, along with several colleagues, proposed the construct of relational competence (Carpenter, Hansson, Rountree, & Jones, 1983; Hansson et al., 1984). We initially defined relational competence as "those skills which contribute to the acquisition, development, and maintenance of relationship" (Hansson et al., 1984, p. 273). Relational competence was hypothesized as important for four relational tasks: (1) initiating relationship; (2) developing and strengthening relationships; (3) using relationships in time of need; and (4) maintaining relationships. Being tied to an individual differences perspective, relational competence is viewed as a personal coping resource. Although originally described from the context of social support, it has become clear to us that relational competence is of broad relevance to the field of relationships. The success of relationships often depends upon it, and the benefits of relationships are often mediated by it. Further, relational competence becomes quite important to the stress and coping process because it both emphasizes competencies which are relevant to coping with interpersonal difficulties and is an important contributor to the quality and effectiveness of those relationships. For example, in discussing coping with relational difficulties among the elderly, Hansson and Carpenter (1990) proposed three ways in which relational competence is relevant to the coping process: (1) many of the stressors encountered by the elderly are interpersonal in nature and best solved with interpersonal skills; (2) a large proportion of coping takes place within an interpersonal context, and those most effective in that setting will have greater coping options available; and (3) positive relationships contribute to effective coping generally. These three contributions of relational competence appear relevant to the nonelderly as well and parallel the three we offered above when discussing the relevance of the entire interpersonal domain. Two-component model. When we consider the great variety of relational tasks it is evident that many social competence variables are of potential relevance. Although we recognize that certain variables are likely to be more central to the construct and of greater utility, most of our studies, and those of our colleagues, have focused on the relational competence variables which held the greatest interest. However, after several studies a pattern emerged which paralleled perspectives from related literatures. For example, in a study of interpersonal correlates of loneliness (Jones, Carpenter, & Quintana, 1985), factor analysis revealed that the largest factor included variables such as assertiveness, masculinity, and shyness. However, several other social competence variables, such as love, trust, and altruism, strongly correlated with loneliness but did not load on this factor. A similar distinction is made in several theories of interpersonal functioning (e.g., Horowitz & Vitkus, 1986; Wiggins, 1982), which propose orthogonal dimensions such as dominance-submission and love-hate or control and affiliation. Based on this evidence, Carpenter (1987, 1989) refined the construct of relational competence to focus on two relatively independent sets of skills. The first component, labeled "Initiation," emphasizes characteristics which are important for making initial contact with others, developing relationships, and utilizing re-
Interpersonal Aspects of Coping
101
lationships in time of need. Initiation skills might include assertiveness and social presence, for example. Initiation skills appear to be socially valued and associated with status, and so are highly correlated with self-esteem. The second component, labeled "Enhancement," includes characteristics which are most relevant to strengthening relationships, fostering closeness, and meeting the needs of others so that the reciprocal demands of a relationship can be met. Examples of enhancement skills might include intimacy, empathy, and altruism. Although enhancement skills are also socially valued, their usefulness exists primarily within a close relationship, and they are therefore less global in their influence. As noted earlier, our emphasis is on a match between resources and coping demands. The two-component model of relational competence allows for distinctions that can clarify such a match. For example, we have already suggested how each appears best suited for achieving different goals. Skill differences in the two sets of competencies might also explain many of the sudden shifts in coping effectiveness experienced by many people over time. For example, one person might have the enhancement skills for an effective marriage, deriving satisfaction from the relationship and coping more effectively overall; whereas another person, lacking enhancement skills, may not derive such benefits from the marriage. However, their relative coping effectiveness might be reversed should they both become widowed and the second individual had stronger initiation skills needed to develop new partnerships. Measurement of Relational Competence Development of the relational competence scale. This two-component model was used to develop the Relational Competence Scale (Carpenter, 1987, 1989). From the potential domain of variables which could be included for the initiation and enhancement components five were selected to represent each component. They were chosen to emphasize skills which are broadly recognized as important in relationships, and which represent the domain of possible variables reasonably well. The five variables included in the initiation component are: assertiveness, dominance, instrumental competence, lack of shyness, and lack of social anxiety. The five variables included in the enhancement component include: intimacy, trust, interpersonal sensitivity, empathy, and perspective taking. The scale was developed using a rational-theoretical approach and the iterative procedures of Jackson (1970), yielding ten subscales of ten Likert-format items each. Each set of five subscales sums to yield a component score. The subscales have good internal consistency reliability (.80 to .88) and test-retest stability over twelve weeks (.76 to .88). Similarly, the initiation and enhancement component have high reliability (.95 and .94, respectively) and stability (.84 and .82, respectively). Both exploratory and confirmatory factor analyses support the twocomponent model. Initial validity. The subscales of the Relational Competence Scale are strongly correlated with validated measures assessing highly similar constructs, offering
102
Personal Coping
evidence for concurrent validity. Similarly, ratings by close others (spouses and best friends) yield substantial correlations with scale scores, suggesting that selfratings are reflected in observable behavior. Several studies show relationships which highlight the usefulness of the relational competence construct generally and the two-component distinction particularly. For example, in a study of married couples (Davis & Carpenter, 1987), marital satisfaction was found to be most strongly related to one's own initiation skills and to the spouse's enhancement skills. Interactions Building social support. Social support has most commonly been viewed as what others do for us. Thus, the individual has largely been regarded as a passive recipient of support. In contrast, several writers have suggested that characteristics of the individual might be important in the process (e.g., Cohen & Syme, 1985b; Schaefer, Coyne, & Lazarus, 1981). Specifically, Hansson et al. (1984) proposed that relational competence would be important in the development and utilization of social support. They reasoned that those with greater relational skills likely fostered a better support network, were more likely to construe relationships as supportive, and were better equipped to access relationships in time of need. In a test of these hypotheses, one of us (Carpenter, 1985) measured both social support and relational skills in college students, relating these measures to indicators of emotional functioning. Although not a true test of the hypothesis that interpersonal competence leads to social support, the study found that such competence was a better predictor of outcomes than was social support, and that the predictive variance of social support was almost completely subsumed by the competence measures. The latter finding emphasizes the close connection between interpersonal skills and social support. This meshes nicely with the finding of Sarason, Sarason, and Shearin (1986) that social support functions in many ways like an individual difference variable, perhaps reflecting a somewhat stable internal state or belief more than a simple observation of the highly variable support behaviors of others. Using social support. Even when social support is available, effective use of that resource may be moderated by characteristics of the individual. The act of asking for help emphasizes many of the problematic dynamics of relationships. The one seeking help becomes a receiver in the relationship, thereby possibly creating a sense of obligation or disrupting the status quo in the balance of exchange. Asking for help can imply an admission that one is not personally up to the task, contributing to a lowered sense of mastery and a fear that others will develop a poorer opinion of the recipient. Similarly, appropriate use of others in times of need can contribute positively to relationships. Sharing one's needs and vulnerabilities can enhance closeness. Using others' strengths can
Interpersonal Aspects of Coping
103
create esteem in the other person, adding a reinforcing quality to the act of providing help. Although help-seeking behavior has been the focus of much research, it has not been well integrated with research on social support and coping. Similarly, Gore (1985) criticized the tendency to separately research individual coping and social support, making it more difficult to understand how help seeking fits into one's coping response. Thus, it remains rather unclear which individual differences lead one to appropriate use of available social support. The issue becomes especially complex when we consider studies which show that help-seeking is related to poorer outcomes (e.g., Pearlin & Schooler, 1978). This leads us to wonder how social support can be beneficial if help-seeking has a negative impact, an apparent contradiction that is not yet explained. We propose two ways in which individual differences might contribute to effective use of available social support. First, individuals may differ in the skills used to ask for help in a way that prompts others to give support. A person may perceive one's own lack of skill for asking, and simply not ask. Alternatively, one might ask in an undesirable way (e.g., whining, demanding, or clumsy), thereby reducing the desire in others to comply. Characteristics that lead to excessive or inappropriate requests for support will likewise often lead to a withdrawal of support. In contrast, some might be so skilled in asking that others gladly comply with the request, or that the help-seeking nature of the exchange (along with the problems of help-seeking mentioned above) is not obvious. Second, personal characteristics might lead one to view help-seeking as fearful or undesirable. Those for whom past support-seeking often resulted in rebuff may anticipate rejection and, consequently, avoid exposing themselves. People who overvalue independence may view asking for help as a sign of weakness or may wish to avoid the unpleasant, potential consequence of obligation to others. We have found that relational competence is related to reported frequency and effectiveness of various coping activities involving the use of others. In particular, even though initiation skills are positively correlated with indices of social support, they are negatively correlated with frequency of help-seeking and reliance on others as coping behaviors. In contrast, enhancement skills have a weaker, although still significant, positive correlation with perceived social support and a positive correlation with preference for coping activities that involve the use of others. It appears, then, that interpersonal skills such as assertiveness and dominance are related to development of social networks and to beliefs that one has good support, but at the same time mostly lead to presenting oneself as independent. This causes us to speculate that those high in initiation skills primarily derive emotional support from their support networks, allowing them to see themselves as not really asking others for help. Alternatively, they may also ask for and receive tangible support, but are more subtle in their requests so that they avoid seeing their behavior as help-seeking.
104
Personal Coping
Figure 7.1 Interpersonal Model of Coping
AN INTERPERSONAL MODEL OF COPING We have developed a model of how social resources contribute to the coping process. Although the model might apply equally well to the process whereby other types of resources impact the coping process, our description focuses on interpersonal features. The model, as shown in Figure 7.1, separates resources into personal and environmental types, with relational competence and social support representing the two types. The process of coping is, of course, not static, and we realize that each component of the process can be affected by changes in other components. Thus, feedback loops are assumed. However, we believe that those noted in our model are the primary pathways in a given situation. Thus, implied feedback pathways are assumed to usually have less powerful effects, except perhaps during early childhood when skills and beliefs about self are still quite flexible and during periods of prolonged or extreme stress when significant disequilibrium and reevaluation of self occur (cf. Schaefer & Moos, Chapter 10).
Interpersonal Aspects of Coping
105
The model proposes that personal and environmental resources lead to beliefs about oneself and about others. It is these beliefs that provide the context in which cognitive evaluation of potential stressors occurs. Coping occurs when the appraisal process yields the conclusion that demands exceed readily available resources, with the form and effectiveness of this coping in part determined by the personal and coping resources of the person. The coping might be an effort to reduce demands by changing the stressor, to increase available resources, or to manage the stress response itself. These coping efforts then lead to outcomes, including the psychological and physiological concomitants of the stress response, as well as broader outcomes like health and well-being. Finally, these outcomes provide a new starting point for the process. We describe below the contribution of social resources to appraisal, coping, and outcomes. Beliefs Although beliefs might be thought of as personal resources, we separate them out in our model to better illustrate how they are affected by other individual differences and by environmental resources. The model focuses on two sets of beliefs, those about self and those about others. These beliefs arise primarily from our observations of ourselves and others, and from the feedback we receive in our interactions with others. These two sets of beliefs are described below, including reasons for our expectation that relational competence will have an especially strong impact on these beliefs. Beliefs about self. Relationships are an important source of feedback with which people assess themselves. Clinicians have long realized that involvement in meaningful and fulfilling relationships is related to self-esteem. Much of the loneliness literature confirms this (e.g., Peplau & Perlman, 1982). Similarly, the child development literature emphasizes the role of others in the development of self-image and beliefs about one's abilities. Although clear data are lacking, a number of theorists have proposed that positive feedback about one's worth is very likely one of the ways in which social support contributes to positive outcomes; those who feel that others care about them are encouraged to view themselves as worthwhile and capable. Similarly, those who have the skill to foster good relationships are more likely than others to perceive their own skill, and hence, view themselves positively. Such relational competencies are also expected to exert through social support an indirect effect on self-beliefs; this occurs because they are more likely to receive positive feedback from the resulting good relationships. Relationships, then, are important for the development of many positive beliefs about oneself. Several such beliefs stand out because of their demonstrated importance in the coping literature. Self-esteem and mastery (e.g., Pearlin & Schooler, 1978) have been shown to be important personal coping resources. Those who have a general sense of self-worth appear to view stressors as less threatening, largely because they believe they are capable of handling the de-
106
Personal Coping
mands of the situation. Similarly, Bandura (1977) has shown that self-efficacy, or the belief in one's ability to deal with a particular demand, is related to a reduced stress response and more positive outcomes. Mastery appears to function much like self-efficacy, but is not necessarily tied to a particular demand. Beliefs about others. It is useful to distinguish beliefs about others from what others actually do for us. Specifically, beliefs about how others will respond and help are what we use to evaluate our situation when demands are placed upon us. Others may or may not respond as we expect, but we can only utilize our expectations about their behavior when evaluating our resources (cf. Ursin & Hytten, Chapter 11). The enduring nature of relationships, however, affords us some predictability. We form our beliefs about how others will contribute to our needs in the coping process largely based on our perceptions of our social networks and the support historically forthcoming or available from those networks. Thus, the nature of one's environmental resources directly contributes to such beliefs. It is not uncommon, however, to find that many people will persist in their beliefs about others' support when evidence suggests they are wrong. For example, some people believe that others will not help and persist in that belief even when considerable support is provided; in contrast, some people may believe that support is available, or even present, in spite of the fact that none is offered in that particular situation. We believe, then, that individual differences contribute to one's perception of the help available from others. This is consistent with the characterization of relational competence presented by Hansson et al. (1984), which they described as affecting how people construe relationships. For example, our data suggest that relational competence is more strongly related to subjective indices of social support, including beliefs about the support that would be available if it was needed, than to objective indices of support. Appraisal Just as beliefs about self and others are not completely distinct from personal resources, these beliefs are difficult to distinguish from appraisal. It is this similarity and overlap in concepts that causes us to see beliefs as the bridge between resources and appraisal. In keeping with Lazarus and Folkman (1984), appraisal is defined as the cognitive evaluation of the stressor's demands and one's ability to respond to the demands. Thus, in appraisal, conclusions are drawn about the adequacy of resources. We suspect that whereas some beliefs about self and others focus on what one can do and what others can do for that person (nature of resources), many beliefs are more evaluative (adequacy of resources). To the extent that the belief is an evaluation of resources, appraisal is distinct from the belief only in that the belief is applied to the situation at hand. That is, beliefs are seen as ongoing, whereas appraisal is the application of beliefs to a particular situation when demands appear excessive. For example, if one holds a belief that the resources for handling conflict are poor, when involved in a major marital disagreement appraisal will likely apply that belief
Interpersonal Aspects of Coping
107
to the present conflict, with the conclusion that the task is unmanageable. The appraisal process is more complex when evaluative beliefs are not already available, in that conclusions about resource adequacy result from comparisons between beliefs about the nature of resources and perception of the nature of demands. Through experience, beliefs about self and others might become associated with particular environmental demands (as in self-efficacy for a certain task). Yet, many evaluative beliefs are general, reflecting a rather global belief in the adequacy of one's resources (more like self-esteem or mastery). Because of their generality these global beliefs appear to be particularly powerful contributors to the appraisal process. Beliefs about one's general adequacy for meeting classes of demanding situations (such as interpersonal difficulties), are easily applied. Perhaps this accounts for the prominence of self-esteem and similar beliefs of a global type; people tend to use them first when evaluating demands, and it is primarily when such appraisals are inadequate that more specific beliefs are evaluated. From this perspective, then, one's beliefs about the two main classes of social resources, social support and relational competence, should strongly influence appraisal. We hypothesize that they are general and readily accessible; thus, people will rely on them heavily when appraising. This should be especially true for appraising social stressors, but because of their generality it seems likely that they will be utilized for appraising nonsocial stressors as well (the relevance of social factors for coping with a wide variety of stressors was discussed earlier). For example, even though many measures of social support inquire about availability of support in a wide variety of situations, such scales tend to have high item intercorrelations. Further, social support is usually defined by some global score, and scales measuring many specific support situations do not appear superior to those measuring a few broad situations. To clarify the role of social resources and beliefs in the appraisal process we have assessed relational competence and appraisal in persons experiencing a variety of stressors. When faced with a clearly relational stressor, relational competence was highly correlated with appraisal (Carpenter & Suhr, 1988). Those with high competence viewed the stressors as salient and their own performance as important, but perceived the task as less demanding and emotion arousing and anticipated better performance. Appraisal and relational competence were also significantly correlated for those facing nonrelational stressors, although the magnitude of the relationship was much weaker. Such findings are consistent with our hypothesis that beliefs about one's relational skills are very important in appraisal and, to varying degrees, are used when appraising a wide variety of stressors.
Coping We define coping as activities engaged in with the intention of reducing or avoiding stress by reducing demands, increasing resources, creating more fa-
108
Personal Coping
vorable beliefs and appraisals, or reducing the emotional reaction to stress. Unlike the view of coping offered by Lazarus and Folkman (1984), for example, we emphasize intent and allow that relatively automatic behaviors may qualify as coping activities. In our model coping is activated by appraisal, not by stress. Interpersonal factors appear to contribute to coping in three primary ways. First, many coping activities are interpersonal behaviors. Second, social resources and beliefs can influence one's choice of coping activities. Third, interpersonal resources affect the skill with which selected coping activities are carried out. Relational coping. Social factors are relevant to coping primarily because so many coping activities are interpersonal. Communication, sharing, social comparison, negotiation, intimacy, and other common interpersonal behaviors become coping behaviors when enacted in response to stressful appraisal. Other activities, such as support seeking and confession, primarily appear in stressful situations but are clearly interpersonal. It is not surprising, then, that measures of coping like the revised Ways of Coping Inventory include items like *'Tried to get the person responsible to change his or her mind," and "I expressed anger to the person(s) who caused the problem." Many interpersonal coping activities are problem-focused—that is, they are designed to reduce the imbalance between demands and resources. According to our definition of coping, this would occur through two processes, reducing demands and raising resources. Relational coping appears best suited to reduce demands when the demands themselves are interpersonal. Thus, one who is tactful and patient may do best at defusing an argument. In contrast, relational coping appears relevant to raising resources whether or not the stressor is interpersonal. For example, one might use friends to access needed information or to raise money; similarly, support seeking might result in improved self-esteem. Relational coping activities might also be emotion-focused, emphasizing the reduction of the stress response and alteration of the original appraisal. For example, many people find social activities relaxing and distracting. The problems don't go away, but attention is redirected and the enjoyment of social interaction is in marked contrast to the stress response. When one talks over problems with friends new ideas may be offered which change neither demands nor resources, but result in a new way of viewing the problems. Coping choice and coping skill. Social resources and beliefs influence our choice of coping activities and affect how well the activities are performed. It is the combination of these two factors, choice and coping skill, which determine coping effectiveness. One's choice of coping activity depends on many variables, including the type of stressor, habit, and one's appraisal of stressor and resources. Thus, finding oneself in a new city naturally encourages one to choose a response of meeting new people and developing friendships. It is also likely that beliefs about the availability of, skill in using, and past success with social resources contribute to one's choice of coping activity. For example, one who has friends and has successfully drawn emotional support from them in the past is more
Interpersonal Aspects of Coping
109
likely to use emotional support when stressed than would someone without either appropriately supportive others or a history of benefiting from emotional support. One might choose a coping response because it has worked in the past even though it is clearly ineffective for the demand at hand. For example, a highly independent individual may persist in trying to solve a problem alone when the help of others is both available and needed. Although the mechanism by which social variables contribute to this choice is probably no different than for other variables, these social factors appear especially influential. Whereas coping effectiveness positively reinforces a coping choice, it is clear that choice and effectiveness are not synonymous. Even though a situation naturally suggests a set of coping activities, if the person has poor skill in using such coping, high effectiveness is unlikely. A match between coping activity and resources is critical to such effectiveness. In light of the prominence of relational coping activities, those with high relational competence and social support are in the best position to effectively engage in a variety of coping activities.
This page intentionally left blank
8
Perceived Control, Personal Effectiveness, and Emotional States Herbert M. Lefcourt
If one wished to create a work of fiction that contained elements of shock and horror with nightmarish overtones, the situations described would obviously be extremely stressful, and more importantly, the victims in those situations would more than likely be frozen in a state of inertia, feeling helpless to avert the inevitable consequences of their dire circumstances. Films such as Alfred Hitchcock's Psycho, for example, present an innocent victim who unknowingly, and thus helplessly, falls prey to inevitable calamity while the audience writhes in vicarious dread, seeing a bizarre and horrible murder unfold. The very sight of old, decrepit, and isolated houses such as that depicted in Psycho probably still has the power to arouse fears among many of those who were terrified by the film some two decades ago. Though shock and surprise could be seen as primary causes of the terror, to the audience, especially those who knew or could guess the plot, the inability of anyone to avert the awful event seen on the screen probably amplified the shock and horror produced by it. In fact, the very passivity and its accompanying helplessness that an audience necessarily experiences while watching most entertainment probably enhance the effects of entertainments that are intended to be frightening. It makes imminent sense that stressful circumstances which cannot be altered will have a strong impact upon the victim's sense of well-being. If one were to suffer some excruciating circumstances and were unable to move, to do something to either end the situation or to lessen its impact, it would seem possible that it could go on forever, or lead inevitably to even worse ends. To stand by helplessly and watch one's loved ones, valued possessions, or one's own life being slowly destroyed could be the ultimate of horrors, something that perpetrators of torture seem to be all too well aware of. This may also account for the greater fear with which we regard slow, degenerative diseases, such as
112
Personal Coping
cancer or acquired immunodeficiency syndrome (AIDS), as opposed to rapid, acute maladies, such as myocardial infarctions. I grant that anticipated pain may also be an important factor. The condition of helplessness or personal externality serves to exacerbate the fears that we are likely to experience in the encounter with stressful experiences. On the other hand, to be able to do something, no matter how futile, would seem to be a better way to encounter dire circumstances, minimizing the horror that passivity in the face of stress might engender. In my monograph (Lefcourt, 1982) and in the first volume of my edited series (Lefcourt, 1983) concerned with locus of control, descriptions have been offered of several real-life stories in which the protagonists underwent extremely stressful experiences, such as airplane crashes, boating disasters, and incarceration, each with its prolonged sequelae. In each story, the protagonist's beliefs about control seemed to have played a major role in determining how that person grappled with the duress that was experienced. In one survival story (Thompson, 1975), the ordeal of three persons who were shipwrecked in the Pacific Ocean was recounted. They had been trapped in a foot and a half of air space in the hull of an overturned trimaran for seventy-three days. The victims were the ship's owner, his brother-in-law, and the brother-in-law's wife. The woman died early during the travail and the two men were eventually rescued. The owner of the boat was a devout Seventh Day Adventist who insisted that their survival and rescue were up to God. Given this self-abnegation, he remained passive throughout the ordeal except for his ' 'challenging" (exhortatory exclamations) God to deliver them. When this approach failed to alleviate their duress, he easily succumbed to hopelessness and despair such that even following their rescue he remained distraught and died shortly thereafter. The brother-in-law, however, survived what had been an even worse calamity for him. His wife, who was in the early months of pregnancy, died by his side in their tiny living space and remained in close proximity during the early stages of bloating and decay until her husband engineered her exit from their tight quarters. Despite the unbearable set of events, this survivor remained an active, deliberate person who refused to capitulate and accept helplessness. As an ardent believer that "God helps those who help themselves" he had organized all of their available resources from the moment that the three of them had become imperiled. In addition to marshaling and organizing their nutritional and physical resources, he had arranged "intellectual" and "social activities" for the three of them so as to counteract the panic, numbness, and depression that threatened to overwhelm the others. This strong believer in self-responsibility mused at the end of his story about the differences between himself and the boat's owner who had already died: "I proved that if a man husbands his energy and uses his power resourcefully, then that man can pass—marked, but basically unharmed—through the most excruciating of ordeals" (p. 238). In a second personal account of the ways in which individuals come to terms with stress, I had described the ordeal of Alexander Dolgun (Dolgun & Watson, 1975), a young American who had been abducted by the Soviet secret police
Perceived Control
113
from the streets of Moscow in 1948 and lived his next twenty-four years in various prisons and labor camps throughout the Soviet Gulag. Dolgun's story contained all of the horrors and tortures that Solzhenitzyn has described. What makes his story unique and relevant for this discussion is the resourcefulness that Dolgun displayed in the face of unrelenting cruelty, along with his recall of the thoughts and fantasies that occupied his mind during captivity. Though many of Dolgun's reminiscences implicate perceived control as a tool for his survival, there is one point in his narrative that does it with a sharp sense of drama. Dolgun was being interrogated nightly by a particularly savage official who took pleasure in kicking him, hitting him with a truncheon, and otherwise mutilating him. He was not permitted to fall asleep at all, though a bed was lowered from the ceiling each night. Lights would be shined in his eyes and he would be beaten if he were to fall asleep. The bed was subsequently returned to its position in the ceiling at daybreak so that he would have no opportunity to rest comfortably during the daytime. Sensing that this particular episode in prison could mark the beginning of the end for him, Dolgun began to save the tiny scraps of coarse paper that were issued to him daily to be used as toilet paper. He hid these bits of paper and chewed them carefully to create a papier-mache wad with which he could fill up the striker plate into which the bed latched when it was placed in the ceiling. He reasoned that he could then replace the bed so that it appeared to be firmly locked in place. These machinations were employed so that if he felt that his agony was increasing beyond endurance, he would be able to use the bed as a quasi-guillotine and end his life. As the author noted: "It was not pleasant... to make such a p l a n . . . yet it gave me a small but psychologically necessary handhold on some elements of control over my future, and it gave me access to an escape route if what was happening got beyond the point where I could bear it" (p. 121). After Dolgun succeeded in rigging up his escape device his mood changed from one of apprehension and fear to one of buoyancy. He described the succeeding interrogation as something of a surprise both for himself and the interrogator. In response to the interrogator's questions, Dolgun answered in ways that were so preposterous that he enraged the official who took to beating him harshly. Dolgun, however, did not cower or exhibit the terror that beatings usually engendered. Rather, he found himself laughing at the interrogator which infuriated the official further. This very sharp turn in affect following the return of a feeling of control and efficacy argues eloquently for the value of control as a determinant of the response to stressful experiences. In each of the disaster stories described in the locus of control books, the protagonists revealed similar faith in their ability to cope with adversities, and described junctures in their experience with anguish wherein this faith was exercised. From these biographical anecdotes, then, it is possible to conclude that beliefs about one's ability to control the events in one's life play a major role in the way one confronts crises. The person who has retained faith in his or her
114
Personal Coping
capability to act will persist in the effort to deal with threats and stresses in his or her life. Among the most frightening stressors that humans are heir to is the sudden realization that their health is beginning to fail, as in the experience of a heart attack or a stroke. Norman Cousins (1979, 1983) has written about these kinds of experiences and the subsequent panic reactions that can exacerbate their effects. In his case, the disorders were ankylosing spondylitis and myocardial infarction. In his first book, Cousins described the very serious treatment that he received for this acute collagen disorder from which recovery was unlikely. The treatment he had received at the hospital compounded his feelings of helplessness and despair such that prognostic indicators became progressively worse. It wasn't until he took an active role in his own treatment and began exerting judgment as to what was to be done to him that his downward spiral became arrested and his recovery began. Likewise, following the myocardial infarction, the very urgent behaviors of the ambulance crew that took him to the hospital, and the hospital staffs' efficient and serious responses to him served to increase his own feelings of helplessness and anxiety, which he attempted to combat with humor and relaxation. Throughout his books, Cousins emphasizes the point that acute illnesses are events that can be interpreted and treated in ways that accentuate or diminish their stressful properties. This segmentation of events per se from the responses we make to them is similar to that promulgated in the psychological framework concerning stress and coping processes by Richard Lazarus and his colleagues (Lazarus & Folkman, 1984). In this theoretical and empirical treatment of stress, events are found to have their impact only as they are mediated by appraisal processes or the victim's interpretations of the stressful event. The suffering potentially involved in physical traumas would seem to be exacerbated when these perils are appraised as being threatening and are experienced passively, helplessly, and seriously. Though testimonies to the importance of beliefs about control are impressive, their importance could be dismissed if there were no substantial empirical data to support their contentions. To this end, several field studies were described in the aforementioned books about locus of control that offered some support for the assumed importance of control for helping to withstand disasters. Among them is a fascinating study of the ways in which new myocardial infarction patients come to terms with their life threatening experiences. Cromwell, Butterfield, Brayfield, and Curry (1977) were able to examine the behavioral correlates of locus of control within an intensive care unit. Patients classified as internal from Rotter's Internal-External Locus of Control Scale (Rotter, 1966) were rated by the professional staff as being more cooperative and less depressed during their stay in the intensive care unit than were externals. On three highly intercorrelated physiological measures (sedimentation rates, serum glutamic oxaloacetic transaminase levels, and lactate dehydrogenase levels), externals were found to have worse prognoses than internals. Additionally, externals had higher
Perceived Control
115
peak temperatures during intensive care and remained longer in the unit and in the hospital than did internals. Though these data may be interpreted in different ways, one tempting conclusion is that internals behave in ways that do not aggravate their fragile conditions. Rather than becoming distressed, with all of the personal and physiological consequences of that state, internals evince greater cooperation and less depression than externals, possibly reflecting their more active participation in the struggle for survival. Thus, it would seem that responses to life endangering threats such as myocardial infarctions may be at least partially determined by personality characteristics such as locus of control. As was the case in Normal Cousins' description of his response to treatment for myocardial infarction, the subjects in this field study who appeared to adapt more readily to the highly stressful circumstances in which they were suddenly immersed were internals who might be expected to behave in a more self-directed fashion when undergoing acute stress. Another suggestive field study (Anderson, 1977) concerned the responses of businessmen to the disaster of flooding which all but wiped out their commercial enterprises in a small Pennsylvania community following Hurricane Agnes in 1972. Anderson (1977) examined the business performance of ninety businessmen over a three-and-a-half year period following the flood. The businessmen were assessed for locus of control beliefs, perceived stress, coping behaviors, and organizational performance. With coping behavior classified as problemsolving versus emotion-focused, externals were found to have used fewer problem-solving coping methods and more emotion-directed coping devices (withdrawal, hostility, etc.) than internals. In addition, externals were more likely to have perceived their circumstances as being highly stressful than were internals. Organizational performance was assessed by credit ratings of their businesses after the three-and-a-half year period had ended. The author concluded that "the task-oriented coping behaviors of internals are apparently associated with a more successful solution of the problems created by the stressful event, since the performance of the internals' organizations is higher" (p. 450). Recently, Solomon, Mikulincer, and Avitzur (1988) have found some corroborative evidence linking the stress of war, locus of control, coping styles, and post-traumatic stress disorders. With a sample of 262 Israeli soldiers who had fought in the 1982 Lebanon war, locus of control (scored in a positive direction) was found to be negatively correlated with the intensity of posttraumatic stress disorders at both two and three years after the war (r = — .38 and —.29, respectively). Veterans of that war who were external with regard to locus of control, then, were more apt than were internals to suffer with stress disorders. Coping style was related to locus of control as well, with externals being more emotion-focused copers than were internals. In turn, emotion-focused coping was positively related to the intensity of post-traumatic stress disorders and, in a regression analysis, proved to be the more potent variable for predicting
116
Personal Coping
that disorder. These findings bear similarity to those of Anderson, implicating both beliefs about control and coping style in the responses made to traumatic events. In another study assessing the stressful impact of participation in the Lebanese war, Hobfoll, London, and Orr (1988) found that Pearlin and Schooler's (1978) Mastery Scale, which bears similarity to the measures of locus of control, was negatively related to anger and anxiety in samples of Israeli male and female university students. Those whose scores indicated that they believed that they could be effective or "masterful" in dealing with their life problems were less apt to be angry or anxious. However, there was no interaction found between the war-related stressful event measure and the mastery scale which led the authors to question the stress moderator role of feelings of mastery. It is interesting to note in this study though, that war-related stressful events did not produce a main effect upon anxiety or anger in and of itself. This scale which queried subjects' closeness to the war, did not actually assess experienced stresses of war. Rather, questions focused upon whether or not subjects had to serve in the war at all or were close to those who did and whether anyone to whom they were close had been injured or died in that war. But the actual experiences of the subjects themselves were not assessed by this scale as they are in the more conventional measures of stress, which may help to account for its failure to produce main effects or interactions in the prediction of anxiety or anger. From these studies and others concerned with stress experienced in medical education (Kilpatrick, Dubin, & Marcotte, 1974), in commuting to work (Novaco, Stokols, Campbell, & Stokols, 1979), in adapting to Marine Corps training (Cook, Novaco, & Sarason, 1980), and in the caring for sick children (Hobfoll & Lerman, 1988), beliefs about control and mastery have been found to play a significant role in predicting the kinds of responses that people make to those stressors, though the results have not always been strong, clear, and free of enigma. The sometimes confusing higher order interactions, and the "now you see them and now you don't" results have been particularly evident when the effects of stress have been studied using aggregate measures of life stress. RESEARCH WITH LIFE EVENTS STRESS SCALES Much research concerning stress has made use of survey devices such as the Social Readjustment Scale (Holmes & Rahe, 1967) or more recent scales such as the Life Experiences Survey (Sarason, Johnson, & Siegel, 1978); and in studying the role of control beliefs some investigators have used published forms of locus of control scales while others have simply queried their subjects as to the controllability of the life experiences that they have acknowledged on surveys of stressful events. For example, researchers such as Husaini and Neff (1980) and McFarlane, Norman, Streiner, Roy, and Scott (1980) have found personal judgments of the controllability of experienced events predictive of psychiatric symptomotology and subjective strain. That is, events regarded by a subject as
Perceived Control
117
having been uncontrollable were found to be associated with greater strain and disturbance than were events which subjects deemed controllable. On the other hand, Manuck, Hinrichsen, and Ross (1975a, 1975b), using independent measures of stress and locus of control, found that externals did not differ from internals with regard to anxiety or help seeking after having experienced stressful events. In the absence of stressful experiences, however, externals had higher anxiety scores and sought help more than their peers. In other words, externals seemed to be characteristically anxious and help seeking whether they were undergoing stressful experiences or not. Internals, however, who were customarily less anxious than externals, became as distressed as externals only when they encountered stressful events. In contrast, other investigators such as Johnson and Sarason (1978) and Kobasa (1979) found some evidence suggesting that locus of control operated as a moderator variable. In these investigations externals have shown stronger relationships between stress and depression, anxiety, and illness than have internals. The affective responses of the latter subjects have seemed less predictable than those of externals from measures of stressful experiences. Along with scales assessing what she has referred to as commitment and challenge, Kobasa (1979) considers control to be a major component of hardiness, a construct indicating buoyance and resilience during encounters with stress. More recent investigations have replicated some of Kobasa's findings though some investigators (Hull, Van Treuren, & Virnelli, 1987) have found support only for the control and commitment components, and have argued that their effects are primarily direct, in that feelings of helplessness or a lack of control, and an absence of commitments are psychologically stressful in and of themselves. Any interactive effects indicative of a buffering of the impact of stress are seen as secondary and likely to be observed only in particular situations. Both Sandier and Lakey (1982) and Lefcourt, Martin, and Saleh (1984) have found higher order interactions between stressful events, locus of control, and social support in the prediction of mood disturbances. In each study, internals who had a high degree of social support proved to be more resilient in the face of stress than internals with less social support or externals with or without social support. In both of these investigations there was little direct support for control as a singular moderator of stress, a finding that has also been reported by Nelson and Cohen (1983). It was only in interaction with social support that control was found to ameliorate the effects of stress in these studies. Similar findings have most recently been reported by Hobfoll and Lerman (1988) who found that mothers of seriously ill children who were high in their sense of mastery and enjoyed general social support or good intimacy with a friend were the least emotionally distressed during their travail. However, in this study, the sense of mastery also produced a significant main effect. A high sense of mastery was associated with less emotional distress in general. In an investigation from our own labs (Lefcourt, Miller, Ware, & Sherk, 1981) further clarifications and complications were uncovered. In this study we had used two different life event measures, the Coddington Life Events Scale (Cod-
118
Personal Coping
dington, 1972) which assesses the occurrence of stressful experiences in each of four stages of development, from early childhood through the high school years; and Sarason's Life Experiences Survey (Sarason et al., 1978) which assesses stressful experiences within the preceding year. In addition we used four different locus of control scales in the attempt to predict mood scores that were to be obtained over a period of several weeks. Given the repeated testing of moods which assumedly should augment the reliability of their assessment, we had felt confident that our results would shed light on the presumed moderator effects of locus of control. What we found in three successive studies was that externals tended to be dysphoric generally, while internals only showed indications of mood disturbance when undergoing current or recent stress as was measured by Sarason's Life Experiences Survey. However, when the stressful events measured by the Coddington scale did not immediately precede the period during which moods were being assessed, the results looked more supportive of the moderator effect model. That is, when the stressful life events had occurred some three to five years prior to the mood assessment period, we found that the current mood states of internals were less likely to be correlated with them. On the other hand, those same relationships between stressful experiences occurring three to five years prior to the present and current mood disturbances were highly significant among externals. These data led us to conjecture that while everyone may show immediate affective responses to stressful experiences, internals are more apt to recover given the passage of time. In turn, we guessed that the greater propensity of internals to become involved in their pursuit of current goals and satisfactions, in other words to be problem-solving focused, would serve to hasten the decay of effects deriving from those earlier events. The more passive, emotion-focused coping that characterizes externals, on the other hand, should lead to less coping with and less acceptance of prior stressful experiences so that these aversive events should have a more enduring impact upon externals than upon internals. More recently, Caldwell, Pearson, and Chin (1987) have also found evidence to the effect that locus of control operates as a stress moderator but only in interaction with other variables such as gender and social support. They did not find any support for a straightforward and simple interaction between locus of control and stress in predicting depression or maladjustment. Only among males did locus of control have an impact upon symptom formation such that internal males were more apt to develop psychosomatic health symptoms under stress while external males were more likely to become depressed under such circumstances. Females showed no such effects as a function of locus of control though they did vary in their responses in accord with social support. These findings bear similarity to those reported by Hunsley (1981), described in Lefcourt (1982). Females in that study showed no moderation by locus of control of the level of reactivity of moods following stress, whereas males showed strong effects, with the externals reporting deflated mood levels following stress in contrast to the more buoyant internals.
Perceived Control
119
The ultimate conclusion to be drawn from these data is that while locus of control does obviously play some role in determining responses to stressful experiences, it does not do so in a simple and straightforward way. This could raise questions about the personal anecdotes and the aforementioned field research that have implicated control beliefs, the sense of responsibility and mastery, and active coping as ameliorative factors in stressful circumstances. On the other hand, it is quite possible that scale derived stress research suffers from certain inadequacies that make it less suitable for studying the moderator effects of control than the field or case study approach discussed earlier. That externality is associated with negative affects such as depression there is little doubt. A recent meta-analysis of the locus of control-depression linkage (Benassi, Sweeney, & Dufour, 1982) affirms that locus of control orientation and degree of depression are significantly related, that the relationship is moderately strong, and that it has been consistent across studies even when control expectancies concern failures and successes separately. Thus we can be assured that externality is associated with depressive affect. What is uncertain, is whether an internal locus of control and its associated coping characteristics serve to minimize the depressive effects that can result from specific stressful experiences. As I had noted earlier, the uncertainty concerning the linkage between stress, locus of control, and the resulting dysphoric affects derives largely from those studies wherein stress has been assessed by aggregate life event measures. When life event scales have been used in stress research, sex of subjects, access to social supports, and beliefs about control seem to interact in complex and not always consistent ways which determine whether or not these variables will have a stress moderating effect. In addition, we have found evidence that the temporal proximity of experienced life events to the time period during which assessment of moods is obtained makes a considerable difference with regard to whether locus of control is found to have stress moderating properties. How might we explain these complexities? Thoits (1983) has reviewed the literature pertaining to the life stresspsychological distress linkage and concluded that much of the uncertainty and the low magnitude relationships found within this research domain may be due to our failure to correctly assess the impact of major life events. That is, similar life events such as death of a loved one or the loss of employment may have different portents for different persons because of the contexts within which these events occur. For example, death of a loved one following a long debilitating illness may be a relief in contrast to a sudden and unexpected death which catapults the survivor into a whole series of new quandaries for which the person may be ill prepared. As Thoits notes, the persistent strain created by the stressor may be more responsible for psychological distress than the stressor per se. This assumption has been at the core of Pearlin's research on stress (Pearlin, 1983). Pearlin contends that it is continuing strains which produce psychological malaise by convincing victims that they do not have the capability to master the problems occurring to them; and that their consequent loss of social roles may lead to
120
Personal Coping
lowered self-esteem. As a further effect, Thoits suggests that lowered self-esteem can leave one feeling shame before others which may eventuate in social withdrawal. This removal from close social encounters may then leave a person bereft of potential social support, thus compounding the impact of the original stressproduced strains. Other researchers such as Brown (Brown & Harris, 1978) have argued that stress can not be adequately assessed by the scales currently in use because of their decontextualized format. Brown has more often employed lengthy interviews to assess stressors occurring in his subjects' lives. His rather timeconsuming and cumbersome procedures, however, are too formidable for most investigators to undertake. As a consequence, researchers have had to make do with scalar procedures that lack contextual background and are subject to memory distortions. Both Brown (1974) and Kessler (1983) have discussed the biasing of recall due to current mood states whereby a person who is depressed, for instance, may be better able to recall negative events than when he or she is in an elated mood, and better able to recall positive events when elated than depressed. Additionally, it is likely that the interpretations drawn as to whether an event was positive or negative in the first place can reflect current mood state; a family gathering may be recalled with warmth and affection when the recaller is feeling elated; however, when depressed, the same person may recall more clearly certain slights and annoyances that had occurred at that gathering and thus cast a sallow glow to the recall of that occasion. This form of recall biasing has most recently been demonstrated even in response to olfactory stimulation. Ehrlichman and Halpern (1988) have found that subjects exposed to pleasant versus unpleasant odors exhibit different ratings of happiness associated with recalled experiences. Pleasant odors tended to be associated with the memory of happier experiences. Therefore, it is obvious that in our use of life event scales, current mood states may dictate some of the recall and evaluation of earlier events which would serve to inflate the relationship between life events and mood disturbances and raise questions pertaining to the causal direction between stress and moods. A second, more obvious point is that much of the research that makes use of aggregate life event measures has been conducted with university age students whose stressors just may not be of the same magnitude or frequency as they are for those persons who have been studied in field settings. The latter comprise primarily adult samples where the intensity and impact of the experience have often been self-evident. Student populations also rarely exhibit the levels of mood disturbance that can characterize the more general public; and in describing their own moods, young students may often exaggerate their degree due to their own lack of perspective and relative inexperience with the more severe and tragic stressors that age can bring. Finally, as is the case with most research wherein scalar measures of personality and psychological states are used, there will always be much room for error deriving from the mismatch of the subjects' and the experimenters' languages,
Perceived Control
121
and a misreading of each others' intentions. That is, the items on any measures are abstractions that may have different referents for the subject and the investigator. As well, the subjects' frankness and willingness to share their personal experiences and viewpoints with a relative stranger will always be questionable. Consequently, we must accept the fact that there will always be much error variance in our measurement which can be attributed to miscomprehension, mistrust, response biases, and to deception, to name but a few causes of confusion in scalar research. Granted these limitations to studies which make use of standardized scales as opposed to the case study and field research described above, it is quite possible that the results with the former are more veridical than are those from the latter. In addition, while locus of control is certainly related to mood states, it may not necessarily operate as a moderating variable in all stress related circumstances. After all, there are occasions where efforts to control are doomed to failure, when it might be better if a person were capable of acceptance or even resignation. LOCUS OF CONTROL, COPING PROCESSES, AND THE APPRAISAL OF THREAT Another way to evaluate the role of locus of control as a stress moderating variable derives from an exploration of the nomological network that encompasses stress and coping processes. Lazarus and his colleagues (Lazarus & Folkman, 1984) have described the process of coping with stress as involving a series of appraisals whereby persons first determine if a given event offers any threat to them (primary appraisal), and then, if the answer to the first appraisal is positive, whether they are capable of doing something to avert the impact of the threat (secondary appraisal). The latter comprise both actions that can alter the nature of the stressor (problem-solving coping) and those that can change the nature of a person's affective responses to the threat (emotion-focused coping). For example, people could become aware that their neighborhood or city has become a less secure and friendly place in which to live. At some point, certain persons will interpret rising crime rates, deteriorating housing conditions, long commuting times, and increasing incidents of hostile or unresponsive behaviors from strangers as indicating that their milieu has become threatening. Following such a primary appraisal some persons who feel capable of mobility may seek out opportunities elsewhere. Others, however, because of either age, limited assets, and/or the belief that they would not be able to start up all over again in a new setting surrounded by strangers, may suffer their deteriorating circumstances with a sense of resignation. Within this example we may see that individual variance can derive for one, from differences in sensitivity to the changes occurring in the milieu; and by sensitivity, we are including an inferred internal dialogue in which the individual weighs and compares his or her observations with what he or she knows about the larger world. Thus, sensitivity implies both perceptual and cognitive processes that underlie the assimilation of information.
122
Personal Coping
Secondly, people will differ in their readiness to make judgements about these perceived changes. However, once a judgment about threat is made, there are a multitude of self-assessments that come to comprise the secondary appraisal whereby a person chooses how he or she is to respond to the threat at hand. One's sense of effectiveness is related to the assets at one's disposal, to the constraints within which one must function, and to perceived opportunities and available options. Given this portrait of the coping process, we may now ask if the literature concerning locus of control contains any information that would reflect upon these stages of the coping process. Among the earliestfindingsthat were reported with the locus of control construct were those indicating that persons with more internal control expectancies seemed to be more alert and attentive to information that had relevance for their well-being. Seeman and Evans (1962), using a twelve-item measure of powerlessness derived from Rotter's I-E scale, were able to predict the amount of knowledge that tuberculosis patients had about their own threatening disease. Externally oriented tubercular patients were found to know less about tuberculosis than internals on both a direct and an indirect (staff ratings) measure of same. The results of this study indicated that internals avail themselves of information even if it has negative connotations for themselves. Externals, on the other hand, seemed to have been less concerned with the gathering of information that would have allowed them to more accurately appraise the level of potential threat that was inherent in their circumstances. If there were some actions that could, in fact, effect the course of that disease, externals, as a consequence of their failure to attend to relevant information, would have been less likely to discover the corrective or beneficial behaviors than would internals. Early research with locus of control had also indicated that externals were more likely to smoke than were internals (Straits & Sechrest, 1963; James, Woodruff, & Werner, 1965). This difference no doubt reflected a similar failure in the primary appraisal process, of assimilating information attesting to the fact that smoking is injurious to health (threatening), and consequently, failing to engage in problem-solving behavior that would involve an alteration of life habits. It is also arguable, however, that the failure of persons to quit smoking despite information about its negative effects may be more reflective of emotion-focused coping devices such as denial, which one would expect from more fatalistic persons, than of the failure to ascertain the dangers involved in smoking. Seeman found corroboration for his findings with tuberculosis in a second study conducted with reformatory inmates (Seeman, 1963) that focused upon vigilance for information pertaining to parole. Seeman had predicted that persons who denied that they were powerless (internals) would be more accurate in their recollection of information concerning the attainment of parole, but not in their recall of information related to reformatory life that was of less personal relevance; that is what he found. His results led him to conclude that an individual's sense of powerlessness governs his attention and information acquisition pro-
Perceived Control
123
cesses. This connection between beliefs about control and information assimilation that would be prodromal to the appraisal of threat has been supported by the findings of several other research investigators. In our own laboratory, we have explored the attention processes of persons varying in their beliefs about control as they underwent mildly stressful experiences. In each of several tasks we "double-crossed" our subjects by misleading them as to the purpose and meaning of what they were doing. For example, in one study (Lefcourt, Gronnerud, & McDonald, 1973) our subjects had undergone a series of very boring verbal tasks and believed that we were interested primarily in developing some arcane measures of verbal facility. The last task, however, proved to be revelatory insofar as our ostensible purposes were concerned. The task was a word association test which began innocently enough but became progressively more focused upon sexual content. The provocative words (rubber, bust, snatch, etc.) were all double entendres, so that the underlying substance and meaning of the task and the consequent suspicion about our intentions could justifiably be allayed for a while at least, until so many double entendres had been presented that it would have required massive denial to remain blind to their sexual content and to our possible duplicity. From an observation room adjacent to the room in which the subject was seated we had recorded their facial expressions on videotape as they performed on the test. In that study we found a range of evidence attesting to the fact that internals were the quickest to "catch on," that is, to note that "something was amiss," and then they were the quickest to act upon that observation. Their response times to the double entendres increased earlier in the sequence of words administered than they did for externals indicating that internals had become "aware" sooner than had externals. In addition, their eye movements, laughter, and subsequent joking indicated that they surely knew "something was up" but that they were not intimidated by their discovery. Externals, by contrast, were slow to "catch on" and less given to expressions of mirth in the process. We have found confirmation of these data in other "double-cross" research studies. Internals have generally been quicker than externals to note the circumstances and contextual cues that help to reveal the "secret" meanings of the experiments in which they have been engaged. Consequently, they have been found to be less embarrassed or surprised by these experimental situations when the purposes of the experiments have become more explicit. In one such study where subjects were presented their own photograph at the end of a series of photographs of rather disreputable looking persons whose crimes they were attempting to deduce from appearances (the photographs were described as police mug shots), internals often rebounded with humorous retorts whereas externals appeared to be offended and in some cases angry (Lefcourt & Martin, 1986). For example, one internal subject leered at the female experimenter and, after a long pause, exclaimed loudly, "He's a rapist," and subsequently dissolved in laughter. In laboratory situations, then, evidence has indicated that internals are more
124
Personal Coping
apt to be aware of changing circumstances, a necessary precondition of stress appraisal. However, this is not to say that they necessarily come to regard those circumstances as threatening. As noted, in the studies by Anderson (1977) and others, externals more often regard their experiences as being stressful than do internals. Given that internals seem to be more aware of changes in their circumstances yet may be less threatened, it is possible that we are observing the effects of secondary appraisals, the judgements individuals make as to whether they can do something that will alter the threatening circumstances that they are encountering. At this point we may ask whether internals are better equipped to deal with potentially stressful experiences than are externals such that they are less potentially disarmed by them. In three studies conducted by social geographers some interesting evidence has been obtained suggesting that internals are both more aware of potential dangers and more planful in their responses to them. Sims and Baumann (1972) in a study that examined responses to tornadoes, found that in communities where beliefs about control were more modally internal (rural Illinois), people were more apt to behave appropriately to that threat, listening to the radio and staying in their basements, than were those who lived in an area where externality was more common (rural Alabama). People in the latter location were more likely to have gone outside to see if and when the suspected tornado was coming their way. Consequently, there were more serious accidents in Alabama than there were in Illinois. A few years later, two studies, one conducted in Canada and the other in New Zealand, offered some corroborative evidence linking locus of control to the responses made to natural disasters. In the former, SimpsonHousley, Lipinski, and Trithardt (1978) reported that internals were more knowledgeable about the potential dangers of flooding river plains around Lumsden, Saskatchewan than were externals. As well, internals were found to live in residences that were further away from the flood plain than were externals, though this latter difference was of borderline significance (.10 level). In the New Zealand study, Simpson-Housley and Bradshaw (1978) reported analogous data concerning earthquake hazards in a suburb of Wellington where there are two active geological faults in the substructure of that city. Residents were queried about their perceptions of earthquake hazards. Internals were more frequently found to have taken preventive measures to help withstand the threat of earthquakes indicating a heightened awareness of the potential danger. Externals, on the other hand, were found more ready to take reparative measures after the occurrence of earthquakes. In other words, either externals were failing to consider the danger of earthquakes or were engaging in emotion-focused coping, denying the threat until after a crisis occurred. In essence, externals could be said to have been operating in an "out of sight, out of mind" style. Internals were also found to expect more disruption from an experience with earthquakes than were externals which suggests a more accurate primary appraisal, which in turn, may help to account for their greater preparedness. Though the methods used in these three studies of natural disasters leave
Perceived Control
125
something to be desired for psychological investigators, the consistency of the findings encourages us to believe that internals are more ready to perceive potential threats and are more prepared to take some kind of action to help ward off the effects of those stressors whether they be tornadoes, floods, or earthquakes. It is reasonable to predict that fatalistic people would be found living in hazardous areas more often than would persons who readily assume responsibility for themselves, and that externals would engage in more emotion-focused than problem-solving coping in order to continue living in such perilous circumstances. A number of studies, in addition to that of Anderson's (1977), have found evidence to the effect that internals are more apt to engage in problem-solving coping processes than are externals, and that externals are more likely to engage in emotion-focused coping than are internals (Silver, Auerbach, Vishniavsky, & Kaplowitz, 1986; Strickland, 1978). In a dramatic four-day simulation of a hostage taking incident, Strentz and Auerbach (1988) found that external airline employees were most distressed in the condition where they had been taught problem-solving coping methods as opposed to either a control condition or one in which they had been taught emotion-focused coping strategies. On the other hand, externals did not differ from internals when they had been taught emotionfocused coping procedures or were in the control condition. Thus, while internals benefitted as much as externals from emotion-focused coping, the latter were not able to benefit from their exposure to problem-focused coping. A problemsolving coping style seems to be more comfortably adopted by internals. Returning to our earlier discussion of awareness and/or cognitive alertness as components of primary appraisal, we have conducted several experiments in our labs which bear upon people's awareness or sensitivity to each other during social interactions. We have examined the ways in which people attend to each other during protracted conversations which could have become stressful if the social interchanges were not handled competently. In one such study (Lefcourt, Martin, Fick, & Saleh, 1985), we observed same-sex students who were strangers to each other, discussing how they had dealt with their feelings while watching "Subincision," a film which Lazarus (1986) had used as a stressor for his laboratory studies of stress. In this anthropological film, Australian aborigines were observed during a rite of passage in which young boys suffered ritualized genital mutilation. The film is generally regarded as gruesome to watch and elicits emotion-focused coping. Consequently, we believed that subjects would be pleased to share their feelings with another if just for social comparison purposes. In talking to a stranger about one's personal responses to an aversive presentation that contains elements of sex, repulsion, and violence, however, a person needs to carefully gauge his or her partner's responses to his or her own disclosures during a discussion to avoid embarrassing the other person or becoming embarrassed oneself; otherwise the conversation could degenerate into unpleasantness or a sullen silence. Therefore, close attention and sensitivity to one's partner should facilitate better communication. The listening behavior or
126
Personal Coping
attentiveness of subjects in one sample was rated globally; in a second sample, subjects were rated on several components of social competence reflecting attention. The latter consisted of time spent talking, length of utterances, eye contact when listening and talking, nonverbal affirmations and encouragements, silences, etc. In both studies, those who had scored as internals with regard to affiliation were more likely to have displayed active listening and responsiveness than were externals for affiliation. This measure, the locus of control for affiliation scale, derives from the Multidimensional-Multiattributional Causality Scale (Lefcourt, von Baeyer, Ware, & Cox, 1979) which also contains a locus of control scale for achievement. As a point of interest, the achievement locus of control scale was much less related to the social criteria than was the affiliation locus of control scale. Here, we found evidence that internality with regard to affiliation was associated with more active listening or attentiveness, a necessary condition or a component of the primary appraisal process. Thus, attentiveness and sensitivity to one's partner's responses which should alert one to potential sources of discomfort in the interaction are associated with an internal of control. In a subsequent study conducted with married couples, these findings were corroborated and made all the more relevant to our concern with locus of control and coping processes. In this latter study (Miller, Lefcourt, Holmes, Ware, & Saleh, 1986) eighty-eight married couples were videotaped as they attempted to deal with conflicts that are fairly common sources of marital discord. The research was originally undertaken in the hope that we might be able to uncover some of the reasons for the high rate of divorce that characterizes contemporary North American life. We had hypothesized that individuals who believed that their spouses' behaviors were mysterious (unpredictable) and uncontrollable would be less attentive to their partners and, therefore, less aware of potential problems that could surface between them than would persons who believed that their actions played a significant role in shaping their spouses' behaviors. In turn, we anticipated that those who believed that they could affect their marriage partners would be more able to cope with quandaries as they arose, given their earlier and greater awareness of emerging difficulties. To assess these beliefs we had developed a Marital Locus of Control Scale (Miller, Lefcourt, & Ware, 1983) for which we had found reliability and some preliminary evidence of validity. Couples were asked to improvise responses to three conflict situations that had been previously constructed by Gottman (1979). While the conflict situations were like a set of standardized tasks with explicit role descriptions, participants were asked to add their own personal information to the role enactment that would individualize the situations, thus increasing their potential involvement in them. The three situations consisted of an "in-law problem" wherein the husband believed that plans had been made to spend Christmas with his family, but the wife did not regard the plans asfinalizedand had reservations about the visit; a "money problem" wherein the wife had spent money on clothing that had been considered part of their long-range savings program; and a "com-
Perceived Control
127
munication problem" wherein a tired wife wants some time to herself while her husband desires conversation and closeness after returning home from work. Throughout the enactments of each role the interactions were videotaped and were later rated for "engagement" and "problem-solving effectiveness" comprising "solution quality" and "solution satisfaction." Engagement consisted of ratings of an individual's tendency to become involved in conflict in an open, direct, and persistent way. Solution quality was rated by observers while solution satisfaction was derived from spouses' ratings. Solutions were judged as more successful when the points of view of both spouses were understood and taken into account in devising some resolution of the problem. If feelings were still ruffled and not addressed by the solution, the quality of the solution would have been adjudged as being less than adequate. Engagement was regarded as a sine qua non of solution quality and satisfaction. Independent raters were used to rate each set of variables so as to preclude rating generalizations. The rated behaviors and solutions of the couples were correlated with one another as we had hypothesized. Engagement did seem to be a precondition of solution quality and satisfaction, and the latter two were strongly related to one another. Most importantly, locus of control for marital satisfaction was related to all three in the expected directions, and, in turn, all of these variables were related to an independent measure of marital satisfaction that husbands and wives had completed separately. In this study, we were able to observe the actual coping behavior of spouses faced with conflicts within their marriages. The veridicality of the role enactments was obvious in the involvement that was evident and in the commentary of the couples at the end of the experiment. These were very real situations for married couples as Gottman (1979) has argued. Our subjects' coping behavior, then, was quite meaningful especially as it was related to their self-ratings of marital satisfaction. In this situation, we were able to observe couples' studying each other, appraising their spouses' feelings to ascertain how serious a problem they were encountering. Ourfindingsattest to the fact that internals are more apt to be observant than are externals, and are better able to resolve difficulties when they were unearthed. Internals proved to be more ready than externals to perceive and encounter potential conflicts with problem-focused coping, working to understand and alter conditions that were sources of grievance for their spouses. CONCLUSIONS Throughout this review of empirical findings with measures assessing locus of control and cognate constructs, it is evident that there is much support for a linkage between beliefs about control and affective states. People who feel helpless, who do not believe that there is anything that they can do to effect those things that matter to them, are more often found to be depressive or anxious. In contrast, persons who believe that they can be effective, who sense that their
128
Personal Coping
efforts are never entirely futile, will more often be found to be enthusiastic, resilient, and, if one may stress the affirmative, happy. While our evidence is unclear whether control serves to attenuate the effects of stress or simply indicates a general state of well-being, it is apparent that feelings of control and effectiveness are preferable to those of helplessness. In our discussion linking beliefs about control to coping processes, we have asserted that internals are more apt to be perceptive and aware of dangers lurking in the future than are externals. As such, one might wonder why internals should not be expected to be more anxious than externals. Awareness of potential dangers, after all, could produce worry and concern, whereas ignorance or innocence assumedly should at least have apathy as its reward. However, empirical data refute this assertion. Internals are not more apt to be worriers than are externals, and in reviewing research with anxiety and dysphoria, it would certainly seem that externality is associated with worry. It is when we turn to Lazarus' conception of secondary appraisal that these findings become more interpretable. If, after deciding that current or forthcoming situations are perilous, a person believes that he or she can prepare for those situations and avert their more negative consequences, he or she will not seem to be as anxious as will the person who fears that there is naught to be done that can influence the threatening situation. In a series of studies by Glass and Singer (1972) this is exactly what had been demonstrated. Subjects who knew that they could turn off a loud, aversive, and unpredictable noise were not as disturbed as were those who had no recourse to control while doing attention demanding tasks. That is, the performances of those with control showed less decrement as a function of loud and cacophonous noise than did the performances of subjects who believed that they could do nothing if the noise were to become distressing or even unbearable. The latter subjects did show the expected deterioration in their concentration abilities. These differences occurred despite the fact that those who could turn off the noise never did so. It was the mere belief that one could act effectively, if one needed to, that seemed to account for all of the differences that were observed. Why, we might ask, should this be the case? What is there about persons who characteristically anticipate being able to affect the important matters in their lives that makes them seem to be more buoyant and happy than are those who are fatalistic? Throughout my own research career with the locus of control construct, I have found repeated evidence to the effect that internals are more cognitively active than are externals when they become engaged in the kinds of enigmatic situations that we concoct for laboratory investigation; the cognitive activity that characterizes internals seems akin to that of a running internal monologue in which a person considers and chooses among alternative positions and constructs as he or she deals with the realities of daily life. Externals, in contrast, seem to take whatever occurs to them as a given, as if there were no other way to regard their circumstances other than from the view that has been provided or is most salient. In Lefcourt (1982), I contended that externals are
Perceived Control
129
more easily persuadable, more easily influenced and shaped by forces occurring around them because they do not engage as readily in an internal monologue in response to their circumstances. Furthermore, I used the display of humor as partial evidence of that kind of internal processing and monologue that I believed to be characteristic of persons with an internal locus of control. In the literature concerned with humor, the term bisociation was coined by Arthur Koestler (1964) to describe the process whereby two mutually exclusive conceptions could be held contemporaneously so that the individual could rapidly shift from one position to another as he or she mused about the meaning of the event. This rapid shifting, when accompanied by arousal due to the importance or excitation associated with the event, was said by Koestler to be responsible for manifestations of humor. Such rapid shifting among perspectives would be one of the legacies of an elaborate internal monologue and would therefore comprise the linkage between beliefs about control and the expression of humor. If seeing the same events or circumstances from contrasting viewpoints is essential for humor, being aware of choices and contrasting viewpoints would seem quintessential for the experiencing of one's self as a source or agent of one's own fate. In my own early research where we attempted to examine this potential relationship between control and humor, it was my contention that internals were less apt than externals to accept the serious face validity of the experimental situations that we had contrived. That is, we thought that internals would become more easily alerted and suspicious than externals as they engaged in our experiments, due to their assumed more extensive internal monologues. For example, in the double entendre word association task described earlier (Lefcourt, Gronnerud, & McDonald, 1973) we had anticipated that as our subjects were exposed to an increasing number of sexual double entendres, internals would have more quickly mused to themselves, "what the hell's going on here?" than would have externals; once alerted in this way, internals would be quicker to hypothesize that this experiment was not what it was cooked up to be. In a follow-up to this experiment we also examined the humorous reactions displayed by subjects during this task (Lefcourt, Sordoni, & Sordoni, 1974). In the inspection of verbal and nonverbal responses occurring during the word association test we found ample evidence that internals were laughing more and earlier than were their external peers, and that the laughter was of the sort which indicated that they had "caught onto the joke" and indeed, were enjoying it. Externals, on the other hand, were more serious, less mirthful, and much slower in catching onto the shift in focus that this task provided from all of those tedious ones that had preceded it. At about the same time that we were completing the double entendre investigation, we were also conducting another study in which subjects were asked to enact a series of role plays that concerned personally meaningful situations for undergraduate students (Lefcourt, Antrobus, & Hogg, 1974). One role concerned academic problems, the other, social. Subjects were asked to interact
130
Personal Coping
with the experimenter as if he were a good friend. All of the subjects enacted one "success" and one "failure" role, with the area (social vs. academic) of the success or failure randomly divided. Videotapes of the interactions were subsequently rated for humorous repartee. While other variables such as field dependence also had some impact, locus of control was found to have a decided influence upon the humorous reactions that we observed. We found that our internal subjects were the most apt to come up with jests that would sometimes catch us by surprise. For example, while discussing academic disappointment, wherein the subject was supposed to have been despondent after having been rejected by eight different graduate schools to which he or she had applied, a subject parried to the commiserating friend, "Oh, it could have been worse... I could have been accepted by eight graduate schools." This is a good example of the "flip side," a behavior or comment that turns a situation on its head. To turn an argument around in this way requires what in the humor literature is referred to as perspective taking. That is, one has to stand back a bit from one's immediate situation and "take stock" of what is transpiring while not losing a sense of involvement in the current moment. In other words, there must be a certain distance or remoteness so that one can engage in an internal monologue during which the bisociative process, from which humor derives, becomes potentiated. Most notable for the focus of this chapter is the contention that humor with its implied remoteness has the function of "preserving the sense of self... it is the healthy way of feeling a 'distance' between one's self and the problem, a way of standing off and looking at one's problem with perspective" (May, 1953, p. 54). If, as May argues, humor is a distancing strategy that lessens the likelihood that we will be overwhelmed by our problems, and if humor derives from bisociative processes which are prominent in internal monologues, and if this proclivity to engage in extensive internal monologues is more common among those persons with an internal locus of control, then we can begin to understand some of the molecular processes that underlie the relationship between locus of control, stress, and mood disturbances. Being more circumspect about their stressful experiences than externals, internals may be less overwhelmed by them and more ready to reconstrue their significance so as to better cope with them. That internals should be more flexible than externals, being able to shift perspectives during their encounters with stressful events may be an essential component of internality. To perceive one's self as responsible for one's outcomes and experiences requires the sensitivity to note one's own actions at the same time as being attentive to those of another. That is, someone would have to perceive both changes in the external world and variations in their own behaviors, mannerisms, and gestures if they were to conclude that they were causing the kinds of experiences they were having. On the other hand, the perception of external causes, though also requiring attention and attributional processes, would not seem to demand the same degree of cognitive activity. To regard chance, powerful others, or circumstances as determinants of one's experiences is to
Perceived Control
131
acknowledge that one does not fully understand how one's experiences can be determined. While in some instances the identification of external sources of influence would require the kind of active sleuthing that is implied in the term cognitive activity, more often it may simply indicate that people have not attended to their interactions carefully enough to discern what role they have played in bringing about their fates. In conclusion, empirical evidence indicates that locus of control is associated with mood states. Additionally, though there are paradoxical and enigmatic findings in the literature, locus of control would also seem to be related to the ways in which people deal with the stressful experiences in their lives. Finally, I have suggested the kinds of molecular processes that may be associated with locus of control which could help account for its role as a determinant of the ways in which people deal with potential stresses in their lives.
This page intentionally left blank
9
Temporal Factors in Stress and Coping: Intervention Implications Stephen M. Auerbach
Time and temporality are intrinsic to the study of human stress. An individual is said to experience stress when particular relationships to the environment are appraised as taxing, dangerous, or exceeding his or her coping resources (Lazarus & Folkman, 1984). Such relationships are almost never experienced in isolation; they are influenced by covarying relationships and perceptions, construals of past experiences and future aspirations. In the broadest sense, all environmental transactions are appraised in the context of the inherent stressfulness of the passage of time given our foreknowledge of the inevitability of aging and death. Temporal symbols such as the hourglass, the scythe, and crutches are indicative of our view of time as a largely hostile and capricious force which brings decay and ultimately an end to our existence (Cohen, 1964). Time and temporality are intrinsic to the study of stress and coping in other more specific ways: (1) exposure to stressful stimuli differs in duration and this variable can significantly affect coping outcome, though in complex fashion (physiological findings indicate impaired immune function following shortrterm stress but enhanced functioning following long-term stress, whereas chronic stress is generally found to have a far greater negative impact on psychosocial functioning than short-term, episodic stress; Aldwin & Stokols, 1988); (2) shortcircuiting stressor effects may require meeting some temporal criterion for emission of an appropriate response (cf. McGrath, 1970); (3) stressful events impinge on us at different points relative to our life cycle, and this affects our interpretation and manner of coping with these events (Blanchard-Fields & Irion, 1988; Folkman, Lazarus, Pimley, & Novacek, 1987); (4) stressors which occur "off-time" (i.e., incompatible with personal and societal expectations as to when they should occur) are coped with less effectively (Neugarten, 1976) as are those that occur at a "bad time" (e.g., concurrent with other stressors); (5) stress may be induced
134
Personal Coping
by the anticipation of harmful circumstances one thinks he or she is likely to confront, by an ongoing harmful stimulus, or by the harmful effects of stressors already encountered and this temporal relationship between the individual and the stress inducing stimulus is an important determinant of appropriate coping tactics and stress management interventions. The subject of the present chapter is the last area. Focus will be on the anticipatory and postimpact periods (rather than the period during which one actively deals with the stressor) because during these intervals there is more time to make decisions regarding coping options and to fine-tune attempts to control the stressor itself or the dysphoric emotions elicited by it. The anticipatory period is subdivided into "distal anticipatory" and "proximal anticipatory" phases, and "proximal postimpact" and "distal postimpact" phases of the postimpact period are considered (see Auerbach, 1986, for an earlier presentation of this framework). It should be noted that similar terminology based on temporal stages has been used to categorize psychological phenomena observed before, during, and after disasters (Kinston & Rosser, 1974). Also in this vein, Lazarus and Folkman (1984) have distinguished between threat (where harm or loss has not yet taken place) and harm/loss (where damage has already been sustained) as two major primary stress appraisals; Matheny, Aycock, Pugh, Curlette, and Cannella (1986) have differentiated between preventive and combative coping. This approach also bears some similarity to the influential tripartite prevention of pathology model adapted by mental health practitioners from public health and applied to crisis intervention and community-based interventions designed to prevent psychopathology (Caplan, 1964). In contrast to the latter model, however, the present formulation is not oriented around the notion of decreasing vulnerability to or limiting damage associated with mental disorder. Rather than starting with the fact of a disease (e.g., schizophrenia) or a major social problem (e.g., racism), making assumptions about its etiology and looking for ways to prevent it or treat it as early as possible, emphasis instead is on pinpointing sources of stress; ascertaining, for each temporal stage, the coping strategies which will most likely be effective in short-circuiting stressors or in minimizing emotional distress and behavioral dysfunction associated with them; and specifying intervention procedures likely to be most useful in eliciting effective coping strategies. Following Lazarus and Folkman (1984), coping processes are distinguished from coping outcomes, and two broad coping process dimensions are differentiated: problem-focused and emotion-focused. Coping processes refer to the strategies people use to deal with stressful situations, whereas coping outcomes refer to the effectiveness of these strategies in defusing environmental demands or the dysphoric emotions elicited by them. Problem-focused coping processes involve short-circuiting negative emotions through the emission of behaviors that modify the stressor or allow one to avoid it or minimize its impact, and cognitive activity that leads to the belief that the stressor can or will be controlled instrumentally. Emotion-focused coping denotes behaviors designed to directly palliate
Temporal Factors
135
or eliminate dysphoric emotions elicited by a stressor, with little attention paid to the characteristics of the situation or the nature of the threats it poses. Coping outcome tends to be better when predominantly problem-focused coping is used in largely controllable situations (or those appraised as such) and when predominantly emotion-focused coping is used in largely uncontrollable situations (Forsythe & Compas, 1987; Strentz & Auerbach, 1988). Effective stress management interventions teach people useful coping skills, when to use them, and how to apply them to the coping demands of the situations with which they are dealing (cf. Auerbach, 1989). DISTAL ANTICIPATORY STRESS As individuals, and also on community and societal levels, we attempt to anticipate the occurrence of negative events and to short-circuit them. We do so in three ways. We arrange conditions in order to minimize the likelihood: (1) of the event occurring, (2) that we will be impacted by the event if the event does occur and, (3) that we will be damaged if we are impacted by the event. Whereas it is logical for us to engage in such behaviors from the standpoint of minimizing exposure to noxious events, our efforts at problem-focused "preventive" coping tend to be sporadic and inconsistent (though there can be great individual differences). This is because preoccupation with potential misfortune exacts a cost in diminished flexibility and spontaneity, delay of gratification, and limitation of one's range of potentially reinforcing activities. Thus, much of our anticipatory coping with events we do not consider to be imminently dangerous or to have a high probability of ever impacting us is of the emotionfocused variety. Distal anticipatory problem-focused coping seems to be reserved for stressors that we are strongly motivated to avoid, though it is unclear what factors predispose such coping behavior. For example, Weinstein (1989) found no consistent relationship between experience with a stressor (e.g., criminal victimization) and subsequent precautionary behavior (measures to minimize likelihood of future victimization or of negative outcomes if victimized). Overall, we tend to compartmentalize and attend to stressors on our more immediate temporal horizon. Activities that have been undertaken to get people to engage in "preventive coping," in the absence of an imminent threat, may be summarized in terms of the threefold classification system suggested above. Preventing Occurrence of Potential Stressors Preventing or lowering the probability of occurrence of potentially aversive events involves problem-focused preventive coping at the most molar level, and includes societal attempts to lower the incidence and magnitude of such universal manmade stressors as war, poverty, crime, torture, rape, and child sexual abuse. One approach to minimizing such stressors is to try to alter attitudes which are
136
Personal Coping
known or suspected to promote them. For example, it is likely that attitudes and beliefs in our culture that encourage or justify sexually assaultive behavior create a climate that is not only hostile to victims but one that also helps disinhibit rape behavior among those prone to engage in it. In this regard, Burt (1980) found that acceptance of such mythical beliefs about rape (e.g., "In the majority of rapes, the victim is promiscuous or has a bad reputation," "Women who get raped while hitchhiking get what they deserve," "A woman who is stuck-up and thinks she is too good to talk to guys on the street deserves to be taught a lesson") is associated with sex-role stereotyping, adversarial sexual beliefs, and acceptance of interpersonal violence. This suggests that intervention programs targeted at potential offenders designed to alter such beliefs (cf. Burt, 19761977), or broader-based long-range strategies focused on fighting, at very young ages, sex-role stereotypy and its extension into the sexual arena (Burt, 1980) should be effective in lowering the incidence of rape in our society. A similar analysis may be made with respect to attitudes that may increase the likelihood of nuclear war. One set of such attitudes reflect defensive resignation in the face of a seemingly minimally controllable catastrophic event ("If it happens, it happens;" "What's so special about humankind?" "Many other species have come and gone, maybe this is our turn;" or a more pervasive "numbing" of all feeling; Lifton, 1982, p. 620). It is also argued by some that our leaders and others in different parts of the world have introduced and reinforced wishful but illusory beliefs about nuclear war (e.g., the idea that tactical, limited, controllable nuclear war is possible; that given foreknowledge of a nuclear attack we will be able to prepare and survive; that stockpiling nuclear weapons makes us more secure and war less likely). These are used by governments as reasons to continue nuclear arms buildups, but in turn actually increase the probability of war (Lifton, 1982). Mental health professionals, it is argued, must therefore "examine what forces produce flawed and dangerous thinking, what forces allow such thinking to be accepted unchallenged by the public" (Goldman & Greenberg, 1982, p. 581), and find ways to change these attitudes that reinforce passivity or actually support factors (i.e., arms buildups) that increase the likelihood of nuclear war. A second way of lowering the probability of occurrence of manmade aversive events involves dealing with the potential initiators of those events. In the case of nuclear war this would involve peace and arms control negotiations among nations, and in general fostering commitments and interdependencies (e.g., in economic areas, closing of gaps in value systems) that would conflict with an act of war and its consequences. On a more molecular level, in the area of sexual child abuse for example, programs have been developed requiring incestuous fathers to participate in professional counseling as a condition of release or probation (cf. Swift, 1986). Hopefully, such intervention would lower the probability of recidivism in this high risk group. Similarly, intervention programs with chronic rapists geared toward diminishing their sexual arousal to rape-related
Temporal Factors
137
stimuli and enhancing heterosexual arousal and heterosocial skills (Abel, Blanchard, & Becker, 1976) if widely implemented should diminish rape incidence. Preventing Exposure to Aversive Events A major task in preventing exposure to an event that may be highly aversive, but that is perceived by people as having a relatively low probability of impacting them, is to break through defensive emotion-focused coping (which is moderating immediate anxiety levels) in order to motivate them to take personal action. For example, where the adaptive coping response is to take problem-focused action to minimize likelihood of exposure to a noxious event like cancer, anxiety is often reduced with rationalizations of the "it can't happen to me" variety. Thus, many persons continue to smoke though they are continually bombarded with information demonstrating its unquestionable relationship to cancer, many women do not examine their breasts for early detection of tumors, persons continue to engage in unsafe sex and drug use practices despite the AIDS epidemic, and so forth. Studies involving the use of fear appeals and persuasion to motivate persons to engage in preventive coping behavior indicate that optimal levels of preventive behavior are not stimulated by communications that simply arouse very high fear levels. Undiluted high-fear provoking recommendations produce or reinforce (maladaptive) defensive minimization or denial of the importance of threats; such recommendations need to be moderated by fear-reducing reassurances to stimulate optimal levels of preventive problem-focused coping (Janis, 1969). Further, there are data that suggest that adaptive coping is maximized when fear is aroused via strengthening long-term beliefs in severity of the danger versus trying to arouse fears based on momentary threats, and when the appeals incorporate information that discourages unrealistic emotion-focused thinking (e.g., belief in magical cures). This is exemplified in recent findings that the best way to motivate women to undertake preventive breast self-examinations is to present information stressing the importance and the efficacy of the examination in early detection of breast cancer and thus in increasing life expectancy, along with information emphasizing a woman's ability to perform the examination correctly and to incorporate it into her health care routine (Rippetoe & Rogers, 1987). These principles should be incorporated into other preventive programs such as currently emerging educational programs designed to minimize the spread of human immunodeficiency virus among high risk groups, such as adolescents, who are likely to be ignorant of or minimally sensitive to the need to engage in preventive coping (cf. Flora & Thoresen, 1988; Melton, 1988). Minimizing Negative Consequences of Potential Stressors Optimizing coping behavior geared at preventing or minimizing negative consequences to us if we are confronted by a stressor subsumes two broad tasks:
138
Personal Coping
(1) development of broad-based coping skills in people to enable them to cope with a wide range of anticipated stressors, and (2) development of intervention procedures designed to prepare persons to handle specific stressors that they may confront or are deemed likely to confront in the foreseeable future. Much work relevant to the first area has been done with children in an effort to detect factors that might make them prone to develop psychopathology when exposed to pervasive stressful circumstances such as extreme poverty or dissolution of the family (cf. Cowen & Work, 1988; Garmezy, 1981), and with adults within the context of models viewing life events as markers for promotion of competence through "life development intervention" (Danish & D'Augelli, 1980). Research programs with children have focused on understanding seemingly "stress-resistant" or "invulnerable" children, who maintain "mastery and competence" despite the "presence of sustained and intense life stresses" (Garmezy, 1981, p. 215). Research which differentiates stress-resistant from stressaffected children suggests that preventively oriented interventions with high risk children should emphasize reinforcing "autonomy, independence, social responsiveness, and positive disposition," and provide "caring adult models who communicate consistent rules and expectations and show strong interest and involvement with the child, and secure sources of extrafamilial support," along with "skills training (e.g., problem solving, anger control, communication) and self-views (e.g., sense of mastery and control)" enhancement (Cowen & Work, 1988, pp. 602-603). Other researchers extol the use of broad-based "affective education" procedures to provide children a wide range of interpersonal and intrapersonal skills to enhance flexibility and adaptability and help them cope more effectively with the stressors that they will inevitably face in the normal process of development (Durlak & Jason, 1984). Preparatory procedures designed to maximize the ability to deal with specific stressors in the event that they do occur include programs which target entire communities as well as those aimed at particular high risk subgroups. Such programs typically involve provision of stress-relevant information (optimally, diluted by reassurances) and training in detection of relevant cues, decision making, and execution of appropriate problem-focused skills when necessary. An example of community-wide programs includes the development of "disaster subcultures" in communities that have been repetitively impacted by natural disasters (e.g., floods, tornadoes). In such communities, in the event of a disaster, individuals and organizations are trained to mobilize immediately and take on disaster-relevant roles and functions in order to minimize the negative physical and psychological impact of the stressor (Taylor, Ross, & Quarentelli, 1976). Approaches that target specific subgroups are exemplified by programs aimed at potential sexual child abuse victims that teach children to say no, along with how to apply nonverbal assertiveness and self-defense skills in potential abuse situations (Conte, Rosen, & Saperstein, 1986). Another example is crisis intervention training given to medical staff personnel who are likely to encounter violent patients in emergency room or psychiatric settings. Such training involves
Temporal Factors
139
teaching specialized task-relevant, problem-focused skills such as how to identify occasions when patients are likely to become agitated, behavioral cues indicating agitation, listening skills for calming patients, self-defense techniques, and application of restraints if necessary (Jacobs, 1983; Rice, Helzel, Varney, & Quinsey, 1985). The findings of studies examining efficacy of coping strategies in a wide range of specific situations suggest that if persons are being trained to cope with a situation that may eventually be manipulable, but where initial stress levels are apt to be very high, they should be instructed in judicious application of both emotion-focused and problem-focused coping strategies. Use of emotionfocused coping (induced by techniques such as muscular relaxation, deep breathing, and calming self-talk) would help moderate initially disabling stress levels to the point where the individual would be capable of assessing cues and making decisions regarding appropriate actions to take. For example, in the case of hostage situations, rapid moderation of incapacitating anxiety levels is crucial because opportunities to escape (which are rarely acted upon) are greatest during the highly stressful early phases of abduction (Strentz, 1985; Strentz & Auerbach, 1988). Retrospective data obtained from former Nazi concentration camp inmates also indicated that ability to use emotion-focused mechanisms (denial, detachment, passive submission) to manage near overwhelming panic experienced initially, followed rapidly by development of "an exquisite sensitivity to the mood of the guards and constant alertness to opportunities for gaining extra food or better work assignments" (Schmolling, 1984, p. 114) was an important factor in enabling them to cope successfully and survive. In summary, stress management procedures that are initiated when an event is perceived as a potential but not an immediate threat include those that focus on the event itself and attempts to diminish its incidence, those designed to get potential victims to take action to minimize the likelihood that they will be exposed to the event, and those that prepare potential victims to cope with stressors effectively if they do confront them. Indicants of success of such procedures encompass incidence of stressor occurrence, measures of attitude change, self-reports of intent to engage in preventive behavior and measures of actual preventive behaviors engaged in, and measures of the degree to which people participating in intervention programs confronted the target stressor or experienced negative consequences if they did encounter it. Synthesis of the diverse research literature subsumed by this category is beyond the scope of this chapter. It has been generally concluded, however, that though clinicians and researchers have been effective in identifying high risk factors and in planning interventions, a number of issues need to be resolved in the area of implementing and evaluating interventions (cf. Auerbach, 1986; Garmezy & Masten, 1986; Rickel, 1986). PROXIMAL ANTICIPATORY STRESS When presented with cues that we are about to confront an aversive event, stress increases to levels which for some stressors equals or exceeds those ex-
140
Personal Coping
perienced during actual stressor exposure (Nomikos, Opton, Averill, & Lazarus, 1968; Spacapan & Cohen, 1983). Anticipated stressors that are imminent are either largely amenable to adaptive problem-focused coping that serves to avoid exposure completely or least minimize damage sustained, or are unavoidable, or a commitment has been made by the individual to confront it.
Avoidable Stressors Janis (1986; see Janis & Mann, 1977) outlines the factors influencing adaptive vs. maladaptive coping patterns in situations in which there is clear warning of impending danger and the individual has some time-limited opportunity to analyze cues regarding potential risks and benefits involved in taking various protective actions. Such stressors are often modifiable or the damage they produce may be mitigated if the potential victim is appropriately vigilant, assimilates information, appraises alternatives in an unbiased manner, and chooses the most logical course of action. The initial response often (as with distal anticipatory stressors) is to use maladaptive emotion-focused coping strategies rather than trying to actively defuse or escape the stressor with appropriate problem-focused behavior. For example, Janis notes that many persons when experiencing acute chest pains and other symptoms signalling possible onset of a heart attack respond with some form of denial. Some take active steps to demonstrate to themselves that their symptoms could not indicate a heart attack by engaging in vigorous activity that actually could increase heart damage. Others simply repress and delay seeking treatment. Similarly, "emotional and personal blocks" impede the detection and reporting by professionals of potential or ongoing sexual child abuse (Finkelhor & Hotaling, 1983, p. 3; cited in Swift, 1986). This tendency to maladaptively distort warnings of impending danger in active avoidance situations has also been observed during the prelude to natural disasters. For example, Fritz and Marks (1954) found that among those exposed to an Arkansas tornado who noted a roaring sound prior to actual impact, less than one-third appraised it as implying a seriously threatening event and many interpreted it as indicating a passing train. Relatively little research has been reported on how to transmit warning information during the immediate preimpact period so that people will attend to it and act on it appropriately. One study assessing reactions during an impending flood (Drabek & Stephenson, 1971) suggests that people are more apt to take definitive evasive action when the source of the warning is perceived to be highly authoritative. But other findings indicate that, regardless of source, when warnings are not followed by negative consequences (as is often the case with natural hazards) subsequent warnings elicit a characteristic "false alarm" reaction— they evoke lower fear levels and tend not to be acted on (Breznitz, 1986; White & Haas, 1975).
Temporal Factors
141
Unavoidable Stressors In contrast to circumstances in which we are presented a signal that we are about to confront something unpleasant and must quickly decide on a course of action, we all occasionally confront aversive situations that we may have anticipated for varying periods of time and over which we may have some control over time of onset, but that are largely unavoidable. Some of these are highly associated with the passage of time itself (e.g., "empty nest," retirement, widowhood after a spouse's long illness). Others are fairly common events in which we choose to submit to short-term unpleasantness in order to receive longerterm anticipated benefits. Among the latter type, perhaps the most prevalent are invasive and dangerous medical and surgical procedures that we undergo in order to increase longevity and enhance health status. Considerable research has been undertaken evaluating the efficacy of stress management interventions, administered during the period just prior to stressor exposure, designed to modify patient anxiety levels and enhance coping outcomes (adjustment and possibly recovery). Much of the impetus for this research was provided by Janis's (1958) study of the relationship between preoperative anxiety and postoperative adjustment in surgical patients. His findings suggested that patients who adjust poorly either make insufficient use of emotion-focused coping mechanisms in the face of a minimally controllable situation, or overuse denial preoperatively to the point where they are unprepared for the problem-focused demands of the recovery period. However, subsequent studies that attempted to validate Janis's findings produced mixed results. In addition, though the extensive treatment literature that emerged following Janis's investigation indicated that interventions such as information, modeling, and cognitive-behavioral approaches are useful in promoting desirable outcomes, and that individual difference variables are significant determinants of patient adjustment in conjunction with some treatments, it also produced a number of contradictory results (Anderson & Masur, 1983; Auerbach & Kilmann, 1977; Schultheis, Peterson, & Selby, 1987). It has been argued that some of the inconsistency in findings may result from the fact that researchers do not obtain independent measures of the coping demands of the stressors as appraised by the patients, nor of the coping strategies elicited by interventions (Auerbach, 1989). Even though medical, dental, and surgical situations in general seem to provide relatively little opportunity for control in a real sense on the part of the patient, such situations are likely viewed differently by patients in terms of the degree to which important coping outcomes can be influenced by them through direct action on their part. One might predict that most patients would perceive greater opportunities for instrumental control during procedures in which they had to be conscious and follow instructions (e.g., many dental procedures, diagnostic examinations), and in those in which probability of return to complete health was high. For such situations, one would expect that stress management pro-
142
Personal Coping
cedures which induce a relatively high degree of problem-focused coping relevant to the demands of the situation (e.g., information about the impending procedure including instruments to be used and sequence of steps to be followed, about how to discriminate among sensations to be experienced, along with instruction in how to interpret events and sensations to be experienced, and practice in the use of these strategies; cf. Martelli, Auerbach, Alexander, & Mercuri, 1987), would be most effective, especially for internally-oriented persons who have a strong desire or propensity to deal with stress using problem-focused mechanisms. In contrast, in procedures involving loss of consciousness and a subsequent impatient convalescent period, and especially those involving tangible threat of death or severe debilitation (e.g., head and neck surgery for cancer; see Parrish, 1976), interventions more geared toward inducing emotion-focused coping (e.g., reinforcing perceptions of interpersonal support and caring, teaching techniques such as relaxation and attention redirection to pleasant fantasies) would likely produce a more positive response in the majority of patients. Though these hypotheses have not been directly tested with medical stressors, partial support is provided by the findings of a recent study by Strentz and Auerbach (1988). In this study, in which subjects were prepared for a highly stressful, minimally controllable, simulated captivity situation from which they were told escape was not an option, those who received preparatory instruction in emotionfocused coping strategies (and who actually used them to the greatest extent) adjusted better than subjects given instruction in problem-focused strategies or a control presentation. In summary, though stressors such as impending surgery or an invasive medical examination are largely not under patient control, they may have controllable elements or be appraised as such (especially by persons with a strong desire for control and involvement in what happens to them). Thus in considering how to best aid people to cope with such imminent stressors, it is reasonable to try to understand what aspects of the situation are most threatening to them and why, and to teach coping strategies which dovetail with the demands of the situation as appraised by them. POSTIMPACT STRESS McGrath (1970) has argued that "stress is defined as the anticipation of the inability to respond adequately (or at a reasonable cost) to perceived demand, accompanied by anticipation of negative consequences for inadequate response" (p. 23). Others have theorized about the likely differential effects of "anticipated relative deprivation" vs. "the gap between past expectations and current outcomes" on stress and strain (cf. Caplan, Tripathi, & Naidu, 1985). But empirical data bearing on both the question of the nature of the cognitions and appraisals which underlie anxiety experienced after exposure to aversive events, and on the comparative impact of anticipated vs. retrospective stressors on the stress response are sorely lacking. There are ample data, however, clearly demonstrating
Temporal Factors
143
that in the wake of exposure to extremely aversive situations both animals and humans experience emotional and motivational disruption and deficits in problem-solving ability, which under certain circumstances may persist for extended time periods (Auerbach, 1986; Maier & Seligman, 1976). In general, there has been relatively more emphasis on developing and evaluating interventions designed to enhance the coping of persons about to confront stressful circumstances than for those dealing with the harmful effects of stressor exposure. This can be attributed to several factors, including the influence of a "prevention" philosophy (particularly among crisis intervention specialists and community psychologists) that emphasizes that intervention take place as early as possible, and the lingering influence of traditional crisis theory (Caplan, 1964) that hypothesizes that most people will re-establish a stable emotional equilibrium within a relatively brief period (usually six to eight weeks) after stressor exposure, even in the absence of intervention. Further, particularly in the health care setting with patients facing invasive medical or surgical procedures, there is the implicit and often incorrect assumption that, compared to the preexposure period, the postexposure period is minimally stressful (cf. Auerbach, 1989). Moreover, emphasis is placed on the need for patients' informed consent—facilitating studies of the effects of information type and specificity as a pretreatment coping device. A final factor is the attraction of being able to implement experimental designs in an often more controlled and homogeneous setting (in terms of patients' physical condition and extraneous environmental inputs) during the immediate prestress impact period than in the more open-ended poststress impact period. The increased attention given recently to enhancing coping outcome during the postimpact period may in part be attributed to a generally increasing focus in our society on helping ameliorate the living conditions of victims of crime, poverty, war, disabling disease, and other untoward circumstances. Within the mental health fields much of this increased activity has coincided with the advent of the diagnostic category of post-traumatic stress disorder (PTSD), recently delineated in the 1980 edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III). This formalization of emotionally dysfunctional response to trauma and elevation of it to the status of a disease entity has led to a spate of work among clinical researchers, much of which however has addressed questions pertaining to validation of the established criteria of PTSD and to the nuances of differentiating between PTSD and other related diagnostic categories (e.g., Breslau & Davis, 1987; Horowitz, Weiss, & Marmar, 1987). While this work is useful, it has to some extent diverted attention from obtaining sound empirical data bearing on the questions regarding the parameters of the poststress response that need to be answered in order to systematically apply stress management procedures during the poststress periods. Specifically, what is the length of the interval between exposure to stress and onset of emotional disturbance? Under what conditions can a "latency" period of seemingly adequate functioning be expected prior to manifestation of overt signs of dysfunction (as observed in
144
Personal Coping
combat veterans, repatriated prisoners of war, and concentration camp survivors), and how long is this interval likely to be? Do qualitatively different stages of response inevitably follow in sequential fashion after exposure to particular stressors, and is experiencing each stage essential for achieving a satisfactory level of adjustment? What is the duration and ultimate severity of emotional distress experienced in response to traumatic stressors? Silver & Wortman (1980), Auerbach (1986), and Wortman and Silver (1989) have summarized the research which bears on these questions and its implications for instituting intervention procedures. Their conclusions are summarized below. As noted by Auerbach (1986), though the notion of the delayed traumatic stress disorder is widely accepted among clinicians, "there is no body of systematic research establishing the existence or parameters of a delayed stress response" (p. 25). The question of the regularity of the occurrence of qualitatively different poststress stages and its relation to long-term adjustment is more complex. Silver and Wortman (1980) and Wortman and Silver (1989) conclude that in general, clinicians and theoreticians hold strong assumptions and we have definite expectations about how the coping process unfolds in response to loss and other undesirable life events, but there is no body of systematic empirical research supporting the existence of regularly occurring, qualitatively distinct poststress stages. Further, there is little support for the notion that people must react in particular ways in order to adapt successfully to particular stressors. For example, there is little evidence supporting the widely held assumption that a period of depression following loss is predictive of superior coping outcome (Wortman & Silver, 1989). There is evidence, however, from both laboratory and field studies of some degree of homogeneity of response and of coping needs in the initial period after confrontation with a traumatic stressor. At this point, stress levels are apt to be disablingly high such that the individual is unable to process or at least is unreceptive to aggressive attempts to foster problem-focused coping strategies (cf. Auerbach, 1986; Strentz & Auerbach, 1988; Suls & Fletcher, 1985). Procedures designed to foster short-term, emotion-focused coping, including encouragement of temporary reconciliation to a situation over which the victim has minimal control, are likely to be most productive in this situation. It has been conjectured that such interventions will likely be delivered most effectively in a manner projecting empathic understanding by kin, or at least by persons of similar social background and life experiences as the victim (Bailey, 1988; Thoits, 1986). Regarding duration and ultimate severity of the stress response, though most victims of aversive life experiences are expected by others to ultimately return to preexposure level of emotional stability, many persons (even after initial gross indicants of behavioral dysfunction and emotional distress have seemingly stabilized) experience long-term adjustment problems that are clearly traceable to exposure to the traumatic stressor. Some seem to "continually re-experience the crisis for the rest of their lives" (Silver & Wortman, 1980, p. 308). This is
Temporal Factors
145
evident in many Holocaust victims who are never able to resolve guilt feelings over having survived (cf. Eitinger, 1980), in a large percentage of persons having lost a spouse or child in a motor vehicle crash who four to seven years after the event "continue to ruminate about the accident or what might have been done to prevent it and [who] appear to be unable to accept, resolve, or find any meaning in the loss" (Lehman, Wortman, & Williams, 1987, p. 218), and some rape victims who seem to maintain a "core of distress" which persists for years (Ellis, 1983). Exposure to traumatic stress thus produces both short-term and often longterm adjustment problems. In addition, recent findings indicate that there are important differences between coping mechanisms associated with good coping outcome immediately after a traumatic event vs. many years later (Schulz & Decker, 1985). These data suggest that it is meaningful to distinguish between proximal and distal phases of the poststress period, though the line of demarcation is not as explicit as it is in the anticipatory period (i.e., clear warning that an aversive event is about to occur). Primary coping needs during the distal period include learning problem-focused skills (e.g., widows or widowers who must for the first time make repairs and manage financial affairs and must initiate new social contacts) and cognitive changes. As noted by Auerbach (1986, p. 27) cognitive changes seem to primarily involve "a) conceptualizing the experience and one's reaction to it as normal and rational and, b) reinstatement of the belief that one is generally in control of one's environment rather than subject to the whims of circumstance." As in the proximal prestress impact period, the opportunity to interact with socially supportive peers appears to be an important source of coping assistance, although there is evidence that well-meaning attempts at support are often not regarded as helpful by victims because of misunderstandings by both parties about the most appropriate coping strategies and the length of the coping process (Wortman & Lehman, 1985). though there continues to be great interest in enhancing the long-term adjustment of traumatic stress victims, and though there has been a recent increase in clinical case reports (especially involving treatment of Vietnam veterans), interventions geared toward enhancing coping outcome during the distal poststress impact period have received little systematic evaluation. SUMMARY AND CONCLUSIONS At any moment in time, we range over the temporal landscape as we consider harms that may befall us (those that we are fairly certain will confront us, or that have begun to impinge on us), and those that have already impacted us— all within the context of our knowledge of past triumphs and unsatisfactory outcomes and the different coping strategies we used in those instances. This chapter has focused on the temporal relationship of stressors to the individual who is appraising them as an important situational moderator that helps us understand the coping tasks posed by these stressors and, by extension, the types
146
Personal Coping
of stress management procedures likely to be used most effectively in a given situation. It is recognized that other situational variables (including some related to time; see above) and person variables may affect stress appraisals and coping processes, and further that coping is inherently a dynamic, sequential process during which emotion-focused and problem-focused coping modes sometimes overlap and become indistinguishable as people deal with complex situations and the substressors they subsume. With stressors that are dealt with sequentially across several temporal stages, the manner and effectiveness of coping with them and with relevant intervening events are important determinants of how one subsequently appraises and copes with those events (or in some cases whether one ever confronts them) at the later stages. Thus, it is important to recognize the limitations inherent in cross-sectioning the flow of conscious experience, and attempting to make generalizations about which coping strategies are best to use with particular stressors at particular moments in time (Auerbach, 1989; Folkman & Lazarus, 1988c; Lazarus & Folkman, 1984). Given this caveat, some tentative conclusions may be drawn regarding the role, at each temporal stage, of different coping strategies and procedures designed to elicit them. Traditionally, suppression of dysphoric emotions has been considered a primitive and maladaptive way of coping with stress. When used in the extreme it inhibits realistic appraisal of stressors and cognitive activity leading to effective problem-solving. But moderate levels of emotion-focused coping suppress disabling anxiety levels and hypervigilance, likely allowing realistic situation evaluation and appropriate planning (cf. Janis, 1958). Thus emotion-focused coping seems to serve different functions, some of which are positive, at different temporal stages. In the distal preimpact period, emotional arousal level is low because the individual is unaware there is a threat or is using emotion-focused mechanisms to distort or suppress the meaning of information that is potentially threatening. In order to stimulate appropriate problem-focused activity, fear levels need to be aroused with accurate information which is presented reassuringly and is oriented toward reinforcing long-term fear beliefs, but which doesn't encourage unrealistic hope (Janis, 1969; Rippetoe & Rogers, 1987). Types of problemfocused coping during this period include efforts at preventing or lowering the probability of occurrence of potentially aversive events, of preventing or lowering the probability of exposure to such events should they occur, and preparations to minimize the damage they might inflict. In the proximal preimpact period, the perceived threat is palpable and arousal level is high. With stressors that may be actively avoided and with those the individual knows he will confront, both defensive avoidance and hypervigilance inhibit performance of appropriate problem-solving behavior (cf. Janis & Mann, 1977). For many potentially avoidable stressors, any coping instruction would of necessity have to have taken place during the distal preimpact period, since there is often a very limited period between warning signal and onset of stressor
Temporal Factors
147
(e.g., as in impending natural disaster or attack by hostile forces). Such instruction would emphasize rapid application of emotion-focused strategies (e.g., deep breathing; positive, antianxiety self-statements), if necessary, to moderate fear level in order to facilitate subsequent performance of appropriate problemfocused behaviors that address aspects of the situation that may be changeable. For largely unavoidable stressors for which there is often ample forewarning (e.g., surgical operation), intervention may take place during the proximal preimpact period. There is some evidence that teaching applications of both emotionfocused and problem-focused strategies (rather than just one or the other) results in optimal coping outcomes in such situations (Martelli et al., 1987). However, individual dispositional differences in desire for control and involvement also play a significant role in determining efficacy, as does in all probability the nature of the impending stressor in terms of degree of actual or perceived opportunities for controlling aspects of it (Auerbach, 1989). As in the proximal prestress impact period, emotion-focused coping likely plays an important positive role in moderating emotional distress in the initial stages of the proximal postimpact period, when both fear levels and confusion are apt to be very high. There is some evidence that attempts to focus on the event itself or to force problem-solving activity at too early a point may be counterproductive (cf. Auerbach, 1986). There is also evidence that social support, which can serve both emotion-focused and problem-focused functions, is an important form of coping assistance throughout the poststress period (cf. Thoits, 1986). In the distal postimpact period, cognitive reappraisal of their reaction to the event and of their control beliefs seems to be a particularly important factor in recovery for many victims. Overall, it is somewhat more difficult to make inferences about coping needs and when to institute interventions during the poststress impact than during the preimpact periods. This is because exposure to traumatic events may have wideranging and highly variable effects on emotionality and on ability to execute both problem-focused and emotion-focused coping functions, and there is a lack of normative data on the quantitative and qualitative parameters of the poststress response (Auerbach, 1986; Silver & Wortman, 1980) that would indicate optimal points of intervention.
This page intentionally left blank
10
Life Crises and Personal Growth Jeanne A. Schaefer and Rudolf H. Moos
More than 20 years ago, Caplan (1964) asserted that life crises are "turning points" and times of opportunity as well as risk, but the growth-promoting aspects of life crises have yet to be examined. We know little about the types of positive outcomes that occur following crises, or how characteristics of a crisis and associated personal and environmental factors affect the likelihood that a person will experience psychological growth. For the most part, clinicians and researchers have sought to identify and prevent the problems associated with life crises and transitions. In studies of adaptation to life crises, investigators typically equate good outcome with the absence of physical symptoms and psychopathology. They usually fail to consider the possibility of a new and better level of adaptation that reflects personal growth rather than a return to the status quo. The difficult aspects of life crises cannot be denied. People in transitions and crises experience painful emotional reactions and potentially devastating psychological and physical effects. Yet many people are remarkably resilient in the face of adversity. They manage to find acceptable and sometimes creative solutions and may become more mature and self-confident. People often emerge from a crisis with new coping skills, closer relationships with family and friends, broader priorities, and aricherappreciation of life. Such outcomes are remarkably common. When asked about the impact of a life crisis they have experienced,
Preparation of the manuscript was supported in part by NIAAA Grants AA02863 and AA06699 and by Department of Veterans Affairs Medical and Health Services Research and Development Service research funds. We thank Charles J. Holahan, David Spiegel, and Christine Timko for their valuable comments on an earlier draft of the manuscript.
150
Personal Coping
a substantial number of people, typically 50 percent or more, report some positive outcomes (Cleveland, 1980; Dhooper, 1983; Wallerstein, 1986; Yarom, 1983). Our aim is to provide a better understanding of the positive consequences that can follow a life crisis or transition. Based on a review of the coping and crisis literature, we formulate a conceptual framework that considers the positive outcomes of life crises and present a preliminary way to categorize them. We provide examples of good outcomes of specific crises such as divorce, physical illness, and bereavement. We then consider the environmental and personal determinants of positive outcomes, suggest ideas for research on the growth-promoting aspects of life crises, and describe some implications for intervention programs. LIFE CRISES AND TRANSITIONS: AN IMPETUS FOR PSYCHOLOGICAL GROWTH Dialectical psychologists such as Riegel (1976) see change as necessary for human development. Life crises are viewed as ''constructive confrontations" that spur development. Personal growth can be fostered by the disruption that crises generate and the subsequent reorganization that occurs in their wake. Stressors are a natural and potentially positive part of life; resilience develops from confronting stressful experiences and coping with them effectively (Rutter, 1986; Selye, 1974). People in crisis often experience disruptions in significant relationships, challenges to their basic values and beliefs, role changes, and new demands. The process of confronting these experiences can promote cognitive differentiation, self-confidence, and a more mature approach to life. A person who experiences pain and loss may develop a deeper understanding and empathy for others with similar problems. Exposure to novel crisis situations may broaden a person's perspective, promote new coping skills, and lead to new personal and social resources. As Haan (1982) succinctly stated, "Stress benefits people, making them more tender, humble, and hardy" (p. 255). At least two longitudinal studies have identified associations between life crises and psychological growth. Consistent with her ideas, Haan (1977) found that men and women who experienced more change in their primary family (divorce, remarriage, death of a family member) during adolescence and the early adult years were more empathic and tolerant of ambiguity in middle age. Elder (1974, 1979) analyzed the long-term effects of economic deprivation on children who grew up during the Depression. Some children from deprived families became more competent and resilient than did their counterparts from nondeprived families. These positive outcomes were more likely to occur among girls and middle-class children. Elder attributed them to the added responsibilities these children assumed and the family cooperation that was required in order to cope with financial problems. Like crises, normative transitions can also spur development. Erikson's (1963) eight life stages encompass new challenges that enable a person to manage sue-
Life Crises and Personal Growth
151
ceeding stages provided they are negotiated successfully. According to Erikson, adequate resolution of a transition creates new coping resources that can help a person confront subsequent crises. Normative transitions and sudden life crises share common features that can facilitate personal growth. In both instances, events may stimulate changes in cognition and behavior, a review of life goals, and a reassessment of values and beliefs. Life crises and transitions both promote new cognitive and personal skills that enhance effective adaptation (Medinger & Varghese, 1981). In a series of longitudinal studies, Stewart and her colleagues (1986) examined the process of adaptation among children and adults who experienced transitions such as entering junior high and high school, beginning and graduating from college, marriage, and parenthood. Life changes precipitated a shift toward a receptive emotional stance, which may foster more sensitivity to the environment and more openness to new information and ideas. Later in the postchange period, individuals gradually became more assertive and independent; during this time they may learn how to use and integrate new knowledge and skills. Thus, personal growth can arise from the process of inner adaptation to life changes. CONCEPTUAL FRAMEWORK We have used a stress and coping framework to understand the process of adaptation to physical illness and other life stressors (Moos & Schaefer, 1984, 1986). Here we adapt this framework to examine the determinants of beneficial consequences of life crises and transitions. According to the framework in Figure 10.1, the environmental and personal systems (panels I and II) jointly affect the likelihood and characteristics of a life crisis or transition (panel III). The environmental system is composed of the individual's ongoing life context, including the relatively stable precrisis aspects of his or herfinancial,home, and community living situation as well as characteristics of relationships with family members, work associates, and friends. The personal system includes an individual's sociodemographic characteristics and such personal resources as cognitive ability, health status, motivation, and self-efficacy. Life crises or transitions typically reflect changes in ongoing personal factors, such as a physical injury or illness, or environmental factors, such as the death of a spouse. We organize the coping responses people typically use to manage life crises and transitions (panel IV) into three domains that reflect their primary focus: appraisalfocused coping involves efforts to define, interpret, and understand a situation. Problem-focused coping covers efforts to resolve or master life stressors by seeking information, taking direct action, and finding alternative rewards. Emotionfocused coping involves attempts to manage emotional reactions to life stressors by regulating one's feelings, expressing anger, and accepting the situation. These coping responses are affected by environmental, personal, and crisis-related factors; in turn, they may enhance the likelihood of positive outcomes. Overall, the model posits that life crises (panel III) and the environmental and
Figure 10.1 A Conceptual Model for Understanding Positive Outcomes of Life Crises and Transitions
Life Crises and Personal Growth
153
Table 10.1 Major Types of Positive Outcomes
Enhanced Social
Resources
Development of a confidant relationship BeUer relationships with family members and friends Formation of new support networks Enhanced PersonalResources Cognitive and intellectual differentiation Self-reliance and self-understanding Empathy, altruism, and maturity Changes in basic values and priorities Development of New Coping Skills Cognitive coping skills Problem solving and help-seeking skills Ability to regulate and control affect
personal factors that foreshadow them (panels I and II) can shape appraisal and coping responses (panel IV) and affect the likelihood of a positive outcome (panel V). The bidirectional paths present these processes as transactional and show that reciprocal feedback can occur at each stage. TYPES OF POSITIVE OUTCOMES We have identified three general categories of positive outcomes that reflect three of the sets of factors in our conceptual model: (1) enhanced social resources (panel I); (2) enhanced personal resources (panel II); and (3) development of new coping skills (panel III). In Table 10.1 we outline ten types of positive changes within these three general categories. These ten types encompass most of the positive outcomes identified in the crisis and coping literature. The ten categories are interrelated. The development of new support networks may be
154
Personal Coping
associated with increased empathy and altruism; new cognitive coping skills may be related to changes in values and priorities. Enhanced Social Resources This category includes the development of a relationship with a confidant, better overall relationships with family members and friends, and the formation of new social networks. Development of a confidant relationship. The demands associated with a crisis often tax a person's coping capacity and deplete existing personal resources. Consequently, people may express their fear and weakness, share their innermost feelings, and seek help from a spouse, partner, or special friend. This type of help seeking can foster a deeper and more meaningful confidant relationship. Such a relationship then becomes a basic resource that can help the individual negotiate future crises. This process sometimes occurs in the aftermath of a crisis, such as when prisoners of war learn the value of interpersonal communication in captivity and then share their experiences with a spouse or other confidant when they return home. Personal relationships with family andfriends.Life crises that place an individual in a dependent or vulnerable position, such as a serious illness or natural disaster, often necessitate an increased reliance on others and a need to work cooperatively. For example, physically ill people interact with their caregivers, victims of a natural disaster establish communication networks, and soldiers in battle cooperate with each other in order to survive. People who undergo such experiences often develop better communication skills and a greater appreciation of the value of interpersonal relationships. Improved personal relationships may also emanate from joint problem-solving and caretaking responsibilities in families where a member is ill or disabled. In addition, the threatened loss of a family member due to illness may draw family members together and lead to more open communication. Formation of new support networks. Individuals in crisis often search for people who can help them to understand and find meaning in their plight. People may nurture informal sources of support with friends or neighbors in order to meet the demands posed by a crisis. Moreover, they often contact counselors or join mutual support groups. Involvement in a self-help group can provide a person with a sense of purpose, a way to find meaning in suffering, and opportunities to learn coping skills from others with similar problems. In the long term, support groups can enhance an individual's social network and provide opportunities to establish new relationships by reaching out to other people. Enhanced Personal Resources This category includes growth of personal qualities such as cognitive and intellectual differentiation, self-reliance, empathy and maturity, and changes in basic values and priorities.
Life Crises and Personal Growth
155
Cognitive and intellectual differentiation. The novel experiences associated with life crises and transition provide an opportunity to gain new information about the world. For example, chronically ill persons may learn about the medical aspects of their illness, ways of managing it, and how to interact effectively with health care providers. Crises involving role changes and new responsibilities often promote cognitive and intellectual skills. Thus, children learn about the adult world when they assume new responsibilities after the death of a parent or separation from a parent following a divorce. As people learn more about the meaning and consequences of a crisis, they often develop more complex and differentiated representations of the world around them. This process may lead to greater tolerance of ambiguity and more openness to new information and novel ideas. Self-reliance and self-understanding. Self-reliance often increases when people acquire new skills, become more independent, and successfully manage new roles or seemingly overwhelming tasks and challenges. Self-confidence may be enhanced when a person is able to cope effectively with the unsettling emotions and practical problems aroused by crises, such as a brutal rape, a dehumanizing experience in war, or the vicissitudes of divorce. In the process of working through a crisis, a person may also develop increased self-understanding, as when a man going through a divorce gains insights about how his excessive focus on work at the expense of family contributed to the breakdown of his marriage. Empathy, altruism, and maturity. A serious illness or the death of a loved one can provide an individual with first-hand knowledge of pain and suffering as well as insight into the plight of others with similar problems. Empathy and compassion spring from a better understanding of the feelings and needs of other people. The altruism that people exhibit following a crisis may be a response to the help they receive. Family members and friends frequently rally around a person in crisis; when individuals recover from the crisis, they may reciprocate by reaching out to others in need. Altruistic behavior also provides a way for bereaved and victimized individuals to manage their emotions and contribute something useful to society. Increased empathy and altruism often lead to greater personal maturity. Changes in basic values and priorities. Exposure to unfamiliar lifethreatening situations such as the unexpected death of a child or a spouse, or being diagnosed with a serious illness, may completely shatter long-held beliefs about one's self and the world. These crises produce emotional turmoil and force people to question basic values and priorities and painfully restructure life goals. In addition, people who come face to face with death often set new priorities and change the direction of their life. This type of crisis triggers the realization that life is short and that there is little time to accomplish one's goals and make an impact on the world. People who gain these insights begin to live their life differently and to experience their life as more meaningful and satisfying.
156
Personal Coping
Development of New Coping Skills As people struggle to come to grips with the demands of a crisis or transition, they often try out new ways of dealing with the problem and thus learn new coping skills. Coping skills developed during one crisis may prove to be useful resources when future problems arise. Cognitive coping skills. Efforts to manage crisis-related problems can lead to better cognitive coping skills, such as the ability to logically analyze a situation and divide it into potentially manageable parts. Moreover, in the process of coping with a crisis, a person may learn to search for the positive aspects of a situation and to find some deeper meaning in it. The cognitive demands of a life crisis are likely to strengthen cognitive coping skills that can be applied in other situations. Problem-solving and help-seeking skills. People who respond to crisis by seeking information and taking direct action often learn new problem-solving coping strategies. This process commonly occurs in managing crises that involve role changes and added responsibilities such as divorce, bereavement, and physical illness. In turn, problem-focused coping can foster new daily living skills, as when a widower learns how to cook or a wife acquires the technical knowledge needed to manage her husband's home dialysis treatment. New coping skills may also result from a trial-and-error approach to a problem or from a person's deliberate effort to acquire new knowledge, such as by seeking advice from family, friends, or counselors. Ability to regulate and control affect. The need to withhold immediate action and work through one's feelings may help an individual develop new skills in regulating affect and handling emotions. Crises over which people have little control, such as a personal illness or becoming a prisoner of war, may stimulate them to learn relaxation and meditation skills and contribute to their ability to regulate painful emotions. SPECIFIC CRISES AND ASSOCIATED POSITIVE OUTCOMES The positive outcomes we have identified occur in a variety of life transitions and crises. These include the common crises of divorce, illness, and death and more unusual situations such as combat and imprisonment in war. Divorce The changes associated with divorce often foreshadow dramatic shifts for women, who encounter new tasks as they become head of the household and assume sole responsibility for parenting. Divorce may stimulate a woman to become more self-reliant by obtaining a full-time job, establishing credit in her own name, or resuming her education. These new skills and accomplishments
Life Crises and Personal Growth
157
signal more independence and may lead to increased pride and self-confidence (Kohen, 1981). For example, low-income, divorced mothers who were able to forge new identities and develop new social resources also became more assertive and self-confident (Miller, 1982). Such positive changes are quite prevalent. In a ten-year follow-up, Wallerstein (1986) found that 58 percent of women in their thirties and 44 percent of women in their twenties at the time of their marital breakup experienced long-term improvements in psychological functioning. The women became more assertive about expressing their feelings and developed more realistic views of themselves and the world. Success in new careers was accompanied by increases in selfesteem, especially among women whose roles had been confined to those of wife and mother during their marriage. Children who assume added responsibilities following a death or a divorce also show increased personal resources. Wallerstein and Kelly (1980) found that some children of divorce were more willing to help with household tasks and held more mature views about finances and the problems of marriage. Similarly, Weiss (1979) found that adolescents in single-parent families assumed more responsibilities than their peers in intact families. They were more involved in household decisions and sometimes became substitute parents for younger siblings. Moreover, many children provided emotional support to their parents. Although the adolescents sometimes longed for a more carefree childhood, they prided themselves on being able to handle responsibilities that their peers could not. They became more independent and self-reliant and reported a heightened sense of competence.
Combat, Captivity, and Rescue Work People who successfully manage unfamiliar or seemingly overwhelming tasks often experience a sense of accomplishment and increased self-esteem. Sledge and his coworkers (1980) found that some Vietnam prisoners of war (POWs) experienced positive changes. Some soldiers felt wiser and more tolerant than before they were imprisoned. They also had more self-confidence, which probably stemmed from having successfully met challenges that tested them to their limits. During their lengthy captivity, the men had a chance to reflect on life and evaluate what was meaningful to them. Consequently, some of the men experienced positive personality changes and changed their views of the relative importance of family, career, and relationships with others. During the months and years of captivity, the POWs were forced to live with fellow soldiers who initially were strangers, to negotiate differences, and to rely on each other for support. The men valued the close personal ties they developed with fellow prisoners. These experiences contributed to improvement in the POWs' relationships with others, increases in their understanding and patience, and a greater appreciation of the value of communication (Sledge, Boydstun, &
158
Personal Coping
Rabe, 1980). Upon their return home, some released prisoners of war reported deepened marital relationships and more open communication. Similarly, Israeli soldiers who fought in the front line during the Yom Kippur war experienced intense demands in combat that tested their psychological and physical strength. Success in handling these situations made the soldiers feel stronger and convinced them that they could cope with other life stressors. Facing death in combat situations made them aware of their vulnerability and mortality. Consequently, they tried to make their lives meaningful. They indicated that they enjoyed life more, formed more intense relationships with other people, and felt closer to their wives and children (Yarom, 1983). People who engage in dangerous activities and confront sudden death on a massive scale also change as a result of their experiences. Raphael and her associates (1980) studied rescue workers who responded to a rail disaster and found that many workers reevaluated their lives, placed more importance on personal relationships, and were less interested in material possessions. After confronting massive numbers of dead, the rescue workers came to see life as more precious. Likewise, the majority of rescue workers involved in the Hyatt Hotel disaster reported that their experiences had changed their view of life. Some of them now understood the fragility of life, while others became more compassionate and altruistic (Miles, Demi, & Mostyn-Aker, 1984).
Chronic Physical Illness Chronic physical illness is especially likely to stimulate significant changes in family relationships. For example, a middle-aged man who became a quadriplegic following an accident found that his relationship with his wife matured. After one bad marriage, his second was marked by a fear that his wife might abandon him. The accident provided an opportunity for him to be reassured of his wife's fidelity. His wife visited him daily during his nine-month hospitalization. He later said the accident was "the best thing that ever happened to me. For the first time in my life, I think I really know what love means" (Cassem, 1975, p. 13). The caregiving tasks associated with chronic illness provide an opportunity for family members to express their love for each other. Increased closeness may arise from attending to the needs of the ill family member and sharing the burdens of care, as well as from joint problem-solving activities, and from the threatened loss of a family member. Palmer and her colleagues (1982) studied families engaged in home dialysis. They anticipated that the burden of home dialysis tasks would have a negative effect on the marriages. However, about half of the patients and their spouses reported that their marital relationship had become closer. Similarly, Dhooper (1983) found that nearly half of the spouses of first-time heart attack victims reported that their family ties were strengthened by the crisis
Life Crises and Personal Growth
159
experience. Spouses noted increased emotional closeness and a better appreciation of their partner (see also Meddin & Brelje, 1983). In families where a child had suffered a spinal cord injury, Cleveland (1980) discovered that the demands of caring for a chronically ill child sparked marital tension. Nevertheless, parents reported increased closeness to their injured child, and injured sons enjoyed the added nurturing they received from their fathers. Siblings reported feeling closer to their parents and communicated more openly with each other after the accident. Besides these positive outcomes, children with a chronically ill sibling may develop increased empathy for their parents and more respect for their ill sibling.
Terminal Illness, Death, and Bereavement Even a loss as profound as the death of a loved family member can result in positive changes. Miles and Crandall (1983) found that some parents whose children had died identified positive consequences associated with their loss. Bereaved fathers changed their priorities and spent more time with their families. Parents became more empathic and compassionate and were better able to help others who had suffered a loss. In addition, the parents became more aware of the preciousness of life and lived life more fully. Like adults, children can emerge from the crisis of the death of a family member with increased personal resources. Balk (1983) noted that some adolescents who lost a sibling became wiser and viewed themselves as more mature and self-confident than their peers. The bereaved adolescents lost their sense of invulnerability and realized that a sudden tragedy can permanently alter one's life. Like the bereaved parents, they learned to value each day. Terminal illness can trigger significant changes in people's behavior, values, and priorities. Cancer patients may view the threat posed by their illness as a "catalytic agent for restructuring their lives" (Taylor, 1983). For example, some women with metastatic cancer who participated in a support group reported improved communication with members of their families. Because time was precious, the women became more assertive and did what they liked rather than trying to please others (Spiegel & Yalom, 1978). Finally, some people may find new freedom in successfully confronting their worst fear—death. For some cancer patients, life becomes more exhilarating and their senses are heightened. Patients speak of really "seeing" flowers for the first time and appreciating the seasons as never before (Bertman, 1983). In the face of pain and the threat of death, people may learn to experience life more fully. For example, Robert Mack (1984) underwent painful surgery and coped with the side effects of chemotherapy and radiation for recurrent lung cancer with widespread metastases. Nevertheless, he writes of a more satisfying life following his illness,
160
Personal Coping
I am not being concerned about my impending death. I enjoy all the wonderful relationships I have. I am happier than I have ever been. These are truly among the best days of my life (p. 1642). Mack's experience and those of others like him raise questions about the nature and consequences of life crises. Why is it that some people experience personal growth following a crisis and others do not? What are the determinants of positive outcomes of a crisis? We use the conceptual framework presented earlier to address these questions and to identify some important research directions. DETERMINANTS OF POSITIVE OUTCOMES As shown in Figure 10.1, we posit four sets of determinants of positive outcomes: characteristics of the life crisis or transition, environmental factors, personal factors, and appraisal and coping responses. These four sets of factors influence each other and may be affected themselves by the short-term outcome of a crisis. Event-Related Factors Why is there so much individual variability in response to life stressors? Rutter (1986) points out that some of the variation in response may reflect differences in what appear to be similar stressors and differences in the experiences that follow. In their study of psychiatric reactions to the Mount St. Helens volcanic eruption, Shore and his colleagues (1986) found a progressive "dose-response" relationship. Compared to people in a matched nonexposed community, those who lived in the exposed community were somewhat more likely to develop depression, anxiety, and post-traumatic stress reactions. People who were more closely affected by the disaster, as indicated by a death in the family or substantial property loss, were much more likely to develop these disorders. Similarly, Gleser and her coworkers (1981) found that people who experienced more severe losses in the Buffalo Creekfloodreported higher levels of depression and anxiety. In addition to severity, life events can vary in their focus (self or other), duration, suddenness of onset or predictability, controllability, extent or pervasiveness (the diversity of life areas that may be affected), whether they are experienced by one individual or by a group of people together, and the specific change-inducing problems associated with them (Rees & Smyer, 1983; Thoits, 1983). By shaping the coping tasks people confront, such problems may determine the nature and extent of positive outcomes. Crises that involve a threat to life (serious illness, death of a significant person) may make one value life more; crises that test a person's physical and moral courage (combat, imprisonment) may result in increased self-reliance. A few studies have tied these event-related factors to crisis outcomes. Parkes
Life Crises and Personal Growth
161
and Weiss (1983) found that surviving spouses whose partner's terminal illness was of a longer duration and who had an opportunity to prepare themselves and discuss impending death with their partner experienced better postbereavement outcomes. Sudden, unexpected deaths were associated with poorer outcomes. Similarly, Rando (1983) noted that the time available for parents to prepare for their child's death from cancer had an impact on adjustment. Parents whose child was ill between six and seventeen months before dying did better than parents whose children died less than six months or more than seventeen months after becoming ill. Parents whose child died soon after diagnosis had little warning and could not prepare themselves for the death. Parents whose child died long after the diagnosis also were ill-prepared for the death, perhaps because numerous remissions gave them a false sense of hope. Our current knowledge about the association between event-related factors and positive outcomes is limited. Why are people who confront expected crises, such as a long-awaited death, better able to assimilate the impact of the event? What is the optimal match of new demands and personal maturity? The added responsibilities assumed by some adolescents in single-parent families may contribute to their accelerated maturity. However, a young child who is saddled with multiple demands may be unable to meet normal developmental tasks and begin to experience psychological problems. To address these issues, we need to develop objective ways to characterize stressful life circumstances. One approach is to look more carefully at the stimulus configuration itself. Paterson and Neufeld (1987) have taken an important step in this direction by describing the situational determinants of the primary appraisal of threat. They show how the severity of a stressor reflects three dimensions: the importance of the goals likely to be blocked, the number of goals blocked, and the intensity of the deprivation. Anticipatory stress is also more likely when the stressor is imminent and when it is very likely to occur. It is hard to know whether this type of analysis can be applied to dimensionalizing complex stressors, but it is likely that variations in the characteristics of stressors will explain part of the variation in individual outcomes. As indicated by our model, however, outcomes are also affected by personal and environmental factors. Personal Factors Personal factors include demographic characteristics such as age, gender, and socioeconomic status as well as temperament, ego strength and self-confidence, philosophical or religious commitments, and prior crisis and coping experience. These factors help to define the crisis and to shape the coping resources used to manage it. Sociodemographic characteristics. Social class is an important moderating factor, probably because it reflects the level of available personal and social resources. Elder (1974) found that deprivation during the Depression stimulated personal growth among middle-class children, but not among working-class
162
Personal Coping
children, who tended to experience negative outcomes. These findings were strongest among middle-class girls from deprived families, who were more goaloriented and assertive in adolescence than were middle-class girls from nondeprived families. Elder (1979) attributed these beneficial outcomes to a strong emotional bond between the girls and their mothers, whose gainful employment and modern values may have helped their daughters develop more independence and self-confidence. In her study of divorce, Wallerstein (1986) noted that both sex and age had an impact on long-term outcomes. Four times as many women as men grew psychologically in the ten-year postdivorce period. Women who were in their twenties or thirties at the time of divorce did better than those who were forty or older. Younger women were more resilient and made more economic and emotional progress than older women. Temperament. Research on resilient children has demonstrated connections between temperament and adaptation. Werner (1986) compared children of alcoholics who showed good long-term outcomes from those who did not. The resilient children had a "cuddly and affectionate" temperament as infants; they later showed higher school aptitude and achievement, an internal locus of control, and more self-esteem. However, some temperamentally difficult infants may demand more special family resources that stimulate their long-term intellectual development. According to Maziade and his associates (1987), when environmental opportunities are available, such as in high socioeconomic status families and families in which there is good communication, temperamentally difficult infants may provoke more interaction, obtain more attention and stimulation, and thus develop better intellectual abilities. This benign sequence of events is less likely to occur among children in deprived families. Self-Reliance and Self-Control. People who are more self-reliant and selfconfident are more likely to employ coping strategies that lead to successful crisis resolution. According to Holahan and Moos (1985), stress-resistant individuals are more self-confident and easygoing and less likely to rely on avoidance coping than persons who do less well when experiencing stressors. Similarly, Kobasa and her colleagues (1982) showed that personal hardiness enables people to resist the negative consequences of life stressors. Self-complexity may also moderate the potentially adverse impact of stressors. Linville (1987) found that people who had a more complex image of themselves were less prone to depression, physical symptoms, and minor illnesses following high levels of stressful events. People with a high degree of self-control and who have a sense of purpose in life may be more likely to experience positive changes. Rosenfeld and Krim (1983) found that a group of families who were dislocated by fire managed as well or better after the fire than before. The heads of these households acquired new skills through educational programs, found employment, and expanded their use of community resources. Overall, the men and women who managed well had a positive orientation to life. In this respect, Antonovsky (1987) believes
Life Crises and Personal Growth
163
that a sense of coherence, a global orientation in which the world is seen as comprehensible, manageable, and meaningful, can help people adapt well under stress. Prior crisis experiences. Several researchers have found that prior experience and success in coping with life stressors strengthen a person's selfefficacy and coping resources. In fact, resilience may stem from direct exposure to and successful coping with a small or modified dose of a noxious agent or stressor. Protection lies in the effective engagement with stressors and the adaptive changes that follow rather than in living a stressor-free life (Rutter, 1987). In support of these ideas, Caspi and his coworkers (1987) found that women who had recently experienced acute life stressors reacted less negatively to stressful daily events. They speculate that prior exposure to stressors may lead to a dampening of emotional reactions to subsequent stressors and that experience in coping with major life events can provide feedback on effective ways to manage stress. These processes may help older adults adapt to bereavement. Norris and Murrell (1987) found that family stress increased as the time of a family member's death approached. After the death, the survivor's health worsened if there had been no family stress preceding the death, but otherwise it improved. Ruch and her colleagues (1980) examined the influence of prior life change on the emotional trauma experienced by women rape victims. They found a curvilinear relationship; women who experienced a moderate level of recent life change were less vulnerable than women who experienced either no changes or many changes. The authors speculate that managing some life changes may enhance self-confidence and facilitate the acquisition of new coping skills. Life crises may ultimately be beneficial when they are relatively manageable and provide an opportunity to learn new coping skills. Personal resources are likely to exert their influence in combination with environmental factors. For example, personal resources may enable people to obtain more support from their social ties. Riley and Eckenrode (1986) found that people with more demographic (more education and income) and psychological (internal locus of control, positive beliefs in the benefits of help seeking) resources were better able to mobilize network support after a stressful event. The mobilization of support by high-resource individuals was associated with higher psychological well-being, whereas more mobilized support among lowresource individuals was associated with less well-being. Thus, people with more personal resources are able to mobilize more postcrisis support and to benefit more from it. We turn now to the environmental factors involved in these stress and coping processes. Environmental Factors Informal social support and family cohesion. Social networks and social support can promote a beneficial effect when people experience stressors. Adequate
164
Personal Coping
social resources may help a person redefine an event as relatively benign and make it possible to focus on potential positive consequences. Social support is associated with better rehabilitation outcomes for hemodialysis and myocardial infarction patients and better adaptation to breast cancer and long-term illness (Moos, 1985). A stable and cohesive family climate is an especially important stressresistance factor, which may protect against disruptive and aggressive responses to stress and facilitate the social engagement of children who are experiencing stress (Garmezy, 1987). Supportive families may help child victims of a flood recover quickly, whereas children from irritable or violent families remain more impaired (Gleser, Green, & Winget, 1981). The impact of the quality of family life on adaptation is evident in Werner's (1986) findings. She noted that children of alcoholics who adapted better enjoyed more attention from primary caregivers during their first year of life and experienced less parental conflict and separation during the first two years of life. Elder (1979) found that the outcome of economic deprivation during the Depression was influenced by the prior strength of the marital bond. Under conditions of marital harmony, economic loss due to the father's inability to earn an adequate income strengthened the relationship between fathers and their children. In the absence of marital harmony, deprived boys obtained less emotional support from their fathers, but deprived girls developed closer ties with their mothers. Supportive services and community resources. Some supportive health services can promote family communication and facilitate positive outcomes following the terminal illness of a family member. Lauer and her colleagues (1985) compared children whose sibling participated in a home-care program and died at home with those whose sibling died in hospital. Among children in the homecare group, 85 percent reported improved relationships with their parents and more family closeness; none reported poorer family relationships. These children received support from their parents and information about their dying sibling. In the hospital group, however, only 18 percent of the children developed closer relationships with their parents and 25 percent described worse relationships. Health and counseling services are more likely to promote positive outcomes when they are integrated with informal sources of support. A stable and supportive community can also facilitate postcrisis adaptation. Kaffman and Elizur (1983) found that the environment of the Israeli kibbutz enabled children to cope with their grief following the death of their fathers. Children became attached to family or friends living in the kibbutz, slept in communal children's homes, and were not as dependent on their relatives. In contrast, non-kibbutz children whose fathers died interacted less with extended family, experienced more stressful family changes and less stability in daily life, and had more behavioral and psychological problems. The kibbutz environment may promote adaptation because it facilitates help seeking to solve group coping tasks.
Life Crises and Personal Growth
165
The postcrisis environment. The quality of the postcrisis environment in likely to have a strong influence on outcome. Rutter (1986) points out that parental illness or divorce can lead to quite different experiences for different children. The long-term effects of acute stressors may be due mainly to the chronic adversity they foreshadow. Conversely, "acute stressors" that lead to an improvement in a person's life circumstances, such as separation from a psychotically depressed mother or a sexually abusive father and placement in a more benign family environment, may be associated with improved adaptation. In this vein, Vaillant (1977) describes how a middle-aged man experienced "a sudden burst of health" after the death of his paranoid mother. Pearlin and Lieberman (1979) note that many events entail both loss of an old role or status and acquisition of a new one. In a four-year follow-up of people who retired, these investigators found that the quality of the postretirement experience was critical. When people found relatively benign conditions in their new roles, they enjoyed as much mental health as did people whose occupational lives were stable. An analysis of the effects of marital separation, divorce, and widowhood resulted in similar findings; positive changes may occur when experiences in the marital situation are especially stressful or those in the new role are especially benign. Oatley and Bolton (1985) point out that people often become depressed when a life event disrupts a primary role that provides a source of self-esteem, but that a renewed sense of self can emerge when a life crisis is followed by a positive change in personal relationships or other life circumstances. We need a better understanding of the role of both precrisis and postcrisis environmental factors in the genesis of positive outcomes. Under some conditions, factors that buffer the potentially negative effects of life stressors, such as social support, are associated with positive outcomes. Caspi, Bolger, and Eckenrode (1987) found that stressful daily events had a positive influence on mood among women who had a high level of social support. Some stressful events may enhance well-being when they provide a context for positive social exchanges with members of one's social network. As shown in our model, a person's social resources may also promote positive outcomes by facilitating the development of more effective coping strategies. Appraisal and Coping Responses Appraisal and coping are closely linked in the stress and adaptation process. Cognitive appraisal refers to the individual's perceptions and interpretations of potentially threatening aspects of the environment and subsequent evaluations about what can or should be done about them. Causal attributions are one aspect of the interpretation stage of the appraisal process. Through their influence on coping, causal attributions and appraisals may have an impact on the type and extent of positive outcomes. Lazarus and Folkman (1984) identify three major types of appraisals of a
166
Personal Coping
potential stressor: harm-loss, threat, and challenge. People who anticipate that a stressor is a chance for mastery or gain, and those who attribute the cause of the stressor to external and transient factors, may cope more actively and be more likely to experience positive outcomes. However, when a stressful situation, such as a rape or an accident, is seen as having been potentially avoidable, an internal attribution may be associated with better outcome (Brewin, 1984). More specifically, people who blame their behavior in that situation (an internal but transient and specific cause) may do better than people who blame a defect in their character, which could predispose them to experience other similar stressors in the future (Janoff-Bulman, 1979). Cognitive-coping processes can help some people concentrate on the beneficial aspects of a crisis. They may employ a process such as cognitive redefinition wherein the basic reality of a situation is accepted and restructured to find something favorable. Taylor and her colleagues (1983) contend that people try to cope with the aversive aspects of irreversible crises by selectively enhancing themselves and their situation. For example, a young man who was accidentally blinded convinced himself that his blindness was a benefit in that other people's appearances would no longer distract him and he could appreciate people for their deeper qualities (Hoehn-Saric, Frank, Hirst, & Seltser, 1981). In another example, Silver and her coworkers (1983) found that some incest victims made sense of their experiences by emphasizing the positive outcomes that resulted from it, such as increased confidence that they could handle future problems because they had already confronted one of the worst situations they could imagine. Cognitive-coping strategies may entail an effort to find meaning in the crisis. Taylor (1983) found that cancer patients' search for meaning helped them to understand the crisis and its implications, to regain mastery over the crisis and over their life, and to restore their self-esteem. For many cancer patients, the meaning they derived from their experience brought a new attitude toward life, reordered priorities, and increased self-knowledge. Some might argue that these cancer patients and the blind man just mentioned are rationalizing or denying the impact of their illness or disability, and that their behavior is defensive rather than indicative of positive change or personal growth. We prefer to construe these cognitive processes as cognitive redefinition and positive comparisons and to emphasize their adaptive value. Cognitive redefinition and positive comparisons can be effective coping strategies that enable individuals to maintain hope, master painful emotions, and bolster their mood and self-esteem. Therefore, we emphasize the positive outcomes that may flow from cognitive strategies that minimize the traumatic aspects of crises and involve a search for the meaning of a difficult situation. The coping strategies people use can alter their adaptation to a crisis and influence the likelihood of positive outcomes. In general, people who use more active coping responses experience better outcomes and report fewer psychological symptoms, whereas those who rely on avoidance coping adapt less well.
Life Crises and Personal Growth
167
Problem-focused coping responses can facilitate adaptation to some life crises; they may be associated with positive outcomes of crisis situations that can be resolved. In situations that cannot be resolved, however, personal growth may be associated with a combination of cognitive redefinition and personal acceptance. More generally, active cognitive-coping strategies may be especially effective because they can be used flexibly in many circumstances; compared to behavioral strategies, they are less under situational control and not as likely to be disrupted by increased life stressors or reduced social support. In any case, it is important to consider the match between situational demands and coping strategies (Roth & Cohen, 1986). FUTURE PROSPECTS We have discussed some of the potential beneficial consequences of life crises and presented a conceptual framework that focuses on how they develop. Positive changes and personal growth occur frequently among people who experience life crises or transitions. We need to learn more about the event-related, environmental, personal, and appraisal and coping factors that facilitate such outcomes. To pursue this aim, we emphasize the need to consider the association between life events and ongoing stressors and resources, to examine the personal and social contexts of coping, and to develop a life course perspective. Life Events and Ongoing Stressors and Resources Most prior studies have focused on how stress and coping factors are related to health and well-being. However, to really understand how positive outcomes may flow from the stress and coping process, we need to consider the interrelationships among the mediating factors involved in this process. One question is whether life stressors and social resources change each other over time (that is, the associations among variables in panel I of the model in Figure 10.1). For example, positive outcomes may be more likely to occur when a person's social resources reduce subsequent stressors. When we focused on this issue in a group of depressed patients, we found that patients who had more close friends at intake to treatment experienced fewer stressors during the ensuing year (Mitchell & Moos, 1984). In contrast, high levels of chronic stressors may act as a vulnerability factor that exacerbates the negative influence of daily stressors on mood. Caspi, Bolger, and Eckenrode (1987) found that daily stressors resulted in poorer mood among women living in poor quality neighborhoods, but had no such effect on women in medium and high quality neighborhoods. A related question is whether positive outcomes occur more frequently among people who are in more benign contexts because they are less likely to experience new negative events (an association between variables in panels I and III of the model). We examined this question in preliminary work on a new Life Stressors and Social Resources Inventory (Moos, 1988). Respondents who reported fewer
168
Personal Coping
ongoing stressors and more ongoing resources tended to experience fewer new stressful events. In turn, people who experienced fewer new stressful events were more likely to report an increase in social resources and a decline in life stressors. We can learn more about the positive influence of life events on adaptation by clarifying the interplay between acute stressful events and a person's ongoing life situation. The Personal and Environmental Context of Coping Future research should also consider the personal and social context of coping, that is, the associations among variables in panels I and II and those in panel IV of the model. We need to understand how personal and social resources influence positive outcomes through their impact on the way in which people cope with stressful circumstances. Consistent with Thoits's (1986) conception of social support as a source of coping assistance, people who have more social resources tend to seek more information and support and engage in more active problem solving, which is related to better adaptation. Thus, we found that high family support was associated with more reliance on active cognitive and behavioral coping and less reliance on avoidance coping. Increases in family support were related to increases in problem-solving coping among women and to a decline in emotional discharge among men (Fondacaro & Moos, 1987). These environmental factors may act together with personal factors to influence the selection of coping strategies. Thus, better educated people are more likely to rely on active problem-solving coping. People who are more self-reliant and higher on internal control also tend to rely more on approach than on avoidance coping in most situations (Fleishman, 1984; Holahan & Moos, 1987). However, the association between personal factors and coping responses may depend on appraisal of the stressor. Compared to pessimists, optimists usually rely more on problem-focused coping and seeking social support. When they construe the stressor as uncontrollable, however, optimists tend to focus more on acceptance and resignation (Scheier, Weintraub, & Carver, 1986). This flexibility may foreshadow positive outcomes. Overall, there is growing evidence that some cognitive- (such as logical analysis and positive comparisons) and behavioral- (such as information seeking and problem solving) coping responses are usually related to good adaptation. The next step is to identify the personal and contextual determinants of reliance on these types of coping responses. The Need for a Life Course Perspective Positive outcomes may develop during the height of a crisis or months and years later. In some instances, change comes quickly, as in the heat of battle when a soldier narrowly survives an enemy attack and suddenly gains a greater appreciation of life. Mostly, however, positive outcomes emerge only during a
Life Crises and Personal Growth
169
long recovery process that follows an initial stage of emotional distress and disorganization. A divorced woman may be depressed in the immediate aftermath of her marital breakup. Yet, in her struggle to improve her financial situation and to establish a new identity she may acquire new competencies, greater independence, and more self-reliance. To understand these gradual processes, we need prospective studies that trace how personal growth evolves over the life span. A few studies point to the value of a longitudinal approach in which people are assessed prior to a life crisis. Norris and Murrell (1987) found that family stress increased prior to the death of a family member and declined afterwards. Block, Block, and Gjerde (1986) noted that the behavior of young boys in intact families was affected as long as eleven years prior to parental separation. With respect to a life course perspective, Elder (1979) noted that economic deprivation had positive developmental consequences for young women during adolescence but not at midlife. Longitudinal studies should examine how positive outcomes early in the adaptation process play a beneficial role in later phases of adaptation. For example, parents of a child with cancer often establish contacts with other parents of ill children. These relationships may help them cope with day-to-day problems such as chemotherapy and surgery as well as manage future crises such as the death of their child. Insights gained about how to seek help and regulate distressing emotions soon after the diagnosis of cancer may prove valuable later as parents struggle to cope with their grief. Parents may also develop a renewed sense of competence as they acquire information that they can use to help other parents in a similar crisis. INTERVENTION PROGRAMS According to crisis theory, a person is especially receptive to outside influence in a time of flux. Such accessibility offers family and friends as well as counselors and psychotherapists a special opportunity to exert a constructive impact. Counselors who are mindful of the issues that arise in normative transitions and life crises can prepare individuals and their families for the experiences they are likely to encounter and help them to expand their coping resources. More generally, intervention programs can teach people to recognize specific problems that are likely to arise in the aftermath of a particular stressor and help them to strengthen their personal competence, social resources, and coping skills. Our framework emphasizes that cognitive appraisal and selection of coping responses are influenced by the person, by the environment, and by aspects of the transition or crisis (Figure 10.1). This framework is useful for identifying foci for intervention programs. For example, programs may focus on helping people in crisis avoid conditions that lead to new stressors, changing the appraisal of a transition or crisis, or providing information to alert people to the tasks they will confront and to potential coping strategies for managing them. By helping
170
Personal Coping
people in crisis master stressors, intervention programs are likely to contribute to positive outcomes. For the most part, interventions have focused on enhancing high-risk individuals' social resources and coping skills. For example, Bloom and his associates (1985) developed an intervention program for newly separated adults. The program provided support and augmented personal competence via study groups on socialization, parenting, employment and career planning, legal and financial matters, and housing. Program participants adjusted better psychologically and reported fewer problems than members of a control group. Other interventions might focus on enhancing the personal and social resources that are closely tied to positive outcomes: self-complexity, mastery and self-reliance, sense of coherence, and such cognitive-coping strategies as positive comparisons and selective attention. Both professionally organized and self-help programs can be enhanced by information about the personal and social factors that promote good outcomes of life crises. It may be possible to identify situations in which people are especially likely to use coping skills that foreshadow positive effects, or to specify phases of the adaptation process when people are most open to change and personal growth. Counselors then could implement interventions to enhance natural recovery processes and facilitate personal growth that otherwise might not occur. In turn, people who have received help in mastering a crisis may choose to provide help to others who are experiencing similar crises. As Riessman (1965) pointed out, this process may create a "positive upward spiral" in which both the helper and the recipient gain self-confidence and personal maturity. CONCLUSION Although we have focused on the positive effects of life crises, our intent is not to minimize the problematic consequences of traumatic events or to exaggerate their benefits. Rather, we aim to provide some balance to the current emphasis on the negative aspects of life crises. We support Antonovsky's (1987) call for a salutogenic orientation that enables researchers to consider factors that promote health as well as those that are related to illness. Life crises are an inherent part of the human condition and often exact a high toll. People commonly experience a painful initial period of turbulent emotions before they resolve a crisis and adapt to it. Yet, life transitions and crises can produce developmental shifts; they are instrumental in stimulating psychological change and maturity. By studying individuals for whom life's vicissitudes bring about positive changes and personal growth, we can gain valuable insights into how we develop and become mature and caring people.
11
Outcome Expectancies and Psychosomatic Consequences Holger Ursin and Karsten Hytte
It seems to be a commonplace idea that scientific usage of terms must go beyond common sense terminology. Even so, this still puzzles writers on coping and stress. There is an abundance of authors who complain about the many definitions of coping and stress, but relatively few who offer strict definitions. In short, no agreement exists on the proper definition. While most or perhaps all authors in this volume agree that coping is important for almost everything, there is less agreement as to what coping really means. To the authors of this chapter it seems there are two principal uses which should be identified and distinguished. In the first case the term "coping" is used for the strategies used in a particular situation. The second case covers acquired expectancies about to what extent these strategies really lead to success. The first definition covers how the individual solves the problems, the second whethe the individual has solved them or not. Particular coping styles may correlate with particular ways of responding physiologically, but not with the physiological state. One and the same behavior may be executed under physiological states of high activation ("affective") or low activation ("instrumental") (Ursin 1985). Our own interest area is the relationship between external challenges and physiological and psychological health. It is, therefore, necessary for us to use a terminology and to study phenomena that relate to the physiological state of the individual. We, therefore, are in absolute disagreement with Lazarus and Folkman (1984), who state that it is unnecessary and even circular to tie definitions of coping and stress to physiological changes. We feel the absence or presence of physiological changes is a crucial validation of these concepts and is the reason for studying the phenomena in the first place. We even believe that such physiological changes may in turn affect behavior through endocrine feedback loops to the brain (de Wied, 1974), and through biochemical changes in
172
Personal Coping
the transmitter systems of the brain (Coover, Murison, Sundberg, Jellestad & Ursin, 1984). Behavior affects physiology, and physiology may also affect behavior. We will briefly outline the definitions we find necessary to handle the data available on the relationships between external loads on the organism ("stressors"), the way the individual handles these challenges, and the pathophysiological consequences that may arise. A formal set of definitions will be offered, which we believe is necessary to arrive at a consistent theoretical framework. We intend to use definitions of the terms that are as close to common sense and common usage as possible. We intend to arrive at a theoretical position which is as noncontroversial as possible, and we attempt to make it clear whenever our definitions clearly differ from other definitions in common use. DEFINITIONS OF COPING, HELPLESSNESS AND HOPELESSNESS Expectancy concepts are used frequently in cognitive psychology and may also be found in biological literature. However, in learning theory and in biological psychology such "soft" terms were either frowned upon, or sometimes used quite loosely. In recent years more systematic attempts to broaden learning theory (see Dickinson, 1980) have proven useful to areas of biological psychology and clinical psychology, even if controversies do exist. We suggest that the terms coping, helplessness, and hopelessness should be defined in terms of expectancy—in particular response outcome expectancies. The underlying theoretical assumption is simply that brains are able to store information and that the stored information has the nature of "expectancies." It seems to us that it is an economical and parsimonious position that brains store relationships either between stimuli, or between responses and stimuli. The first phenomenon has been referred to as stimulus expectancy, the second as response outcome expectancy (Bolles, 1972). In traditional learning theory the corresponding terms are classical conditioning and instrumental conditioning. However, we believe that we need more degrees of freedom than are available within traditional learning theory. On the other hand, we attempt to achieve the same rigor in terms, but without the assumptions of associative learning theory, and, admittedly, with considerable less empirical backup for the theoretical positions at which we arrive. We believe that the theoretical structure at which we arrive has considerably more relevance for clinical applications than the strict versions and formulations of learning theory, simply because it seems easier to use and does not require any deep familiarity with learning theory. Traditional learning theory used to deal with very simple stimulus situations. Since the early 1970s, the complexity of any conditioning situation is taken care of by the development of a theory for compound conditioning with the association strength shared by all salient cues in a conditioning paradigm (Rescorla & Wagner, 1972). This has made learning theory more useful. Still more complex functions
Outcome Expectancies
173
are required to account for even very simple behaviors in very simple neuronal circuits. When a frog plans to hit a flying insect with his tongue, it is necessary to use terms like "attention," "preparedness," and "expectancy." To perform a complex act—catching a fly—the frog must direct its movements to where the prey is expected to be in the next time interval. To account for biological adaptations to the environment in species with even more brains than a frog, it seems useful to at least utilize the same software terminology. Finally, to account for functions of brain structures like the hippocampus, even more complex terms have been introduced, like learning of spatial maps (O'Keefe & Nadel, 1978). When the brain has established that one thing precedes another, the brain may be said to expect the second stimulus once the first stimulus has been presented. This is stimulus expectancy (BoUes, 1972), and is a reformulation of what happens during classical conditioning. Brains may, of course, also register what happens after a certain response has been emitted. This is the essence of instrumental conditioning, which therefore may be referred to as response outcome expectancy (Bolles, 1972). These phenomena have also been referred to as stimulus-stimulus learning versus response-stimulus learning. In Pavlovian and Konorskian literature the corresponding terms are Type 1 and Type 2 conditioning (Konorski, 1967). Expectancies may be quantified along several dimensions. Bolles (1972) operated with two dimensions, the strength of the expectancy and the perceived probability of the outcome. He also discussed the value of that outcome. In this chapter, we will use three dimensions: the strength of the acquisition of the expectancy, the perceived probability, and the affective value of the expected outcome. This is in the line with human motivation theory in which there is frequent reference to the incentive value of the expected outcome (Irwin, 1971). Further discussion of these dimensions, and their use in formal definitions of coping, helplessness, hopelessness, and defense have been worked out in greater detail elsewhere (Ursin, 1988). Briefly, the three dimensions are assumed to be continuous variables, and independent. Perceived probability is assumed to have values from 0 to 1, just as for true probabilities. The strength of the acquisition is accounted for by the "habit value," which is also assumed to have values from 0 (no acquisition) to 1 (maximum acquisition). Finally, the affective value of any stimulus is assumed to have values from — 1 (maximal unpleasantness) oes from — 1 (maximal unpleasantness) to + 1 (maximum pleasantness). Zero would then indicate a neutral stimulus, the type preferred as the conditioning stimulus in a conditioning paradigm.
STIMULUS EXPECTANCIES: DEFENSE Before we deal with the response expectancies, we have to deal with the signals that indicate that the states of coping, helplessness, or hopelessness may be initiated. In animal studies of fear and avoidance there are two stages, a classical conditioning phase, and an instrumental phase (Gray, 1975). In the first classical conditioning phase, and an instrumental phase (Gray, 1975). In the first phase, the subjects are taught that they are in some danger, by pairing a neutral
174
Personal Coping
stimulus with a negative stimulus (i.e., through classical conditioning) or, in our framework, through establishment of a stimulus expectancy. In studies of stress in humans, this stage is often bypassed by studying phenomena that are assumed to be threatening and stress-inducing to all humans, but some of the inconsistency and variance in results is simply due to the lack of control of the first conditioning phase. With their impressive cognitive abilities, humans possess another varianceproducing capability manifest in the way they face threats. They may distort and even deny signals giving information about dangers. This ability may be very efficient in producing safety and reduce the physiological responses to threat, but the price may be rather inadequate behavior. The concept is an important part of psychodynamic thinking. It may also be used in a more restricted or conservative sense, relating to the perceptual distortions of reality or true stimulus contingencies, as is the cas