Research in the Social Scientific Study of Religion: Volume 15 (2004)

Research in the Social Scientiﬁc Study of Religion Research in the Social Scientiﬁc Study of Religion Series Editors R...

Author: Ralph L. Piedmont | David O. Moberg

17 downloads 1322 Views 1MB 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

DOWNLOAD PDF

Research in the Social Scientiﬁc Study of Religion

Research in the Social Scientiﬁc Study of Religion Series Editors RALPH L. PIEDMONT DAVID O. MOBERG

VOLUME 15

Research in the Social Scientiﬁc Study of Religion Volume 15

Edited by

Ralph L. Piedmont and David O. Moberg

BRILL LEIDEN • BOSTON 2005

This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data Library of Congress Cataloging-in-Publication Data is also available on the Library of Congress website: catalog.loc.gov LC control number 89650738

ISSN 1046–8064 ISBN 90 04 14146 4 © Copyright 2005 by Koninklijke Brill NV, Leiden, The Netherlands All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. RSSSR is indexed in Sociological Abstracts, Social Services Abstracts, Religion Index Two: Multi-Author Works, Religions and Theology: Religions and Theology, Social Sciences: Comprehensive Works Authorization to photocopy items for internal or personal use is granted by Brill provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. printed in the netherlands

CONTENTS

Preface ........................................................................................ Acknowledgements ...................................................................... Manuscript Invitation ................................................................

vii ix xi

Assessing the personality of clergy: Abbreviated Eysenck Personality Proﬁler (EPP-A) .................................................. Leslie J. Francis, Chris Jackson, and Susan H. Jones

1

Killing as trauma: The religious implications of perpetration-induced traumatic stress .................................... Rachel M. MacNair

17

The impact of changing marital status on religious attendance in Australia .......................................................... John M. Armstrong

41

Religious switching: Does parents’ education matter? ............ Amy Adamczyk Burnout among male Anglican parochial clergy in England: Testing a modiﬁed form of the Maslach Burnout Inventory ................................................................................ Christopher F.J. Rutledge and Leslie Francis Religious coping moderates the relationship between early maladaptive schema origins and dysphoria .......................... Christian T. Racine and John J. Cecero

51

71

95

Gender diﬀerences in stress among Protestant clergy: An exploratory study .............................................................. 117 Stephen J. Fichter Religious choice and the meaning of Church in the lives of evangelical seekers .............................................................. 137 Richard W. Christopherson

vi

contents

Spirituality and religious variables as predictors of well-being in sex oﬀenders ...................................................................... 167 Brendan Geary, Joseph W. Ciarrocchi, and N.J. Scheers Mainline evangelical renewal movements: A preliminary inquiry ...................................................................................... 189 Sean F. Everton Happiness and the varieties of religious experience: Religious support, practices, and spirituality as predictors of well-being ............................................................................ 209 Joseph W. Ciarrocchi and Erin Deneke Temptation bias: Seeing oneself as better able than others to resist temptation ................................................................ 235 Sheila Garos, James K. Beggan, and Annette Kluck The Logoplex as a paradigm for understanding spiritual transcendence .......................................................................... 263 Ralph L. Piedmont Authors’ Biographies .................................................................. 285 Manuscript Reviewers ................................................................ 291 Index of Names .......................................................................... 293 Subject Index .............................................................................. 303

PREFACE

Religion and spirituality cut across all of the activities and underlying ideologies of every domain of human life, whether people recognize their pervasive presence and inﬂuence or not. No single profession, academic specialty, or research methodology can completely grasp the richness of their obviously visible occurrences and, much less, their ineﬀably invisible co-existence. They cannot be wholly comprehended by either empirical sensory observations or rational mental exercises, both of which usually are further limited by occurring inside the narrow conﬁnes of only one society with its religious traditions and cultural adaptations of beliefs and practices. Therefore, despite the richness of the store of knowledge already accumulated, our scientiﬁc investigations of spirituality and religion still are merely scratching the surface of the intricate entanglements of religion and spirituality that are infused into everything that concerns people collectively and as individuals. This edition of Research in the Social Scientiﬁc Study of Religion continues its rich heritage of oﬀering original reports of research, theoretical studies, and innovative social scientiﬁc analyses of religion and spirituality. Its articles are by both extensively published and beginning authors who reside on three continents. Besides sociology and several branches of psychology, their professional appointments and educational backgrounds include mental health, spiritual direction, marital and sexual therapy, practical theology and the ministry, religious education, pastoral counseling, peace studies, social work, statistics, and other areas of specialization. The articles report studies of the role of religion and spirituality in relationship to many topics of current popular interest, among them happiness, burnout, coping with problems, temptation, the rehabilitation of sex oﬀenders, evangelicalism, the inﬂuence of marital status on religious attendance, and experiences in, sources of, and social and mental turmoil associated with the changing of a person’s religious aﬃliation. On a more technical level are analyses of the posttraumatic stress disorder that is related to killing, simpliﬁcation of the methods for personality proﬁling of the clergy, similarities and diﬀerences in the stress of male and female clergy, and development

viii

preface

of the Logoplex as a method and model for interpreting spiritual transcendence. The references to related studies that are found in every article help to facilitate the use of RSSSR as a basis for additional research. They also are a beneﬁcial tool for teaching, exploring alternative clinical applications, and other professional work. The Subject Index will lead readers to a large number of topics that are not immediately apparent from a casual survey of only article titles and abstracts, and the Index of Names can help readers locate reports of the work done by numerous researchers and scholars who study religion.

ACKNOWLEDGEMENTS

We are grateful for the contributions and cooperation of a large number of people without whom publication of RSSSR would not be possible. Most obvious among them are the 21 authors and coauthors of the published articles. A brief biographical sketch of each is provided on pp. 285–289. There were 14 researchers, scholars, and clinicians who have served as anonymous reviewers (see p. 290) of the manuscripts that were received for current publication. They not only have functioned as professional referees evaluating the appropriateness of the respective manuscripts for publication, but they also have given the authors signiﬁcant suggestions to improve the quality and scope of their future research in the social scientiﬁc study of religion and spirituality. Their eﬀorts help to insure a high quality among those reports that are published. Ruth Dennison-Tedesco has valiantly and eﬃciently served as the editorial assistant for this volume. Her painstaking work has contributed immeasurably to high standards of production. Our academic institutions have provided many critical necessities for the production of this volume. Loyola College in Maryland, especially its graduate Department of Pastoral Counseling, has provided oﬃce space, funds for the editorial assistant, telephone services, computer technology, postage, access to its admirable support infrastructure, and related services. Marquette University’s Department of Social and Cultural Sciences similarly has contributed postal expenses and other support. We are grateful, too, to Joed Elich and the production staﬀ of Brill Academic Press who have eﬃciently published this attractive and useful volume. Brill’s marketing department (www.brill.nl or E-mail [email protected]) is eager to ﬁll orders for either single volumes or on-going subscriptions to RSSSR. Please recommend RSSSR to your professional and academic colleagues. Also support its addition or continuation in your academic, religious, research, and public libraries, for its rich contents are relevant to everyone, both lay and professional, who is interested in keeping up with the rapidly expanding frontiers of scientiﬁc knowledge about spirituality and religion. David O. Moberg, Ph.D., and Ralph L. Piedmont, Ph.D., Co-editors

MANUSCRIPT INVITATION

For future volumes we welcome the submission of manuscripts that report on research contributing to the behavioral and social science understanding of religion, whether done by members of those disciplines or other professions. RSSSR is an annual interdisciplinary and international volume that publishes original reports of research, theoretical studies, and other innovative social scientiﬁc analyses of religion. (However, we do not include studies that are purely historical or theological.) Manuscripts should be original contributions (not reprints) based upon any of the quantitative or qualitative methods of research or the theoretical, conceptual, or meta-analytical analysis of research on religion in general or on any speciﬁc world religion. They should not be under consideration for publication by any other journal or publication outlet and should comply fully with the professional ethical standards of psychology, sociology, and other social science professions. Manuscripts may be submitted at any time during the year, although those received within the calendar year have the best chance of inclusion in the next volume. Send four copies, double-spaced on standard size paper, to: Ralph L. Piedmont, RSSSR Co-editor Dept. of Pastoral Counseling Loyola College in Maryland 8890 McGaw Road, Suite 380 Columbia, MD 21045, USA Manuscripts that are judged by the editors as relevant to the coverage of RSSSR are reviewed anonymously for quality and then either accepted (usually along with constructive suggestions for revision) or rejected. Those accepted for publication must conform to the style guidelines of the Publication Manual of the American Psychological Association, 5th edition. (Authors of accepted manuscripts who lack access to it may contact Dr. Piedmont for sample materials to help in the ﬁnal preparation of their papers.) Manuscripts relevant to our subject that are not accepted for publication also receive the beneﬁt

xii

manuscript invitation

of critiques and suggestions that can aid their improvement for submission elsewhere. Inquiries about the suitability of potential contributions or about opportunities to review manuscripts may be sent to Ralph Piedmont at [email protected].

ASSESSING THE PERSONALITY OF CLERGY: ABBREVIATED EYSENCK PERSONALITY PROFILER (EPP-A) Leslie J. Francis, Chris Jackson, and Susan H. Jones *

Abstract Several recent studies have demonstrated the usefulness of the Eysenck Personality Proﬁler in assessing the personality characteristics of clergy. However, the instrument, which contains 440 items, takes a long time to complete. The present study analyzed responses provided by 1,671 Anglican clergy and by 400 ﬁrst-year undergraduate students in order to develop an abbreviated form of the Eysenck Personality Proﬁler (EPP-A) in which each of the 21 primary scales is assessed by a 6-item scale. The data supported the internal reliability and concurrent validity of these 6-item measures and of the 42-item scales of extraversion, neuroticism, and psychoticism to which the 6-item measures contribute. Conﬁrmatory factor analysis supported the view that the 6-item scales of the EPP-A were similar to the full version.

A series of studies has begun to demonstrate the usefulness of Eysenck’s dimensional model of personality (Eysenck & Eysenck, 1985) for understanding individual diﬀerences among the clergy, concentrating on such factors as: conﬂict and dissatisfaction with ministry (Francis & Rodger, 1994a; Francis & Robbins, 1999), choice of stipendiary or nonstipendiary ministry (Francis & Robbins, 1996), AngloCatholic or Evangelical preferences (Francis & Thomas, 1996a), mystical orientation (Francis & Thomas, 1996b; Francis & Louden, 2000), charismatic predisposition (Francis & Thomas, 1997; Robbins, Hair, & Francis, 1999; Kay, 2000; Louden & Francis, 2001), preference for rural ministry (Francis & Lankshear, 1998; Francis & Littler, 2001), perceptions of the selection process (Robbins & Francis, 1999), role prioritization (Robbins & Francis, 2000), stress and burnout (Francis & Rutledge, 2000), baptism policy (Francis, in press), and the clerical persona (Francis, Louden, Robbins, & Rutledge, 2000; Francis, 2003). Another series of studies has employed Eysenck’s * Author Note: All versions of the Eysenck Personality Proﬁler (EPP) are available from Chris Jackson: e-mail, [email protected] Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

2

francis, jackson, and jones

model of personality to explore the ways in which clergy stand out from the population as a whole, including studies of Anglican clergy (Francis & Pearson, 1991; Francis, 1991; Francis & Thomas, 1992; Francis & Rodger, 1994b; Robbins, Francis, & Rutledge, 1997; Robbins, Francis, & Fletcher-Marsh, 2000), Methodist ministers ( Jones & Francis, 1992; Robbins, Francis, Haley, & Kay, 2001), Pentecostal pastors (Francis & Kay, 1995), Catholic priests (Louden & Francis, 1999; Francis & Louden, 2001), and Evangelical clergy (Francis, 2002). These studies have employed one of three instruments: the Eysenck Personality Questionnaire (EPQ ), developed by Eysenck and Eysenck (1975), the Revised Eysenck Personality Questionnaire (EPQR) developed by Eysenck, Eysenck, and Barrett (1985), or the short-form Revised Eysenck Personality Questionnaire (EPQR-S) also developed by Eysenck, Eysenck, and Barrett (1985). These instruments, which comprise 90, 100, and 48 items respectively, measure the three higher order orthogonal dimensions of personality, characterized as extraversion, neuroticism, and psychoticism. According to the most recent edition of the test manual, Eysenck and Eysenck (1991) deﬁne these three higher order dimensions of personality in the following terms, concentrating in each case on those who record high scores. Individuals who score high on the extraversion scale are sociable, like parties, have many friends, need to have people to talk to, and do not like reading or studying by themselves. They crave excitement, take chances, often stick their necks out, act on the spur of the moment, and are generally impulsive. They are fond of practical jokes, always have a ready answer, and generally like change; they are carefree, easy going, optimistic, and like to “laugh and be merry.” They prefer to keep moving and doing things, tend to be aggressive and lose their temper quickly. Altogether their feelings are not kept under tight control, and they are not always reliable. Individuals who score high on the neuroticism scale are anxious, moody, frequently depressed, and worry a lot. They are likely to sleep badly and suﬀer from various psychosomatic disorders. They are overly emotional, excessively reactive to all sorts of stimuli, and ﬁnd it diﬃcult to get back on an even keel after each emotionally arousing experience. Their strong emotional reactions interfere with their proper adjustment, making them react in irrational, sometimes rigid ways. Their main characteristic is constant preoccupation with possible negative outcomes and consequent anxiety.

the personality of clergy

3

Individuals who score high on the psychoticism scale are often solitary, unempathic, and troublesome, not ﬁtting in anywhere. They may be cruel, inhumane, and insensitive. They are hostile to others, even their own kith and kin, and aggressive, even to loved ones. They have a liking for odd and unusual things, and a disregard for danger; they like to make fools of other people and to upset them. Socialization is a concept which is relatively alien to them; empathy, feelings of guilt, sensitivity to other people are notions which are strange and unfamiliar to them. A diﬀerent strand of Eysenck’s work has focussed on identifying and assessing independently the primary scales which comprise the higher order scales that assess the three major dimensions of personality. In an early study, Eysenck and Wilson (1975) argued that each of the three dimensions of personality could be disaggregated into seven primary scales each, and they proposed 30-item measures for the 21 resultant constructs. More recently, Eysenck, Barrett, Wilson, and Jackson (1992) developed the work of Eysenck and Wilson (1975) further to produce the Eysenck Personality Proﬁler (EPP). The Eysenck Personality Proﬁler proposed 20-item measures for each of the 21 primary scales, together with a 20-item lie scale, making an instrument of 440 items. According to the Eysenck Personality Proﬁler, the seven primary scales comprising extroversion were deﬁned as activity, sociability, expressiveness, assertiveness, achievement-orientation, dogmatism, and aggressiveness. The seven primary scales comprising neuroticism were deﬁned as low self-esteem, unhappiness, anxiety, dependency, hypochondriasis, guilt, and obsessiveness. The seven primary scales comprising psychoticism were deﬁned as risk-taking, impulsiveness, irresponsibility, manipulation, sensation-seeking, masculinity, and unreﬂectiveness. Six studies have now reported on the use of the Eysenck Personality Proﬁler among clergy. In the ﬁrst two studies Francis, Robbins, Jackson, and Jones (2000) and Francis, Jones, Jackson, and Robbins (2001) examined the personality proﬁle of male clergy against the population norms for men and for women in general. Their data demonstrated that male clergy recorded a characteristically feminine proﬁle in terms of 16 of the 21 primary scales. In the third study, Francis, Jones, Robbins, and Jackson (2003) examined the personality proﬁle of female clergy against the population norms for men and for women in general. Their data demonstrated that the female

4

francis, jackson, and jones

clergy tended to be less extraverted than women in general, less neurotic than women in general, and less tough-minded than women in general. These ﬁndings help to clarify the way in which women clergy tend to project a characteristically masculine personality proﬁle in respect to one major dimension of personality (neuroticism), but a characteristically feminine personality proﬁle in respect to the other two major dimensions of personality (psychoticism and extraversion). In the fourth study, Jones and Francis (2000) demonstrated the theoretical and empirical potential in concentrating in depth on just one of the 21 primary scale measures, by focussing on the 20-item index of guilt. Their data demonstrated that male clergy recorded higher scores on the index of guilt in comparison to men in general, while female clergy recorded levels of guilt comparable to women in general. In a ﬁfth study, Jones, Francis, and Jackson (2003) focussed on the primary trait of anxiety assessed by the 21-item index. Their data demonstrated that male clergy recorded higher scores on the index of anxiety than did men in general. Female clergy recorded lower scores on the index of anxiety than did women in general. In a sixth study, Jones and Francis (2003) focused on the primary trait of self-esteem assessed by the 21-item index. Their data demonstrated that male clergy recorded higher scores on the index of low selfesteem than men in general. Female clergy also recorded higher scores on the index of low self-esteem than women in general. The low self-esteem recorded by both male and female Anglican clergy is discussed in the light of the marginalization of the clerical profession within contemporary society. Two important conclusions can be drawn from these six studies. The ﬁrst conclusion is that there is clearly potential for further studies employing the Eysenck Personality Proﬁler among clergy in order to understand more about individual diﬀerences in clergy aptitude, performance and susceptibility to stress. The second conclusion is that a 440-item personality inventory not only presents a formidable task to the potential respondents, but also seriously weakens the opportunity to employ a personality measure alongside other crucial indices, like measures of job satisfaction, role performance, stress, and burnout. The aim of the present study, therefore, is to propose an abbreviated form of the Eysenck Personality Proﬁler with satisfactory psychometric properties for use among clergy. There is a well-established history for the development of short forms of other editions of Eysenck’s family of personality tests, although

the personality of clergy

5

there is some variation in what counts as a short form. For example, Eysenck, Eysenck, and Barrett (1985) proposed a short form of the Revised Eysenck Personality Questionnaire comprised of 12-item scales. Floderus (1974) proposed a short-form of the Eysenck Personality Inventory comprised of 9-item scales. The more general strategy to address this problem, however, focuses on 6-item scales. For example, Eysenck (1958) developed 6-item scales based on the Maudsley Personality Inventory. Eysenck and Eysenck (1964) developed 6-item scales based on the Eysenck Personality Inventory. Francis, Brown, and Philipchalk (1992) developed 6-item scales based on the Revised Eysenck Personality Questionnaire. Francis and Pearson (1988) developed 6-item scales based on the Junior Eysenck Personality Questionnaire; Francis (1996) developed 6-item scales based on the Revised Junior Eysenck Personality Questionnaire. It seems appropriate, therefore, for the present study to follow this precedent of 6-item scales, which would lead to an instrument of 132 items including the 21 primary scales and the lie scale. There have been two earlier attempts to produce a short form of the Eysenck Personality Proﬁler. First, Eysenck, Wilson, and Jackson (1996) employed a diﬀerent rationale from the one generally employed in the development of short forms. Instead of reducing the number of items within each of the 21 scales, they reduced the number of primary scales within each of the three dimensions of personality, with the intention of deriving a better three-factor model. Second, Francis and Jackson (in press) recommended a 12-item short form of the Eysenck Personality Proﬁler for use among students. The resultant instrument remains overly long at 264 items. The wider literature has been critical of short forms for a number of good reasons (Levy, 1968; Smith & McCarthy, 1995; Smith, McCarthy, & Anderson, 2000). The point is well taken that short forms may be inappropriate in clinical assessment procedures when the classiﬁcation and treatment of individuals is at stake. With respect to survey style studies, however, the case against short forms is much less substantial. What is required of short forms used in this way is clear evidence of their concurrent validity alongside the parent full form of the scale, and good reliability (both independently and in comparison to the full form). In this sense, short forms need to function as good predictors of scores recorded on the full form. Recent psychometric defenses of short forms have been provided by Scholte and De Bruyn (2001) and by Archer, Tirrell, and Elkins (2001).

6

francis, jackson, and jones

Although the primary aim of this study is to propose an abbreviated form of the Eysenck Personality Proﬁler for use among clergy, it would be a mistake not to ensure that the instrument functioned with satisfactory psychometric properties in other, more general populations. Consequently, the present analysis includes a second sample of students within which the stability of the new abbreviated scales can be checked. Method Sample and Procedures Sample one The full 440-item Eysenck Personality Proﬁler (Eysenck, Barrett, Wilson, & Jackson, 1992) was mailed in December 1992, 1993, 1994, 1995, and 1996 to all men and women ordained as deacons during each respective year within the Anglican Church in England, Ireland, Scotland and Wales. The response rate for the various years varied between 62% and 72%, generating thoroughly completed questionnaires from 1,148 clergymen and 523 clergywomen. Of the male respondents, 18% were in their twenties, 38% in their thirties, 26% in their forties, 13% in their ﬁfties, and 4% were sixty years of age or older. Of the female respondents, 8% were in their twenties, 20% in their thirties, 42% in their forties, 25% in their ﬁfties, and 4% were sixty years of age or older. Sample two The full 440-item Eysenck Personality Proﬁler (Eysenck, Barrett, Wilson, & Jackson, 1992) was administered to all the incoming students as part of the induction program within a college of higher education in Wales. Almost all the students agreed to participate, generating thoroughly completed questionnaires from 290 females and 110 males. Nearly two-thirds (65%) of the respondents were under the age of twenty, 27% were in their twenties, and the remaining 8% were thirty years of age or older. Measures In the present study, unrotated principal component analysis and item-total correlational analysis were employed to identify the best

the personality of clergy

7

sets of 6-items to provide a short form for each of the 21 primary scales. Using these methods, items which recorded low loadings on the ﬁrst factor of principal component analysis or which had low correlations with the sum total of the other items in the scale were progressively dropped. The short-form scales were developed on the clergy sample and then tested on the student sample. Conﬁrmatory factor analysis was then employed to compare the goodness of ﬁt of the full form with that of the abbreviated form among both samples. Results Table 1 assesses the internal reliability of the 20-item scales and the 6-item scales in terms of the alpha coeﬃcient (Cronbach, 1951), and assesses the concurrent validity of the short form in terms of the correlation with the full form. Taking Kline’s (1993) suggestion that the alpha coeﬃcient should reach .70 for a normal length scale, the following full-length scales failed to reach that baseline: from extraversion, expressiveness, dogmatism, and aggressiveness; from neuroticism, hypochondriasis; and from psychoticism, manipulation, and masculinity. It is unlikely that reliable short forms can be generated from unreliable full forms. While a minimum alpha of .70 might be expected for a full form, a level of .60 is a more realistic target for a 6-item short form. For extraversion, the activity, sociability, assertiveness, achievement-orientation, and aggressiveness scales have alphas above .60, while expressiveness and dogmatism do not. Within neuroticism, only hypochondriasis has an alpha below .60. Within psychoticism, risk-taking, impulsiveness, irresponsibility, sensation-seeking and unreﬂectiveness, have acceptable reliability while manipulation and masculinity do not. Table 2 presents the same information as Table 1, but in respect of the student sample. A very similar picture emerges from the two samples. According to both studies, satisfactory short-form scales have been produced to assess activity, sociability, assertiveness, achievement-orientation, aggressiveness, low self-esteem, unhappiness, anxiety, dependency, guilt, obsessiveness, risk-taking, impulsiveness, irresponsibility, sensation-seeking and unreﬂectiveness. According to both studies, less satisfactory short-form scales have been produced to assess expressiveness, dogmatism, hypochondriasis and manipulation. The one diﬀerence between the two samples concerns the

active sociable expressive assertive ambitious dogmatic aggressive

inferiority unhappy anxious dependence hypochondriasis guilt obsessive

risk-taking impulsive irresponsible manipulation sensation-seeking toughminded practical

Extraversion activity sociability expressiveness assertiveness achievement-orientation dogmatism aggressiveness

Neuroticism low self-esteem unhappiness anxiety dependency hypochondriasis guilt obsessiveness

Psychoticism risk-taking impulsiveness irresponsibility manipulation sensation-seeking masculinity unreﬂectiveness

Lie scale

high

scale

scale descriptors

careful controlled responsible empathy unadventurous tenderminded reﬂective

self-esteem happy calm autonomy sense of health guilt freedom casual

inactive unsociable inhibited submissive unambitious ﬂexible peaceful

low

Table 1. Scale properties among clergy

0.67 0.73 0.71 0.50 0.77 0.56 0.67 0.67

0.74

0.78 0.79 0.79 0.73 0.56 0.75 0.66

0.71 0.76 0.48 0.70 0.71 0.55 0.60

6-item alpha

0.71 0.77 0.71 0.60 0.79 0.61 0.74

0.85 0.86 0.83 0.70 0.64 0.76 0.70

0.78 0.82 0.53 0.75 0.74 0.59 0.62

20-item alpha

0.84

0.80 0.78 0.82 0.64 0.87 0.79 0.79

0.90 0.89 0.87 0.81 0.75 0.87 0.83

0.84 0.84 0.64 0.82 0.84 0.63 0.74

r

8 francis, jackson, and jones

self-esteem happy calm autonomy sense of health guilt freedom casual careful controlled responsible empathy unadventurous tenderminded reﬂective

inferiority unhappy anxious dependence hypochondriasis guilt obsessive risk-taking impulsive irresponsible manipulation sensation-seeking toughminded practical

low inactive unsociable inhibited submissive unambitious ﬂexible peaceful

scale descriptors

active sociable expressive assertive ambitious dogmatic aggressive

high

Note r: = correlation between 20-item scale and 6-item scale.

Extraversion activity sociability expressiveness assertiveness achievement-orientation dogmatism aggressiveness Neuroticism low self-esteem unhappiness anxiety dependency hypochondriasis guilt obsessiveness Psychoticism risk-taking impulsiveness irresponsibility manipulation sensation-seeking masculinity unreﬂectiveness Lie scale

scale

Table 2. Scale properties among students

0.74 0.75 0.70 0.64 0.76 0.71 0.75 0.75

0.89 0.90 0.85 0.74 0.73 0.80 0.74

0.77 0.82 0.53 0.78 0.79 0.57 0.76

20-item alpha

0.69 0.71 0.69 0.46 0.74 0.64 0.64 0.65

0.78 0.82 0.80 0.74 0.58 0.74 0.65

0.71 0.75 0.53 0.70 0.75 0.35 0.69

6-item alpha

0.78 0.81 0.80 0.62 0.84 0.81 0.74 0.83

0.92 0.92 0.89 0.83 0.80 0.85 0.85

0.83 0.84 0.66 0.84 0.87 0.58 0.83

r

the personality of clergy 9

francis, jackson, and jones

10

6-item scale of masculinity which produced an alpha coeﬃcient below the .60 threshold among the clergy and above this threshold among the students. So far the analysis has concentrated on generating and testing the twenty-one 6-item primary scales. Eysenck’s theory predicts that these 21 primary scales will combine into the three higher order dimensions of extraversion, neuroticism, and psychoticism (see Eysenck & Wilson, 1975). Table 3, therefore, presents the alpha coeﬃcients for the three 42-item indices of the major dimensions of personality. In both samples, all three higher order scales demonstrate highly satisfactory levels of internal reliability. Table 3. Internal reliability of the higher order dimensions among clergy and students dimension

clergy alpha

student alpha

42-item scale of extraversion 42-item scale of neuroticism 42-item scale of psychoticism

0.80 0.90 0.79

0.79 0.92 0.82

Note r: = correlation between 20-item scale and 6-item scale.

We then used conﬁrmatory factor analysis, by means of structural equation modelling, to determine the goodness of ﬁt of the abbreviated and full-forms of the EPP (see table 4). We compared the three-factor model proposed by Eysenck, Barrett, Wilson, and Jackson (1992) in which seven primary scales composed of extraversion, neuroticism, and psychoticism, and the three factor EPP-S model proposed by Eysenck, Wilson, and Jackson (1996) in which just three primary scales comprise each of the higher order factors. We used ﬁve diﬀerent goodness of ﬁt measures, which each have advantages and disadvantages. The chi-squared goodness of ﬁt is widely utilized, but one assumption is that the variables are normally distributed and the EPP scales do not satisfy this assumption. Moreover, the chisquared test is sensitive to sample size with large samples tending to lead to rejection of otherwise satisfactory models. All models show a relatively poor goodness of ﬁt using the chi-squared test. The Goodness of Fit Index (GFI) measures how much better the model ﬁts the data compared with no model at all. The EPP-S model (both 6- and 20-item versions with both data sets) ﬁts very well (GFI

the personality of clergy

11

approx .90), whereas the seven primary scale models ﬁt more poorly poorer (all between .55 and .81). The Adjusted Goodness of Fit Index is similar to the GFI, but takes into account the degrees of freedom and thereby rewards simpler models. Again, the EPP-S model is satisfactory for both 6- and 20-item versions (AGFI approx .81), whereas the full seven-scale version is much less satisfactory in all cases. The Root Mean Square Error of Approximation (RMSEA) is a diﬀerent kind of measure that concentrates on how precisely the model ﬁts the data. All the models show a relatively poor ﬁt and to a similar degree. In general, we can draw these two conclusions: ﬁrst, that the 6-item version of the EPP has very similar characteristics to the 20-item version of the EPP from the perspective of conﬁrmatory factor analysis; and second, that the ﬁt of the most simple structural equation model to the EPP is generally poor and in need of improvement. Table 4. Conﬁrmatory factor analysis by structural equation modelling

x2 df p RMSEA p GFI AGFI

clergy EPP EPP-A EPP-S EPP-SA

students EPP EPP-A EPP-S EPP-SA

5748 186 0.00 .13 0.00 .78 .72

1588 186 0.00 .17 0.00 .55 .52

4103 186 0.00 .14 0.00 .81 .76

791 24 0.00 .14 0.00 .90 .82

773 24 0.00 .14 0.00 .91 .83

1040 186 0.00 .13 0.00 .73 .67

153.5 24 0.00 .14 0.00 .89 .79

125 24 0.00 .13 0.00 .91 .82

Note: EPP presents the 7 primary scale solution for each higher order factor comprising 140 items (7 × 20 items) EPP-A presents the 7 primary scale solution for each higher order factor comprising 42 items (7 × 6 items) EPP-S presents the 3 primary scale solution for each higher order factor comprising 60 items (3 × 20 items) EPP-SA presents the 3 primary scale solution for each higher order factor comprising 18 items (3 × 6 items) x 2 = chi-squared goodness of ﬁt test where a non-signiﬁcant value indicates a good ﬁt RMSEA = Root mean Square Error of Approximation where values <.05 indicate a good ﬁt and levels of signiﬁcance should be greater than 0.5 GFI = Goodness of Fit Index where values close to one are desired for a good ﬁt AGFI = Adjusted Goodness of Fit Index where values close to one are desired for a good ﬁt

12

francis, jackson, and jones Discussion

Overall, this study suggests a reasonable level of redundancy at the item level in the full form of the EPP. At the same time, it should be noted that the full version of the EPP generally possesses higher reliability, covers a wider domain of behaviors consistent with each of the primary scales and therefore continues to be useful. Conclusions Four main conclusions emerge from this study. First, it is clear that some of the scales in the full 440-item Eysenck Personality Proﬁler fail to meet Kline’s (1993) stringent criterion for internal reliability coeﬃcients of at least 0.70. This is true for expressiveness, dogmatism, and manipulation among both the clergy and the students. Additionally this is true for aggressiveness, hypochondriasis and masculinity among the clergy, but not among the students. Given the clear success of the other scales within the Eysenck Personality Proﬁler, the challenge should now be to improve and to enhance these weaker scales in order to operationalize more adequately the overall theory according to which each of the major three dimensions of personality comprise seven distinct primary scales. Second, it is clear that wherever the full scale achieved a satisfactory alpha coeﬃcient it was possible to create a 6-item short-form (EPP-A) which possessed satisfactory internal reliability and satisfactory concurrent validity in terms of a high correlation with the parent scale. Although there are weaknesses with some of the 6-item scales, these weaknesses are no more serious than weaknesses found in the parent scales. In other words, the short forms all provide a satisfactory approximation of the longer form and can be recommended for further use when time constraints render the longer form impracticable. Third, the 42-item scales of extraversion, neuroticism and psychoticism all achieve very satisfactory internal reliability among both the clergy and the students. This is particularly signiﬁcant in light of the consistent problems of reliability found with various editions of the psychoticism scale. The present 126 personality items promise not only a broad description in terms of the 21 primary scales, but also a highly robust and reliable assessment of the three major dimensions of personality.

the personality of clergy

13

Fourth, our results support those by Jackson, Furnham, Forde, and Cotter (2000) which suggest a relatively poor ﬁt of the full 21scale version of the EPP to the three-factor model, but a reasonably satisfactory ﬁt for the reduced EPP-S solution. Although we could have modiﬁed the structure of the EPP from the simple form proposed in its manual to improve the goodness of ﬁt, this was not the purpose of this study and, furthermore, the multi-scale complexity of the EPP might make meeting all of the assumptions of these goodness of ﬁt indicators very diﬃcult. However the conﬁrmatory factor analysis does provide evidence that the ﬁt of a 6-item per scale version of the EPP or EPP-S is very similar to the 20-item per scale version. In short, this study has shown that a 6-item per primary scale instrument (EPP-A) retains many of the best features of the 20-item per primary scale instrument (EPP), but with the advantage of being much shorter. We believe that the EPP-A will be especially valuable when administration time is limited, but a relatively complete description of personality is required. References Archer, R. P., Tirrell, C. A., & Elkins, D. E. (2001). Evaluation of an MMPI A short form: Implications for adaptive testing. Journal of Personality Assessment, 76, 76–89. Cronbach, L. J. (1951). Coeﬃcient alpha and the internal structure of tests. Psychometrika, 16, 297–334. Eysenck, H. J. (1958). A short questionnaire for the measurement of two dimensions of personality. Journal of Applied Psychology, 42, 14–17. —— Barrett, P., Wilson, G., & Jackson, C. (1992). Primary trait measurement of the 21 components of the PEN system. European Journal of Psychological Assessment, 8, 109–117. Eysenck, H. J., & Eysenck, M. W. (1985). Personality and individual diﬀerences: A natural science approach. New York: Plenum Press. Eysenck, H. J., & Eysenck, S. B. G. (1975). Manual of the Eysenck Personality Questionnaire (adult and junior). London: Hodder and Stoughton. —— (1991). Manual of the Eysenck Personality Scales. London: Hodder and Stoughton. Eysenck, H. J., & Wilson, G. D. (1975). Know your own personality. Harmondsworth: Penguin. Eysenck, S. B. G., & Eysenck, H. J. (1964). An improved short questionnaire for the measurement of extraversion and neuroticism. Life Sciences, 3, 1103–1109. Eysenck, S. B. G., Eysenck, H. J., & Barrett, P. (1985). A revised version of the psychoticism scale. Personality and Individual Diﬀerences, 6, 21–29. Eysenck, H. J., Wilson, G. D., & Jackson, C. J. (1996). Manual of the Eysenck Personality Proﬁler (short). Guildford, Wales: Psi-Press. Floderus, B. (1974). Psycho-social factors in relation to coronary heart disease and associated risk factors. Nordisk Hygienisk Tidskrift, supplementum, 6.

14

francis, jackson, and jones

Francis, L. J. (1991). The personality characteristics of Anglican ordinands: Feminine men and masculine women? Personality and Individual Diﬀerences, 12, 1133–1140. —— (1996). The development of an abbreviated form of the Revised Eysenck Personality Questionnaire ( JEPQR-A) among 13–15 year olds. Personality and Individual Diﬀerences, 21, 835–844. —— (2002). The personality characteristics of male Evangelical clergy: Denominational diﬀerences in the UK. Mental Health, Religion and Culture, 5, 175–181. —— (in press). The impact of age, personality and churchmanship on baptism policy among Anglican clergy in Wales. Modern Believing. —— (2003). Living without the clerical persona: Lie scale scores among male Evangelical clergy. Research in the Social Scientiﬁc Study of Religion, 14, 103–111. Francis, L. J., Brown, L. B., & Philipchalk, R. (1992). The development of an abbreviated form of the Revised Eysenck Personality Questionnaire (EPQR-A): Its use among students in England, Canada, the USA and Australia. Personality and Individual Diﬀerences, 13, 443–449. Francis, L. J., & Jackson, C. J. (in press). Which version of the Eysenck Personality Proﬁler is best? Six, twelve or twenty items per scale. Personality and Individual Diﬀerences. Francis, L. J., Jones, S. H., Jackson, C. J., & Robbins, M. (2001). The feminine personality proﬁle of male Anglican clergy in Britain and Ireland: A study employing the Eysenck Personality Proﬁler. Review of Religious Research, 43, 14–23. Francis, L. J., Jones, S. H., Robbins, M., & Jackson, C. J. (2003). The personality proﬁle of female Anglican clergy in Britain and Ireland: A study employing the Eysenck Personality Proﬁler. Archiv für Religionspsychologie, 25, 221–231. Francis, L. J., & Kay, W. K. (1995). The personality characteristics of Pentecostal ministry candidates. Personality and Individual Diﬀerences, 18, 581–594. Francis, L. J., & Lankshear, D. W. (1998). Personality and preference for rural ministry among male Anglican clergy. Pastoral Psychology, 46, 163–166. Francis, L. J., & Littler, K. (2001). Personality and preference for rural ministry among Church in Wales clergymen. The Psychologist in Wales, 11, 3–5. Francis, L. J., & Louden, S. H. (2000). The Francis-Louden Mystical Orientation Scale (MOS): A study among Roman Catholic priests. Research in the Social Scientiﬁc Study of Religion, 11, 99–116. —— (2001). Parish ministry and Roman Catholic regular clergy: Applying Eysenck’s dimensional model of personality. International Journal of Practical Theology, 5, 216– 226. Francis, L. J., Louden, S. H., Robbins, M., & Rutledge, C. F. J. (2000). Unmasking the clerical persona: Interpreting the correlation between neuroticism and lie scale scores among Roman Catholic and male and female Anglican clergy. Mental Health Religion and Culture, 3, 133–141. Francis, L. J., & Pearson, P. R. (1988). The development of a short form of the JEPQ ( JEPQ-S): Its use in measuring personality and religion. Personality and Individual Diﬀerences, 9, 911–916. —— (1991). Personality characteristics of mid-career Anglican clergy. Social Behaviour and Personality, 19, 81–84. Francis, L. J., & Robbins, M. (1996). Diﬀerences in the personality proﬁle of stipendiary and non-stipendiary female Anglican parochial clergy in Britain and Ireland. Contact, 119, 26–32. —— (1999). The relationship between personality and satisfaction/dissatisfaction with ministry among female stipendiary Anglican clergy in the UK. Pastoral Psychology, 47, 439–444. Francis, L. J., Robbins, M., Jackson, C. J., & Jones, S. H. (2000). Personality theory and male Anglican clergy: The EPP. Contact, 113, 27–36. Francis, L. J., & Rodger, R. (1994a). The inﬂuence of personality on clergy role

the personality of clergy

15

prioritization, role inﬂuences, conﬂict and dissatisfaction with ministry. Personality and Individual Diﬀerences, 16, 947–957. —— (1994b). The personality proﬁle of Anglican clergymen. Contact, 113, 27–32. Francis, L. J., & Rutledge, C. (2000). Are rural clergy in the Church of England under greater stress? A study in empirical theology. Research in the Social Scientiﬁc Study of Religion, 11, 173–191. Francis, L. J., & Thomas, T. H. (1992). Personality proﬁle of conference-going clergy in England. Psychological Reports, 70, 682. —— (1996a). Are Anglo-Catholic priests more feminine? A study among male Anglican clergy. Pastoral Sciences, 15, 15–22. —— (1996b). Mystical orientation and personality among Anglican clergy. Pastoral Psychology, 45, 99–105. —— (1997). Are charismatic ministers less stable? A study among male Anglican clergy. Review of Religious Research, 39, 61–69. Jackson, C. J., Furnham, A., Forde, L., & Cotter, T. (2000). The structure of the Eysenck Personality Proﬁler. British Journal of Psychology, 91, 223–239. Jones, D. L., & Francis, L. J. (1992). Personality proﬁle of Methodist ministers in England. Psychological Reports, 70, 538. Jones, S. H., & Francis, L. J. (2000). Assessing guilt among non-stipendiary Anglican clergy in Britain and Ireland: Gender diﬀerences. Pastoral Sciences, 19, 175– 189. —— (2003). The pastoral care of the Anglican clergy today: A matter of low selfesteem. Journal of Empirical Theology, 16 (1), 20–30. Jones, S. H., Francis, L. J., & Jackson, C. (in press). The relationship between religion and anxiety: A study among Anglican clergymen and clergywomen. Journal of Psychology and Theology. Kay, W. K. (2000). Pentecostals in Britain. Carlisle, Wales: Paternoster. Kline, P. (1993). The handbook of psychological testing. London: Routledge. Levy, P. (1968), Short-form tests: A methodological review. Psychological Bulletin, 69, 410–416. Louden, S. H., & Francis L. J. (1999). The personality proﬁle of Roman Catholic parochial secular priests in England and Wales. Review of Religious Research, 41, 65–79. —— (2001). Are Catholic priests in England and Wales attracted to the charismatic movement emotionally less stable? British Journal of Theological Education, 11, 65–76. Robbins, M., & Francis, L. J. (1999). The selection process: Comparing the experiences of stipendiary and non-stipendiary Anglican clerygwomen in the UK. In I. Lukatis and W. Lukatis, Education for leadership: Special aspect, gender (pp. 151–166). Hanover: Verlag. (Eds.) —— (2000). Role prioritization among clergywomen: The inﬂuence of personality and church tradition among female stipendiary Anglican clerics in the UK. British Journal of Theological Education, 11, 7–23. Robbins, M., Francis, L. J., & Fletcher-Marsh, W. (2000). Personality proﬁle of Anglican clergywomen in Canada and England: A cross cultural comparison on the EPQR-S. Pastoral Sciences, 19, 79–90. Robbins, M., Francis, L. J., Haley, J. M., & Kay, W. K. (2001). The personality characteristics of Methodist ministers: Feminine men and masculine women? Journal for the Scientiﬁc Study of Religion, 40, 123–128. Robbins, M., Francis, L. J., & Rutledge, C. (1997). The personality characteristics of Anglican stipendiary parochial clergy in England: Gender diﬀerences revisited. Personality and Individual Diﬀerences, 23, 199–204. Robbins, M., Hair, J., & Francis, L. J. (1999). Personality and attraction to the charismatic movement: A study among Anglican clergy. Journal of Beliefs and Values, 20, 239–246.

16

francis, jackson, and jones

Scholte, R. J. H., & DeBruyn, E. J. (2001). The Revised Junior Eysenck Personality Questionnaire ( JEPQ-R): Dutch replications of the full-length, short and abbreviated forms. Personality and Individual Diﬀerences, 31, 615–625. Smith, G. T., & McCarthy, D. M. (1995). Methodological considerations in the reﬁnement of clinical assessment instruments. Psychological Assessment, 7, 300–308. Smith, G. T., McCarthy, D. M., & Anderson, K. G. (2000). On the sins of shortform development. Psychological Assessment, 12, 102–111.

KILLING AS TRAUMA: THE RELIGIOUS IMPLICATIONS OF PERPETRATION-INDUCED TRAUMATIC STRESS Rachel M. MacNair Abstract Posttraumatic stress disorder (PTSD) is a pattern of symptoms which has been studied primarily from the perspective of victims of trauma. More recently, some research has focused on the act of killing as a trauma, across various groups, with diﬀering levels of social approval. Symptoms of PTSD are commonly found in these individuals, often with greater severity. This form of PTSD is called perpetration-induced traumatic stress or PITS. Religious implications include: What do negative psychological consequences of perpetrating violence say about the human spirit? How does this ﬁt into pastoral care and therapy? What about the prevention of further violence? How does the biological component of PTSD ﬁt into theoneurology? What does this say about justice or sympathy, forgiveness and reconciliation? Are there ethical implications for public policy?

Religious thought has generally regarded killing as wrong, although there have been variations on when it might be allowed, such as war and executions. Psychology leaves to religion and ethics the question of whether it is wrong, but rather asks the question: Is it traumatic? Is only victimhood and rescue work traumatic, or can the post-trauma reactions associated with victims and their helpers also be applied to people who actually caused the trauma to occur? If perpetrators have post-trauma symptoms, are they less severe, equally severe, or more severe? Psychology and religion together ask the question: If killing is indeed traumatic to those who kill, what does this mean? What are the applications and implications? This article will examine the concepts of posttraumatic stress disorder in general and perpetration-induced traumatic stress in particular, while considering some of their possible religious implications. The concept of posttraumatic stress disorder (PTSD) originated with combat veterans. In World War I, it was called “shell shock” and was thought to be a physical problem. The term used in World War II was “battle fatigue” or “combat fatigue,” and experts decided it was psychological in origin. Thousands of articles and many books have covered various aspects of posttraumatic stress since it was ﬁrst oﬃcially deﬁned in 1980. Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

18

rachel m. macnair

Most of the literature has focused on people who have had a trauma inﬂicted upon them by outside forces. Those forces could be accidents, hurricanes, or tornadoes, but it has been observed that when the trauma originates with an act of human intent, the symptoms are likely to be more severe. Very little of the literature has considered PTSD symptomatology among perpetrators. Most of the literature that has addressed this has analyzed participation in unlawful war atrocities, not the act of killing expected and demanded of a soldier in war. Yet during World War I, social worker and sociologist Jane Addams noted aftereﬀects of having killed. She went to Europe and did interviews with nurses and wounded patients in hospitals. In the language of the time, she talked of insanity among the soldiers in various places, and of their being dazed after participating in attacks. She talked of hearing “from hospital nurses who said that delirious soldiers are again and again possessed by the same hallucination—that they are in the act of pulling their bayonets out of the bodies of men they have killed” ( Johnson, 1960, p. 273). This is a classic symptom of PTSD. Definition of PTSD and PITS There are two oﬃcial deﬁnitions of PTSD: one from the American Psychiatric Association which ﬁrst deﬁned it in 1980 in the third edition of the Diagnostic and Statistical Manual; the other from the World Health Organization in its International Statistical Classiﬁcation of Diseases and Related Health Problems (1992), currently in its tenth revision and normally referred to as ICD-10. According to the DSM-IV, TR (2000), section 309.81 (p. 465), the symptoms necessary for making a psychiatric diagnosis of PTSD are listed here (with notes on children removed). A. The person has been exposed to a traumatic event in which both of the following were present: (1) the person experienced, witnessed or was confronted with an event or events that involved actual or threatened death or serious injury, or a threat to the physical integrity of self or others. (2) the person’s response involved intense fear, helplessness, or horror.

killing as trauma

19

B. The traumatic event is persistently reexperienced in one (or more) of the following ways: (1) recurrent and intrusive distressing recollections of the event, including images, thoughts, or perceptions. (2) recurrent distressing dreams of the event. (3) acting or feeling as if the traumatic event were recurring (includes a sense of reliving the experience, illusions, hallucinations, and dissociative ﬂashback episodes, including those that occur on awakening or when intoxicated). (4) intense psychological distress at exposure to internal or external cues that symbolize or resemble an aspect of the traumatic event. (5) physiological reactivity on exposure to internal or external cues that symbolize or resemble an aspect of the traumatic event. C. Persistent avoidance of stimuli associated with the trauma and numbing of general responsiveness (not present before the trauma), as indicated by three (or more) of the following: (1) eﬀorts to avoid thoughts, feelings, or conversation associated with the trauma. (2) eﬀorts to avoid activities, places, or people that arouse recollections of the trauma. (3) inability to recall an important aspect of the trauma. (4) markedly diminished interest or participation in signiﬁcant activities. (5) feeling of detachment or estrangement from others. (6) restricted range of aﬀect (e.g., unable to have loving feelings). (7) sense of a foreshortened future (e.g., does not expect to have a career, marriage, children, or a normal life span). D. Persistent symptoms of increased arousal (not present before the trauma), as indicated by two (or more) of the following: (1) diﬃculty falling or staying asleep. (2) irritability or outbursts of anger. (3) diﬃculty concentrating. (4) hypervigilance. (5) exaggerated startle response. E. Duration of the disturbance (symptoms in Criteria B, C and D) is more than 1 month. F. The disturbance causes clinically signiﬁcant distress or impairment in social, occupational, or other important areas of functioning.

20

rachel m. macnair

In short, criterion A requires a stressor—a trauma—as a causal mechanism. The same symptoms from an imaginary event or from an event most people would not regard as traumatic would be a diﬀerent disorder. Sections B, C, and D are symptom clusters—re-experience; avoidance and numbing; and arousal, respectively. Criterion E distinguished PTSD from an acute stress reaction, which normally subsides with time and is therefore treated diﬀerently. With PTSD, the passage of time has not been suﬃcient to alleviate the condition. People are stuck. Criterion F is intended to specify that this is a diagnosable disorder. For purposes of this discussion, however, subclinical levels of symptomatology are also of interest. The deﬁnition in ICD-10, F43.1 (p. 344) is in more of a narrative form. Arises as a delayed or protracted response to a stressful event or situation (of either brief or long duration) of an exceptionally threatening or catastrophic nature, which is likely to cause pervasive distress in almost anyone. Predisposing factors, such as personality traits (e.g. compulsive, asthenic) or previous history of neurotic illness, may lower the threshold for the development of the syndrome or aggravate its course, but they are neither necessary nor suﬃcient to explain its occurrence. Typical features include episodes of repeated reliving of the trauma in intrusive memories (“ﬂashbacks”), dreams or nightmares, occurring against the persisting background of a sense of “numbness” and emotional blunting, detachment from other people, unresponsiveness to surroundings, anhedonia, and avoidance of activities and situations reminiscent of the trauma. There is usually a state of autonomic hyperarousal with hypervigilance, and enhanced startle reaction, and insomnia. Anxiety and depression are commonly associated with the above symptoms and signs, and suicidal ideation is not infrequent. The onset follows the trauma with a latency period that may range from a few weeks to months. The course is ﬂuctuating but recovery can be expected in the majority of cases. In a small proportion of cases the condition may follow a chronic course over many years, with eventual transition to an enduring personality change. I propose the term “perpetration-induced traumatic stress” (PITS) to describe a sub-category of PTSD. The term “perpetration-induced” was inspired by the following quotation, in which authors are discussing PTSD as a legal defense in criminal trials: “It must be able to be established . . . that PTSD existed at the time of the violent

killing as trauma

21

crime and did not stem from it, as in some perpetrator-induced trauma” (Hall & Hall, 1987, p. 49). PITS should not necessarily be considered a “disorder” because the symptomatology is of psychological, sociological, historical, and theological interest even if it does not rise to the level of a disorder. Another diﬀerence is the fact that the former is a chronic condition, not an acute one. However, for perpetrators with PITS, the condition could be chronic and yet still be ongoing, as with people responsible over several years for carrying out executions. Hence, the term perpetration-induced traumatic stress, (PITS) denotes any portions of the symptomatology of PTSD, at clinical or sub-clinical levels, which result from situations that would be traumatic if someone were a victim, but situations for which the person in question was a causal participant. The National Vietnam Veterans Readjustment Study Of all the diﬀerent kinds of perpetrators, veterans are the one group in which suﬃcient quantitative data is available to allow for some generalizability at least to similar veterans. The largest study of any at-risk population for PITS was a United States government-commissioned survey of American veterans of the war in Vietnam, done in the 1980s (Kulka et al., 1990). The National Vietnam Veterans Readjustment Study (NVVRS) used a stratiﬁed random sample of Vietnam-era veterans and a comparison group of veterans. MacNair (2002) did a secondary analysis of this data, using the 1,638 Vietnam theater veterans. In that study, PTSD scores were signiﬁcantly more severe among the perpetration groups (those who said they killed, or those who said they were directly involved in killing civilians or prisoners) as compared to the control groups (those who said they did not kill, or those who “only saw” the killing of civilians). This is consistent with the consensus of previous literature (Strayer & Ellenhorn, 1975; Yager, Laufer, & Gallops, 1984; Hendin & Haas, 1984; Laufer, Gallops, & Frey-Wouters, 1984; Green, 1990; Grossman, 1995). Eﬀect sizes for these diﬀerences were .82, .57, and .75 for the light, medium, and heavy combat experienced groups respectively. This included scores limited to those who were exposed to situations in which civilians were killed, so that it was known that they did at least have traumatic experience, unlike those who might have answered “no” to killing because they were doing paperwork.

22

rachel m. macnair

Those “directly involved” had much higher PTSD scores than those who “only saw” killing. An obvious objection is that killing could simply be a marker for having been in more intense combat. The intensity of the combat has always been associated with more severe PTSD. This would be expected if PTSD is a response to trauma. The more intense the trauma, the stronger the response. However, those who had killed in light combat had a higher mean score than those who had not killed in heavy combat. Discriminant function analysis also showed that there were some pattern diﬀerences between those who said they killed and those who said they did not. Those who said yes on killing were especially high in violent outbursts and in intrusive symptoms (nightmares, ﬂashbacks, and unwanted thoughts). They were also higher in hyperarousal, alienation, survivor guilt, and a sense of disintegration. How they would have scored on a sense of guilt other than survivor guilt is not known, because it was essentially not asked. This would be a point of particular interest in this population as well as a matter of interest for considering religious implications. Those answering “no” to killing, surprisingly, scored high on concentration and memory problems. The explanation for this is still at the speculative stage. More details of the extensive statistics done on this data set is described elsewhere (MacNair, 2002), along with discussions of the evidence for PITS in various groups, such as those who perform executions, criminal homicide, Nazis, and police who shoot in the line of duty. Areas for further research and implications in various disciplines are also included. Here the speciﬁcally religious and spiritual implications are examined and expanded. I have divided them into seven categories, but interconnections abound. What Does PITS Suggest About the Human Spirit? The concept that perpetration is itself a form of suﬀering is not unusual in religious literature. For example, in the Vatican’s Pastoral Constitution on the Church in the Modern World, a list of diﬀerent kinds of violence against people is oﬀered with this conclusion: “They poison human society, and they do more to harm those who practice them than to those who suﬀer from the injury” (Gaudium et Spes, 22).

killing as trauma

23

In a non-religious, but soul-based argument, Socrates in Plato’s Gorgias has a lengthy discourse on this point. Socrates summarizes it by saying, “You deemed Archelaus happy, because he was a very great criminal and unpunished. I, on the other hand, maintained that he or any other who like him has done wrong and has not been punished, is, and ought to be, the most miserable of all men; and that the doer of injustice is more miserable that the suﬀerer.” The construct of perpetration-induced traumatic stress could be interpreted as a psychological, scientiﬁc way of operationalizing this view. This allows a test of it as a hypothesis. As discussed above, while much more research needs to be done, that which has been done so far indicates that PTSD symptomatology does appear to be more severe among those who have killed than among those whose symptoms arise from a diﬀerent form of trauma. The idea of perpetration-induced traumatic stress suggests that the human mind, contrary to certain political ideologies, is not well suited for killing. This ﬁts well with the concepts of natural law or ideas that God created us in such a way as to have our minds at their best when they are in harmony with a moral order. When we contemplate the depths of destructiveness in which human beings have engaged, this information is more heartening. This is the positive side of knowing about an otherwise distressing subject: the human mind and the human spirit are not designed to engage in killing without a negative aftermath. Pastoral Care and Therapy What happens to pastoral and professional therapeutic counseling when it becomes clear that active participation in traumatic circumstances was the etiological mechanism for a speciﬁc client? A knowledge that killing is a stressor and that killing has certain common features in its psychological aftermath (violent outbursts, intrusive imagery, a sense of disintegration) is necessary for pastors and therapists who want to do the best work for those suﬀering. Any diﬀerences in what constitutes eﬀective treatment need to be understood. For example, Foa and Meadows (1997) note one treatment that might diﬀer: “In particular, PTSD suﬀerers whose traumatic memories are about being perpetrators rather than victims may not beneﬁt from [prolonged exposure as a treatment] and perhaps will

24

rachel m. macnair

even deteriorate from such treatment” (p. 475). The article they cite is a set of six case studies involving the ﬂooding technique. This uses intense reminders of the trauma for the purpose of desensitization. Participation in abusive violence was one of the times when this appeared to be counterproductive (Pitman, et al., 1991). On the other hand, Kruppa (1991) cites a case study in which she used ﬂooding therapy with a person who had committed a criminal homicide to treat the ﬂashbacks. The treatment was successful for that speciﬁc symptom, but had no eﬀect on the feelings of remorse and disgust. A therapist also has to be careful that a technique involving desensitization to acts of violence is not going to have dire consequences for the commission of further acts. Kruppa, Hickey, and Hubbard (1995) address this point of whether untreated intrusive imagery provides a constant reminder of the oﬀense, and thereby inhibits further oﬀending. Would the imagery instead increase the chance of oﬀending again? What happens if the intrusive imagery is treated and eliminated, but the sense of estrangement or detachment from others is left intact? Does it make a diﬀerence whether the person has remorse, or whether the person will again be in similar situations? The ethical considerations are much more complicated when the etiological stressor is one that the client is likely to repeat rather than one that the client always wished to avoid. When therapists consider the implications of this idea, it is not uncommon that traditional religious concepts of dealing with wrongdoing arise. Foa and Meadows (1997), for example, suggest that when guilt is justiﬁed, “alternative strategies . . . [include] exploring ways of making reparations and bearing witness” (p. 475). Atonement, repentance and forgiveness, bearing witness, and re-identifying one’s self as a diﬀerent person than the one who did the killing (as in being “born again”) have been suggested in many verbal discussions I have had with therapists. This has been one of the responses of the human community in diverse cultures and through many historical periods to the common phenomenon of dealing with killing. They have remained because of extensive experience that they are, in fact, helpful. Glover (1985), however, suggests both a help and a hindrance: Religious activity is one way in which the veteran might ﬁnd atonement and a renewed sense of purpose. However, consideration should be given to whether the chosen activity serves primarily as an escape for the veteran from ever understanding his past. This would appear

killing as trauma

25

to be the case, for example, if religion was primarily used as a vehicle to assure personal salvation in the next world. (p. 18).

The extent to which it is in fact helpful for someone who has killed to understand the past rather than avoiding memories of it is another therapeutic question, along with religion’s role in either contingency. Both Haley (1974) and Shatan (1978) have pointed out that when patients report having committed atrocities, the therapists have more trouble listening. Killing diﬀers from the category of “atrocities” may well cause the same predicament. These authors were primarily stating this as a problem in doing a proper treatment for the client. This is likely to be one of the diﬀerences between perpetrators and pure victims that need to be kept in mind. Feelings of revulsion or punitiveness need to be controlled for the client’s beneﬁt. In addition, however, there are implications for the well-being of the counselor. People who do counseling and therapy can suﬀer from a condition called “secondary traumatization.” This has been noted in those who deal with rape victims and other crime victims. Archbishop Desmond Tutu reported it as a problem with the commissioners and journalists who were spending so much time hearing the brutal details of the repression under apartheid as part of the South African Truth and Reconciliation Commission: We thought we had been reasonably well prepared for the traumatic experience. Despite all this we were shattered at what we heard and we did frequently break down or were on the verge of it . . . Many reported disturbed sleep patterns; some were deeply concerned that they were more short-tempered, quarreled far too easily with their spouses, or were drinking far more than they should (Tutu, 1999, p. 286).

If intensive sympathetic listening to victims can have this eﬀect, how much greater might the eﬀect be when the client caused the trauma? It is not only that such a memory is more traumatizing for the counselor to imagine. There is the additional complication of a sense of revulsion which is devoid of sympathy for the suﬀerer. The expression or lack of expression of remorse by the client becomes a crucial variable in how well the counselor or therapist can handle the interaction. Much more study needs to be done not only on how to best treat the client, but how to best support the counselor. What are the best ways to assist the people who must carry on this work? What help do the helpers need?

26

rachel m. macnair Dreams

Initial studies suggest that the dreams of those who have killed may include diﬀerent motifs, especially motifs which bring in religious questions. In addition to the eidetic dreams—that is, replays of the original event, almost like seeing video of it—there are a couple of common motifs found in group therapy with American veterans of the war in Vietnam (Glover, 1985, 1988; Lifton, 1990). One is that the dreaming person reverses roles and is the person who is killed in a similar situation. Another motif is being accused by those killed, having those killed demanding an explanation, or having them suggest that bad things will happen to the dreamer. To illustrate the ﬁrst motif, one veteran reported a dream with an enemy soldier that had shot him in the leg and whom he had apparently killed. This soldier kept showing up vividly in a recurrent dream, a replay of the event with the major diﬀerence that the veteran himself was killed by his antagonist. On the same page Robert Jay Lifton reports a double self illustrated with this veteran’s dream: “I was arguing with myself. Then there were two separate selves, and one of them ﬁnally shot the other, so that I shot myself.” (Lifton, 1990, p. 429) In this case, the veteran is simultaneously killer and victim. Another illustration is reported in the case of a police oﬃcer who was diagnosed with PTSD resulting from his having shot someone in the line of duty, a shooting that was later ruled justiﬁed. The content of dreams as told to a television reporter was that “over and over, in his dreams, Sal faced a man with a knife. In one dream, he faced him alone. In another, alone and unarmed. And in the worst of all, alone, stark naked and terriﬁed” (Dateline, 2000). This suggests religious questions about the boundaries of the self and its relation to other people. Both Glover (1985) and Lifton (1990) attribute the dreams about being killed to a sense of guilt, and both report veterans objecting to this interpretation. Another possible interpretation has to do with the psychology of distinguishing the self from that which is not the self. We learn as babies that when we choose to move our ﬁngers, they move. When we choose to move someone else’s ﬁngers, they do not. In this way, we learn to distinguish the self from the not-self. This clear-cut distinction becomes blurred when we imagine being someone else or get absorbed in a story or in a task. The distinction becomes blurred when we go to

killing as trauma

27

sleep. It is common in a dream state of mind for images to mingle self and not-self together. When an act of killing takes place, the diﬀerence between the self which is doing it and the not-self which is killed is quite clear. It may be more blurred in a dreaming state of mind. This is not the kind of mature self-transcendence that is seen as positive in many religious traditions, but a form of losing self in a more negative way. Yet since it does involve viewing the self as if from the outside, it shares similarities with the concept of self-transcendence that deserve exploration. To illustrate the second motif, Glover oﬀers a case in which the veteran “frequently hears the voices of the Vietnamese women and children he killed, accusing him (in English) of killing them or demanding of him the explanation of why he killed them. The voices warn him that bad things will happen to himself and his family” (Glover, 1988, p. 70). The motif of being accused by those killed was also found by a medical reporter at a conference of the National Abortion Federation: “They talk about . . . their dreams, in which aborted fetuses stare at them with ancient eyes and perfectly shaped hands and feet asking, ‘Why? Why did you do this to me?’” (Gianelli, 1993). Religious themes can be helpful in understanding the psychological component of these dreams. One of those themes is dealing with guilt. Another is repairing badly torn relationships with the human community. All this suggests tie ins with pastoral counseling and other professional therapies. It also suggests another way to understand the psychological dynamics of causation in PTSD symptoms, which in turn helps us to understand what the existence of PITS has to say about the human spirit. Prevention of Further Violence The psychological eﬀects of violence can become a cause of further violence. PITS is one psychological model which can help us to understand cycles of violence and the maxim, “violence breeds violence.” While there are also other models or perspectives, understanding this one can generate interventions to help break those cycles. In situations where violence is generated by single individuals, such as domestic abuse and street crime, some investigation has shown

28

rachel m. macnair

precise mechanisms for PTSD symptoms to be a risk. Silva and colleagues (2001) oﬀer four ways that PTSD symptoms can cause violent behavior. The ﬁrst is sleep disturbance with physical thrashing around in the bed. Anyone in the same bed is at the receiving end of blows without warning and with no intention by the sleeper. The second way is associated with outbursts of anger, irritability, and hostility. The third is associated with ﬂashbacks, and comes from ﬂashbackinduced misidentiﬁcation of others. The fourth is combat-addiction violence. Flashback experiences, just like the dreams of PTSD, can seem quite real. This is especially so since they are based on experiences the person has actually had. On occasion, the ﬂashback can be severe enough that it leads to a loss of ordinary understanding of reality. Authors oﬀer as an illustration the case of Mr. A: “On one occasion, he had just won several games of pool and, in frustration, the losing party verbally attacked Mr. A. At that point, Mr. A noted that the man’s face had transformed into the face of a Vietnamese foe who was wearing the traditional black clothes of the Viet Cong. Mr. A. stated that for several minutes he was convinced that he was dealing with a dangerous Viet Cong soldier and therefore attacked the man perceived as the enemy with his hands” (Silva et al., 2001, p. 309). Other forms of dissociation can also lead to confusion that contributes to violent events. The lessened ability to distinguish real threats from circumstances that others would not regard as threatening can be debilitating when the goal is avoiding unnecessary violence. As for the combat-addiction violence, any kind of “repetition compulsion,” or desire to get another adrenaline rush in the way one has done in the past, or re-enacting for any positive goal of excitement or post-action calmness, can lead directly to violent acts. As an example, Mr. D. would frequent Chinatown and other areas of the city with signiﬁcant Asian populations. There, he would engage in numerous physical ﬁghts with those reminding him of his former Vietnamese enemy. He sought these physical confrontations in order to “feel alive”. . . . After those ﬁghts, he would welcome the sense of calmness. As is the case for many Vietnam veterans with PTSD, he often had feelings of emptiness and numbness that contributed to a lack of meaning in his life. His engaging in frequent physical ﬁghts

killing as trauma

29

brought about a sense of excitement that he described as “being alive again.” His repeated violent addictive behaviors resulted in decreases in anxiety, tension, and other negative emotions (Silva et al., 2001, pp. 310, 313). (Addiction to trauma as a concept will be discussed in more detail in the next section.) These four categories (thrashing in bed, outbursts, ﬂashback-induced misidentiﬁcation, and combat addiction) deal only with the kind of chaotic violence not planned even by the individuals engaged in it. Most violence is more socially organized than this, and all forms of massive violence, such as wars, have among their causes some social psychological components. The symptoms of PTSD can be applied to the theories about causation of such violence from The Psychology of Peace to show how people with such symptoms make such violence more likely to occur (MacNair, 2003). From the American Psychiatric Association’s (2000) deﬁnition of PTSD, symptom C(5) is a “feeling of detachment or estrangement from others,” and symptom C(6) is a “restricted range of aﬀect (e.g., unable to have loving feelings).” These symptoms can exacerbate and, sometimes initially cause, the practice of using dehumanizing language about the targets of violence and euphemisms about the actions against them (Brennan, 1995). They support minimizing or ignoring the eﬀects of the actions. Those two symptoms along with symptom D(2), “irritability or outbursts of anger,” render the occurrence of scapegoating more likely. All of these have been labeled “mechanisms of moral disengagement” which facilitate the commission of acts of massive violence (Bandura, Barbanelli, Caprara, & Pastorelli, 1996). Putting a mental distance between the doer and the deed can be facilitated by the “numbing of general responsiveness” that helps deﬁne symptoms cluster C. The numbing together with the sense of estrangement from others can easily interact with these mechanisms, making them worse or more likely to occur. Those who already have PTSD may be more susceptible to destructive demands of authority. The famous experiments by Stanley Milgram (1974) showed that the majority of experimental participants “delivered” the highest level of shock upon “subjects” when authority ﬁgures continually instructed them to do so, in spite of obvious (though simulated) distress on the part of the person being “shocked,” and in spite of real distress experienced by the person

30

rachel m. macnair

doing the shocking as well. Milgram had hypothesized that nativeborn Germans would comply more frequently than those Americanborn individuals. He then intended to search for the cultural diﬀerences, but instead found that in fact Americans also complied (Blass, 2000). One of the best conditions for achieving greater noncompliance was having another person across the room, who was apparently engaged in the same experiment, show noncompliance. Yet the estrangement from others, blocked emotions, and numbing take away one of the major resources available to cause noncompliance. Sympathy for the victim is curtailed, as well as attention to other people resisting. Additionally, if it is the leaders who suﬀer from PTSD symptoms, this could help explain why an authority expects violent behavior and gives the demands for compliance. The same symptoms that make compliance more likely may also make the issuance of the orders more likely. When the dissociation that can make individuals unable to distinguish real threats is present in policymakers, it can contribute to groupthink, war hysteria, and other forms of being unable to appreciate complexities of a particular situation. Detachment or estrangement from others would also reduce motivation to try to understand diﬀering perspectives, which has been found to be an important component in decisions to go to war, or take action to avoid war (Conway, Suedfeld, & Tetlock, 2001). The Biological Component and Theoneurology Using Positron Emission Tomography (P.E.T.) scans, Newberg, D’Aquili, & Rause (2001) demonstrated how the “orientation area” of the brain becomes less active during Buddhist meditation and Franciscan nuns’ prayers, providing a neurobiological basis for the resultant sense of oneness with the divine. However, because brain activity changes, that does not invalidate the spiritual experience. As Newberg, D’Aquili, & Rause (2001) noted, brain changes in response to an apple pie do not diminish the fact that the pie has an aroma. Indeed, Newberg, et al. data provide support for the case that spiritual experiences are not ethereal, but have a concrete physiological foundation. This is the positive side of the spirituality of the brain— the part that deals with heavenly mystical experience. What of the more hellish aspects? For the negative side, knowledge of the impact

killing as trauma

31

of actions that separate people from the divine can be found in the biological aspects of PTSD. The most clear biological component, the one which was ﬁrst known, is that there are measurable physiological reactions to reminders of the trauma. The increased heart rate, blood pressure, and galvanic skin response can be measured and in some cases have even been used to diagnose PTSD. Suﬀerers are hooked up to the measuring machines and given some sort of pictures or sounds or script that reminds them of the trauma. There is a high correlation between those who have a physiological reaction and those who have been diagnosed by other methods. In hormonal and other biochemical phenomena, PTSD can be distinguished from the general stress response. The unique hormonal proﬁle in PTSD is characterized by low cortisol levels, continuously high levels of epinephrine and norepinephrine—also known as adrenaline and noradrenaline (the catecholamines)—a high norepinephrine/cortisol ratio, high testosterone levels, and high total thyroxine (T4) levels in the face of lower free T4 levels. This is an oversimpliﬁed summary in a very complicated area, and this proﬁle is still being explored. It has been found that there are possible brain diﬀerences in those that have PTSD, particularly a smaller volume in the right hippocampus (Bremner, et al., 1995). Longitudinal studies are currently trying to determine whether these brain diﬀerences are the result of PTSD, or whether people with such brain characteristics are more predisposed to PTSD. The hippocampus is associated with memory. Investigations have also focused on the amygdala, associated with the hippocampus and with stress, and on Broca’s area, which deals with language. Language processing or lack of processing may have something to do with the intrusive imagery or numbing common among PTSD suﬀerers. It may be that further study will show that there is actually little or no biological diﬀerence between those who are victims and those who are perpetrators in the traumatic circumstances. That would be an interesting ﬁnding because the absence of a diﬀerence would mean perpetration is not precluded from being accounted a trauma. Biological similarities would bolster the case that PITS is a form of PTSD. It might also be found that the biological patterns will be mostly the same, but with some pattern diﬀerences. We should be familiar with these diﬀerences for therapeutic pharmacological

32

rachel m. macnair

interventions and for understanding of causation—that is, for therapy and for understanding what PITS says about the human mind and the human spirit, as discussed previously. As mentioned above in the section on further violence, there is a theory of a possible addictive nature of trauma—a sense of thrill, of exhilaration, that accompanies the act. The role of endogenous opioids would be involved here. Stress situations, especially highly traumatic ones, can lead to endogenous opioid analgesia (van der Kolk et al. 1985; Southwick, Yehuda, & Morgan, 1995) and related complicated biochemical reactions. In artiﬁcial form, opioids are related to morphine, heroin, and cocaine—drugs well known for their addictive qualities. Also related is the role of dopamine release in the brain. Koepp et al. (1998) found that PET scans showed endogenous dopamine release during “goal-directed” video games. A large portion of video games are not merely goal-directed, but have as a goal the simulated killing of humans or ﬁctitious aliens. This is about as close to studying the body’s chemical reactions to perpetration as the laboratory will ethically allow. Dopamine would also have an eﬀect that could lead to a sense of pleasure at the activity, and therefore a possibly addictive eﬀect. For both the opioids and the dopamine, the stress response normally helps with pain avoidance in a crisis. This is adaptive, since running away from the danger may require that injuries be ignored. In this case, however, as with much of the stress response’s eﬀect on health, the impact usually is not at all adaptive. Grossman (1995) characterizes the exhilaration as a common stage in the killing process. He quotes a Rhodesian veteran: Combat addiction . . . is caused when . . . the body releases a large amount of adrenaline into your system and you get what is referred to as a ‘combat high.’ This combat high is like getting an injection of morphine—you ﬂoat around, laughing, joking, having a great time, totally oblivious to the dangers around you. . . . Problems arise when you begin to want another ﬁx of combat, and another, and another, and, before you know it, you’re hooked. As with heroin or cocaine addiction, combat addiction will surely get you killed. And like any addict, you get desperate and will do anything to get your ﬁx. (pp. 234–237)

Several have reported psychiatric interviews showing this rush, this sense of thrill, a high as with cocaine, in veterans suﬀering from PTSD (Solursh, 1988, 1989; Nadelson, 1992; Wikler, 1980). Since

killing as trauma

33

these were selected as men with chronic PTSD, this “rush” does not appear to protect against PTSD. It may well aggravate it. Solursh and Nadelson both suggest that the symptoms of intrusive images in nightmares and ﬂashbacks may be accompanied by this “rush.” This “rush” could contribute to maintaining the symptom and complicating the treatment. They also report the symptoms to be especially strong when clients were experiencing demands in the workplace or authorities under whom they felt powerless. The re-enactment is a mental assertion of power as well as excitement. However, as is common with “highs,” a letdown period follows afterward. Powerlessness and frustration returns. Ironically, a sense of thrill may still be seen as a reaction to trauma. Those pain-relieving opioids produced by the brain when it senses danger is a beneﬁcial adaptation developed for our survival; unfortunately, becoming addicted to the chemicals produced is not beneﬁcial. In historical terms, it may oﬀer some insight into the term “bloodthirsty.” The concept can be applied across other groups besides combat veterans. For example, some of those working in occupations which involve killing may in fact be ensnared in a situation which is not easy for them to leave, in the same way that it is diﬃcult for an addict to stop taking drugs. It also sheds new light on the thrill associated with blood sports such as cockﬁghts, bullﬁghts, and hunting. As Grossman (1995) says in this context, “What hunter or marksman has not felt a thrill of pleasure and satisfaction upon dropping his target?” (p. 234). A bullﬁghter interviewed by the American newsmagazine 60 Minutes (air date January 11, 1998) said: When you come out of this experience and—you appreciate everything you have around you; the skies look bluer, the birds sound better, the food tastes better . . . I mean, if I could tell you what it was, maybe we could bottle it and sell it and save a lot of people—you know, if we could bottle the adrenaline, if we could bottle that feeling a matador has after a ﬁght and sell—and it’ll be wonderful.

Is there a resemblance between this statement and others made about those times when the feeling is in fact put in a bottle, a syringe, or a powder? If there are people who have an addiction, either biochemical or behavioral, then therapy needs to be altered accordingly. As one discussion of this aspect put it, “a therapeutic approach based on the assumption that the PTSD symptomatology was aversive to

34

rachel m. macnair

the veterans would be unlikely to be successful” (Hodge, 1997, p. 96). Since there are clearly times when the symptoms are aversive, a clear distinction would need to be made. Nor does the lack of aversion mean that the symptoms require no treatment, since this is a common attribute of addictions that do require treatment. The dangers to other people in failing to treat someone with an addiction to violence are also a major concern, if the idea of such an addiction turns out to have merit after further research. Justice, Sympathy, Forgiveness, Reconciliation Some people see PITS as a good thing. They see those who kill as monsters who deserve the consequences of killing. Others look at the same information and object that it provides too much misplaced sympathy to those who kill, when sympathy should be reserved for victims. However, when there is already sympathy, as with soldiers or executioners or police who kill in the line of duty, then those who suﬀer PITS from doing something seen as necessary are often considered heroic. The circumstance of the killing, whether it was socially approved or required, may have a great deal to do with how much sympathy there is for the suﬀerer of PITS. There are those who suggest that therapy and healing are necessary in all cases, no matter the level of sympathy. Several years ago when I ﬁrst approached the subject of how the human mind responds to the act of killing, I knew that “battle fatigue” was called “posttraumatic stress disorder” in the psychological literature. I therefore assumed that this is what I would need to study. It was an assumption derived from the bias of my background as a Quaker paciﬁst, in a tradition similar to that of Jane Addams. While I was making the same kind of assumption that she had in her observation of World War I soldiers, I found upon delving into the literature that other people were not making the same assumptions. Even in the case of veterans, people were primarily thinking of the risks of combat and the stresses of seeing friends killed as being the traumas causing the problems. There was fear from the dangers of being shot at, grief from seeing the results of others being shot, but little consideration of the act of shooting as being traumatizing. At least four major blind spots have kept this from being more widely considered. One is sympathy for people such as veterans, and

killing as trauma

35

denial they have done anything for which to suﬀer psychological consequences. Related to this is the possibility that the nation which sends soldiers into combat has responsibility for them, and that responsibility will feel much more like guilt if the activities necessary to combat have such an aftermath. The fact that practices are sometimes controversial is a third reason. People on the side of advocating a particular practice may be defensive, and opponents may not think that those engaged in the practice warrant any sympathy. Finally, there are those for whom people in general have no sympathy at all, such as Nazis, slave-catchers, torturers, those practicing genocide, and those guilty of criminal homicide. Acknowledging “villains” as having pain does not occur to many, and again, they are seen as undeserving of any sympathy and lacking humanity to any degree. Yet for some victims, or families of those killed, this knowledge may actually be an aid to forgiveness. Not all perpetrators will suﬀer from the symptoms, of course; it is only a possible consequence, not a certainty. Yet those with remorse who express that they have suﬀered in these ways can be understood to be credible. If the phenomenon is suﬃciently widespread, then there is reason to believe them when they say it is so in their particular case, and is not merely created as a plea for sympathy. Victims and relatives of victims will vary in how much they are willing to give any weight to this. Yet for those inclined toward forgiveness, but concerned that this not be interpreted to mean allowing people to get away with outrageous acts, interactions which include knowledge of PITS can help in reconciliation eﬀorts. This may be especially so in society-wide reconciliation eﬀorts, such as those in post-apartheid South Africa, where the number of perpetrators was so great that they were oﬀered amnesty in exchange for full disclosure as a process of avoiding the harm of recriminations and the harm of forgetting what had happened (Tutu, 1999). The reactions of people to the concept of PITS will vary widely according to their own thoughts about justice. Religious doctrine, practices, and understanding of the human condition can play a major role in what these responses may be, and which ones are the most spiritually healthy.

36

rachel m. macnair Ethical Implications for Public Policy

What implication is there for jobs which require killing, and therefore obligate workers to be at risk for a mental disorder? At the very least, informed consent about this risk is a minimal ethical standard when employment requires killing. Additionally, just as people involved in emergency rescue work require extra attention to possible posttrauma reactions, so might those whose work involves killing, such as executioners. This also strengthens the legal case for adding a “conscience clause” that could be invoked to avoid participation. This information could add valuable insight into debates over whether such socially approved violence should be a public practice at all. For the death penalty, for example, many of the arguments oﬀered never take into account the eﬀect of executions on those who carry them out. Even with this added information, there will be those who still favor the death penalty. Nevertheless, this information should be taken into account. The public debate over abortion is another area where PITS can add insight. Can we shed light on whether or not abortion is killing of a human being, as one side of the debate claims, or merely a medical procedure, as some on the other side claim, by studying the psychological aftermath of abortion doctors and nurses for PTSD symptoms? If we do, does it matter to the debate? Studies so far have been inconclusive and open to alternative explanations (MacNair, 2002), but are certainly suggestive that the PTSD symptoms may be prevalent. There have been two studies done on fairly large samples, studies done by people who favor abortion availability. They have noted the symptoms and suggested that this must be dealt with in order to keep enough abortion practitioners available (Such-Baer, 1974; Roe, 1989). Therefore, even if PTSD symptoms were ﬁrmly established, proponents and opponents of abortion would still have diﬀering views as to what this means. Nevertheless, as a part of the reality, it can add an important dimension to the debate. In war, there is no controversy as to whether there is killing, but only as to whether it is justiﬁed. David Grossman, an oﬃcer in the United States Army, argued (1995) that facing up to what the act of killing really does to well-intentioned people is necessary to realistic social policies. Veterans are not helped if the real causes of their PTSD are assumed to be more benign or due to their individual predispositions. The same is true for other classes of socially sanc-

killing as trauma

37

tioned killers. Society as a whole has made the arrangements that caused this individual aﬄiction. Social forces can heal it, and social forces with greater knowledge and understanding, can stop making those arrangements. The belief that violence is necessary in certain instances may also be tempered by the knowledge of these additional consequences to those who carry out the policy. The casualty count cannot be known merely because the war, or other violent policy, has ended. The casualty count is yet to be ascertained. It will be borne by the society for years to come. Conclusion As is common in theoretical discussions of this sort, there are many interconnections between the implications. For example, the peculiar therapy and treatment needs of those with PITS can also be a public policy implication, as it becomes necessary for the society to provide for those who need the therapy. Reconciliation eﬀorts and violence prevention are also matters of public policy, and therapy and treatment are connected to both. Understanding the biological aspects is crucial for treatment, and indicates something about the human spirit and how it was created to be. Diﬀering religious orientations will also do diﬀerent things with this information. Jews, Christians, Muslims, Hindus, Buddhists, and members of other religions may come up with diﬀering commentaries on what this concept might mean. Because the well-deﬁned concept of PTSD is relatively new, and that of PITS is even more recent, what that commentary will be has yet to unfold. References American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (text revision.). Washington, DC: Author. Bandura, A., Barbanelli, C., Caprara, G. V., & Pastorelli, C. (1996). Mechanisms of moral disengagement in the exercise of moral agency. Journal of Personality and Social Psychology, 71, 364–374. Blass, T. (2000). Obedience to authority: Current perspectives on the Milgram paradigm. Mahwah, NJ: Lawrence Erlbaum Associates. Bremner, J. D., Randall, P., Scott, T. M., Bronen, R. A., Seibyl, J. P., Southwick, S. M., et al. (1995). MRI-based measurement of hippocampal volume in patients with combat-related posttraumatic stress disorder. American Journal of Psychiatry, 152, 973–981.

38

rachel m. macnair

Brennan, W. (1995). Dehumanizing the vulnerable: When word games take lives. Chicago: Loyola University Press. Conway, L. G., Suedfeld, P. & Tetlock, P. E. (2001). Integrative complexity and political decisions that lead to war or peace. In Christie, D. J., Wagner, R. V. & Winter, D. D. (Eds.). Peace, conﬂict, and violence: Peace psychology for the 21st century. Upper Saddle River, NJ: Prentice-Hall. Dateline NBC. (2000). NYPD Blues (transcript). December 26, 2000. Livingston, NJ: Burrelle’s Information Services. Foa, E. B., & Meadows, E. A. (1997). Psychosocial treatments for posttraumatic stress disorder: A critical review. Annual Review of Psychology, 48, 449–480. Gaudium et Spes [Constitution of the Church in the Modern World]. (1965). Rome: Second Vatican Council. Gianelli, D. M. (1993). Abortion providers share inner conﬂicts. American Medical News, July 12. Glover, H. (1985). Guilt and aggression in Vietnam veterans. American Journal of Social Psychiatry, 1, 15–18. —— (1988). Four syndromes of post-traumatic stress disorder: Stressors and conﬂicts of the traumatized with Special Focus on the Vietnam combat veteran. Journal of Traumatic Stress, 1, 57–78. Green, B. L. (1990). Deﬁning trauma: Terminology and generic stressor dimensions. Journal of Applied Social Psychology, 20, 1632–1642. Grossman, D. (1995). On killing: The psychological cost of learning to kill in war and society. Boston: Little, Brown and Company. Haley, S. A. (1974). When the patient reports atrocities. Archives of General Psychiatry, 30, 191–196. Hall, H. V., & Hall, F. L. (1987). Post-traumatic stress disorder as a legal defense in criminal trials. American Journal of Forensic Psychology, 5, 45–53. Hendin, H., & Haas, A. P. (1985). Posttraumatic stress disorders in veterans of early America. Psychohistory Review, 12, 25–30. Hodge, J. E. (1997). Addiction to violence. In J. E. Hodge, M. McMurran, C. R. Hollin (Eds.). Addicted to crime? New York: John Wiley & Sons. Johnson, E. C. (1960) Jane Addams: A centennial reader. New York: The Macmillan Company. Koepp, M. J., Gunn, R. N., Lawrence, A. D., Cunningham, V. J., Dagher, A., Jones, T., et al. (1998). Evidence for striatal dopamine release during a video game. Nature, 393, 266–268. Kruppa, I. (1991). The perpetrator suﬀers too. The Psychologist, 4, 401–403. Kruppa, I., Hickey, N., & Hubbard, C. (1995). The prevalence of posttraumatic stress disorder in a special hospital population of legal psychopaths. Psychology, Crime & Law, 2, 131–141. Kulka, R. A., Schlenger, W. E., Fairbank, J. A., Hough, R. L., Jordan, B. K., Marmar, C. R., et al. (1990). Trauma and the Vietnam war generation: Report on the ﬁndings from the National Vietnam Veterans Readjustment Study. New York: Brunner/ Mazel. Laufer, R. S., Gallops, M. S., & Frey-Wouters, E. (1984). War stress and trauma: The Vietnam veteran experience. Journal of Health and Social Behavior, 25, 65–85. Lifton, R. J. (1990). Adult dreaming: Frontiers of form. In R. A. Neminoﬀ & C. A. Colarusso (Eds.), New dimensions in adult development (pp. 419–442). New York: Basic Books. MacNair, R. M. (2002). Perpetration-induced traumatic stress: The psychological consequences of killing. Westport, CT: Praeger Publishers. —— (2003). The psychology of peace: An introduction. Westport, CT: Praeger Publishers. Milgram, S. (1974). Obedience to authority: An experimental view. New York: Harper & Row.

killing as trauma

39

Nadelson, T. N. (1992). Attachment to killing. Journal of the American Academy of Psychoanalysis, 20, 130–141. Newberg, A., D’Aquili, E., & Rause, V. (2001). Why God won’t go away. New York: Ballantine Books. Pitman, R. K., Altman, B., Greenwald, E., Longpre, R. E., Macklin, M. L., Poire, R. E., et al. (1991). Psychiatric complications during ﬂooding therapy for posttraumatic stress disorder. Journal of Clinical Psychiatry, 52, 17–20. Roe, K. M. (1989). Private troubles and public issues: Providing abortion amid competing deﬁnitions. Social Science and Medicine, 29, 1191–1198. Shatan, C. (1978). Stress disorders among Vietnam veterans: The emotional context of combat continues. In C. R. Figley (Ed.), Stress disorders among Vietnam veterans: Theory, research, and treatment. New York: Brunner/Mazel. Silva, J. A., Derecho, D. V., Leong, G. B., Weinstock, R., & Ferrari, M. M. (2001). A classiﬁcation of psychological factors leading to violent behavior in posttraumatic stress disorder. Journal of Forensic Sciences, 46, 309–316. Solursh, L. (1988). Combat addiction: Posttraumatic stress disorder re-explored. Psychiatric Journal of the University of Ottawa, 13, 17–20. Solursh, L. (1989). Combat addiction: Overview of implications in symptom maintenance and treatment planning. Journal of Traumatic Stress, 2, 451–462. Southwick, S. M., Yehuda, R., & Morgan, C. A. (1995). Clinical studies of neurotransmitter alterations in postraumatic stress disorder. In M. J. Friedman, D. S. Charney, A. Y. Deutch (Eds.). Neurobiological and clinical consequences of stress (pp. 335–350). Philadelphia: Lippincott-Raven. Strayer, R. & Ellenhorn, L. (1975). Vietnam veterans: A study exploring adjustment patterns and attitudes. Journal of Social Issues, 31, 81–93. Such-Baer, M. (1974). Professional staﬀ reaction to abortion work. Social Casework, July, 435–441. Tutu, D. M. (1999). No future without forgiveness. New York: Doubleday. van der Kolk, B. A., Greenberg, M., Boyd, H., & Krystal, J. (1985). Inescapable shock, neurotransmitters, and addiction to trauma: Toward a psychobiology of posttraumatic stress. Biological Psychiatry, 20, 314–325. Wikler, N. (1980). Hidden injuries of war. In C. R. Figley & S. Leventman, Strangers at home: Vietnam veterans since the war. Philadelphia: Brunner/Mazel, pp. 87–106. World Health Organization. (1992). International statistical classiﬁcation of diseases and related health problems (10th revision). Geneva, Switzerland. Yager, T., Laufer, R., & Gallops M. (1984). Some problems associated with war experience in men of the Vietnam generation. Archives of General Psychiatry, 41, 327–333.

THE IMPACT OF CHANGING MARITAL STATUS ON RELIGIOUS ATTENDANCE IN AUSTRALIA John Malcolm Armstrong Abstract The period 1976 to 1996 was a period of rapid transition for religious institutions which was marked by decline in attendance at religious services. This paper addresses the issues related to cohabitation of couples prior to marriage and its aﬀect on patterns of religious attendance at Christian churches in Australia. In particular it analyzes how patterns of partnering have changed during this period by comparing patterns of religious attendance of couples in the Australian Family Formation Project, a longitudinal study conducted 1981–1991 by the Australian Institute of Family Studies in Melbourne, Australia.

One of the most signiﬁcant demographic changes to occur in Australia during the period between 1966 and 1996 was the increase in cohabitation associated with delay or postponement of marriage. Two social factors seem to be closely associated with these phenomena: the Family Law Act was passed in 1975 which permitted couples to more easily divorce their current partner, and people in younger age groups started in the 1970’s to enter into cohabiting relationships prior to marriage. This suggests that there was a profound change in the way people related to each other, but also how they related to Christian religious institutions which formed the nexus between marriage, church attendance and family values. Keyes in particular has noted how the decline in this nexus has caused a change in civil society where there appears to be less concern for others whereas “Religious attendance was a relatively robust mediator of the relationship between marriage and social civility (Keyes, 2002, p. 407).” A change in patterns of partnering suggests a weakening of the inﬂuence of religious institutions in providing an environment in which people could express the signiﬁcance of these relationships to themselves and to society. Garﬁnkel and Bradford Wilcox also point to how couples who distance themselves from a religious congregation are more likely to experience social isolation and have less access to institutions which can support them in raising a family (Garﬁnkel, I. and Bradford Wilcox, W., 2004). Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

42

john malcolm armstrong

This decline in the nexus between marriage and religious attendance has led to various theories being advanced about the causal relationship between the decline in religious inﬂuence of organized religion and a greater privatization of belief and morality. Peterson and Donnenwerth have argued that this has caused people to view their religious commitments in individualistic terms rather than in terms of obligation and institutional loyalty, people are looking for meaning rather than rules to control their lives (Peterson and Donnenwerth, 1997, p. 1072). This weakening of traditional religious values has led Christians to rely more on secular rather than religious values in forming their opinion about such issues as the appropriateness of premarital sex. This has a signiﬁcant eﬀect on the levels of a person’s attendance at a religious service as Inskeep points out, drawing from data from the 1986–1996 General Social Survey, that people who have sex before marriage are less likely to attend church on a regular basis (Inskeep, 2002, p. 3). Brooks would add that the religious inﬂuence on couples in forming family would naturally decline if there are lower levels of attendance and this would weaken the normative expectations of family behaviour and social support leading to family decline (Brooks, 2002, p. 193). Given that these changes in marital status have taken place, it has led to corresponding changes in patterns of religious attendance. Most religious institutions would believe that sexual relationships between men and women should occur within marriage and that they would discourage other forms of relationships outside of marriage such as cohabitating relationships. Religious institutions would consider that most relationships between men and women should be monogamous and lifelong. The institution may well also apply sanctions that would prevent people exhibiting such behaviours from participating fully in the life of the religious group. As a consequence, these sanctions may well have an eﬀect on a person’s pattern of attendance if they are living in a cohabiting relationship. However, against this background of religious institutions promoting the importance of marriage, greater proportions of young people remained unmarried for longer periods of time in 1996 when compared to 1976 (Australian Bureau of Statistics, Marriages, 1996). This delay in marriage is of particular signiﬁcance to this study of religious attendance, because of the longer period of time that a person might not have formal contact with a religious institution between leaving school and entering into a married relationship. This in itself

marital status and religious attendance

43

may indicate a weakening in associations between attendance and religious aﬃliation. This would be more signiﬁcant for those who attended schools aﬃliated to religious institutions. Second, there is a greater likelihood that those who are unmarried will experience a cohabiting relationship prior to marriage. This can be evidenced by the fact that over the period 1976 to 1992 there was a higher proportion of people who cohabited prior to marriage (Table 1). Table 1. Percentage of currently married persons who cohabited prior to marriage (In percentages) 1976 Cohabited prior to marriage Did not cohabit prior to marriage

Year of Marriage 1986 1992

23.0 77.0

45.6 54.4

56.3 43.7

Source: Australian Bureau of Statistics, Marriages 1993

This pattern of cohabitation seemed to be highlighted by a higher proportion of people in younger age groups entering into cohabiting relationships prior to marriage. Table 2 shows that a signiﬁcant number of younger people in the 15–24 age group who were partnered were living in cohabiting relationships. What is signiﬁcant to this study is the level of cohabitation shown by young people aﬃliated to Anglican, Presbyterian, Catholic and Uniting religious groups. The fact that a greater proportion of the 15–24 age group are co-habiting when compared to the older age groups suggests a certain diﬀerence in patterns of behaviour from that which would be taught by each of these religious institutions. This is especially signiﬁcant when it is compared with those who have no religion who have the highest levels of cohabitation among each of the age groups. Similarly, it is signiﬁcant to notice that the Pentecostal, Islamic and Orthodox religious groups have the lowest levels of cohabiting relationships in the 15–24 age group. These diﬀerences may indicate not only the cultural diﬀerences which exist in the Islamic and Orthodox groups, but also the strong connection that the Pentecostal churches retain between aﬃliation and practice.

john malcolm armstrong

44

Table 2. Proportion of population aged 15–44 in couple relationships who were in cohabiting relationships in 1991 (In percentages) Religious Group No Religion Anglican Presbyterian and Reformed Salvation Army Catholic Uniting Church Lutheran Judaism Baptist Seventh Day Adventist Buddhism Churches of Christ Hinduism Pentecostal Orthodox Islamic Total Australia

15–24

Age Groups 25–34

35–44

58.0 47.8 45.2 38.1 37.8 37.6 36.8 35.0 26.8 23.9 22.1 21.0 12.3 10.7 9.0 4.9 41.3

22.1 14.7 14.4 11.9 11.3 9.3 11.9 9.8 7.5 7.7 7.5 6.6 4.7 3.0 4.0 2.9 13.3

11.4 6.8 6.9 6.4 5.2 4.3 5.7 4.6 3.8 3.6 4.3 3.7 2.5 1.7 2.1 1.8 6.4

Source: Australian Bureau of Statistics, Social Trends 1994, p. 188

The Australian Bureau of Statistics 1994 Social Trends also showed in table 3 that people who are married to their partner are more likely to have the same religious identity as their partner than those living in a cohabiting relationship. This would suggest that the sharing of aﬃliation may be a strong determining factor in how people form relationships with each other. As a consequence, the pattern of a person’s religious attendance might also be aﬀected by the nature of their relationship. Table 3. 1991 Proportion of partners with same aﬃliation (In percentages) Religious Group Catholic Anglican Presbyterian and Reformed Uniting Church Orthodox Baptist

Marital Status 67.9 65.3 45.9 64.0 87.3 66.7

Married Co-Habiting 39.8 47.9 19.5 28.2 39.5 19.3

marital status and religious attendance

45

Table 3. (cont.) Religious Group Churches of Christ Lutheran Pentecostal Salvation Army Seventh Day Adventist Total Australia

Marital Status 73.1 59.4 88.5 58.8 80.3 68.9

Married Cohabiting 25.6 19.8 45.0 13.2 32.2 49.0

Source: Australian Bureau of Statistics Social Trends 1994, p. 188

The signiﬁcance of the establishment of cohabiting relationships among young people can be expected to have an eﬀect on their patterns of religious attendance. Similarly, it could also be expected that diﬀerences in religious aﬃliation may aﬀect patterns of religious attendance. What can be assumed from other studies are that marriage and religious attendance are mutually supportive of each other and that partners with similar religious belief are more likely to attend religious services. However, what was not known prior to this research was the level of eﬀect that cohabitation and changes to religious aﬃliation might have on patterns of attendance at religious services. The longitudinal data provided by the Australian Family Formation Project (AFFP) between 1981 and 1991 made it possible to examine and compare the transition of marital status and also the aﬀect of changes in religious aﬃliation on the level of religious attendance. The longitudinal nature of the AFFP data used made it possible to compare the attendance patterns of those who changed their marital status with those who did not. Table 4 shows that those who were married or never married in 1981 and had the same marital status in 1991 maintained similar levels of attendance in both periods, with married men having higher levels of attendance than men who had never married. Among females, the reverse can be seen with the attendance of women who had never been married being a little higher in 1981 than it was for married women. In 1991, the levels of attendance of married women and women who had never been married were almost the same. This can also be seen by the group of men and women who move from being never married to married retaining very similar patterns of attendance. It appears that in this case it could be assumed that marriage has the eﬀect of maintaining attendance patterns already established (Armstrong, 2001).

46

john malcolm armstrong

Table 4. Comparison of weekly attendance levels in 1981 and 1991 AFFP Study based on change in marital status of those belonging to Christian Groups Marital Status 1981–1991

Never Married Married Never Married-Married Never Married-Cohabit

Male Female N Stage 1 Stage 2 Stage 1 Stage 2 Males Females % % % % 16.8 18.4 21.8 11.8

18.7 22.3 22.2 7.1

28.1 24.0 23.0 14.1

24.6 25.3 20.0 12.5

80 219 163 57

61 397 143 39

However, it is signiﬁcant to note that attendance declines or remains distinctly lower among those who enter into a cohabiting relationship than among married and never married people. Here it can be seen that the level of attendance is lower in 1981 when compared with the group who were never married for both periods. This would suggest that they already have a pattern of lower attendance prior to their decision to enter into a cohabiting relationship. This suggests that this group has already started to distance themselves from patterns of religious inﬂuence which would endorse a married relationship. It can also be observed that the attendance of males and females declines moderately for people who cohabit in the period between 1981 and 1991. This decline is greater among males than among females. This would indicate that where young people enter into a cohabiting relationship there is an adverse eﬀect on levels of attendance. Therefore, it can be assumed that cohabitation does have a signiﬁcant impact in distancing a person from the inﬂuence of a religious institution. When a person in that relationship has a diﬀerent religious aﬃliation from his/her partner, this may also lower the level of religious attendance (Armstrong, 2001). Table 5 compares attendance in 1981 and 1991 based on the compatibility of a respondent’s aﬃliation in 1981 with that of their partner in 1981 and 1991. This examines the changing patterns of attendance based on the similarity or diﬀerence in a couple’s religious aﬃliation and what eﬀect a change in the partner’s aﬃliation had on the level of attendance of the respondent. What can be seen in table 5 is that among males and females who retained the same aﬃliation as their partner in 1981 and 1991 there is not much diﬀerence in attendance. When the group of males and females who had diﬀerent patterns of aﬃliation in 1981, but

marital status and religious attendance

47

where the partner’s aﬃliation in 1991 was the same as the respondent’s, there are signs that this had the eﬀect of increasing a respondent’s level of attendance. This increase is greater among males than females, but has the result of producing similar levels of attendance for both sexes. Table 5. Weekly attendance levels of respondents in 1981 and 1991 based on a comparison of a respondent’s religious aﬃliation in 1981 and its similarity with partner’s religious aﬃliation in 1981 and 1991 Attendance of Respondent Males Females 1981 1991 1981 1991 Religion of respondent in 1981 and 1991 and religion of partner in 1991 are all the same Religion of respondent in both 1981 and 1991 is diﬀerent from religion of partner in 1991 Religion of partner and respondent are the same in 1991 but respondent’s religion diﬀerent in 1981 Religion of partner and respondent are diﬀerent in 1991 1981 but respondent’s religion in was the same as partner’s religion in 1991 N

N

25.8

27.6

32.2

33.0

256

7.8

9.1

14.0

13.5

268

9.6

29.8

23.8

30.5

83

21.5

12.4

24.5

24.6

26

218

415

633

When this situation is reversed and the partner’s religious aﬃliation in 1991 was similar to the respondent’s aﬃliation in 1981 but diﬀerent in 1991 the level of attendance among males drops to similar levels as those males who had diﬀerent aﬃliations from their partners in 1981 and those males who had diﬀerent aﬃliations than their partners in both periods. However, among females where the partner’s aﬃliation in 1991 is something diﬀerent from the respondent but was the same in 1981, this does not lower the attendance of the respondent. Also where females partner’s aﬃliation diﬀers from that of the respondents in both periods, the level of attendance remains at similar low levels. This comparison would indicate that a woman’s pattern of attendance is not inﬂuenced by the religious aﬃliation of her partner.

48

john malcolm armstrong

However, a man’s pattern of attendance is inﬂuenced by the religious aﬃliation of his partner. Where the religious aﬃliation is the same, there are indications that both men and women have very similar patterns of attendance. However, where religious aﬃliations are diﬀerent, men’s attendance will be much lower. This again indicates that men’s attendance is more inﬂuenced by the aﬃliation of their partner. It may also indicate that women have a greater inﬂuence in maintaining higher levels of attendance where aﬃliations of a partner are similar (Armstrong, 2001, p. 152). What is signiﬁcant for couples who cohabit is that their level of attendance may be lowered by two signiﬁcant factors, the ﬁrst by the fact that they have experienced a level of social isolation from the institution which views marriage as the preferred form of partnering and second by the fact that couples in cohabiting relationships are less likely to share similar religious beliefs. These two factors seem to indicate that the isolation for cohabiting couples is more likely to be twofold: The ﬁrst, that they are socially isolated from the normal form of partnering approved by a religious institution, and second, by the fact that there is less likely to be a common sharing of belief which would link them to a religious institution. Summary The importance of these ﬁndings gives substance to explaining the decline in patterns of religious attendance especially among young people. While much eﬀort is put forward by religious institutions in addressing the declining levels of attendance, little attention seems to be paid to the important nexus that exists between marriage and religious attendance. Therefore, various causes are put forward which blame secularism and materialism for a decline in religious belief and attendance when it would appear that changes in partnering and social isolation from religious institutions have a much bigger inﬂuence in determining a person’s involvement with a religious institution. The reality of a greater level of cohabitation among young people provide some important challenges for religious institutions in providing an environment which acknowledges this social reality and which aids in forming relationships which are life-giving to the couple and society. The idea that marriage can be stressed as the ideal without acknowledging the existence of these cohabiting rela-

marital status and religious attendance

49

tionships prior to marriage would ignore the changing ways that people seek to form and create lives together. For religious institutions in particular it is a matter of looking at how they seek to inﬂuence young people in the formation of life-giving relationships. In an age where many young people are sexually and relationally literate, it also needs to be shown how they can also become sacramentally literate in forming relationships which have meaning for them and for society. The rebuilding of this nexus between preparation for marriage and the support network that religious institutions provide would strengthen both the couple and society. References Armstrong, J. (2001). Religious attendance and aﬃliation patterns in Australia 1966–1996. The dichotomy of religious identity and practice, PhD. Thesis, Department of Demography and Sociology, Research School of Social Sciences, Australian National University, Chapter 7. Australian Bureau of Statistics, (1996). Marriages 1996, Canberra. Australian Institute of Family Studies, (1984). Family Formation Study-Stage 1, 1980–81– Study Description, Social Science Data Archives, Australian National University, Canberra. Brooks, C. (2002). Religious Inﬂuence and the Politics of Family Decline: Trends, Sources and U.S. Political Behavior. American Sociological Review, 67, Issue 2, 191–211. Castles, I. (1994). Australian Social Trends 1994, Australian Bureau of Statistics, Canberra. Edgar, D. (1985). Family Formation Study, Stage 1, 1980–81, [machine readable data File]. Principal Investigator D. E. Edgar, Institute of Family Studies; data collected by Reark Research Pty. Ltd. Melbourne: Institute of Family Studies [producer] 1982. Canberra: Social Science Data Archives, The Australian National University [distributor], 1985. 2 data ﬁles (106848 physical records) + accompanying user’s guide 105 p). Garﬁnkel, I. and Bradford Wilcox W. (2004). Religion, Ideology and Marriage among Fragile Families http://crcw.princeton.edu/CRCW/papers/religionandfamily/ garﬁnkelpresentation.pdf accessed 13th April 2004. Inskeep, K. (2002). An Analysis of the Church attendance in the general population. Department for Research and Evaluation, Evangelical Lutheran Church in America. Keyes, C. (2002). Social civility in the United States. Sociological Inquiry, 72, 393–408. Peterson, L. and Donnenwerth, G. (1997). Secularization and the inﬂuence of religion on beliefs about premarital sex. Social Forces, 75, 1071–1089.

john malcolm armstrong

50

Appendix In order to compare the data between surveys it was necessary to arrive at a measure of average weekly attendance that can be applied to the data across surveys. To arrive at such a measure, it is necessary to ascribe values to each of the responses made to the question on religious attendance. It is proposed that the measure will be based on the following assumed frequency of attendance: Survey Response Never Once a year Twice a year Every two months Once a month Twice a month Three times a month Every week

Assumed Annual Frequency 0 1 2 6 12 24 36 48

The level of average weekly attendance is applied to each of the data-sets by the application of the following formula. ∑(Vp*F)/T Tp Where Vp F T Tp

= = = =

Population for a particular level of attendance Assumed frequency of attendance on a principal day of worship 52 Principal days of worship in a year Total population of religious group in the survey

The beneﬁt of this measure was that it not only enabled comparison between surveys which used alternate responses to the attendance question but it also enables comparison with data collected by religious organisations. Thus the actual possible attendance on a particular day of worship is measured rather than the attendance level of those who are attending each week. The measure adopted is thus a more reliable guide as to levels of attendance on any particular day of worship.

RELIGIOUS SWITCHING: DOES PARENTS’ EDUCATION MATTER? Amy Adamczyk* Abstract How are switching to more liberal religious preferences and apostasy aﬀected by educational attainment? Previous research has focused on the relationship between religious adherence and college attendance. In this paper, I extend this work by looking at whether diﬀerences between parent and child educational attainment aﬀect later religious devotion among respondents who at one time were Protestant. Using data from the 1998 General Social Survey, I test whether conservative religious commitment among college-educated children with less educated parents is diﬀerent from commitment among the college-educated children of educated parents. Findings reveal that respondents who have earned an associate’s degree, but have less educated parents are not more likely than people with college-educated parents to switch from a conservative faith to a more liberal one. In addition, this research shows that diﬀerences between parental and child education are not signiﬁcantly related to religious apostasy.

In the year 2000, over 60% of high school graduates in the United States went to college. At college, people are exposed to a variety of new perspectives (Aronowitz 2000). Most Americans grow up in Catholic, Protestant, or Jewish homes in which few are exposed to alternative religions or meet people who have disaﬃliated from JudeoChristian faiths. At college, they will be released from the sacred canopies of their parents’ homes and are likely to encounter these other views (Wuthnow 1999). In addition, if their parents did not attend college, they may be the ﬁrst in their families to experience the intellectual challenges that the college classroom provides, which may cause them to question their religious beliefs (Wuthnow 1998). Many studies have examined the eﬀect of college attendance on religious belief (Feigelman, Gorman, and Varacalli 1992; Roozen 1980; Caplowitz and Sherrow 1977). There is also much research on the inﬂuence of parental religiosity on college attendance (Darnell and Sherkat 1993; Lehrer 1999). This study combines aspects of

* Authors Note: I would like to acknowledge Jacob Felson for his editorial help in putting this paper together. Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

52

amy adamczyk

these prior studies in a novel way, proposing that the eﬀect of college education on faith might depend on parental education. Many religious, college-educated parents may have found a way to reconcile the values of religion with secular education, presenting each to their children as compatible with the other. Less educated religious parents may be more likely to see aspects of college education as hostile to faith. Thus, socialization in less-educated, religious homes may not prepare children for challenges to their faith, which they experience in college. Thus, to understand the inﬂuence of education on faith, we should ascertain the inﬂuence of diﬀerences between parents’ and children’s educational attainment. Are people who went further in school than their parents more likely to switch denominations, or disavow religion altogether? The present study addresses this question. Speciﬁcally, it uses data from the 1998 General Social Survey (GSS) to assess the inﬂuence of educational diﬀerences between respondents and their parents on change in religious preferences. Sociologists have long believed that parents have an important inﬂuence on the attitudes and values of their children. Traditional theorists like Freud (1933, 1928) and Erikson (1950) have posited that the principal socializing agent of childhood are parents, and they remain inﬂuential long after children leave home. This idea is especially true in terms of religion: the strongest predictor of religious aﬃliation is parental religious aﬃliation (Acock and Bengtson 1978; Hayes and Pittelkow 1993; Myers 1996; Pearce and Axinn 1998; Okagaki and Bevis 1999). Parents can have a strong inﬂuence on their children’s preferences for various reasons. Their strong inﬂuence may be attributed to successful parental socialization or it may have to do with intergenerational transmission of class, race, religious aﬃliation, and other structural life experiences (Glass, Bengtson, and Dunham 1986; Myers 1996). Children inherit parents’ social status, as well as their religious values. Mothers and fathers transmit their values to their children partly through psychosocial conditioning and partially through social structural characteristics. A number of studies have shown that social position is related to religious preferences (Demerath 1965; Roof and McKinney 1987; Stark and Glock 1968). Although there is much socio-economic variation within denominations, researchers still ﬁnd that denominations with high mean levels of belief tend to have lower mean levels of

religious switching

53

education (National Bureau of Economics Research 2001). As they raise their children, parents in high-belief Protestant denominations may pass on their (typically) high levels of Christian belief and their membership in a lower-status church. By contrast, parents in lowbelief denominations, like the Episcopal Church, may pass on (typically) low levels of belief and membership in a high-status church. The cultural environment at church may be a function of its social status and mean levels of belief. What happens to children who leave lower-status, high-belief church environments to attend college? Do they remain in lower status denominations, or switch to denominations with higher status? Through exposure to new perspectives and development of critical thinking, education may threaten faith. College students may be encouraged by professors to base their opinions on demonstrable evidence. College students may apply this empiricist mindset to religious beliefs, thereby undermining them. In addition, Sherkat and Wilson (1995) argue that members of the upper educational strata prefer an intellectual ethos that emphasizes socially aﬃrming elements of religious ideology and accentuates the value of individual autonomy over respect for authority (1995). Conservative Protestants may oppose this intellectual ethos, since they tend to favor more obedience in raising children (Ellison and Sherkat 1993a). In contrast, submission to authority and faithful acceptance of Biblical “truths” attains primary signiﬁcance in the lives of many conservative Protestants. Thus, many conservative Protestants believe that secular education undermines divine authority by promoting “humanism” and denigrating faith (Darnell and Sherkat 1993). Other research has found that conservative Protestant parents harbor distinctively authoritarian parenting orientations (Ellison, Bartkowski and Segal 1996). Although learning how to be critical may challenge previous religious beliefs, it may also encourage people to consider religious alternatives. In other words, critical thinking is not necessarily antithetical with religious ideas, and it is not clear that religious preferences will necessarily change because of educational attainment. Paradoxically, Hammond and Hunter (1984) found that the religious beliefs of Evangelical college students who attend secular universities were strengthened relative to Evangelicals who attend Christian colleges. Evangelical students at Christian colleges reported fewer opportunities to witness to others outside their faith than Evangelicals attending secular universities. The authors suggested that greater opportunity

54

amy adamczyk

for evangelizing actually strengthened the faith of Evangelicals at secular campuses. The secular context helped strengthen Evangelical beliefs because in-group solidarity was greater when out-groups were present. By contrast, in Christian colleges, where religious beliefs are more homogenous, religion may become less critical as an identity marker. Thus, Evangelical beliefs may be weakened in Evangelical settings. Although there is much research on the relationship between social status and religious background, little research has examined how diﬀerences between parents’ and children’s educational attainment may be related to change in children’s religious preferences. Results from existing research are inconclusive. Sherkat and Wilson (1995) used the 1965 and 1973 waves of the Youth-Parent Socialization Panel Study to examine religious switching and apostasy. They found that upwardly mobile individuals are not more likely than others to join religiously liberal groups. Conversely, in their study of Americans who gave up religion, Feigelman, Gorman, and Varacalli (1992) found that religious disaﬃliation was related to increased levels of education. And in their study on attitude similarity in three generations of families, Glass, Bengtson, and Dunham (1986) found that, although weak, children’s education was signiﬁcant in predicting religious ideological diﬀerences between parents and children. However, they also found other variables (number of children, marital status, labor force status, and income) to be signiﬁcant and stronger predictors than college attendance. In summary, research seems mixed about the relationship between education and change in religious preferences. Education may inﬂuence change in religious preferences because college attendance enhances social status and exposure to critical ideas. However, parents have also been shown to have a very important inﬂuence on children’s religious preferences. If parents have integrated the critical thinking skills, which often comes with a college degree, with their conservative denominational orientation, their children may not be so inclined to disaﬃliate or switch to a more liberal religious preference. To understand the inﬂuence of education, particularly diﬀerences between parents’ and their children’s educational attainment, I ﬁrst examine similarities and diﬀerences in parents’ and children’s educational attainment on oﬀspring change from conservative to more liberal Protestant denominational aﬃliations. Of particular importance for this research is whether the college-educated children of

religious switching

55

uneducated parents are more likely to switch to liberal religious preferences than the college-educated children of educated parents. Second, I compare the inﬂuence of similarities and diﬀerences in parents’ and children’s education on oﬀspring change for individuals who remained Protestant and those who were Protestant but are now apostates (religiously unaﬃliated). Method Participants The data employed for this study were drawn from the 1998 General Social Survey (GSS), a biennial cross-sectional personal interview survey of U.S. households conducted by the National Opinion Research Center (NORC). NORC draws a representative sample of Englishspeaking households, and interviews one person from each. The median length of the interview is about one and one-half hours. Because only one interview is completed in each household, persons living in large households are less likely to be interviewed. Thus, interviews are weighted proportionally to the eligible number of people in each household.1 The full sample included 2,832 cases. However, I examined only the 753 people who were classiﬁed as belonging to a fundamentalist Protestant denomination while growing up (table 1 and 2) and the 1,492 people who said that they belonged to any Protestant denomination while growing up (table 1 and 3). The GSS was chosen because it is a nationally representative survey, which includes questions about religion as a youth and current religion. In addition, the GSS has classiﬁed religious preferences into three diﬀerent categories; conservative, moderate, and liberal. Because it is fairly common for Americans to change denominational aﬃliation, assessing switching between denominations would not be the best assessment of change in religious preferences. As Roof and McKinney (1987) point out, people tend to switch to denominations that are similar to the ones they left in worship style, beliefs, structure, and emphasis. Therefore, it is preferable to assess change between groups like liberals and conservatives, rather than similar Protestant denominations like Church of the Nazarene and Southern Baptist Convention. 1 For more information on the interview weighting procedure see “Appendix A— Sampling Design and Weighting” in Davis, Smith and Marsden (2001).

56

amy adamczyk Measures

Dependent Variables To assess the inﬂuence of diﬀerences in children’s and parents’ college education on change in Protestant religious preferences, two analyses were completed; one that examined apostasy for individuals reared in Protestant families, and one that looked at change from fundamentalism to a more religiously liberal preference within Protestantism. The main variable for assessing change within Protestantism was computed from two variables in the GSS that classiﬁed Christian denominational aﬃliation into conservative, moderate, and liberal categories. These variables were computed by ﬁrst asking respondents, “In what religion were you raised,” and “What is your religious preference?” Both questions had an identical list of responses, which included Protestant, Catholic, Jewish, or none. If the respondent answered that he or she was raised or now prefers Protestant, he or she was asked to name the speciﬁc denomination. The GSS then classiﬁed denominations into three religious preference categories (fundamentalist, moderate, or liberal) using ﬁve diﬀerent techniques: (1) prior classiﬁcations schemes, (2) membership in theologically oriented ecumenical associations, (3) surveys of denominational members, (4) surveys of denominational clergy, and (5) theological beliefs of denominations. (For more information on this classiﬁcation scheme see Smith 1990.) Using this classiﬁcation scheme, the GSS created two variables that measured the fundamentalism or liberalism of the Protestant denomination in which the respondent was raised, and another that measured the fundamentalism or liberalism of respondents’ current denominational preference. For both variables the categories were fundamentalist, moderate, and liberal. Because this study was interested in whether or not respondents change to more liberal religious preferences, the ﬁrst analysis examined switching from a fundamentalist Protestant denomination to a more liberal Protestant preference.2 Respondents who said that they

2 In a separate analysis not shown here, I included people with moderate preferences while growing up in the same category as those who were fundamentalist. I then looked at how discrepancies in parents and children’s educational attainment inﬂuenced change to a more liberal religious preference, either a switch from moderate to liberal, or conservative to moderate or liberal. I did not ﬁnd any diﬀerences in the signiﬁcance levels between this analysis and the one presented in this paper. Also, the coeﬃcients were very similar.

religious switching

57

now adhere to a denomination or religion other than Protestantism were excluded in the ﬁrst analysis. A large proportion of those excluded reported “none” when asked about their current religious preference. For analyzing change in religious preference excluding those people who said “none” is problematic. Parent and child diﬀerences in educational attainment may be related to apostasy. If respondents were coded as Protestant while growing up and are still Protestant or claim no religious aﬃliation, they were included in a second analysis. To assess the inﬂuence of similarities and diﬀerences in parents’ and children’s educational attainment on apostasy or change within Protestantism, I developed a set of dummy variables to be analyzed using logistic regression techniques. I looked at people who remained Protestant aﬃliated and those who switched from Protestantism to no religious aﬃliation. To get a large enough number of apostates I had to include all respondents who were reared in Protestant homes, regardless of the denomination’s level of liberalism. Independent Variables To determine whether respondents changed their religious preferences, I examined similarities and diﬀerences in educational attainment between respondents and their parents. To measure parents’ educational attainment, I dichotomized two variables, which asked respondents to select the highest degree that each parent had received. The response categories were “Less than high school,” “High school,” “Associate/Junior College, Bachelor’s, and Graduate,” “Don’t know,” “No answer.” Because few respondents (15%) had two parents with a college degree, if either parent had received a post-high school degree, the respondents’ parents were coded as having a degree. If the respondent did not know whether a parent had received a degree, I used the available information to categorize the respondent’s parents. If one parent had not received a degree and no information was available on the second parent, I classiﬁed the respondent’s parents as not having a degree. To calculate children’s level of education, I examined answers to the question, what was the highest degree that you received? The response categories were “Less than high school,” “High school,” “Associate/Junior college,” “Bachelor’s,” and “Graduate,” “Don’t know,” “No answer.” The majority of respondents had received a high school degree, but only 32% had received a post-high school

58

amy adamczyk

degree. Because of this distribution and to be consistent with the coding of parents’ education, I dichotomized responses into categories for those who did and did not receive at least a two-year college degree. Using the two dichotomized variables for respondents’ and parents’ education, I developed a set of four dummy variables, which were calculated according to diﬀerences and similarities between the respondent’s education and that of his or her parents. The ﬁrst dummy variable, Neither, included respondents who did not have at least a two-year degree and did not have parents with a degree. The second, child more, included college-educated respondents (at least a two-year degree) with non-college educated parents. Parents more included respondents who did not have a college degree, but had college-educated parents. Finally, both, included respondents who, like their parents, had received at least a two-year college degree. Three of these four dummy variables were entered into the regression equation at one time. When, for example, neither, child more, and parent more were entered into the equation, the reference category was both, the excluded variable. Signiﬁcant diﬀerences between groups were determined by changing the omitted category. In addition to the key independent variables, previous research has suggested that there are some important indicators that might also be related to change in religious preferences. These are age (Fowler 1981), race (Sullins 1993; Roof and McKinney 1987), gender (De Vaus and McAllister 1987), marital status, income (Clydesdale 1997) and region (Kellstedt and Smidt 1991). All of these were included in the analysis. Age was a continuous variable that asked respondents how old they were and ranged from 18 to 89. Some respondents may not be old enough to have completed their degrees. They may be working towards a degree now, but have not been out of high school long enough to have completed it. To make sure that my ﬁndings were not inﬂuenced by my decision to look at all people 18 and over, I re-ran my analysis examining only respondents 20 and over, which would be enough time after high school to receive a two-year degree. The signiﬁcance levels were the same for the two groups, and the coeﬃcients were very similar. Because it increased my sample size, I included all respondents 18 and over. Because 310 of the respondents in the full sample reported not knowing their income or did not answer the question, I created two

religious switching

59

income variables, income and income missing. For income, respondents were asked, “In which of these groups did your total family income, from all sources, fall last year before taxes, that is?” Income had 23 categories that ranged from less than $1,000 to over $110,000. If any of the respondents did not give an answer to this question, they were given the value “0” and included with respondents who answered the income question. I then created a dummy variable, income missing, which distinguished between people who report their income and those who did not give an answer. Both variables were then entered together, which allow me to maintain my sample size while controlling for respondents who did not answer the income question. Race is measured with two dummy variables: black and other race. White is the omitted category. A dummy variable for female was included. Because people who reside in the Bible Belt tend to be more religious than people in other parts of the country (Waters, Heath, and Watson 1995), I included region of interview in the analysis. The GSS divided regions of the nation into nine categories: New England, Mid-Atlantic, East North Central, West North Central, South Atlantic, East North Central, West North Central, South Atlantic, East South Central, West South Central, Mountain, and Paciﬁc. Respondents who resided in the South Atlantic, East South Central and West South Central parts of the United States were categorized as living in the Bible Belt. Descriptive information is given in table 1 for the total sample used in this paper, which consists of respondents who were reared in Protestant families. It also includes descriptive information on individuals who were classiﬁed as adhering to a fundamentalist denomination while growing up, which is a subset of the ﬁrst sample. Procedures Using logistic regression techniques, I regressed change in preferences within Protestantism on diﬀerences and similarities in respondents and parents’ educational attainment. I then included control variables and compared the two models to see if the signiﬁcance levels and the coeﬃcients had changed. After this, I employed logistic regression to examine diﬀerences and similarities in education on change from Protestantism to apostasy. Control variables were then included in the analysis to see if the results remained the same.

amy adamczyk

60

Table 1. Descriptive Statistics for Total Sample and Individuals Reared in Fundamentalist Familiesa Total Sample N = 1492 Percentagesc

Reared in a Fundamentalist Family b n = 763 Percentagesc %

Key Independent Variables Neither (N) Both (B) Parents more (P) Child more (C)

62.7 11.4 08.8 17.1

71.1 06.8 06.7 15.4

Control Variables Income Missing Black Other Female Married Bible Belt

11.4 19.8 1.70 55.5 57.9 46.4

11.3 25.6 2.00 56.0 58.4 55.1

Minimum d

Maximum

Mean

S.D.

Total Sample Age Income

18 0

89 23

46.8 13.9

16.9 6.92

Reared in a Fundamentalist Family Age Income

18 0

89 23

45.7 13.6

16.6 6.81

a

Descriptive information is based on a list wise deletion of cases retained for analysis in the second and third tables. b The 763 respondents in this sample are a subset of the 1,492 in the total sample. c Percentages are given for discrete variables. d For continuous variables the minimum, maximum, means, and standard deviations are given. Neither (N) = Neither child nor parent has a two-year degree Both (B) = Both child and parent have at least two-year degrees Parents degree (P) = Parent has at least two-year degree, child does not Child’s degree (C) = Child has at least two-year degree, parent does not

religious switching

61

Results In Table 2, I present the ﬁndings for the inﬂuence of the diﬀerence between child’s and parents’ educational attainment on change within Protestantism. In the ﬁrst model, only the target variables were entered, and signiﬁcant diﬀerences were found between respondents who had at least a two-year college degree and those who did not. In contrast to uneducated respondents, college-educated respondents were signiﬁcantly more likely to switch to liberal religious preferences over time. In the second model, I included all of the control variables. Like the ﬁrst model, I found that college-educated children with educated parents were more likely to become religiously liberal over time, compared to the uneducated children of uneducated parents. I also found a signiﬁcant diﬀerence between the uneducated respondents of educated parents and the educated children of educated parents. Children who went to college and had parents with a college degree were more likely to become religiously liberal over time than the uneducated children of college-educated parents. Table 2. Change from Fundamentalism to More Liberal Religious Preferences Estimated by Diﬀerences in Parents’ and Respondents’ Education Logistic Regression Coeﬃcients (Odds ratios reported in parentheses)

Key Variables Neither (N) Child more (C) Parents more (P) Both (B) reference category Summary of Signiﬁcant Diﬀerencesa

Control Variables No answer to income

Model 1

Model 2

–1.174*** (.309) –.515 (.598)

–.977** (.376) –.514 (.598

–1.654** (.191) – B>N*** B>P** K>N** not shown K>P* not shown

–1.177* (.308) – B>N** B>P+

.853 (2.346)

amy adamczyk

62 Table 2. (cont.)

Model 1

Model 2

–.842** 763

.060* (1.062) –.927** (.396) –.939 (.391) –.012* (.988) .024*** (1.024) –.468* (.626) .163 (1.177) –2.795*** 763

Income Black Other Female Age Bible belt region Married Constant N

Signiﬁcance +<.10 *<.05, **<.01 ***<.001 a

Signiﬁcant diﬀerences between groups are determined by changing the omitted category. Neither (N) = Neither child or parent have a two-year degree Both (B) = Both child and parent have at least two-year degrees Parents degree (P) = Parent has at least two-year degree, child does not Child’s degree (C) = Child has at least two-year degree, parent does not

In addition to changing preferences within Protestantism, I also looked at the inﬂuence of education on apostasy from Protestantism. Table 3 indicates statistical diﬀerences between the diﬀerent groups. Before adding controls (Model 1), neither of the groups were signiﬁcantly diﬀerent from each other. The second model shows that the control variables are not suppressing this relationship. Table 3. Change from a Protestant Religious Preference to Apostasy Estimated by Diﬀerences in Parents’ and Respondents’ Education Logistic Regression Coeﬃcients (Odds ratios reported in parentheses)

Key Variables Neither (N)

Model 1

Model 2

–.184 (.832)

.096 (1.101)

religious switching

63

Table 3. (cont.) Model 1 Child more (C) Parents more (P) Both (B) reference category Summary of Signiﬁcant Diﬀerencesa Control Variables No answer to income

–.079 (.924) .062 (1.064) — None

Income Black Other Female Age Bible belt region Married Constant N

–2.113*** (.121) 1492

Model 2 .214 (1.293) –.201 (.818) — None .739+ (2.094) .009 (1.009) –.180 (.853) –.986 (.373) –.717*** (.488) –.028*** (.972) –.800*** (.449) –.482* (.617) –.335 (.715) 1492

Signiﬁcance +<.10 *<.05, **<.01 ***<.001 a

Signiﬁcant diﬀerences between groups are determined by changing the omitted category. Neither (N) = Neither child or parent have a two-year degree Both (B) = Both child and parent have at least two-year degrees Parents degree (P) = Parent has at least two-year degree, child does not Children degree (C) = Child has at least two-year degree, parent does not

Discussion The ﬁrst aim of this study was to assess whether college-educated children of uneducated parents were more likely to switch to liberal religious preferences than the college-educated children of educated

64

amy adamczyk

parents. Findings show that there are no signiﬁcant diﬀerences between the two groups. The results presented here suggest that college for the children of uneducated parents does not have any more of an impact on switching than it does for the college-educated children of educated parents. Although parents’ education did not have an inﬂuence on switching for educated children, respondent’s education clearly seems to have some inﬂuence when we compare respondents who went to college to those who did not. The educated children of educated parents were signiﬁcantly more likely to switch to a liberal religious preference than uneducated respondents. Previous research has suggested that religiously liberal preferences are related to educational attainment (Feigelman, Gorman, and Varacalli 1992). The relationship between education and switching is, in part, due to the high correlation between education and income. Previous research has pointed out that members of lower-status groups are socialized to prefer stricter and more conservative religious ideas (Sherkat and Wilson 1995; Ellison and Sherkat 1993a; 1993b). As one’s education increases so also does income and social class. Through interaction with higher status groups, college-educated respondents come to prefer a higher status denomination, which may be more liberal. Before income was included, signiﬁcant diﬀerences were found between those respondents who went to college and those who did not, regardless of parents’ education. When income was included in the model, the odds of switching to a more religiously liberal preference increased by a factor of 1.06. Income is thus suppressing the inﬂuence of diﬀerences between the college-educated children of uneducated parents and respondents who did not receive a degree. We know then that respondents’ education has some role in switching from conservative to more religiously liberal preferences, particularly as education is related to income. However, if the added income that comes with a college education is the only thing driving a switch to more liberal religious preferences, income should eliminate all diﬀerences between educated and uneducated respondents, which was not found. Even with income included, diﬀerences between uneducated respondents and the educated children of educated parents remained. Parental education has some inﬂuence on switching to more religiously liberal preferences, but only when we compare the educated children with educated parents to respondents who did not go to college. For changes within Protestantism this

religious switching

65

ﬁnding is consistent with previous research that children tend to follow the denominational preference in which they were raised (Ozorak 1989; Stark and Finke 2000; Hayes and Pittelkow 1993), but a college degree can make an important diﬀerence (Lehrer 1999, Hammond and Hunter 1984; Hunsberger and Brown 1984). Diﬀerences between the college-educated children of educated parents and respondents without at least a two-year degree may be related to the intellectual ethos of the home environment when parents are not only educated, but also sending their children to college. Further research is needed to determine how exactly the home environment diﬀers for these groups and the way it is related to children’s level of education. Speciﬁcally, the results presented here raise the question of why the college-educated children of parents without a degree are not signiﬁcantly diﬀerent from any of the other groups when the college-educated children of educated parents are signiﬁcantly diﬀerent from the uneducated respondents. The second aim of this study was to determine if diﬀerences in education inﬂuence respondents to switch away from religion. The results presented here suggest that it makes little diﬀerence. The lack of signiﬁcant results in examining change toward apostasy suggests that although college may have an inﬂuence on switching to religiously liberal preferences, it is not related to apostasy. Lack of signiﬁcant ﬁndings for explaining apostasy adds to the debate about secularization. To assess the inﬂuence of secularization, many researchers have examined the inﬂuence of education on people’s religious preferences. On the one hand, some researchers have found that an increase in education is correlated with disaﬃliation (Caplowitz and Sherrow 1977; Hunsberger and Brown 1984; Feigelman, Gorman, and Varacalli 1992; Roozen 1980). Conversely, others have found that upwardly mobile individuals are not more likely than others to join religiously liberal groups (Sherkat and Wilson 1995). The research presented here supports the latter ﬁndings and also corroborates existing literature on general religious switching. As both Roof and McKinney (1987) and Sullins (1993) report, movement within Christianity is common among people in the United States, but individuals typically do not switch from Christianity to another religion or no preference. The ﬁndings presented here show that children who, like their parents, get at least a two-year degree may become more religiously liberal over time, but are not more likely to disaﬃliate than other people.

66

amy adamczyk Conclusion

Because of the prevalence of Protestantism and high levels of college enrollment in the United States, the persistence of religion has been a priority for social scientists of religion. Previous research has shown that both parents’ religious adherence and respondent’s education inﬂuence religious preferences. However, we have not known how diﬀerences in parents’ and children’s educational attainment contribute to Protestant switching and apostasy. Many religious and college-educated parents may have reconciled religious values with secular education, which they can share with their college-bound children. But uneducated parents may see college as a challenge to their children’s religious faith. Because they haven’t attended, they may feel unable to prepare their children for the religious challenges of higher education. The research presented should ease their worries. First-generation college students are no more likely than second, third and fourth-generation college students to quit the faith or switch to more liberal Protestant preferences. Diﬀerences between parents’ and children’s college education do not seem to seriously disrupt Protestant aﬃliation. This study’s mostly null ﬁndings point to the possible strength of parental religious socialization and religious communities. At college, students’ are exposed to a diverse array of perspectives and ideas, which is meant to prepare young people to be good citizens, equipped workers, and informed individuals. However, encountering new ideas may challenge students’ previous perspectives. Beckwith (1985) argues, for example, that historical knowledge of the instability and plurality of religions can erode religious conﬁdence and the anti-religion rhetoric of universities can challenge religious beliefs. It seems that parents and religious communities have been able, in part, to make salient the religious identities of ﬁrst-generation college students in such a way that they cannot easily be shaken by higher education. Another reason for the ﬁndings presented here is that America’s college environments may not be particularly challenging to individuals’ religious identities. The United States is one of the most religious nations in the industrialized world. With over 75% of Americans believing in God, American college students may simply not face much anti-religious sentiment, which suggests why religious beliefs among Americans are quite strong.

religious switching

67

The research presented here only examined changes in Protestant aﬃliations, which tells little about the expression and importance of Protestant identities. Religion is multi-faceted and diﬀerences in parents’ and children’s education may be related to change in religious attendance, saliency of belief, level of prayer, and interaction with other religious individuals. Previous research (Roof 1976; Funk and Willits 1987; National Bureau of Economics Research 2001) has, for example, found that education is related to increases in religious participation and changes in social attitudes. However, researchers do not know how diﬀerences in parents’ and children’s education inﬂuence religious attitudes and participation, which would further clarify our understanding of college’s inﬂuence. Although making an important contribution to what we already know about religion and education, this study has some important limitations. First, the research presented here does not use longitudinal data, which would give a clearer indication of people’s preferences when they were still living with their parents. In addition, the data did not contain information on the type of college respondents attended. Previous research (Hammond and Hunter 1984) has indicated that this can have an important inﬂuence on change in preferences. Likewise, many individuals who were reared in fundamentalist homes may have attended fundamentalist primary and secondary schools. Unfortunately, the GSS does not have information on the type of college or school attended. This sort of information would help us determine how much religious oriented schools and colleges contribute to the maintenance of religious identities. Finally, classiﬁcation of respondents into fundamentalist denominational categories was not based on self-identiﬁcation. Rather, the research presented here relies on the General Social Survey’s classiﬁcation of respondents into fundamentalist and non-fundamentalist Protestant categories, which is based on respondents’ denominational aﬃliation. However, as Kellsted and Smidt (1991) point out, classiﬁcation systems such as these may not be consistent with where people would place themselves. If, for example, Missouri Synod Lutherans, which are classiﬁed as a fundamentalist with the GSS system, were asked whether they would consider themselves fundamentalist, they may disagree. Denominational-based classiﬁcation systems may not capture the respondent’s Protestant orientation. Additionally, the educational status of people within groups like

68

amy adamczyk

fundamentalists varies considerably, which may interfere with signiﬁcant ﬁndings that use these variables. Unfortunately, the GSS has not, except for a couple of years, included any current or previous religious self-identiﬁcation measures. In addition to self-identiﬁcation, future research would beneﬁt from information about the respondent’s perspective on religious doctrine (e.g. biblical literalism), which tends to divide denominations. It would be particularly helpful if additional questions were asked while respondents were still living at home and then again after college graduation or when established in a career so that a change score could be calculated. Because the research presented here measures religious switching only by change in denominational classiﬁcation and not the respondent’s stated religious beliefs, it is a conservative test. Basic to religion is maintaining a worldview that governs the meaning of life, morality, and ethics prescribing and proscribing certain types of behaviors. For the religious person, some of these beliefs and behaviors must be maintained by faith without empirical evidence. Given the popularity of Judeo-Christian religions in the United States, it is common to have little interaction with people of other faiths, or with those who do not maintain a particular preference, especially for young people still living in their parents’ home. However, college oﬀers not only exposure to critical thinking, but also interaction with people from other denominational backgrounds. With more people attending college, the inﬂuence of this experience on religious beliefs is of particular importance for assessing religious trends in American culture. The ﬁndings presented here add to this by showing that diﬀerences between parents’ and children’s educational attainment for respondents with a college degree does not inﬂuence apostasy or change from conservative to more liberal Protestant preferences. References Acock, A. C., & Bengtson, V. L. (1978). On the relative inﬂuence of mothers and fathers: A covariance analysis of political and religious socialization. Journal of Marriage and the Family 40: 519–530. Aronowitz, S. (2000). The knowledge factory: Dismantling the corporate university and creating true higher learning. Boston: Beacon Press. Beckwith, B. P. (1985). The decline of U.S. religious faith, 1912–1984. Palo Alto, CA: B. P. Beckwith. Caplowitz, D., & Sherrow, F. (1977). The religious dropouts: Apostasy among college graduates. Beverly Hills, CA: Sage.

religious switching

69

Clydesdale, T. T. (1997). Family behaviors among early U.S. baby boomers: Exploring the eﬀects of religion and income change, 1965–1982. Social Forces 76: 605–635. Darnell, A., & Sherkat, D. E. (1993). The impact of Protestant fundamentalism on educational attainment. American Sociological Review 62: 306–315. Davis, J. A., Smith, T. W., & Marsden, P. V. (2001). General Social Surveys, 1972–2000: Cumulative Codebook. Chicago: National Opinion Research Center. Demerath, N. J., III. (1965). Social class in American Protestantism. Chicago: Rand McNally. De Vaus, D., & McAllister, I. (1987). Gender diﬀerences in religion: A test of the structural location theory. American Sociological Review 52: 472–480. Ellison, C. G., & Sherkat, D. E. (1993). Obedience and autonomy: Religion and parental values reconsidered. Journal for the Scientiﬁc Study of Religion 32: 313–329. —— (1993 b). Conservative Protestantism and support for corporal punishment. American Sociological Review 58: 131–144. Ellison, C. G., Bartkowski, J. P., & Segal, M. L. (1996). Conservative Protestantism and the parental use of corporal punishment. Social Forces 74: 1003–1029. Erikson, E. (1968). Identity youth and crisis. New York: Norton. —— (1950). Childhood and society. New York: Norton. Feigelman, W., Gorman, B. S., & Varacalli, J. A. (1992). Americans who give up religion. Social Science Research 76: 138–144. Fowler. J. W. (1981). Stages of faith: The psychology of human development and the quest for meaning. New York: HarperCollins. Freud, S. (1933). New introductory lectures in psychoanalysis. New York: Norton. —— 1928. The future of an illusion. (W. D. Robson-Scott, Trans.). Hogarth Press: London. Funk, R. B. & Willits, F. K. (1987). College attendance and attitudes change: A panel study, 1970–1981. Sociology of Education 60: 224–231. Glass, J., Bengtson, V. L., & Dunham, C. C. (1986). Attitude similarity in threegeneration families: Socialization, status inheritance, or reciprocal inﬂuence? American Sociological Review 51: 685–698. Hammond, P. E. & Hunter, J. D. (1984). On maintaining plausibility: The worldview of Evangelical college students. Journal for the Scientiﬁc Study of Religion 23: 221–238. Hayes, B. C. & Pittelkow, Y. (1993). Religious belief, transmission, and the family: An Australian study. Journal of Marriage and the Family 55: 755–766. Hunsberger, B. & Brown, L. B. (1984). Religious socialization, apostasy, and the impact of family background. Journal for the Scientiﬁc Study of Religion 23: 239–251. Kellstedt, L. & Smidt, C. (1991). Measuring fundamentalism: An analysis of diﬀerent operational strategies. Journal for the Scientiﬁc Study of Religion 30: 259–278. Lehrer, E. L. (1999). Religion as a determinant of educational attainment: An economic perspective. Social Science Research 28: 358–379. Myers, S. M. (1996). An interactive model of religiosity inheritance: The importance of family context. American Sociological Review 61: 858–866. National Bureau of Economics Research. (2001). Education and religion (Working Paper No. w8080.) Cambridge, MA: Author. Okagaki, L. & Bevis, C. (1999). Transmission of religious values: relations between parents’ and daughters’ beliefs. The Journal of Genetic Psychology 160: 303–318. Ozorak, E. W. (1989). Social and cognitive inﬂuences on the development of religious beliefs and commitment in adolescence. Journal for the Scientiﬁc Study of Religion 28: 448–463. Pearce, L. D. & Axinn, W. G. (1998). The impact of family religious life on the quality of mother-child relations. American Sociological Review 63: 810–828. Roof, W. C (1976). Traditional religion in contemporary society: A theory of localcosmopolitan plausibility. American Sociological Review 41: 195–208.

70

amy adamczyk

Roof, W. C. & McKinney, W. (1987). American mainline religion: Its changing shape and future. New Brunswick, NJ: Rutgers University Press. Roozen, D. A. (1980). Church dropouts: Changing patterns of disengagement and reentry. Review of Religious Research 21: 427–450. Sherkat, D. E. & Wilson, J. (1995). Preferences, constraints, and choices in religious markets: An examination of religious switching and apostasy. Social Forces 73: 993–1026. Smith, T. W. (1990). Classifying Protestant denominations. Review of Religious Research 31: 225–245. Stark, R. & Finke, R. (2000). Acts of faith: Explaining the human side of religion. Berkeley: University of California Press. Stark, R. & Glock, C. Y. (1968). The new denominationalism. Review of Religious Research 7: 17–28. Sullins, P. D. (1993). Switching close to home: Volatility or coherence in Protestant aﬃliation patterns. Social Forces 72: 339–419. Waters, M. S., Heath, W. C. & Watson, J. K. (1995). A positive model of the determination of religious aﬃliation. Social Science Quarterly 76: 105–123. Wuthnow, R. (1999). Growing up religious: Christians and Jews and their journeys of faith. Boston: Beacon Press. —— (1998). After heaven: Spirituality in America since the 1950s. Berkeley: University of California Press.

BURNOUT AMONG MALE ANGLICAN PAROCHIAL CLERGY IN ENGLAND: TESTING A MODIFIED FORM OF THE MASLACH BURNOUT INVENTORY Christopher F.J. Rutledge and Leslie Francis Abstract The present study proposes and tests a modiﬁed form of the Maslach Burnout Inventory, which has been Anglicized and adapted for use among clergy engaged in parochial ministry. Each of the three components of burnout (emotional exhaustion, depersonalization, and personal accomplishment) are assessed by 10-item subscales and also by 6-item short forms of these subscales. Normative data are provided by age groups, based on a random sample of 1,071 male Anglican clergy in England. Extroversion, neuroticism, and psychoticism scores are shown to be signiﬁcant predictors of self-assessed burnout.

Herbert Freudenberger (1974) is attributed with identifying what is now termed the burnout syndrome. Having experienced what he described as feelings of “burn out” (a consequence of working in the free clinic movement), he describes “burn out” in the following way: The dictionary deﬁnes the verb “burn out” as “to fail, wear out, or become exhausted by making excessive demands on energy, strength, or resources.” And that is exactly what happens when a staﬀ member in an alternative institution burns out for whatever reasons and becomes inoperative to all intents and purposes (Freudenberger, 1974, pp. 159–160).

According to this understanding, the physical signs of burnout include exhaustion, fatigue, a lingering cold, frequent headaches, gastrointestinal disturbances, sleeplessness, and shortness of breath. In terms of behavior, there are signs of anger and irritation. Emotions become diﬃcult to control and the slightest pressures may create feelings of being overburdened. Individuals can become excessively rigid, stubborn, and inﬂexible, and they may exhibit negative attitudes toward new ideas. Those who are prone to burnout are dedicated and committed, working too much, too long and too intensely (but are not workaholics; Gill, 1980), often dealing with chronicity (Mendel, 1979), and receiving little reward for success (Neiderhoﬀer, 1969). Burnout is signalled by workers who merge themselves and their lives with Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

72

rutledge and francis

the institution (Meadow, 1981). Those in the front line of the caring services are more prone to burnout than those administrators one step removed from the clients (Riggar, Godley, & Hafer, 1984; Murphy-Hackett & Ross, 1984; Wallace, Roberg, & Allen, 1985). Subsequent attempts in deﬁning burnout, and the diﬃculties encountered, have been well documented (Daley, 1979; Mendel, 1979; Doohan, 1982; Einsiedel & Tully, 1982; Zastrow, 1984; Beemsterboer & Baum, 1984; Burns, 1986; Dolan, 1987), but providing an adequate deﬁnition of what is meant by burnout has not proved easy. Assessing Burnout A major contribution to the understanding, deﬁnition, and assessment of burnout has been made by the pioneering work initiated by Christina Maslach and Susan Jackson through the development of the Maslach Burnout Inventory. They conceptualize burnout syndrome in the following way: Burnout is a syndrome of emotional exhaustion, depersonalization, and reduced personal accomplishment that can occur among individuals who do “people work” of some kind (Maslach and Jackson, 1986, p. 1).

The ﬁrst aspect of the burnout syndrome is increased feeling of emotional exhaustion, where workers, through their own emotional giving, ﬁnd that they can no longer continue to give at an emotional level. Emotional exhaustion will often be associated with such expressions as “I don’t care any more,” and “I don’t have any feelings left” (Maslach, 1978). As emotional resources are depleted, members of the caring professions feel that they are no longer able to give of themselves at a psychological level. It is this dimension which many researchers consider the key to understanding burnout (Cordes & Dougherty, 1993). The second aspect of the burnout syndrome is the development of depersonalization. As their work begins to take its psychological toll, members of the caring professions can begin to adopt negative, cynical, and dehumanizing attitudes towards, and feelings about, their clients, which will include certain types of language, compartmentalism, intellectualism and other withdrawal techniques (Maslach & Pines, 1977). Clients are often viewed as somehow deserving of their problems, and are often blamed for their own victimization (Ryan, 1971).

burnout among anglican clergy

73

The third aspect of the burnout syndrome is the experience of reduced personal accomplishment. Alongside emotional exhaustion and depersonalization, members of the caring professions may begin to feel unhappy about themselves, and dissatisﬁed with their accomplishments on the job. In a more recent conceptualization of this three-component model of burnout, Maslach, Schaufeli, and Leiter (2001) employ the following descriptions. The three key dimensions of this response are an overwhelming exhaustion, feelings of cynicism and detachment from the job, and a sense of ineﬀectiveness and lack of accomplishment. The exhaustion component represents the basic individual stress dimension of burnout. It refers to feelings of being overextended and depleted of one’s emotional and physical resources. The cynicism (or depersonalization) component represents the interpersonal context dimension of burnout. It refers to a negative, callous, or excessively detached response to various aspects of the job. The component of reduced eﬃcacy or accomplishment represents the self-evaluation dimension of burnout. It refers to feelings of incompetence and a lack of achievement and productivity at work (Maslach, Schaufeli, & Leiter, 2001, p. 399).

In order to measure experienced burnout Maslach and Jackson (1981a) devised what may be termed a preliminary form of the Maslach Burnout Inventory. This consisted of 47 items, each demanding two responses, one on a frequency of occurrence scale and the other on an intensity scale, both using a six-point anchored response format. The resultant data were subjected to a factor analysis, which led to a reduction in the number of items to 25. The resultant 25-item inventory was administered to a new sample and the combined results of the new sample and the initial sample were subjected to a further factor analysis which yielded a four factor solution. The factors were similar for both frequency and intensity: nine items loaded on the emotional exhaustion scale, ﬁve items loaded on the depersonalization scale, and eight items loaded on the personal accomplishment scale. The remaining three items were included as an optional dimension, which was designed to measure involvement. These four subscales, in this format, were used by Maslach and Jackson (1979) on a sample of 130 police couples. Burnout is conceptualized as a continuous variable ranging from low to high, and is not to be seen as a dichotomous variable which is either present or absent. Scores are considered high if they fall in

74

rutledge and francis

the upper third of a normative distribution. For both the emotional exhaustion and depersonalization subscales, a high score corresponds to higher degrees of experienced burnout. A moderate correlation was found between the emotional exhaustion and the depersonalization subscales, in accord with the theoretical expectations that the two are separate but related aspects of burnout. For the personal accomplishment subscale, a low score corresponds to a higher degree of perceived burnout. The correlations between the personal accomplishment subscale and the other two subscales were found to be low. In 1986, the Maslach Burnout Inventory was revised. In the ﬁrst edition, two questions relating to intensity and frequency were asked about each item. However, it was found that there was a positive correlation between the two (on average a correlation of .56), and Maslach and Jackson (1986) decided that it was no longer necessary to measure both dimensions. The new inventory, therefore, measures frequency only. Empirical support for such a decision is to be found in a number of studies. For example, Iwanki and Schwab (1981) concluded that the high correlations between the ratings for intensity and frequency make the two-dimensional format of the questionnaire unnecessary. The ﬁndings of Klausner and Green (1984) on burnout among dental educators clearly demonstrated the correlation which exists between frequency and intensity of experienced burnout. In addition, in a study on 462 teachers, Gold (1984) concluded that: the two scoring systems can be expected to yield comparable factor structures and thus essentially equivalent levels of construct validity. In fact, it would appear that either scoring system would suﬃce in identifying teachers who are becoming burned out in terms of self-perceptions (Gold, 1984, p. 1016).

Three of the four factors identiﬁed by Maslach and Jackson (1981b) had eigenvalues greater than one, while the fourth dimension— involvement—had an eigenvalue of less than one. Moreover, this factor of involvement accounted for only 3 of the 25 items, was found to be related to the emotional exhaustion scale, and had appeared as an optional dimension in the initial Maslach Burnout Inventory. In their study of teacher stress and burnout, Belcastro and Gold (1983) did not use these three optional items, arguing that the fourth subscale did not contribute signiﬁcantly to the ability of the MBI to provide a valid criterion measure of burnout. The revised

burnout among anglican clergy

75

Maslach Burnout Inventory omitted this fourth optional dimension (Maslach and Jackson, 1986). Reliability of the Maslach Burnout Inventory The reliability of the Maslach Burnout Inventory has been determined by measuring the internal consistency, where each item is measured against every other item, using coeﬃcient alpha (Cronbach, 1951). It is generally considered that an alpha of 0.70 or higher indicates an acceptable level of internal reliability (Nunnaly, 1978). In the initial sample (n = 1,316) Maslach and Jackson (1986) found the reliability coeﬃcients for the subscales as following: 0.90 for emotional exhaustion; 0.79 for depersonalization, and 0.71 for personal accomplishment. This level of internal reliability has received much support (for example, Ceslowitz, 1989; Abu-Hilal & Salameh, 1992; Vealey, Udry, Zimmerman, & Soliday, 1992; Schaufeli & van Dierendonck, 1993; Gupchup, Lively, Holiday-Goodman, Siganga, & Black, 1994; Buunk, Schaufeli, & Ybema, 1994; Burke, 1994; Pretorius, 1994; Schaufeli, Bakker, Hoogduin, Schaap, & Kladler, 2001). In addition the test-retest method has also been used. Maslach and Jackson (1986) used a small sample (n = 53) and test-retested over a period of two to four weeks, with the reliability coeﬃcients for the subscales being 0.82 for emotional exhaustion, 0.60 for depersonalization, and 0.80 for personal accomplishment. Jackson, Schwab, and Schuler (1986) also carried out a test-retest on 248 teachers, with the two tests separated by a year. The results were 0.60 for emotional exhaustion, 0.54 for depersonalization, and 0.57 for personal accomplishment. In addition, Wade, Cooley, and Savicki (1986) in their year-long longitudinal study of helping professionals, Capel (1991) in her study of teacher burnout, and Greenglass, Fiksenbaum, and Burke (1994) in their study of men and women employed within a single school board, all found that burnout scores were relatively stable over a period of time. Validity of the Maslach Burnout Inventory For Maslach and Jackson (1986) the validity of their instrument, measuring burnout as a discrete phenomenon, is demonstrated in

76

rutledge and francis

two ways. The ﬁrst is the discriminant validity. Is the Maslach Burnout Inventory measuring a syndrome distinct from other psychological constructs? One set of studies has examined the relationship between burnout and job satisfaction. Empirically there is support to show that burnout is not simply an expression of dissatisfaction with work (Maslach & Jackson, 1981b; Jackson, Turner, & Brief, 1985; Riggar, Godley, & Hafer, 1984; Leiter, 1985), and that high levels of job satisfaction can coexist with high levels of stress and burnout (Farber, 1983). Lawler and Hall (1970) also found that levels of satisfaction were not related to self-rated eﬀort or performance, dimensions which are similar to some items in the personal accomplishment subscale. Nevertheless, there is also empirical evidence to support the inverse relationship between job satisfaction and burnout (LeCroy & Rank, 1987; Berwick, 1992; Stearns & Moore, 1993). Dolan (1987) found that burnout was intricately linked with work and that the degree of satisfaction experienced will have some bearing on experienced burnout. He concluded from his study of nursing and administrative staﬀ that “the results of this study conﬁrm the thesis that job satisfaction is a reliable indicator of burnout. As was expected there was an inverse relationship between job satisfaction and burnout (Dolan, 1987, p. 9).” This conclusion has been echoed by a number of researchers. Maslach and Jackson (1984) found that burnout correlated signiﬁcantly (p > .001) with greater job dissatisfaction. Raﬀerty, Lemkau, Purdy, and Rudisill (1986) found a signiﬁcant correlation between job satisfaction and the Maslach Burnout Inventory subscales. Maslach and Jackson (1986), using the Maslach Burnout Inventory and a measure of job satisfaction on the Job Diagnostic Survey (Hackman & Oldham, 1974, 1975), found moderate negative correlations for both emotional exhaustion and depersonalization and a slight positive correlation for personal accomplishment. The studies by Riggar, Godly, and Hafer (1984) and by Wolpin, Burke, and Greenglass (1991) also lend support to this conclusion, demonstrating that burnout and job satisfaction are related, but distinct constructs. Another set of studies has examined the relationship between burnout and depression. Meier (1984), using a total burnout score (a practice not recommended by Maslach and Jackson, 1986 and reinforced by Huebner, 1993), found that there was a moderately strong correlation between burnout and certain measures of depres-

burnout among anglican clergy

77

sion. This result is supported further by the work of Firth, McKeown, McIntec, and Britton (1987) in their study of nursing staﬀ working in psychiatric hospitals, where they found strong similarities between burnout and professional depression. The correlation between the two constructs, however, is not that high, leading commentators like Walsh (1987) to conclude that “burned-out persons may be depressed, though depression tends to be all enveloping rather than situation speciﬁc. Certainly, not all depressed people are burned out as well. For some, work actually becomes a point of relief and pleasure in an otherwise depressing existence (Walsh, 1987, pp. 279–280).” The distinction between burnout and depression has also been established empirically by Leiter and Durup (1994), Glass and McKnight (1996), Bakker, Schaufeli, Demerouti, Janssen, van der Hulst, and Brouwer (2000) and Brenninkmeyer, van Yperen, and Buunk (2001). The second area to be explored is that of convergent validity. Maslach and Jackson (1986) identify three main areas: external validation (evidence given by an observer which corroborates an individual’s self-rating), dimensions of the job experience (where hypotheses are conﬁrmed about certain job characteristics and burnout), and personal outcomes (where hypotheses between experienced burnout and expected personal reactions are conﬁrmed). In terms of external validation, support for the Maslach Burnout Inventory is cited by Maslach and Jackson (1979) from their study on policemen. They found that police who scored high on the emotional exhaustion subscale were rated by their wives as coming home upset and angry, tense or anxious, physically exhausted, and complaining about problems at work. In terms of dimensions of the job experience, the validity of the Maslach Burnout Inventory is demonstrated by data which conﬁrm the relationship between certain job characteristics and experienced burnout. For example, it is hypothesized that the greater the number of clients one had to deal with, the greater the perceived burnout. Maslach and Pines (1977), in their study of staﬀ at day centers, note that the staﬀ from the high-ratio centers liked their jobs less, and they gave a lower evaluation of the center. Within the sphere of personal outcomes, support for the Maslach Burnout Inventory is given in a variety of studies. For example, in a study among 180 nurses Maslach (1976) demonstrates that people experiencing burnout would be dissatisﬁed with opportunities for

78

rutledge and francis

personal growth and development on the job. In a study among 142 policemen, Maslach and Jackson (1979) demonstrated that burnout scores were highly predictive of intention to leave the force. Factor structure of the Maslach Burnout Inventory Subsequent empirical research among a number of diﬀerent populations has generally recovered the hypothesized three-factor structure of emotional exhaustion, depersonalization, and personal accomplishment, including the following studies: Belcastro, Gold, and Hays (1983) among teachers, Powers and Gose (1986) among university students, Raﬀerty, Lemkau, Purdy, and Rudisill (1986) among family practice physicians, Pierce and Molloy (1989) among teachers, Gold, Bachelor, and Michael (1989) among student teachers, Lee and Ashforth (1990) among supervisors and managers from a large public welfare agency, Rowney and Cahoon (1990) among employees of large organizations, Gold, Roth, Wright, Michael, and Chen (1992) among elementary and secondary teachers, Sarros and Sarros (1990, 1992) among Australian teachers, and Scherer, Cox, Key, Stickney, and Spangler (1992) among employees in small business ﬁrms. Other studies supporting the three-factor structure of the Maslach Burnont Inventory include Golembiewski and Munzenrider (1981), Green and Walkey (1988), Pelsma, Roland, Tollefson, and Wigington (1989), Byrne (1991), Cordes and Dougherty (1993), Glass, McKnight, and Valdimarsdottir (1993), Papadatou, Anagnostopoulos, and Monos (1994), de Rijk, Le Blanc, Schaufeli, and de Jonge (1998), and Whitehead, Ryba, and O’Driscoll (2000). Green, Walkey, and Taylor (1991) published the results of a study where the Maslach Burnout Inventory was administered to four groups of diﬀerent nationalities. The data were subsequently subjected to principal component analysis, followed by a two-factor varimax rotation of all 22 items. This produced a two-factor solution, with items for emotional exhaustion and depersonalization forming one factor and the personal accomplishment items forming the other factor. Principle components analysis of the 14 emotional exhaustion items and depersonalization items, followed by a two-factor varimax rotations, produced a clear and replicable separation of these two subscales. Potential overlap between emotional exhaustion and depersonalization was also reported by Whitehead, Ryba, and O’Driscoll (2000).

burnout among anglican clergy

79

Walkey and Green (1992) tested the validity of the three-factor structure of the Maslach Burnout Inventory by examining the data obtained from six separate studies (Nagy, 1985; Brookings, Bolton, Brown, & McEvoy, 1985; Allie, 1983; Gentilini, 1982; Iwancki & Schwab, 1981; Green & Walkey, 1988). The data, when subjected to a principal components analysis, produced loadings that fell into three distinct groups, related to their subscale clusters. Green, Walkey, and Taylor (1991) concluded that the Maslach Burnout Inventory is extremely robust across a wide diversity of subjects. However, it needs to be noted that some researchers have questioned the necessity of certain items in the Maslach Burnout Inventory. Byrne (1993), using a large sample to determine the validity of the Maslach Burnout Inventory, found that items 12 (personal accomplishment) and 16 (emotional exhaustion) might well be excluded. The revised 20-item scale proved psychometrically sound. This really conﬁrmed what Byrne (1991) had found in an earlier study when substantial improvements were made in model ﬁt with the deletion of three items measuring emotional exhaustion (items 2, 16, and 20) and one item measuring personal accomplishment (12). Strong criticism for the three-factor solution has been made by Koeske and Koeske (1993) who did not consider that the tri-component conceptualization of burnout had fostered a rich conceptual framework on work stress. They felt that there was a need to reconceptualize, and that the burnout element should be perceived as lying within the emotional exhaustion element of the Maslach Burnout Inventory. Interestingly, in a study on female social workers, Corcoran (1985) only used the emotional exhaustion and depersonalization subscales because it was felt that the personal accomplishment subscale was not suﬃciently sound. Assessing burnout among clergy Although the Maslach Burnout Inventory has been used with some success among clergy and religious professionals (see Warner & Carter, 1984; Rodgerson & Piedmont, 1998), there are ways in which some of the items might seen strange or even oﬀensive to some clergy. As preparation for a series of major research studies among clergy in the United Kingdom, Rutledge (1999) proposed to modify the Maslach Burnout Inventory in four ways. These modiﬁcations were undertaken through a series of individual interviews and focus groups with

80

rutledge and francis

clergy discussing their reactions to the items in the original Maslach Burnout Inventory. As a consequence of this process, the original American items were Anglicized. The original items were shaped to reﬂect the experience and language of parochial ministry. Additional items were developed reﬂecting further aspects of parochial ministry in order to bring the three subscales to the same length of 10 items each. The response scale was changed from a seven-point measure of frequency to a ﬁve-point measure of attitudinal intensity, following the convention of Likert (1932), ranging from agree strongly, through agree, not certain, disagree, and disagree strongly. Permission was obtained from the Consulting Psychologists Press to adopt the Maslach Burnout Inventory in these ways for use among clergy (under license and at a cost). Against this background, the aim of the present paper is to report on the reliability and scale properties of the three subscales of the modiﬁed form of the Maslach Burnout Inventory among male Anglican parochial clergy in England. The construct validity of the independence of the three subscales of the instrument will be examined against measures of Eysenck’s three-dimensional model of personality (Eysenck & Eysenck, 1985). This model of personality is now well established in clergy studies in the United Kingdom (Francis, 1991; Francis and Rodger, 1994). Method Sample The questionnaire was mailed to a 15% random sample of all male clergy identiﬁed on the Church Commissioners’ payroll as being engaged in stipendiary parochial ministry in England (see Rutledge, 1999). A total of 1,476 questionnaires were mailed resulting in 1,071 usable responses, representing an overall positive response rate of 72.6%. Just 3% of the respondents were under the age of 30, 19% were in their thirties, 31% were in their forties, 31% were in their ﬁfties, 15% were in their sixties, and 1% were in their seventies. Measures Burnout was assessed by a modiﬁed form of the Maslach Burnout Inventory (Maslach & Jackson, 1986) in which each of the three sub-

burnout among anglican clergy

81

scales comprised 10 items arranged for scoring on a ﬁve-point Likert scale: agree strongly, agree, not certain, disagree, and disagree strongly. The modiﬁed emotional exhaustion subscale included the following item: “I feel burned out from my parish ministry.” The modiﬁed depersonalization subscale included the following item: “I don’t really care what happens to some of my parishioners.” The modiﬁed personal accomplishment subscale included the following item: “I have accomplished many worthwhile things in my parish ministry.” The modiﬁed form of the Maslach Burnout Inventory for use among parochial clergy has been modiﬁed and reproduced by special permission of the publisher, Consulting Psychologists Press. The publishers denied permission to publish the actual items from the modiﬁed Maslach Burnout Inventory here. Personality was assessed by the Eysenck Personality Questionnaire (Eysenck & Eysenck, 1975) which proposes a 21-item measure of extraversion, a 23-item measure of neuroticism, and a 25-item measure of psychoticism. It also contains a 21-item lie scale. Each item is assessed on a dichotomous scale: yes and no. Results The three sub-scales of the modiﬁed Maslach Burnout Inventory achieved the following alpha coeﬃcients (Cronbach, 1951): emotional exhaustion, .89; depersonalization, .81; and personal accomplishment, .78. All three scales display a highly satisfactory degree of internal consistency reliability. As is consistent with previous research there were signiﬁcant correlations between all three subscales: emotional exhaustion and depersonalization, r = .6229; emotional exhaustion and personal accomplishment, r = –.3745; depersonalization and personal accomplishment, r = –.3970. The four scales of the Eysenck Personality Questionnaire achieved the following alpha coeﬃcients: extraversion, .84; neuroticism, .86; psychoticism, .52; and lie scale .78. The lower alpha coeﬃcient recorded by the psychoticism scale is consistent with the known diﬃculties in operationalizing this construct (Francis, Philipchalk, & Brown, 1991).

rutledge and francis

82

Table 1. Correlation matrix for personality and burnout Extroversion

Neuroticism

Psychoticism

Lie scale

Emotional exhaustion

–0.2211***

+0.5003***

Depersonalization

–0.1568***

+0.3857***

+0.1484*** –0.0567 +0.2779*** –0.1789***

Personal accomplishment

+0.4500***

–0.3094***

–0.0510 +0.0937**

Note: ** p < .01; *** p < .001

Table 1 presents the correlation coeﬃcients between the four scales of the Eysenck Personality Questionnaire and the three subscales of the modiﬁed Maslach Burnout Inventory. These data demonstrate that high extroversion scores are associated with lower emotional exhaustion, lower depersonalization, and higher personal accomplishment; that higher neuroticism scores are associated with higher emotional exhaustion, higher depersonalization, and lower personal achievement; and that higher psychoticism scores are associated with higher emotional exhaustion, and higher depersonalization. Eysenck’s fourth construct, the lie scale, is generally interpreted as an index of social conformity. The data demonstrate that high scores on the lie scale are associated with lower depersonalization, and higher personal accomplishment. Table 2. Mean scale scores for 10-item scales Age Group

30–39 40–49 50–59 60–69 All clergy

Emotional Exhaustion Mean SD 23.1 23.3 22.1 19.7 22.3

6.8 7.4 7.6 6.8 7.3

Depersonalization Mean

SD

Personal Accomplishment Mean SD

20.8 20.8 19.4 17.7 19.9

5.2 5.3 5.3 5.1 5.3

24.4 24.6 24.8 24.8 24.7

4.6 4.6 4.7 4.9 4.7

Note: The personal accomplishment scores have been reversed so that a high score represents low personal accomplishment.

burnout among anglican clergy

83

Given the high reliability of the three subscales of emotional exhaustion, depersonalization, and personal accomplishment, and the representative nature of the sample, it is legitimate now to publish normative data for male Anglican parochial clergy. Table 2 presents these data for clergy as a whole and for separate age groups. For these normative data the personal accomplishment scores have been reverse coded so that higher scores represent lower personal accomplishment and a higher level of burnout. These data demonstrate that higher levels of burnout are associated with the younger cohort of clergy across two of the three subscales: emotional exhaustion, F = 9.9, p < .001; and depersonalization, F = 15.4, p < .001. On the other hand, there is no signiﬁcant relationship between age and levels of personal accomplishment. Table 3. Multiple regression signiﬁcance tests predictor variables

r2

Emotional exhaustion age 0.0170 neuroticism 0.2542 psychoticism 0.2652 extraversion 0.2874 lie scale 0.2902 Depersonalization age 0.0345 neuroticism 0.1722 psychoticism 0.2252 extraversion 0.2429 lie scale 0.2451 Personal accomplishment age 0.0009 neuroticism 0.0999 psychoticism 0.1012 extraversion 0.2604 lie scale 0.2666

Increase r2

F

P <

Beta

t

P <

0.0170 0.2373 0.0110 0.0221 0.0029

17.8 327.1 15.4 31.9 4.1

.001 .001 .001 .001 .05

–0.0904 +0.4592 +0.1314 –0.1509 +0.0573

–3.3 +16.8 +4.8 –5.6 +2.0

.001 .001 .001 .001 .05

0.0345 0.1376 0.0530 0.0177 0.0022

36.8 170.9 70.3 23.9 3.0

.001 .001 .001 .001 NS

–0.1193 +0.3238 +0.2360 -0.1395 –0.0502

–4.2 +11.5 +8.4 –5.0 –1.7

.001 .001 .001 .001 NS

0.0009 0.0990 0.0013 0.1592 0.0062

0.9 113.1 1.5 220.9 8.6

NS .001 NS .001 .01

0.0387 +0.2193 +0.0594 –0.4157 –0.0841

1.4 +7.9 2.1 –15.0 –2.9

NS .001 .05 .001 .01

Note: The personal accomplishment scores have been reversed so that a high score represents low personal accomplishment.

Table 3 presents the use of multiple regression in order to test whether the bivariate relationships between burnout (emotional exhaustion, depersonalization, and personal accomplishment) age and personality are cumulative. In these equations, the predictor variables

84

rutledge and francis

were entered in the ﬁxed order of age, neuroticism, psychoticism, and the lie scale. These data conﬁrm that age, neuroticism, psychoticism, and extroversion function as independent and cumulative predictors of both emotional exhaustion and depersonalization. In the case of personal accomplishment, neuroticism and extroversion are key predictors, while psychoticism emerges as of only marginal signiﬁcance. Age is not a signiﬁcant predictor of personal accomplishment. For research purposes, there are times when even a 30-item instrument may overload a questionnaire which is intent on generating information across a wide range of variables. For this reason, it was proposed to develop a short form of the modiﬁed Maslach Burnout Inventory in which each of the three subscales of emotional exhaustion, depersonalization and personal accomplishment were comprised of those six items in the larger subscales which achieved the highest item rest-of-scale correlations. The six-item scales achieved the following alpha coeﬃcients: emotional exhaustion, 0.87; depersonalization, 0.78; and personal accomplishment, 0.75. All three scales display a highly satisfactory degree of internal reliability for instruments of this length. Table 4. Mean scale scores for the 6-item scales Age Group

30–39 40–49 50–59 60–69 All clergy

Emotional Exhaustion Mean SD 13.1 13.3 12.4 11.1 12.6

4.5 4.9 4.9 4.3 4.7

Depersonalization Mean

SD

Personal Accomplishment Mean SD

12.1 11.9 10.9 9.9 11.3

3.4 3.6 3.4 3.2 3.4

15.1 14.9 15.2 15.3 15.1

3.0 3.0 3.2 3.4 3.1

Note: The personal accomplishment scores have been reversed so that a high score represents low personal accomplishment.

Finally, Table 4 presents the mean scale scores recorded by the clergy on the short forms of the scales of emotional exhaustion, depersonalization and personal accomplishment.

burnout among anglican clergy

85

Discussion Six main features of these data deserve attention. First, all three scales of the 10-item modiﬁed form of the Maslach Burnout Inventory achieve high levels of internal consistency reliability and good levels of item-rest-of-test correlations. The highest of these item-rest-of-test correlations provide a good conﬁrmation of the face validity of the three scales. By this technique, the scale of emotional exhaustion is characterized by the item, “I feel emotionally drained from my parish ministry;” the scale of depersonalization is characterized by the item, “I have become more callous toward people since working in parish ministry;” and the scale of personal accomplishment is characterized by the item, “I gain a lot of personal satisfaction from working with people.” The three shorter 6-item scales of the modiﬁed form of the Maslach Burnout Inventory also achieve high levels of internal consistency reliability. Second, the levels of endorsement given to each of the 30-items of the modiﬁed Maslach Burnout Inventory provides an insight into the proportions of clergy who are reporting individual indicators of burnout. In respect to emotional exhaustion, nearly a third of the clergy feel that they are working too hard in their parish ministry (31%), and say that they feel used up at the end of the day in parish ministry (29%). One in ﬁve feel frustrated by their parish ministry (21%). One in eight feel emotionally drained from their parish ministry (13%), and say that working with people all day is a real strain for them (12%). A signiﬁcant minority of clergy report that they feel fatigued when they get up in the morning and have to face another day (9%), that they feel burned out from their parish ministry (8%), that they feel like they are at the end of their tether (8%), that they would feel a lot better if they could get out of parish ministry (6%), and that working with people directly puts too much strain on them (5%). With regard to depersonalization, more than one in ten of the clergy feel that parishioners blame them for some of their problems (16%), and say that they are less patient with parishioners than they used to be (11%). A signiﬁcant minority of the clergy report that they worry that parish ministry is hardening them emotionally (9%); that they ﬁnd it diﬃcult to listen to what some parishioners are really saying to them (7%); that they do not really care what happens to some parishioners (5%); that they feel they treat some parishioners

86

rutledge and francis

as if they were impersonal objects (4%); that they wish parishioners would leave them alone (4%); that they feel nowadays that most people cannot be really helped with their problems (3%); that they have become more callous toward people since working in parish ministry (3%); and that they cannot be bothered to understand how some people feel about things (1%). In respect to personal accomplishment, four out of every ﬁve clergy gain a lot of personal satisfaction from working with people (82%), and say that if they could have their time all over again they would still go into parish ministry (80%). Over half of the clergy feel that they can easily create a relaxed atmosphere with their parishioners (57%), claim that they feel exhilarated after working closely with their parishioners (52%), and say that in their parish ministry they deal with emotional problems very clearly (51%). Between one-third and one-half of the clergy feel that they have accomplished many worthwhile things in their parish ministry (48%), that they are positively inﬂuencing people’s lives through their parish ministry (44%), and that they can easily understand how their parishioners feel about things (36%). A quarter of the clergy report that they feel very energetic (24%). A ﬁfth of the clergy feel that they deal very eﬀectively with the problems of their parishioners (18%). Third, age emerged as a signiﬁcant predictor of emotional exhaustion and depersonalization, but not of personal accomplishment. The signiﬁcant negative correlation between age and scores both on the emotional exhaustion subscale and on the depersonalization subscale, but not on the personal accomplishment subscale is consistent with ﬁndings among other professional groups as reported by Bartz and Maloney (1986), Lee and Ashforth (1991), Jackson, Barnett, Stajich, and Murphy (1993), Cook and Banks (1993), and Price and Spence (1994). Kilfedder, Power, and Wells (2001) found a negative correlation between age and depersonalization, but not between age and emotional exhaustion or personal accomplishment. Thus, according to the majority of studies, older clergy are less likely than younger clergy to suﬀer from either emotional exhaustion or depersonalization. Two theories may account for these diﬀerences between younger and older clergy. Maslach, Schaufeli, and Leiter (2001) suggest that those who burn out early in their careers are likely to quit their jobs, leaving behind the survivors who consequently exhibit lower levels of burnout. In other words, younger clergy who suﬀer from emotional exhaustion or depersonalization may decide to leave paro-

burnout among anglican clergy

87

chial ministry either because of ill health or to seek alternative employment. Older clergy may have learned how to pace their work better so as to avoid such signs of burnout. Fourth, neuroticism scores emerged as the strongest and most consistent predictor of individual diﬀerences over the three dimensions of burnout. Clergy who scored higher on the neuroticism scale were more likely to suﬀer from emotional exhaustion, to display signs of depersonalization, and to enjoy fewer feelings of personal accomplishment. This ﬁnding is consistent with the ﬁnding of Francis and Rodger (1994) that clergy who score high on Eysenck’s neuroticism scale are likely to be more dissatisﬁed with their ministry. It is also consistent with the ﬁnding of Manlove (1993) that scores on Eysenck’s neuroticism scale are positively associated with burnout among child care workers and with the ﬁnding of Rodgerson and Piedmont (1998) that neuroticism is the strongest predictor of burnout among clergy in terms of the ﬁve-factor model of personality proposed by Costa and McCrae (1992). Fifth, psychoticism scores emerged as signiﬁcant predictors of emotional exhaustion and depersonalization, but not of personal accomplishment. Clergy who scored higher on the psychoticism scale were more likely to suﬀer from emotional exhaustion and to display signs of depersonalization. This is consistent with the wider theory that tough-minded individuals are less likely to show empathy to others, less likely to be properly in tune with their own feelings, and less likely to be at ease with themselves or with other people (Eysenck & Eysenck, 1976). Sixth, extroversion scores emerged as signiﬁcant predictors of emotional exhaustion, depersonalization, and personal accomplishment. Introverted clergy were particularly likely to enjoy fewer feelings of personal accomplishment. Introverted clergy were also more likely to suﬀer from emotional exhaustion and to display signs of depersonalization. This ﬁnding is also consistent with the ﬁnding of Rodgerson and Piedmont (1998) who employed the Costa and McCrae (1992) index of extroversion among Baptist clergy in the United States of America. The relationship between introversion and burnout lends weight to the view that many aspects of the clerical profession presuppose a predisposition toward extroversion. At the same time, it has to be recognized that ministry candidates tend to be more introverted than the general population (Francis, 1991). On the other hand, no relationship was found between burnout and Eysenck’s

88

rutledge and francis

measure of extroversion among child care workers by Manlove (1993) or among teachers by Capel (1992). Conclusion This study set out to develop and test a modiﬁed form of the Maslach Burnout Inventory and to propose a short form of the modiﬁed instrument. The Maslach Burnout Inventory was modiﬁed in four ways. First, the items were adapted in order to ﬁt the experience and terminology of parochial ministry. Second, the language was Anglicized in order to be more appropriate for use within the United Kingdom. Third, the opportunity was taken to develop new, additional items in order to bring all three subscales to the same length. Fourth, the response scale was conformed to the conventional Likert ﬁve categories in order to assess attitudinal intensity. Data generated among a random sample of 1,071 Anglican clergy engaged in parochial ministry supported the psychometric properties of this modiﬁed instrument and enabled the generation of normative scores according to age groups. On the basis of this study, both the full form and the short form of the modiﬁed Maslach Burnout Inventory can be recommended for further use. Building on the present study, future research would be beneﬁcial in two main directions. First, important information about the level of burnout experienced among clergy could be derived from similar comparative studies among clergy serving in other denominations in England and among Anglican clergy in other countries. A replication of the present study among another random sample of male parochial clergy serving in the Church of England would help to establish whether perceived levels of burnout among Anglican clergy were increasing, decreasing, or remaining stable over time. Second, important information about the predictors of burnout among clergy could be derived from studies which incorporate the modiﬁed Maslach Burnout Inventory alongside other key variables. A start has already been made in this direction by Francis and Rutledge (2000) who established that clergy serving in rural ministry report lower levels of personal accomplishment in comparison with clergy serving in other geographical areas. In particular, future studies might wish to explore the relationship between burnout and such strategies as peerassisted clergy review, hierarchical appraisal, and other recommended support mechanisms.

burnout among anglican clergy

89

References Abu-Hilal, M. M., & Salameh, K. M. (1992). Validity and reliability of the Maslach Burnout Inventory for a sample of non-western teachers. Educational and Psychological Measurement, 52, 161–169. Allie, S. M. (1983). Organizational and personal life stress and the role of moderator variables in the prediction of burnout, performance and serious illness. Dissertation Abstracts International, 43, 2430–A. Bakker, A. B., Schaufeli, W. B., Demerouti, E., Janssen, P. M. P., van der Hulst, R., & Brouwer, J. (2000). Using equity theory to examine the diﬀerence between burnout and depression. Anxiety Stress Coping, 13, 247–268. Bartz, C., & Maloney, J. P. (1986). Burnout among intensive-care nurses. Research in Nursing and Health, 9, 147–153. Beemsterboer, J., & Baum, B. H. (1984). Burnout: deﬁnitions and health-care management. Social Work in Health Care, 10, 97–109. Belcastro, P. A., & Gold, R. S. (1983). Teacher stress and burnout. Journal of School Health, 53, 404–407. Belcastro, P. A., Gold, R. S., & Hays, L. C. (1983). Maslach Burnout Inventory: Factor structures for samples of teachers. Psychological Reports, 53, 364–366. Berwick, K. R. (1992). Stress among student aﬀairs administrators: The relationship of personal characteristics and organizational variables to work-related stress. Journal of College Student Development, 33, 11–19. Brenninkmeyer, V., van Yperen, N. W., & Buunk, B. P. (2001). Burnout and depression are not identical twins: Is decline of superiority a distinguishing feature? Personality and Individual Diﬀerences, 30, 873–880. Brookings, J. B., Bolton, B., Brown, C. E., & McEvoy, A. (1985). Self-reported job burnout among female human-service professionals. Journal of Occupational Behavior, 6, 143–150. Burke, R. J. (1994). Stressful events, work-family conﬂict, coping, psychological burnout, and well-being among police oﬃcers. Psychological Reports, 75, 787–800. Burns, R. C. (1986). Endodontic burnout. International Journal of Psychosomatics, 33(2), 17–23. Buunk, B. P., Schaufeli, W. B., & Ybema, J. F., (1994). Burnout, uncertainty, and the desire for social comparison among nurses. Journal of Applied Social Psychology, 24, 1701–1718. Byrne, B. M. (1991). The Maslach Burnout Inventory: Validating factorial structure and invariance across intermediate, secondary, and university educators. Multivariate Behavioral Research, 26, 583–605. —— (1993). The Maslach Burnout Inventory: Testing for factorial validity and invariance across elementary, intermediate, and secondary teachers. Journal of Occupational and Organizational Psychology, 66, 197–212. Capel, S. A. (1991). A longitudinal study of burnout in teachers. British Journal of Educational Psychology, 61, 36–45. —— (1992). Stress and burnout in teachers. European Journal of Teacher Education, 15, 197–211. Ceslowitz, S. B. (1989). Burnout and coping strategies among hospital staﬀ nurses, Journal of Advanced Nursing, 14, 553–557. Cook, B. B., & Banks, S. R. (1993). Predictors of job burnout in reporters and copy editors. Journalism Quarterly, 70, 108–117. Corcoran, K. J. (1985). Measuring burnout: A reliability and convergent validity study. Journal of Social Behavior and Personality, 1, 107–112. Cordes, C. L., & Dougherty, T. W. (1993). A review and an integration of research on job burnout. Academy of Management Review, 18, 621–656. Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory and NEO Five Factor Inventory: Manual. Odessa, Florida: Psychological Assessment Resources.

90

rutledge and francis

Cronbach, L. J. (1951), Coeﬃcient alpha and the internal structure of tests. Psychometrika, 16, 297–334. Daley, M. R. (1979). Preventing worker burnout in child welfare. Child Welfare, 58, 443–450. de Rijk, A. E., Le Blanc, P., Schaufeli, W. B., & de Jonge, J. (1998). Active coping and need for control as moderators of the job demand-control model: Eﬀects on burnout. Journal of Occupational and Organizational Psychology, 71, 1–18. Dolan, N. (1987). The relationship between burnout and job satisfaction in nurses. Journal of Advanced Nursing, 12, 3–12. Doohan, H. (1982). Burnout: a critical issue for the 1980’s. Journal of Religion and Health, 21, 352–358. Einsiedel, A. A., & Tully, H. A. (1982). Methodological considerations in studying the burnout phenomenon. In J. W. Jones (Ed.). The burnout syndrome (pp. 89–106). Park Ridge, IL: London House Press. Eysenck, H. J., & Eysenck, M. W. (1985). Personality and individual diﬀerences: A natural science approach, New York, Plenum Press. Eysenck, H. J., & Eysenck, S. B. G. (1975). Manual of the Eysenck Personality Questionnaire (adult and junior). London: Hodder and Stoughton. —— (1976). Psychoticism as a Dimension of Personality. London: Hodder and Stoughton. Farber, B. A. (1983). Stress and burnout in the human services professions. New York: Pergamon Press. Firth, H., McKeown, P., McIntec, J., & Britton, P. (1987). Professional depression: Burnout and personality in longstay nursing. International Journal of Nursing Studies, 24, 227–237. Francis, L. J. (1991). The personality characteristics of Anglican ordinands: Feminine men and masculine women? Personality and Individual Diﬀerences, 12, 1133–1140. Francis, L. J., Philipchalk, R., & Brown, L. B. (1991). The comparability of the short form EPQ-R with the EPQ among students in England, the USA, Canada and Australia. Personality and Individual Diﬀerences, 12, 1129–1132. Francis, L. J., & and Rodger, R. (1994). The inﬂuence of personality on clergy role prioritization, role inﬂuences, conﬂict and dissatisfaction with ministry. Personality and Individual Diﬀerences, 16, 947–957. Francis, L. J., & Rutledge, C. (2000). Are rural clergy in the Church of England under greater stress? A study in empirical theology. Research in the Social Scientiﬁc Study of Religion, 11, 173–191. Freudenberger, H. J. (1974). Staﬀ burn-out. Journal of Social Issues, 30, 159–166. Gentilini, J. M. (1982). The relationship between the characteristics of job burnout and the perceived and personal job characteristics of vocational rehabilitation counsellors. Dissertation Abstracts International, 43, 1046–A. Gill, J. J. (1980). Burnout: A growing threat in ministry. Human Development, 1, 65–71. Glass, D. C., & McKnight, J. D. (1996). Perceived control, depressive symptomatology, and professional burnout: A review of the evidence, Psychology and Health, 11, 23–48. Glass, D. C., McKnight, J. D., & Valdimarsdottir, H. (1993). Depression, burnout, and perceptions of control in hospital nurses. Journal of Consulting and Clinical Psychology, 61, 147–155. Gold, Y. (1984). The factoral validity of the Maslach Burnout Inventory in a sample of California elementary and junior high school classroom teachers. Educational and Psychological Measurement, 44, 1009–1016. Gold, Y., Bachelor, P., & Michael, W. B. (1989). The dimensionality of a modiﬁed form of the Maslach Burnout Inventory for university students in a teacher-training program. Educational and Psychological Measurement, 49, 549–561. Gold, Y., Roth, R. A., Wright, C. R., Michael, W. B., & Chen, C. Y. (1992). The factorial validity of a teacher burnout measure (educators survey) administered to

burnout among anglican clergy

91

a sample of beginning teachers in elementary and secondary schools in California. Educational and Psychological Measurement, 52, 761–768. Golembiewski, R. T., & Munzenrider, R. (1981). Eﬃcacy of three versions of one burn-out measure: MBI as total score, sub-scale scores, or phases? Journal of Health and Human Resources Administration, 4, 228–246. Green, D. E., & Walkey, F. H. (1988). A conﬁrmation of the three-factor structure of the Maslach Burnout Inventory. Educational and Psychological Measurement, 48, 579–585. Green, D. E., Walkey, F. H., & Taylor, A. J. (1991). The three factor structure of the Maslach Burnout Inventory: A multicultural, multinational conﬁrmatory study. Journal of Social Behavior and Personality, 6, 453–472. Greenglass, E. R., Fiksenbaum, L., & Burke, R. J. (1994). The relationship between social support and burnout over time in teachers. Journal of Social Behavior and Personality, 9, 219–230. Gupchup, G. V., Lively, B. T., Holiday-Goodman, M., Siganga, W. W., & Black, C.D. (1994). Maslach Burnout Inventory: Factor structures for pharmacists in health maintenance organizations and comparison with normative data for USA pharmacists. Psychological Reports, 74, 891–895. Hackman, J. R., & Oldham, G. R. (1974). The Job Diagnostic Survey: An instrument for the diagnosis of jobs and the evaluation of job redesign projects. JSAS Catalogue of Selected Documents in Psychology, 4, 148 (Ms. No. 819). —— 1975). Development of the Job Diagnostic Survey. Journal of Applied Psychology, 60, 159–170. Huebner, E. S. (1993). Burnout among school psychologists in the USA: Further data related to its prevalence and correlates. School Psychology International, 14, 99–109. Iwancki, E. F., & Schwab, R. L. (1981). A cross validation study of the Maslach Burnout Inventory. Educational and Psychological Measurement, 41, 1167–1174. Jackson, R. A., Barnett, C. W., Stajich, G. V., & Murphy, J. E. (1993). An analysis of burnout among school of pharmacy faculty. American Journal of Pharmaceutical Education, 57, 9–17. Jackson, S. E., Schwab, R. L., & Schuler, R. S. (1986). Toward an understanding of the burnout phenomenon. Journal of Applied Psychology, 71, 630–64. Jackson, S. E., Turner, J., & Brief, A. P. (1985). Burnout among public service lawyers. Unpublished manuscript, University of Michigan. Kilfedder, C. J., Power, K. G., & Wells, T. J. (2001). Burnout in psychiatric nursing. Journal of Advanced Nursing, 34, 383–396. Klausner, L. H., & Green, T. G. (1984). What does professional burnout mean to the dental educator? Journal of Dental Education, 48, 91–94. Koeske, G. F., & Koeske, R. D. (1993). A prelimary test of a stress-strain outcome model for reconceptualizing the burnout phenomenon. Journal of Social Service Research, 17, 107–135. Lawler, E. E., & Hall, D. T. (1970). Relationship of job characteristics to job involvement, satisfaction and intrinsic motivation. Journal of Applied Psychology, 54, 305– 312. Le Croy, C. W., & Rank, M. R. (1987). Factors associated with burnout in the social services: An exploratory study. Journal of Social Service Research, 10, 23–39. Lee, R. T., & Ashforth, B. E. (1990). On the meaning of Maslach’s three dimensions of burnout. Journal of Applied Psychology, 75, 743–747. —— (1991). Work-unit structure and processes and job-related stresses as predictors of managerial burnout. Journal of Applied Social Psychology, 21, 1831–1847. Leiter, M. P. (1985). Burnout as a function of communication patterns in a multidisciplinary mental health setting. Paper presented at the annual convention of the Canadian Psychological Association, Halifax, Nova Scotia.

92

rutledge and francis

Leiter, M. P. & Durup, J. (1994). The discriminant validity of burnout and depression: A conﬁrmatory factor analytic study. Anxiety Stress Coping, 7, 357–373. Likert, R. (1932). A technique for the measurement of attitudes. Archives of Psychology, 140, 1–55. Manlove, E. E. (1993). Multiple correlates of burnout in child care workers. Early Childhood Research Quarterly, 8, 499–518. Maslach, C. (1976). Burned-out. Human Behavior, 5, 16–22. —— (1978). Job burnout: How people cope. Public Welfare, 36, 56–58. Maslach, C., & Jackson, S. E. (1979). Burned-out cops and their families. Psychology Today, 12, 59–62. —— (1981a). Maslach Burnout Inventory. Palo Alto, California: Consulting Psychologists Press. —— (1981b). The measurement of experienced burnout. Journal of Occupational Behavior, 2, 99–113. —— (1984). Patterns of burnout among a national sample of public contact workers. Journal of Health and Human Resources Administration, 7, 189–212. —— (1986). Maslach Burnout Inventory (2nd ed.). Palo Alto, CA: Consulting Psychologists Press. —— (1986). MBI—Human Services Survey. Palo Alto, CA: Consulting Psychologists Press. Maslach, C., & Pines, A. (1977). The burn-out syndrome in the day-care setting. Child Care Quarterly, 6, 100–113. Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology, 52, 397–422. Meadow, K. P. (1981). Burnout in professionals working with deaf children. American Annal of the Deaf, 126, 13–22. Meier, S. T. (1984). The construct validity of burnout. Journal of Occupational Psychology, 57, 211–219. Mendel, W. M. (1979). Staﬀ burn-out: Diagnosis, treatment, and prevention. New Directions for Mental Health Systems, 2, 75–83. Murphy-Hackett, E. S., & Ross, N. R. (1984). How one agency is ﬁghting burnout. Public Welfare, 42, 23–27. Nagy, S. (1985). Burnout and selected variables as components of occupational stress. Psychological Reports, 56, 195–200. Neiderhoﬀer, A. (1969). Behind the shield: The police in urban society. New York: Anchor Books. Nunnaly, N. C. (1978). Psychometric theory. New York: McGraw-Hill. Papadatou, D., Anagnostopoulos, F., & Monos, D. (1994). Factors contributing to the development of burnout in oncology nursing. British Journal of Medical Psychology, 67, 187–199. Pelsma, D. M., Roland, B., Tollefson, N., & Wigington, H. (1989). Parent burnout: Validation of the Maslach burnout inventory with a sample of mothers. Measurement and Evaluation in Counseling and Development, 22, 81–87. Pierce, C. M. B., & Molloy, G. N. (1989). The construct validity of the Maslach Burnout Inventory: Some data from down under. Psychological Reports, 65, 1340–1342. Powers, S., & Gose, K. F. (1986). Reliability and construct validity of the Maslach Burnout Inventory in a sample of university students. Educational and Psychological Measurement, 46, 251–255. Pretorius, T. B. (1994). Using the Maslach Burnout Inventory to assess educators’ burnout at a university in South Africa. Psychological Reports, 75, 771–777. Price, L., & Spence, S. H. (1994). Burnout symptoms among drug and alcohol service employees: Gender diﬀerences in the interaction between work and home stressors. Anxiety Stress and Coping, 7, 67–84. Raﬀerty, J. P., Lemkau, J. P., Purdy, R. R., & Rudisill, J. R. (1986). Validity of

burnout among anglican clergy

93

the Maslach Burnout Inventory for family practice physicians. Journal of Clinical Psychology, 42, 488–492. Riggar, T. F., Godley, S. H., & Hafer, M. (1984). Burnout and job satisfaction in rehabilitation administrators and direct service providers. Rehabilitation Counselling Bulletin, 27, 151–160. Rodgerson, T. E., & Piedmont, R. L. (1998). Assessing the incremental validity of the Religious Problem-Solving Scale in the prediction of clergy burnout. Journal for the Scientiﬁc Study of Religion, 37, 517–527. Rowney, J. I. A., & Cahoon, A. R. (1990). Individual and organizational characteristics of women in managerial leadership. Journal of Business Ethics, 9, 293–316. Rutledge, C. J. F. (1999). Exploring burnout among male Anglican parochial clergy. Unpublished PhD dissertation, University of Wales, Lampeter. Ryan, W. (1971). Blaming the victim. New York: Pantheon Books. Sarros, A. M., & Sarros, J. C. (1990). How burned out are our teachers: A cross cultural study. Australian Journal of Education, 34, 145–152. Sarros, J. C., & Sarros, A. M. (1992). Social support and teacher burnout. Journal of Educational Administration, 30, 55–69. Schaufeli, W. B., Bakker, A. B., Hoogduin, K., Schaap, C., & Kladler, A. (2001). On the clinical validity of the Maslach Burnout Inventory and the Burnout Measure. Psychology and Health, 16, 565–582. Schaufeli, W. B., & van Dierendonck, D. (1993). The construct validity of two burnout measures. Journal of Organizational Behavior, 14, 631–647. Scherer, R. F.; Cox, M. K; Key, C. C.; Stickney, F. A., & Spangler, E. (1992). Assessing the similarity of burnout dimensions in two business samples. Psychological Reports, 71, 28–30. Stearns, G. M., & Moore, R. J. (1993). The physical and psychological correlates of job burnout in the Royal Canadian Mounted Police. Canadian Journal of Criminology, 35, 127–147. Vealey, R. S., Udry, E. M., Zimmerman, V., & Soliday J. (1992). Intrapersonal and situational predictors of coaching burnout. Journal of Sport and Exercise Psychology, 14, 40–58. Wade, D. C., Cooley, E., & Savicki, V. (1986). A longitudinal study of burnout. Children and Youth Services Review, 8, 161–173. Walkey, F. H. and Green, D. E. (1992). An exhaustive examination of the replicable factor structure of the Maslach Burnout Inventory. Educational and Psychological Measurement, 52, 309–323. Wallace, P. A., Roberg, R. R., & Allen, H. E. (1985). Job burnout among narcotic investigators: An exploratory study. Journal of Criminal Justice, 13, 549–559. Walsh, J. A. (1987). Burnout and values in the social-service profession. Social Casework—Journal of Contemporary Social Work, 68, 279–283. Warner, J. and Carter, J. D. (1984). Loneliness, marital adjustment and burnout in pastoral and laypersons. Journal of Psychology and Theology, 12, 125–131. Whitehead, A., Ryba, K., & O’Driscoll, M. (2000). Burnout among New Zealand primary school teachers. New Zealand Journal of Psychology, 29, 52–60. Wolpin, J., Burke, R. J., & Greenglass, E. R. (1991). Is job satisfaction an antecedent or a consequence of psychological burnout? Human Relations, 44, 193–209. Zastrow, C. (1984). Understanding and preventing burn-out. British Journal of Social Work, 14, 141–155.

RELIGIOUS COPING MODERATES THE RELATIONSHIP BETWEEN EARLY MALADAPTIVE SCHEMA ORIGINS AND DYSPHORIA Christian T. Racine and John J. Cecero Abstract According to the model of schema therapy, early maladaptive schemas (EMS) that originate in frustrating or traumatic childhood experiences with parents and siblings are associated with a lifetime predisposition to dysphoria, problematic interpersonal relationships, and maladaptive coping styles. Results from this study lend support to this model by demonstrating that the childhood origins of many EMS are positively associated with dysphoria. Moreover, these ﬁndings support the moderating eﬀect of negative religious coping on the relationship between the EMS domainorigins and dysphoria. Limitations and suggestions for future research are discussed.

Early maladaptive schemas (EMS) are associated with a variety of negative life outcomes such as psychiatric disorders (Carine, 1997; Harris & Curtin, 2002), substance abuse (Ball & Cecero, 2001; Decouvelaere, Graziani, Gackiere-Eraldi, Rusinek, & Hautekeete, 2002), and adjustment disorders (D’Andrea, 2003). Early maladaptive schemas are “broad, pervasive themes, comprised of memories, emotions, cognitions, and bodily sensations, regarding oneself and one’s relationships with others, that are developed during childhood or adolescence, elaborated throughout one’s lifetime, and dysfunctional to a signiﬁcant degree” (Young, Klosko, Weishar, 2003). Given Pargament’s (1997) assertion that positive religious coping buﬀers the eﬀects of stress on well-being, we hypothesized that religious coping may function as a moderating variable between parental interactions believed to underscore the formation of the EMS and dysphoria in later life. Early Maladaptive Schemas: A Brief Overview Young et al. (2003) have described EMS as inﬂexible and enduring beliefs about oneself, which originate in dysfunctional experiences with caregivers, siblings, or peers during the ﬁrst few years of life. Young suggested that if the child’s needs for connection, autonomy, Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

96

racine and cecero

reasonable expectations and limits, unconditional acceptance, and spontaneity are not adequately met, the child will develop a negative view of self and others which will adversely impact social and occupational relationships as well as emotional well-being. These aﬀective and relational blueprints will then persist throughout later life, ultimately forming schemas that will be positively associated with dysphoria or anxious and depressive mood. A growing body of research has supported this theory. Ball and Cecero (2001) examined the relationship between the EMS and depressive personality disorder diagnosis in a sample of outpatients. They found that depressive personality disorder was positively related to EMS endorsement. Harris and Curtin (2002) examined the relationship between EMS and scores on the Beck Depression Inventory, second edition (BDI-II; Beck, Steer, & Brown, 1995) in a sample of undergraduates. They found that BDI-II scores were signiﬁcantly related to EMS endorsement. Schmidt and colleagues (Schmidt et al., 1995) examined the relationship between the EMS and depressive symptoms in a study of undergraduates. They found that higher overall scores of EMS correlated positively with several measures of depressed mood. In a sample of undergraduates, D’Andrea (2003) found the EMS correlated signiﬁcantly and positively with negative aﬀect. Young et al. (2003) proposed the existence of 18 EMS that can be clustered around 5 “need domains” described below. Though the EMS are described as separate constructs, people often experience several of them simultaneously (Young et al., 2003). For example, multiple schemas within the same domain are often activated at the same time, however, EMS across the ﬁve domains are also activated together. This typically produces a complex emotional and cognitive response to stressful situations. The Disconnection/Rejection domain originates in an individual’s unmet needs for security, nurturance, empathy, and respect. Young et al. (2003) proposed that these schemas stem from childhood experiences of cold, rejecting, withholding, unpredictable, or abusive families. This domain consists of ﬁve EMS: abandonment, which is the perception that others will not provide a consistent or reliable source of emotional support; mistrust/abuse, which is the expectation that signiﬁcant others will be abusive, humiliating, or manipulative; emotional deprivation, which is the expectation that signiﬁcant others will never meet one’s emotional support needs; defectiveness/shame, which

religious coping moderates the relationship

97

is a persistent sense of being defective, inferior, or unlovable; and social isolation/alienation, which is the feeling of isolation from the rest of the world. The Impaired Autonomy and Performance domain originates in childhood experiences of parental enmeshment or overprotection, where the child’s self-conﬁdence is undermined and underdeveloped. The four EMS within this domain are: dependence/incompetence, which is the feeling that one is unable to handle everyday responsibilities; vulnerability to harm, which is an exaggerated fear of unpredictable disaster; enmeshment/undeveloped self, which is an excessive emotional involvement with another person that inhibits proper social development; and failure, which is the feeling that one has failed and is fundamentally inadequate relative to others. The Impaired Limits domain originates in childhood experiences of permissiveness or overindulgence on the part of parents, depriving children of necessary realistic limits, and undermining frustration tolerance. The two EMS in this domain are: entitlement/grandiosity, which is the expectation of being able to act without regard for others, and insuﬃcient self-control/self-discipline, which is the feeling that self-control is unimportant and that emotions and impulses need little restraint. The Other-Directedness domain originates in childhood experiences of excessive focus on meeting the needs of others, primarily parents, at the expense of adequate attention to one’s own needs or wishes, in order to gain attention, approval, and love. The three EMS in this domain are: subjugation, which is the perception of coercion to surrender to another’s will in order to avoid anger, retaliation, or abandonment; self-sacriﬁce, which is the disregard for one’s own personal gratiﬁcation in favor of meeting others’ needs; and approval-seeking/recognition-seeking, which is the disregard for the development of one’s own secure sense of self in favor of gaining others’ recognition. The Over-vigilance and Inhibition domain originates in experiences with parents who emphasize performance and duty at the expense of pleasure, joy, and spontaneity. The four EMS in this domain are: negativity/vulnerability to error, which is the feeling of helplessness to prevent negative aspects of life; over-control/emotional inhibition, which is the feeling that the expression of emotion will lead to embarrassment or other negative consequences; unrelenting standards/hypercriticalness, which is the feeling that one must meet unrealistically high standards in order to avoid criticism; and punitiveness, which is the belief that people should be harshly punished for making mistakes.

98

racine and cecero Religious Coping

The concept of religion has taken on a wide variety of deﬁnitions from the concrete to the abstract and from the philosophical to the scientiﬁc. Consequently, the diverse scholarly disciplines that draw from these deﬁnitions each adopt their own applications (Zinnbauer, Pargament, & Scott, 1999). Researchers in the social sciences (Zinnbauer et al., 1999; Piedmont, 1999a; 1999b) have focused on religion and spirituality as important factors in explaining individual diﬀerences and personality. This renewed interest is borne out in the general public as well. Whereas the 1960s and 1970s saw a 60% decrease in religious involvement (Roof, 1993), recent studies have (Shorto, 1997) indicated that upwards of 90% of surveyed Americans report a belief in God and 80% believe in the power of prayer to aﬀect the course of major life events such as serious illness (Wallis, 1996). Early methods of describing and assessing religious approaches to coping with stress relied on a single construct such as church attendance or frequency of prayer (Bahr & Harvey, 1979; Sherkat & Reed, 1992). Such methods, however, left unanswered many questions as to the speciﬁc ways in which people actually used (or misused) their religion in their day-to-day lives. Pargament (2002) has addressed these questions through his studies on positive and negative religious coping. “Coping, like religion, is a process, a search for signiﬁcance” (Pargament, 1997, p. 90) as well as a means by which people orient themselves during diﬃcult times. Though Pargament (1997) refers to religious coping as having positive and negative aspects, their distinction is not simply a good-bad dichotomy. Rather the distinction is based on a diﬀerential between the ways in which people perceive a divine power operating in their lives, and how much of an active collaboration they have with that power (Pargament, 2002). Pargament (1997) has described an immerging trend in the coping literature wherein participants utilize combinations of positive and negative religious coping methods during brief periods of distress. Pargament (1997) has suggested that the combined approach of positive and negative religious coping might reﬂect an active struggle within the individual (between the two coping styles) which may beneﬁt the coping process in subtle ways. From this perspective, positive religious coping is “derived from a secure relationship with God, a sense of spirituality, a belief that there is

religious coping moderates the relationship

99

meaning to be found in life, and a sense of spiritual connectedness with others” (Pargament, 2002, p. 171). Negative religious coping methods come from a “general religious orientation that is, itself, in tension and turmoil, marked by a shaky relationship with God, a tenuous and ominous view of the world, and a religious struggle in the search for signiﬁcance” (Pargament, 2002, p. 171). Positive religious coping methods include spiritual support and benevolent religious reframing, whereas negative religious coping entails religious turmoil, frustration, and punishment. Given such a deﬁnition, it is hard to imagine any beneﬁt from negative religious coping. But perhaps coping, like religion, is made up of a constant waxing and waning of perceived comfort and happiness, fear and shame. When some religious negotiate such challenges, they believe their faith is strengthened, so perhaps a mix of positive and negative religious coping methods likewise strengthens the long-term coping process. Regardless of whether or not this combined approach does actually beneﬁt the coping process, the literature is clear that in the face of high positive coping, people can make better use of small amounts of negative religious coping (Pargament, 1997). Numerous studies have demonstrated the beneﬁts of positive religious coping on adjustment during times of acute distress (Neighbors, Jackson, Bowman, & Gurin, 1983; Veroﬀ, Douvan, & Kulka, 1981; and Bijur, Wallston, Smith, Lifrak, & Friedman, 1993). In a recent meta-analysis of 147 studies involving religiousness and depressive symptoms, Smith, McCullough, and Poll (2003) found that negative religious coping was signiﬁcantly, positively associated with depressive symptoms while positive religious coping was signiﬁcantly, negatively associated with depressive symptoms. Several cross-sectional studies (Grosse-Holtforth, Pathak, Koenig, Cohen, Pieper, & VanHook, 1996; Koenig, 1995; Koenig et al., 1992; Koenig et al., 1995) have established negative correlations between self-reports of religious coping and depressive symptoms. Koenig and colleagues (Koenig et al., 1992) conducted longitudinal research with elderly medical inpatients and reported that increased use of positive religious coping at baseline was negatively associated with depressive symptoms at six-month follow-up (beta = –0.18, p = 0.01), while controlling for baseline depression levels. In a series of studies with 577 medically ill elderly people, Pargament and colleagues (Koenig, Pargament, & Nielsen, 1998; Pargament et al., 2000) found that positive religious coping methods

100

racine and cecero

were negatively related to depressive symptoms and negative religious coping methods were positively related with depressive symptoms (Koenig, Pargament, and Nielsen, 1998). These studies provide support for a connection between religious coping and depressed mood in medically ill populations, but say little about the generalizability to non-medically ill populations. In a sample of undergraduates, Pargament et al. (1994) found that positive coping methods were associated with positive aﬀect, and Pargament et al. (1995) found similar results in a study of ﬂood victims. In a study of undergraduates, Belavich and Pargament (1995) found that positive religious coping was signiﬁcantly correlated with positive aﬀect. Pargament and colleagues (Pargament, Ishler, Dubow, Stanik, Rouiller, Crowe, Cullman, Albert, & Royster, 1994) examined the relationship between positive and negative religious coping methods and aﬀect in a sample of undergraduates two days prior to the 1990 Gulf War ground assault, and one week after suspension of hostilities. Prior to the ground assault, the authors found that positive religious coping was signiﬁcantly and positively associated with positive aﬀect, while negative religious coping was signiﬁcantly and positively associated with negative aﬀect. After cessation of ﬁghting, the authors found that positive coping was signiﬁcantly and negatively associated with negative aﬀect. These studies suggest that the relationship between religious coping and mood generalizes to nonmedically ill populations but does not fully address the question of generalizability to non-acutely distressed populations. Speciﬁcally, two questions arose: Does religious coping have a similar relationship to mood at non-acute levels of stress; and If there is a relationship, does it vary with the level of stress? In an eﬀort to examine the eﬀects of religious coping on nonacutely distressed populations, researchers (Wheaton, 1985, and Pargament, 1997) have presented three models for the potential moderating eﬀects of religious coping. In the religious stress moderator model, religious coping has no impact on positive outcomes at low levels of stress (Pargament, 1997). As stress levels increase, however, individuals utilizing high levels of positive religious coping fare much better than do those utilizing lower levels of religious coping. In the second model, labeled religious stress deterrent, individuals beneﬁt equally from higher levels of positive religious coping in both low and high stress situations. The third model, labeled the combined religious moderator-deterrent model, is an amalgam of the ﬁrst and second models

religious coping moderates the relationship

101

wherein individuals beneﬁt from higher levels of positive religious coping at both high and low levels of stress (similar to the second model), however, the beneﬁts of higher religious coping increase as stress increases (similar to the ﬁrst model). Pargament (1997) reported empirical support for each of the three models when tested in situations involving acute stress. Though it is not yet clear which of the three models best represents the actual relationship between religious coping and mood, the literature supports the general idea that positive religious coping is beneﬁcial at both high and low levels of stress (Pargament, 1997). In the present study, we attempted to lend support to this idea by examining the moderating eﬀects of religious coping on self-reports of dysphoria in a non-acutely distressed population. We intended to show that religious coping does have eﬀects on mood, even at low levels of distress, however, our paradigm is not contingent on any of the three models. Goals of the Present Study in Examining the Relationships Between EMS Domain-Origins, Dysphoria, and Religious Coping Young (1994, 1999) originally based his EMS theory on clinical observations and practice, but recent studies, already discussed, have provided empirical support for his theory (Lee, Taylor, & Dunn, 1999; Leung, Waller, & Thomas, 1999; Schmidt, Joiner, Young, & Telch, 1995). Our ﬁrst objective in the present study was to add to this support, speciﬁcally for Young’s hypothesized relationship between the ﬁve domain-origins and dysphoria. Researchers have demonstrated connections between EMS and psychiatric disorders (Carine, 1997) as well as depressive symptoms in both clinical (Ball & Cecero, 2001; Welburn, Coristine, Dagg, Pontefract, & Jordan, 2002) and undergraduate samples (Harris & Curtin, 2002; Schmidt, Joiner, Young, & Telch, 1995) but little empirical work has examined the relationship between the EMS and less acute forms of distress in non-clinical samples. We predicted that parental interactions that are hypothetically consistent with the formation of early maladaptive schemas would be positively correlated with dysphoria. Speciﬁcally, we predicted that the EMS domain-origins (as measured by the Young Parenting Questionnaire: YPQ ; Young 1994) would correlate positively with a dysphoric aﬀect variable (as measured by the

102

racine and cecero

Multiple Aﬀect Adjective Check List—Revised: MAACL-R, Zuckerman & Lubin, 1985). Further, we predicted that religious coping would moderate the relationship between the EMS domain-origins and dysphoria such that negative religious coping would intensify the relationship while positive religious coping would attenuate the relationship. Such ﬁndings would contribute to the understanding of the eﬀects of religious coping in non-acutely stressful situations. Method Participants Participants in this experiment were 152 female (67.3%) and 67 male (29.6%) undergraduates from a private, Catholic, northeastern U.S. university. Participants volunteered for a study examining emotional well-being and spirituality in exchange for introductory psychology course credit. Participants’ ethnicities were: Caucasian, n = 167 (73.9%); Latino, n = 21 (9.3%); Asian, n = 17 (7.5%); African American, n = 14 (6.2%); Native American, n = 2 (0.9%); and Indian, n = 1 (0.4%). Participants identiﬁed religious aﬃliations were: Catholic, n = 143 (63.3%); Protestant, n = 28 (12.4%); Jewish, n = 7 (3.1%); Orthodox, n = 5 (2.2%); Muslim, n = 4 (1.8%); Buddhist, n = 3 (1.3%); and Other, n = 32 (14.2%). Though the sample was comprised of diverse ethnic, cultural, and religious backgrounds, the sample was predominantly Catholic Caucasians. Procedure Upon arriving at the testing location, participants gave informed consent and completed a series of self-report measures which took approximately 60 minutes to complete. Upon completion of the measures, each participant was given a debrieﬁng form and was given the opportunity to ask questions of the experimenter present. Materials Demographic Questionnaire This scale asked participants about their age, gender, and religious aﬃliation. Participants endorsed Likert-type items concerning their

religious coping moderates the relationship

103

frequency of prayer, reading of Biblical and/or other religious literature, and the extent to which they have a unifying relationship with God. Young’s Parenting Questionnaire (YPQ ) (Young, 1994) The YPQ is a 72-item questionnaire designed to assess the hypothetical origins of the 5 schema-domains previously described. Participants rated their mothers and fathers on a 6-point Likert-type scale (1 = completely untrue; 2 = mostly untrue; 3 = slightly more true than untrue; 4 = moderately true; 5 = mostly true; 6 = describes him/her perfectly). Internal consistency for domains ranged from a = 0.47 to a = 0.91 (Milewski et al., 2000, unpublished masters’ thesis). In the present study, internal consistencies ranged from a = .86 to a = .90, n = 202. Though the YPQ has not yet been used frequently in the literature, it is a variation of the Young Schema Questionnaire (YSQRE2R; Young & Brown, 2001) which has been validated in several studies (Lee, Taylor, & Dunn, 1999; Schmidt, Joiner, Young, & Telch, 1995; Welburn, Coristine, & Dagg, 2002; Young, 1998; Young & Brown, 1990). Multiple Aﬀect Adjective Check List—Revised (MAACL-R) (Zuckerman & Lubin, 1985) The MAACL-R is a 132-item, self-administered checklist, which provides information regarding the participants’ aﬀect. The directions can be adjusted to instruct the respondents to check adjectives that describe how they “feel right now” (trait aﬀect) or how they “generally feel” (trait aﬀect). This study utilized the trait format of the scale. The MAACL-R contains ﬁve unipolar scales (Anxiety, Depression, Hostility, Positive Aﬀect, and Sensation Seeking), as well as two composite scales of dysphoria (obtained by summing Anxiety, Depression, and Hostility subscales) and PASS (obtained by summing Positive Aﬀect and Sensation Seeking) (Lubin & van Whitlock, 1998). In the present study, we were, however, solely concerned with the Dysphoria composite variable. These scales were derived from factor analyses with two groups of undergraduates (n = 112 and n = 424). Both groups were given the MAACL-R on two occasions, ﬁve days apart. Results from the factor analysis (FA) produced ﬁve factors (Anxiety, Depression, Hostility, Positive Aﬀect, and Sensation seeking) which accounted for 51% of the total variance. All factors loaded 0.30 or higher on a single factor. The Anxiety, Depression, and Hostility

104

racine and cecero

subscales were highly intercorrelated which indicated that they were not independent (Zuckerman, Lubin, & Rinck, 1983). Zuckerman and colleagues (Zuckerman, Lubin, Rinck, Soliday, Albott, & Carlson, 1985) examined the discriminant validity of the MAACL-R trait form with three samples (118 psychiatric outpatients, 82 psychiatric inpatients, and 200 non-clinical people with the same demographic breakdown as the patient samples in terms of age, race, and education). Diagnosis for the patient samples was based on DSMIII (APA, 1980). The authors utilized discriminant function analysis to assess the diagnostic discrimination between the groups. Participants meeting criteria for Major Depression scored signiﬁcantly higher than all diagnostic groups on the Anxiety, Depression, and Dysphoria scales of the MAACL-R. The MAACL-R correctly identiﬁed depressive individuals 72% of the time and non-clinical people 60% of the time supporting the discriminant validity of the trait format of the scale. Researchers (Lubin, Cain, & van Whitlock, 1992) have demonstrated the MAACL-R’s reliability and validity with a sample of 43 undergraduates. The authors reported signiﬁcant correlates between the MAACL-R Anxiety, Depression, Hostility, and Dysphoria scales and the “Unfavorable Adjectives” scale of the Adjective Check List (ACL; Gough & Heilbrun, 1965). In a study with 1,179 undergraduates, pre-college adolescents, and outpatients, Lubin and van Whitlock (1998) examined the internal consistency and test-retest reliability of the MAACL-R trait form. The authors reported that internal consistency reliabilities for the combined Anxiety/Depression scales were: a = 0.86 for the undergraduate sample (n = 433); a = 0.82 for the adolescent population (n = 746); and a = 0.86 for the outpatient population (n = 202). The internal consistency reliabilities for the Hostility scales were: a = 0.74 for the undergraduate sample (n = 433); a = 0.71 for the adolescent population (n = 746); and a = 0.76 for the outpatient population (n = 202). The authors reported that the test-retest reliability for the Anxiety/Depression scale at two weeks and eight weeks was r = 0.61 and r = 0.44, p < .001 respectively. The test-retest reliability for the Hostility scale at two weeks and eight weeks was r = 0.62 and r = 0.50, p < .001 respectively. The authors did not report degrees of freedom. In the present study, the internal consistency reliability for the Dysphoria composite variable (comprised of the Depression, Anxiety, and Hostility scales) was a = .73, n = 223.

religious coping moderates the relationship

105

Brief Religious Coping Questionnaire (Brief-R-COPE) The Brief-R-COPE asked participants to ﬁrst consider a negative event in their lives and then to rate the extent to which they engaged in speciﬁc religiously oriented coping methods. The scale is made up of 14 Likert-type items (1 = not at all; 2 = somewhat; 3 = quite a bit ; 4 = a great deal ). The Brief-R-COPE, was originally written to study the Oklahoma City federal building bombing victims (Pargament et al., 1998). Pargament and colleagues (Pargament et al., 1998) conducted several principal-components analysis (PCA) with three samples, including 296 Oklahoma City bomb victims, 540 undergraduates who reported having experienced a signiﬁcant stressor in the past three years, and 551 medically ill, adult inpatients. All three PCAs produced two-factor solutions (consistent with the positive and negative religious coping constructs) with signiﬁcant chi-squares and adequate reliabilities (internal consistency ranges from a = .69 to a = .90; Pargament, Koenig, & Perez, 2000). Pargament and colleagues (Pargament et al., 2000) then conducted a conﬁrmatory factor analysis (CFA) to verify the factor structure obtained from the principal-components analyses. The CFA model produced a solution with acceptable ﬁt (chi-square = 2406.46, df = 1133, p < .05). In the present study, internal consistency reliabilities were a = .93 and a = .83 for the positive and negative religious coping subscales respectively (n = 222). Results Correlations of EMS Domain-Origins and Dysphoria We examined the relationship between the diﬀerent EMS domainorigins by separately correlating the mother and father EMS domainorigins with the Dysphoria composite variable (See Table 1). Three of the ﬁve Mother EMS domain-origins signiﬁcantly correlated with Dysphoria (Disconnection/Rejection-Mother (DR-M), r = 0.17, p < .05; Impaired Autonomy-Mother (IA-M), r = 0.19, p < .05; Impaired Limits-Mother (IL-M), r = 0.15, p < .05). Two of the Mother EMS domain-origins, Other Directedness (OD-M), and Overvigilance/ Inhibition (OV-M), were positively correlated with Dysphoria, though they failed to reach signiﬁcance (OD-M, r = 0.11, p < .05; OV-M, r = 0.08, p < .05). Four of the ﬁve Father EMS domain-origins were signiﬁcantly correlated with Dysphoria (Disconnection/Rejection-

racine and cecero

106

Father (DR-F), r = 0.35, p < .05; Impaired Autonomy-Father (IAF), r = 0.16, p < .05; Impaired Limits-Father (IL-F), r = 0.24, p < .05; and Other Directedness-Father (OD-F), r = 0.19, p < .05). One of the Father EMS domain-origins, Overvigilance/InhibitionFather (OV-F), had virtually no correlation to Dysphoria (OV-M, r = 0.01). This supported Young’s (1999) hypothesis that the parental origins of EMS are associated with negative trait aﬀect in later life. Table 1. Correlation of Mother and Father EMS Domain-Origins and the Dysphoria Composite Variable EMS domain-origin

Dysphoria r

1. 2. 3. 4. 5.

Mother ratings Disconnection Rejection (DR-M) Impaired Autonomy (IA-M) Impaired Limits (IL-M) Other Directedness (OD-M) Overvigilance and Inhibition (OV-M)

0.17* 0.19** 0.15* 0.11 0.08

1. 2. 3. 4. 5.

Father ratings Disconnection Rejection (DR-F) Impaired Autonomy (IA-F) Impaired Limits (IL-F) Other Directedness (OD-F) Overvigilance and Inhibition (OV-F)

0.35** 0.16* 0.24** 0.19** 0.01

* p < .05 ** p < .01

Regression Analysis as a Test of Moderation To examine possible moderator eﬀects, we performed multiple regressions with each of the seven EMS domain-origins (DR-M, IA-M, IL-M, DR-F, IA-F, IL-F, and OD-F) that correlated signiﬁcantly with the Dysphoria outcome variable. In each of the regressions, the EMS domain-origin was entered ﬁrst, followed by an interaction term created by multiplying the EMS domain-origin by the moderating variable (i.e. the positive or negative religious coping score from the Brief R-COPE). Moderation was found to occur when the interaction term produced a signiﬁcant beta score.

religious coping moderates the relationship

107

Negative Religious Coping Consistent with the hypotheses, negative religious coping signiﬁcantly moderated the relationship between all three of the mother EMS domain-origins and the Dysphoria composite variable (Disconnection/Rejection-Mother: beta = .282, F (1,207) = 3.42, p < .05; Impaired Autonomy-Mother: beta = .236, F (1,217) = 4.15, p < .05; Impaired Limits-Mother: beta = .235, F (1,217) = 2.72, p < .05; See Table 2). Table 2. Moderation Eﬀects of Negative Religious Coping on Dysphoria Moderator

Beta

F

Degrees of Freedom

DR-M IA-M IL-M DR-F IA-F IL-F OD-F

.282** .236** .235** .148 .238** .220* .256**

3.42 4.15 2.72 14.0 2.63 6.47 3.96

(1,207) (1,217) (1,217) (1,200) (1,212) (1,212) (1,210)

* p < .05 ** p < .01 DR = Disconnection-Rejection IA = Impaired Autonomy IL = Impaired Limits OD = Other Directedness M = Mother F = Father

Consistent with the hypotheses, negative religious coping signiﬁcantly moderated the relationship between three of the four father EMS domain-origins and the Dysphoria composite variable (Impaired Autonomy-Father: beta = .238, F (1,212) = 2.63, p < .05; Impaired Limits-Father: beta = .220, F (1,212) = 6.47, p < .05; Other Directedness-Father: beta = .256, F (1,210) = 3.96, p < .05). Negative religious coping did not signiﬁcantly moderate the relationship between the Disconnection/ Rejection-Father EMS domain-origin and Dysphoria (beta = .148, F (1,200) = 14.0, p > .05). Positive Religious Coping Contrary to the hypotheses, positive religious coping did not signiﬁcantly moderate the relationship between any of the EMS domain-origins and the Dysphoria composite variable (See Table 3). Though this lack of signiﬁcance may accurately represent the relationship between positive religious coping and the predictor and

racine and cecero

108

Table 3. Moderation Eﬀects of Positive Religious Coping on Dysphoria Moderator

Beta

F

Degrees of Freedom

DR-M IA-M IL-M DR-F IA-F IL-F OD-F

.081 .049 .640 .750 .005 .069 .033

6.63 7.87 6.44 15.3 6.50 9.53 7.96

(1,207) (1,217) (1,217) (1,200) (1,212) (1,212) (1,210)

* p < .05 ** p < .01 DR = Disconnection-Rejection IA = Impaired Autonomy IL = Impaired Limits OD = Other Directedness M = Mother F = Father

outcome variables, it may also have been aﬀected by other factors (e.g. small sample size as well as the measures used to assess EMS). To further explore the relationship with positive religious coping, we conducted a more liberal test of moderation. We ﬁrst performed a median split of the positive religious coping values. This creates two groups representing high positive religious coping (i.e. the 50 percent above the median) and low positive religious coping (i.e. the 50 percent below the median). In accordance with the method described by Baron and Kenny (1986), we then entered the EMS domainorigins with signiﬁcant positive correlations (DR-M, IA-M, IL-M, DR-F, IA-F, IL-F, and OD-F) into a regression analysis with the dysphoria composite variable while controlling for R-COPE values. In this method, moderation is suggested (though more tentatively) when the unstandardized beta weights for the low positive religious coping group are signiﬁcant, but the high positive religious coping group is not signiﬁcant. Consistent with our hypotheses, unstandardized beta weights for all three mother EMS domain-origins (DR-M, IA-M, and IL-M) and Dysphoria were signiﬁcant at low levels of positive religious coping, but were not signiﬁcant at high levels of religious coping which suggests a moderation relationship (See Table 4). Also consistent with our hypotheses, unstandardized beta weights for three of the four father EMS domain-origins (IA-F, IL-F, and OD-F) and Dysphoria

religious coping moderates the relationship

109

were signiﬁcant at low levels of positive religious coping, but were not signiﬁcant at high levels of religious coping which suggests a moderation relationship (See Table 4). Contrary to our hypotheses, the unstandardized beta weights for the DR-F domain-origin were signiﬁcant at both low and high levels of positive religious coping which does not suggest a moderation relationship. It is important to note that the median split approach of exploring moderator eﬀects is only tentative, and we do not know if positive religious coping acted as a signiﬁcant moderator in the relationship. It does, however, raise the question of moderation that can be explored further. Table 4. Moderation Eﬀects of Positive Religious Coping on Dysphoria Utilizing a Median Split of the Religious Coping Data EMS Domain-Origins

(DR-M) (IA-M) (IL-M)

(DR-F) (IA-F) (IL-F) (OD-F)

Degree of (+) R-Cope

Dysphoria B b

Mother ratings Low (n = 110) High (n = 98) Low (n = 117) High (n = 101) Low (n = 117) High (n = 101)

0.22* 0.12 0.23* 0.14 0.24** 0.02

0.58* 0.22 0.43* 0.20 0.24** 0.02

Father ratings Low (n = 104) High (n = 97) Low (n = 113) High (n = 100) Low (n = 113) High (n = 100) Low (n = 111) High (n = 100)

0.34** 0.36** 0.19* 0.12 0.30** 0.11 0.20* 0.18

0.86** 0.93** 0.27* 0.17 0.33** 0.17 0.27* 0.27

* p < .05 ** p < .01 DR = Disconnection-Rejection IA = Impaired Autonomy IL = Impaired Limits OD = Other Directedness M = Mother F = Father

110

racine and cecero Discussion

Signiﬁcant positive correlations between 7 of the 10 EMS domainorigins and negative trait aﬀect support Young et al.’s (2003) theory that Early Maladaptive Schema origins relate to negative life outcomes such as negative mood. Speciﬁcally, parental interactions that were consistent with schema formation positively correlated with dysphoria (see Table 1). The Disconnection-Rejection, Impaired Autonomy, and Impaired Limits domain-origins for both mothers and fathers were all signiﬁcantly correlated with dysphoria, supporting our hypotheses that the parental styles associated with EMS formation would be associated with negative trait aﬀect. Our hypotheses concerning this connection were not wholly supported, however, in that the Other-Directedness domainorigin was signiﬁcantly correlated only for the father variables, whereas in the mother variables, the positive correlation was non-signiﬁcant. This might suggest that the kinds of maladaptive parenting styles associated with the formation of the Other-Directedness schema (such as subjugation of will) are more toxic when associated with the father than with the mother. These results might also suggest something particular to our sample such as the eﬀects of a restricted age range (largely college sophomores). Given their ages, our sample might be especially sensitive to perceived paternal over-control (e.g. recalling dad’s authority as being especially harsh while remembering mom as the warmer, more nurturing ﬁgure). Results involving the moderating eﬀects of religious coping were curious in that negative religious coping appeared to have a more pronounced eﬀect on the relationship between the EMS domain-origins and dysphoria than did positive religious coping. In a standard test of moderation (i.e. regression with the interaction variable), negative religious coping signiﬁcantly moderated the relationship between six of the seven EMS domain-origins (Disconnection/Rejection-Mother, Impaired Autonomy-Mother, Impaired Autonomy-Father, Impaired Limits-Mother, Impaired Limits-Father, and Other DirectednessFather) and dysphoria, while positive religious coping did not signiﬁcantly moderate any of the relationships. This provides partial support for our second hypothesis which proposed that negative religious coping would intensify the damaging eﬀects of the EMS domainorigins on mood. It did not, however, support our hypothesis that positive religious coping would attenuate the damaging eﬀects.

religious coping moderates the relationship

111

The parenting styles and childhood experiences that underscore the formation of the Early Maladaptive Schemas are likely echoed in the unstable relationships and tenuous support-seeking (from God or a spiritual being) that is associated with negative religious coping. Young and colleagues (Young et al., 2003) deﬁne the Disconnection/ Rejection, Impaired Autonomy, and Impaired Limits EMS domains as being the most severe and intractable. Therefore it is not surprising that they would have the strongest connections with dysphoria and be the most eﬀected by negative religious coping. Similarly, it is possible that those people who formed EMS through problematic or chaotic parental interactions would be reluctant to view God (whom they may view as just another potential caregiver) as a stable source of support. This may partially explain why positive religious coping did not appear to signiﬁcantly buﬀer the eﬀects of the EMS domain-origins. It is unclear, however, why the Disconnection/ Rejection-Mother Domain-Root was moderated while the Disconnection/Rejection-Father Domain-Root was not. It is possible that the support seeking expectations for mother and father are diﬀerentially represented in the EMS domain-roots. However, the more liberal test of moderation did suggest that positive religious coping might moderate the relationship between six of the seven EMS domain-origins and dysphoria. The same six domainorigins in fact that were moderated by negative religious coping. Based on these tentative results, it is possible that further examination might reveal a signiﬁcant moderating relationship with positive religious coping. For example, it is possible that moderating eﬀects may have been obscured by the measurement of schema origins rather than direct measures of the schemas themselves. Parental interactions believed to underlie EMS formation are less clearly deﬁned and assessed than are the EMS. Further research utilizing direct measures of the EMS (such as the Young Schema Questionnaire) might contribute further to the results of the present study. Though numerous studies have examined similar moderator relationships with religious coping and aﬀect (see Pargament, 1997 for a review), most of these studies have relied on speciﬁc, acute lifestressors as criterion variables. Results of the present study are therefore somewhat unique in that they replicate the moderator eﬀect with a non-acute, trait variable. This further supports the integrative moderator-deterrent model of religious coping described previously wherein religious coping is thought to operate just as well at

112

racine and cecero

lower levels of stress as it does at high levels of stress (Pargament, 1997). Limitations and Future Research Directions Young’s (1994, 1999) original formulation was based on work with a clinical sample, whereas our sample was comprised entirely of a non-clinical and generally homogeneous sample of students from a private, Catholic university. Though the religious aﬃliation of the subject pool would not necessarily be an issue (religious coping being a religion-nonspeciﬁc concept), the “religiousness” of the environment might be. Total reliance on self-report measures of spirituality, early maladaptive schemas, religious coping, and dysphoria might also present a problem in interpreting these results. Corroborative measures or third-party measures from parents, family members, or friends might improve reliability. This also highlights the limitations of our cross-sectional, correlative design from which no causal inferences can be drawn. It would be more informative to examine longitudinal assessments of the formation and ongoing activation of the EMS with measures of negative life outcomes. Participants volunteered for this study knowing in advance that it was going to examine their views on spirituality and their own adaptive functioning. It is very possible that this prior knowledge may have led some participants to bias their responses in order to present their families, themselves, and their methods of coping with distress in a more favorable light. Furthermore, even the statistically signiﬁcant correlations were moderate at best (ranging from r = .11 to r = .35). As we discussed above, a potential explanation for this might be that we measured the proposed origins of EMS domains and not the actual schemas themselves. Though the schemas are logically hypothesized to relate directly to the domain-origins, it is possible that the current self-report instruments (i.e. the YPQ ) might not be sensitive enough to the schematic origins to validly assess them in a retrospective format. Further replication of the present paradigm using direct schema measures (such as the YSQ-RE2R: Young & Brown, 2001) as well as a heterogeneous, clinical sample is indicated. Additionally, Pargament (1997) suggests that closed-ended assessment measures of religious coping tend to produce higher percentages of endorsement than do

religious coping moderates the relationship

113

open-ended measures. Replication of this paradigm with multiple open- and closed-ended measures is therefore indicated. References American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. Bahr, H. & Harvey, C. (1979). Correlates of loneliness among widows bereaved in a mining disaster. Psychological Reports, 44, 367–385. Ball, S., & Cecero, J. (2001). Addicted patients with personality disorders: Traits, schemas, and presenting problems. Journal of Personality Disorders, 15(1), 72–83. Baron, R. M. & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Beck, A. T., Steer, R., & Brown, G. (1995). Beck Depression Inventory—II manual. San Antonio, TX: The Psychological Corporation. Belavich, T., & Pargament, K. (1995). The role of religion in coping with daily hassles. Paper presented at the meeting of the American Psychological Association, New York, NY. Bijur, P., Wallston, K., Smith, C., Lifrak, S., & Friedman, S. (1993). Gender diﬀerences in turning to religion for coping. Paper presented at the meeting of the American Psychological Association, Toronto, Ontario. Carine, B. E. (1997). Assessing personal and interpersonal schemata associated with Axis II Cluster B personality disorders: An integrated perspective. Dissertation Abstracts International, 58, 1B. D’Andrea, J. (1999). Early maladaptive schemas, religious problem solving, and psychological adjustment. Unpublished doctoral dissertation, Fordham University, New York. Decouvelaere, F., Graziani, P., Gackiere-Eraldi, D., Rusinek, S., Hautekeete, M. (2002). Hypothesis of existence and development of early maladaptive schemas in alcohol-dependent patients. Journal de Therapie Comportementale et Cognitive, 12(2), 43–48. Gough, H., & Heilbrun, A., Jr. (1965). The adjective check list: Manual. Palo Alto, CA: Consulting Psychologists Press. Grosse-Holtforth, M., Pathak, A., Koenig, H., Cohen, H., Pieper, C., & VanHook, L. (1996). Medical illness, religion, health control and depression of institutionalized medically ill veterans in long-term care. International Journal of Geriatric Psychiatry, 11, 613–620. Harris, A., & Curtin, L. (2002). Parental perceptions, early maladaptive schemas, and depressive symptoms in young adults. Cognitive Therapy and Research, 26(3), 405–416. Harvey, C., Barnes, G., & Greenwood, L. (1987). Correlates of morale among Canadian widowed persons. Social Psychiatry, 22, 65–72. Koenig, H., Cohen, H., Blazer, D., Pieper, C., Meador, K., Shelp., Goli, V., DiPasquale, R. (1992). Religious coping and depression in elderly hospitalized medically ill men. American Journal of Psychiatry, 149, 1693–1700. Koenig, H., Cohen, H., Blazer, D., Kudler, Krishnan, K., & Sibert, T. (1995). Cognitive symptoms of depression and religious coping in elderly medical patients. Psychosomatics, 36, 369–375. Koenig, H. (1995). Religion and older men in prison. International Journal of Geriatric Psychiatry, 10, 219–230. Koenig, H., Pargament, K., & Nielsen, J. (1998). Religious coping and health status in medically ill hospitalized older adults. Journal of Nervous and Mental Disease, 186, 513–521.

114

racine and cecero

Lee, C. W., Taylor, G. & Dunn, J. (1999). Factor structure of the schema questionnaire in a large clinical sample. Cognitive Therapy and Research, 23(4), 441–451. Leung, N., Waller, G. & Thomas, G. J. (1999). Core beliefs in anorexic and bulimic women. Journal of Nervous & Mental Disorders, 187(12), 736–741. Lubin, B. & Van Whitlock, R. (1998). A grade four reading level key for the multiple aﬀect adjective check list-revised. Perceptual and Motor Skills, 86, 119–125. Maton, K. (1989). The stress-buﬀering role of spiritual support: Cross-sectional and prospective investigations. Journal for the Scientiﬁc Study of Religion, 28, 310– 323. Menaghan, E. (1983). Individual coping eﬀorts: Moderator of the relationship between life stress and mental health outcomes. In H. B. Kaplan (Ed.), Psychosocial stress: Trends in theory and research (pp. 157–191). New York: Academic Press. Milewski, G. B., Cecero, J. J., & Bernstein, D. P. (2000). The psychometric properties of the young parenting inventory (YPI): A new measure of early maladaptive schemas. Unpublished master’s thesis, Fordham University, New York. Neighbors, H., Jackson, J., Bowman, P., & Gurin, G. (1983). Stress, coping, and black mental health: Preliminary ﬁndings from a national study. In R. Hess & J. Hermalin (Eds.), Innovation in prevention (pp. 5–29). New York: Haworth Press. Pargament, K. (1997). The psychology of religion and coping: Theory, research, practice. New York: Guilford Press. —— (2002). The bitter and the sweet: An evaluation of the costs and beneﬁts of religiousness. Psychological Inquiry, 13, 168–181. Pargament, K., Ishler, K., Dubow, E., Stanik, P., Rouiller, R., Crowe, P., Cullman, E., Albert, M., & Royster, B. (1994). Methods of religious coping with the Gulf War: Cross-sectional and longitudinal analyses. Journal for the Scientiﬁc Study of Religion, 33, 347–361. Pargament, K. Koenig, H., & Perez, L. (2000). The many methods of religious coping: Development and initial validation of the RCOPE. Journal of Clinical Psychology, 56(4), 519–543. Petrocelli, J., Glaser, B., Calhoun, G., & Campbell, L. (2001). Early maladaptive schemas of personality disorder subtypes. Journal of Personality Disorders, 15(6), 546–559. Piedmont, R. L. (1999a). Does spirituality represent the sixth factor of personality? Spiritual transcendence and the ﬁve-factor model. Journal of Personality, 67, 985–1013. —— (1999b). Spiritual Transcendence and the Scientiﬁc Study of Spirituality. Unpublished manuscript. Roof, W. (1993). A generation of seekers: The spiritual journeys of the baby-boomer generation. San Francisco: Harper Collins. Schmidt, N. B., Joiner, T. E., Young, J. E., & Telch, M. J. (1995). The Schema questionnaire: Investigation of psychometric properties and the hierarchical structure of a measure of maladaptive schemata. Cognitive Therapy and Research, 19(3), 295–321. Sherkat, D. & Reed, M. (1992). The eﬀects of religion and social support on selfesteem and depression among the suddenly bereaved. Social Indicators Research, 26, 259–275. Shorto, R. (1997, December 7). Belief by the numbers. The New York Times Magazine, pp. 60–61. Smith, T., McCullough, M., & Poll, J. (2003). Religiousness and depression: Evidence for a main eﬀect and the moderating inﬂuence of stressful life events. Psychological Bulletin, 129, 614–636. Veroﬀ, J., Douvan, E., & Kulka, R. (1981). The inner American: A self-portrait from 1957 to 1976. New York: Basic Books. Wallis, C. (1996, June 24). Faith and healing. Time, 147, 58–63. Welburn, K., Coristine, M., & Dagg, P. (2002). The schema questionnaire short form: Factor analysis and relationship between schemas and symptoms. Cognitive Therapy & Research, 26, 519–530.

religious coping moderates the relationship

115

Wheaton, B. (1985). Models for the stress-buﬀering functions of coping resources. Journal of Health and Social Behavior, 26, 352–364. Young, J. (1994). Young parenting inventory (YPI). New York: Cognitive Therapy Center of New York. —— (1999). Cognitive therapy for personality disorders: A schema-focused approach (3rd ed.). Sarasota, FL: Professional Resource Press. Young, J., & Brown, G. (2001). Young schema questionnaire, 2nd edition, revised. New York. —— (1990). Young schema questionnaire. New York: Cognitive Therapy Center of New York. Young, J., Klosko, J., & Weishaar, M. (2003). Schema therapy: A practitioner’s guide. New York: Guilford Press. Zinnbauer, B. J., Pargament, K. I., & Scott, A. B. (1999). The emerging meanings of religiousness and spirituality: Problems and prospects. Journal of Personality, 67, 889–919. Zuckerman, M., & Lubin, B. (1985). Manual for the multiple aﬀect adjective check Listrevised. San Diego, CA: Educational & Industrial Testing Service. Zuckerman, M., Lubin, B., & Rinck, C. M. (1985). Construction of new scales for the multiple aﬀect adjective check list. Journal of Behavioral Assessment, 5, 119–129.

GENDER DIFFERENCES IN STRESS AMONG PROTESTANT CLERGY: AN EXPLORATORY STUDY Stephen Joseph Fichter* Abstract The purpose of this study was to identify more clearly the factors associated with stress among clergy of both genders. Protestant clergy from Bergen County (N = 70) were sent surveys which included the Religious Occupations and Stress Questionnaire (ROS). Thirty clergy responded, of which 13 were female and 17 were male. While there was no statistical diﬀerence between female and male clergy with respect to overall stress scores, other factors associated with stress resulted in statistically signiﬁcant ﬁndings, for example, number of years in ministry, membership in clergy support groups, and perceived aﬃrmation by parishioners.

In particularly stress-ﬁlled moments, from the personal and familial to the societal and national, many people turn to their clergy in their search for meaning and comfort. Whether the rabbis, priests, or ministers try to counsel an individual in the aftermath of a tumultuous divorce, or attempt to reassure their communities about the meaning of life after a tragedy of the caliber of Sept. 11, 2001, clergy of all denominations frequently endure at times high levels of stress. Numerous eﬀorts have been made to identify the sources of stress among the clergy (Dewe, 1987; Hatcher & Underwood, 1990; Malony, 1988) and yet as Birk, Rayburn, and Richmond (2001) have pointed out in their explanation of the study of religious occupations and stress, for the most part, clergy of both genders are often an overlooked segment of the general population that experiences stress. In addition to the ordinary stressors that come as an integral part of their pastoral ministry, clergywomen often experience diﬃculties * Stephen Joseph Fichter plans to pursue a Ph.D. in the sociology of religion at Rutgers University. He earned his master of social work from Fordham University and studied spirituality, philosophy, and theology in Europe for 12 years in preparation for his priestly ordination. He currently serves at Saint Gabriel’s Parish in Saddle River, New Jersey while teaching philosophy and religion at Bergen Community College. He is the grandnephew of Joseph Fichter, Jesuit pioneer in the Sociology of Religion. A special word of thanks is due to Dr. Janna Heyman and Dr. Carole Rayburn for their guidance and support. Correspondence concerning this article should be addressed to Stephen Joseph Fichter at stephen ﬁ[email protected] Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

118

stephen joseph fichter

due to their minority status in a profession that has been dominated by males for so many centuries (Charlton, 1997; McDuﬀ & Mueller, 1999; Piper, 1995; Royle, 1987). Myriad articles abound concerning the division that the topic of women’s ordination has caused within local congregations and entire churches. It is not only the Roman Catholic Church that adamantly reserves priestly ordination to men alone as Pope John Paul II laid out clearly in his 1994 letter, Ordinatio Sacerdotalis. In 1997, for example, two Reformed Presbyterian denominations (the Presbyterian Church in America and the Orthodox Presbyterian Church) formally discontinued cooperation with the Christian Reformed Church due to, among other important issues, a diﬀerence in opinion concerning the ordination of women. From a purely historical perspective, female clergy are a recent phenomenon, but their novelty has not stopped them from becoming a rapidly growing reality within major sectors of organized religion in the United States. The number of women entering seminaries has more than tripled during the past three decades. In 1972, women counted for only 10.2%, but by 1996 they made up 33.9% of the total enrollment in Protestant seminaries. In the Evangelical Lutheran Church of America alone, the number of ordained women surged from 76 in 1977 to almost 1,500 in 1997 (Charlton, 1997). Although theological debates concerning the validity of ordaining women may continue for many decades to come, female clergy now occupy a permanent place in society. In his ﬁnal address as president of the Religious Research Association, Edward C. Lehman, Jr., a widely recognized leader in the ﬁeld of research on women in the ministry, spoke at length about his experience during the last quarter of a century, which witnessed the arrival of large numbers of clergywomen into the pulpits of Protestant churches throughout the United States. Lehman described the opposition these women faced when they ﬁrst expressed their conviction that God was calling them to serve in roles that up until then had been reserved to men. Once that initial hurdle was cleared, women entered seminaries en masse in the 1970s and forever changed the minister-in-training atmosphere with their feminine presence. Although there were some professors who opposed the admission of women into their theological programs, in general, the semi-sheltered environment of the seminary was more welcoming than the world that awaited many of these women when they were assigned to their ﬁrst parishes after graduation (Lehman, 2001).

gender differences in stress

119

Female ministers, both the pioneers and those of subsequent generations, and the stress they experience in their ministry were two of the main foci of this study. In order to understand the issue better, it was necessary to highlight what previous research has uncovered in the area of gender diﬀerences between male and female clergy, and to follow the general course of Christian women ministers’ diﬃcult journey into spiritual leadership positions during the past 25 years. The information gathered from these two preliminary arenas set the stage for a greater comprehension of how stress aﬀects female clergy diﬀerently from their male counterparts. Musson (2001) found in a random grouping of 900 Church of England parochial clergy that “contrary to the usual sex diﬀerences in personality proﬁle found in general population samples, female clergy were shown to be less outgoing, more emotionally stable, more dominant, less rule-conscious, less emotionally sensitive, less apprehensive, and more open to change than male clergy” (p. 175). Musson cites examples from similar studies ( Jones & Francis, 1992; Ruppart, 1985) where gender reversal was evident. One such example comes from applying the Myers and Briggs Type Indicator (MBTI) test in which 61% of the general male population states that they prefer the Thinking characteristic as deﬁned by the MBTI, while the other 39% prefers the Feeling characteristic. Ruppart (1985) found that among male clergy, the preferences –23% for Thinking versus 77% for Feeling—are reversed and are more in line with those of the general female population. In their study of 1,819 full-time ministers in the United Church of Christ and the Disciples of Christ Church, McDuﬀ and Mueller (1999) found that clergywomen generally receive more support than clergymen, not only on the social and congregational levels, but also from their colleagues and their church hierarchies. In support of their “compensating diﬀerentials” argument, these two researchers found that even though women ministers generally receive lower wages than their male counterparts, they are compensated for this deﬁciency through a system of informal beneﬁts and social supports. Using data collected from the same preceding research project, McDuﬀ (2001) found that a gender paradox exists for clergy job satisfaction as indeed it does across many other occupations and professions. She states that although “female clergy receive lower pay, serve less desirable churches, and have fewer opportunities for advancement,

120

stephen joseph fichter

yet they are signiﬁcantly more satisﬁed with their jobs than male clergy in comparable positions” (p. 15). Not only do male and female Protestant clergy diﬀer in regard to job satisfaction, but they also employ signiﬁcantly diﬀerent ministerial styles. Using data from 1,688 clergy from eight denominations, Perl (2002) found support for the argument that clergywomen tend to be more people-oriented, while their male counterparts are more interested in administration and in moving up the clerical ladder. This ﬁnding is based upon the proportion of time that male and female clergy allocated to the various tasks that are part of a minister’s life. Women ministers reported spending much more time in one-on-one pastoral counseling and very little time in administration and supervision, which is a major reversal from the modus operandi of male clergy. Through their nurturing disposition female clergy have helped to shift the focus of the pastoral ministry from a bureaucratic focus toward a more personal approach. Their trailblazing sisters in the military also brought about enormous changes in their own milieu. In a fascinating comparative study of the rapid integration of women into these two previously male-dominated spheres of public life, Royle (1987) points out how the arrival of women in these two arenas helped to call attention to three conﬂictual “blind spot” areas that went undetected previously: conﬂicts between jobs and careers, between careers of husbands and of wives, and between specialists and generalists. As in the United States military, the entry of women into positions of leadership in the churches has not been easy, especially for the pioneers. In a longitudinal panel study of 30 United Methodist and Lutheran female ministers, ﬁrst interviewed while in the seminary during the late 1970s and then interviewed again 15 years later, it was found that “of the original 30 women in the sample, 24 of them either never graduated, were never ordained, never took a parish position or had left the parish ministry at some point in their careers, either temporarily or permanently. At the time of the second interview, only 12 were in parish or denominational administrative positions” (Charlton, 2000). Having used the same data set three years earlier, Charlton pointed out that when she ﬁrst interviewed these women they were among the 1% female minority of ministers in their denominations. When she met them 15 years later, their minority status had increased sub-

gender differences in stress

121

stantially to 10%. The particularly hard uphill climb of this ﬁrst generation deﬁnitely took its toll. Many of these women complained of contradictory job descriptions, blurred boundaries between workplace and home, and a certain lack of privacy, leading many of them to forego their ministry in mid-career or to seek more specialized ministries such as hospital chaplaincies and service agencies instead of the more standard and traditional parish service (Charlton, 1997). Konieczny and Chaves (2000) brought to light another factor that may add to the stress of clergywomen. Using data from the 1998 National Congregations Study, they discovered that churches led by women tend to be smaller than those served by men and, more to the point, tend to have fewer economic and organizational resources at their disposal. The data also revealed that female-headed congregations are most numerous among liberal and moderate Protestant denominations which led to the decision in this survey to limit participation to American Methodist Episcopal (AME) Zion, Baptist, Congregational (United Church of Christ), Episcopal, Evangelical Lutheran, Presbyterian, Reformed, and United Methodist Churches all of which pertain to this category and are well represented in Bergen County, New Jersey, the geographical area chosen for the research. In these denominations nationwide, more than 20% of their congregations are administered by women, which is in stark contrast to the less than 5% representation in the more conservative and Evangelical Protestant denominations (Konieczny & Chaves, 2000). Even in liberal and moderate Protestant denominations that wholeheartedly approve of women’s ordinations, female clergy face some form of discrimination as Sullins (2000) discovered upon close examination of 15,056 priests in the Episcopal Church. The results of his extensive study showed that 2,089 female Episcopal priests (representing 14% of the sample) were consistently over represented in subordinate and low status positions within the church. In order to know whether this was something peculiar to the Episcopal Church, Sullins compared his ﬁndings with data collected from a 1989 study of 1,505 Presbyterian ministers and found the results to be very similar. Perhaps the frustration that is a byproduct of being (or at least feeling) “stuck” in subordinate positions, or the existence of some residual rejection from the people whom they try to serve, is what causes Protestant clergywomen to experience more depression than

122

stephen joseph fichter

nuns, their closest ministerial counterparts in the Roman Catholic Church. In a broad ecumenical study of 254 women active in religious occupations, Rayburn (1991) found that Catholic nuns experienced the least amount of depression and had better coping resources than did Protestant female ministers and female rabbis. While some would say that such relatively positive results for nuns advocate for a more traditional place for women in the Church, others would say that the depression in this ﬁrst generation of female ministers is a price that had to be paid in order to liberate the ﬁeld of religious leadership for women. In a candid moment of joint reﬂection and reminiscence between two female leaders of research on women in ministry, Adair Lummis admitted that: A traumatic example of disjunction between my research and my value stance occurred when I found that the more strongly that clergy women endorsed feminist values of increasing women’s presence and power in churches and using inclusive language in church services, the worse their overall health (spiritual, physical, emotional, professional). Because I wanted empirical support for my belief that feminist clergywomen are among the healthiest people on earth, I had diﬃculty with these ﬁndings but believed they need to be reported (Lummis & Nesbitt, 2000, p. 446).

The results from Zikmund and Lummis’ study (1998) highlighted the correlation between the opinion of laypeople towards their clergy and the clergy’s sense of overall job satisfaction. If this is true and Lehman’s study (1987) concerning the generally positive attitude of lay church members toward women clergy is also true, then we should have found that female ministers experience much job satisfaction. One recommendation Lehman makes concerning future research is to analyze how acceptance or rejection plays out among the various denominations that allow for ordained women. In an analogous manner, this concept was ﬁtting with one of the original foci of this project which was to detect nuances of stress levels among the clergy, both female and male, from the various Protestant churches represented. The hypothesis was, where there was greater acceptance by the lay members of the diﬀerent denominations, there should be less stress in their ministers. For example, among the United Methodists, where acceptance has been almost universal, there should be greater satisfaction than among the Lutherans who have split into two major camps concerning this issue. Unfortunately, the sample size of this

gender differences in stress

123

current study was not large enough to be able to analyze such an interesting point. In his study of 280 New Zealand Protestant ministers of whom 8% were women, Dewe (1987) found that the sources of stress for clergy could be divided into six broad subgroups: work overload, role conﬂicts, role ambiguity, dealing with grief and people in need, relationships with parishioners and parish, and self pressures. Many of these sub-groups can overlap and as Birk, Rayburn, and Richmond (2001) also noted “although they are placed on a pedestal by ecclesiastical authorities and congregants, clergy often struggle to maintain an image of perfectionism and saintliness, but they also want to be accepted as the erring human beings that they are” (p. 136). Dewe also sought (as did Birk, Rayburn, and Richmond) to investigate the coping strategies that clergy generally employ when trying to deal with stress in their lives and in their ministries. Dewe (1987) developed a three-stage research project in which 38 ministers attending a general meeting of an unidentiﬁed Protestant Church in New Zealand were asked a series of open-ended questions on the topic of stress (stage one), from which he developed a questionnaire (stage two), which was then mailed to all the clergy of the aforementioned church (stage three). A three-tiered Likert-scale was used on all 48 items in the stress part of Dewe’s questionnaire (1987): Respondents were then instructed to think about the occurrence of each of the items (events) during the last six months and to rate each on a ﬁve-point scale, ﬁrst in terms of frequency (‘never’ to ‘frequently’), then in terms of anxiety (‘not at all anxious’ to ‘extremely anxious’), and, ﬁnally, in terms of tiredness (‘not at all tiring’ to ‘extremely tiring’) (p. 353).

A 60% response rate was obtained when 280 out of a total of 468 ministers responded to the mailed questionnaire. As was stated above, 8% of the respondents were women. The average age of all of the respondents was 46, and 90% of them were married. Four out of ﬁve (82%) were involved in parish work, while the other 18% were engaged in more specialized ministries. Among other things, what Dewe discovered was that signiﬁcant relationships were present between two of the coping strategies (social support and rationalizing the problem) and the three main stress factors of parish conﬂicts and church conservatism, diﬃculties involving parish commitment and development, and emotional and time diﬃculties surrounding

124

stephen joseph fichter

crisis work. No analysis was even mentioned concerning the diﬀerent ways in which the male and female clergy diﬀered from each other in their experience of stress. In their study of a group of Southern Baptist ministers, Hatcher and Underwood (1990) underlined the fact that during times of duress many people seek counseling from their ministers before and/or instead of anyone else in the other helping professions. Also, the high stress level of this type of work is compounded by the reality that oftentimes the clergy have no one in whom to conﬁde except God. In the same study, Hatcher and Underwood also pointed out that the subject of stress has been an important issue to the leaders of the Southern Baptist Convention as seen in the numerous research projects (Bailey & Bailey, 1979; Drakeford, 1981; Faulkner, 1981a, 1981b; McSwain & Treadwell, 1981) that they have commissioned. There were 103 participants in this particular study of whom 95 were men and 8 were women. Those ministers with lower self-concepts were also the ones with higher “trait anxiety” levels (i.e., greater proneness to apprehension) and hence also experienced greater diﬃculty in coping with stress in their lives and in their ministries (Hatcher & Underwood, 1990). No attempt was made to study potential gender diﬀerences due to the lack of the number of women ministers participating in the project. In contrast to what has been said so far about the high level of stress experienced by clergy, Malony (1988) found that, in general, clergy experience less on-the-job stress and personal strain than the general population. The basis of his study was a comparison of the results from Blackmon’s investigation (1984) with those of Rayburn, Richmond and Rogers (1983 & 1984). His ﬁndings do not imply that ministers lead stress-free lives, rather that in the midst of their many responsibilities and the often unrealistic expectations placed on them they feel able to handle all of the diﬃculties inherent in their career description. Blackmon’s (1984) survey of 300 active ministers from four Protestant churches in Southern California found that those ministers generally did not experience much conﬂict with their parishes, felt that they were paid well, had happy marriages, and sensed that their ministries were actually of beneﬁt to their families. Focusing only on how female clergy experience stress, Piper (1995) compared work stress among Lutheran clergywomen in the United States (n = 64) and in Norway (n = 70). Facing the “double jeopardy” diﬃculty of working in a formerly male-dominated profession,

gender differences in stress

125

she found that the women clergy who participated in her survey experienced not only the conﬂicts that the men faced, but also carried the added burden of being trailblazers. Piper used the Coping and Stress Proﬁle developed by Olson, Stewart and Wilson (1992) to measure how these women from diverse continents experienced stress in the four main areas of their lives: personal, work, couple, and family. The results from this study showed that Lutheran clergywomen in both Norway and the United States reported very high levels of work stress. Through multiple regression analysis it was found that self-esteem was the best predictor value for work stress. More recently, a detailed examination concerning the extent to which work-family conﬂict among United Methodist female clergy is mediated by career prominence and job satisfaction was conducted by Shehan, Cody, and Wiggins-Frame (2002). These three researchers found that although the women ministers who participated in their project exhibited high degrees of work satisfaction, the majority also showed many signs of non-clinical depression. While most research in the speciﬁc ﬁeld of stress among women clergy has focused on one particular denomination, a study by Rayburn, Richmond, and Rogers (1994) embraced a wide cross-section of the main groupings of clergywomen presently working in the United States including Catholic nuns, Episcopal priests, Jewish rabbis, and Presbyterian and United Methodist ministers. On the one end of the spectrum, Roman Catholic nuns were found to have the lowest overall scores on stress while women rabbis, on the other end, were reported to be at serious risk for stress. In a previous, but very similar study, the same three researchers found that female rabbis experience the most stress (Rayburn, Richmond & Rogers, 1988). As was stated earlier, this study will focus exclusively on the wide range of denominations within Protestantism which has by far the largest number of ordained female clergy in this country. The purpose of this study was to identify more clearly the factors associated with stress among both female and male Protestant clergy. Gender diﬀerences were identiﬁed and analyzed with care. Through thoughtful and thorough exploration of the various factors associated with clergy stress, useful insights were brought to the fore and, ultimately, shared with those who have dedicated their lives to attending to the spiritual needs of others. Perhaps some of the ﬁndings will also serve to enlighten those mental health workers who oﬀer their clients the opportunity to incorporate spirituality in their work together.

126

stephen joseph fichter

Using the total adjusted score of each participant on the Religious Occupations and Stress questionnaire as the dependent variable, the following ﬁve hypotheses were tested: (1) Female ministers experience more stress than their male counterparts. (2) Those ministers who belong to a clergy support group will experience less stress than those who do not take advantage of this opportunity. (3) There is an inverse correlation between years in ministry and overall stress. (4) Those ministers who feel more aﬃrmed by their parishioners will experience less stress than those who do not perceive as much aﬃrmation. (5) There is an inverse correlation between time spent in daily personal prayer and overall stress. Method Participants The sample was drawn from the Bergen County Council of Churches database whose entire list included 279 Protestant ministers. Since gender was the key variable in this study, all 35 women ministers were chosen and then an equal number of males from the same denominations were randomly selected. The self-administered questionnaires were then mailed to all 70 clergy who were advised about the voluntary and conﬁdential nature of the study. By choosing to answer and return the questionnaire, the participants gave their informed consent. Due to the joint nature of the project the envelope carried the logo of the Christian Health Care Center and the cover letter was printed on Fordham University letterhead. A postage-paid return envelope was provided as well. Three weeks after the initial mailing, a follow-up post card from the Christian Health Care Center was sent. In total, there was a 43% response rate (n = 30), 13 of which were female and 17 male. Measures Formal permission was obtained to use the Religious Occupations and Stress Questionnaire (ROS) which was developed speciﬁcally in

gender differences in stress

127

order to measure stress among clergy. The authors, Rayburn, Richmond, and Birk (1996), have done extensive reliability and validity testing through their many research projects involving Catholic, Jewish, and Protestant religious leaders of various denominations as was mentioned previously. Not only did the Religious Occupations and Stress Questionnaire have face and content validity (attested to by a panel nine experts), but over the course of time it has been proven to have high internal consistency, test-retest reliability, and construct validity. It evolved during a 20-year time span from a brief 16-item scale to 178 items which, after a series of tests, were pared down to 102 items and then ﬁnally to 92. In two separate tests for convergent validity comparing results from the ROS to the Vulnerability to Stress Scale and the Susceptibility to Stress Scale, it was found that the Pearson product-moment correlation was .70 between the two measures on both occasions (Birk, Rayburn, & Richmond, 2001). The ﬁve subscales that make up the ﬁnal version of the ROS are: general stress (25 items), occupational stress (12 items), identity/ image stress (29 items), gender-related stress (10 items), and coping strategies (16 items). Sample items from each one of the subscales are as follows: (1) General Stress: “I live in a ‘ﬁshbowl’ with my life on display.” (2) Image or Identity Stress: “As a religious professional, I must keep peace.” (3) Occupational Stress: “On my job, I am expected to complete too many tasks in too little time.” (4) Gender Issues Stress: “I am comfortable with female images of God.” (5) Coping (all scored in reverse): “I get adequate sleep and feel rested.” For the purpose of this study, all 92 items were used. The theoretical instrument scores range from extremely low stress (92) to extremely high stress (460). When the results did not meet skewness standards both on the subscale level and the total stress score, the adjustment method proposed by Mertler and Vannatta (2001) was employed. That is, since the number of outlying cases was fairly small (only one or two in each instance), the outliers were adjusted to a value that is within the extreme tail of the accepted distribution by replacing them with a maximum or minimum value (depending on the

stephen joseph fichter

128

direction of outliers) that fell within that normal distribution. This was done by using a Stem-and-Leaf plot within the “Explore” function of SPSS. Results Descriptive statistics Data was entered into SPSS following standard procedures. Slightly more than one out of three females responded (37%) whereas almost half of the males (49%) did. The mean age was 53 years old (SD = 7.5). The ethnicity of participants was determined by self-report; there were ﬁve ethnic groups from which to choose. Only one respondent did not answer this question and therefore was grouped among “Other.” The ethnic breakdown of the study sample was as follows: 86.7% Euro-American/White (n = 26) and 13.3% (n = 4) Other. All of the respondents (n = 30) obtained at least a college education with some graduate school experience and 60% (n = 18) studied at the post-graduate level. The mean age for females was 50.1 years (SD = 6.0) and for the males 55.3 years (SD = 7.9). The number of years in ministry was 21.5 (SD = 103) for the entire group, 16.6 (SD = 7.7) for the women and 26 (SD = 10.4) for the men. Table 1 shows the results concerning marital status, denomination, and ethnicity. Table 1. Characteristics of Bergen County Protestant Clergy By Gender Variable Marital status (n = 29) Single (never married) Married Separated or divorced Denomination (n = 30) Episcopal Presbyterian Other Protestant Ethnicity (n=30) Euro-American/White Other

Female Percent

Male Percent

Total Percent

25.0 50.0 25.0

0.0 88.2 11.8

10.3 72.4 17.2

23.1 23.1 53.8

29.4 23.5 47.1

26.7 23.3 50.0

84.6 15.4

88.2 11.8

86.7 13.3

Note: Percentages based on responses. “Other Protestant” in the denomination category includes AME Zion, Baptist, Congregational, Lutheran, Methodist, and Reformed.

gender differences in stress

129

Analysis An internal consistency estimate of reliability computed for the ROS in this study produced a total coeﬃcient alpha of .92 which reveals a very high reliability. The coeﬃcient alpha for the general stress subscale was .88, followed by identity/image stress (.86) and occupational stress (.70). The coping strategies subscale yielded an alpha of .61 while the gender-related stress was the weakest at .45. Tables 2 and 3 report the mean and standard deviation for each one of the subscales in general and according to gender. Table 2. Descriptive Statistics on Stress Subscales and Total Score Mean General Stress Image or Identity Stress Occupational Stress Gender Issues Stress Coping skills Total score (Range 92 to 460)

45.17 57.05 26.93 12.17 46.00 187.46

Standard deviation 8.99 7.00 5.87 3.39 7.19 23.95

Table 3. Descriptive Statistics on Stress Subscales and Total Score By Gender

General Stress Image or Identity Stress Occupational Stress Gender Issues Stress Coping skills Total score

Female M

Female SD

Male M

Male SD

45.62 56.63 27.69 12.96 47.16 190.17

10.34 7.58 6.17 4.13 8.73 29.84

44.84 57.37 6.35 11.57 45.10 185.36

8.13 6.74 5.76 2.69 5.86 19.01

As there is no point of comparison with a national mean score on the ROS (either in general or by gender), it is impossible to know whether or not this particular group of clergy is more or less stressed than their colleagues in other parts of this country. In the future, it would be worthwhile to administer the ROS to more clergy of both genders in order to establish a normative population against which new research could be compared.

stephen joseph fichter

130

Inferential Results In order to evaluate the ﬁrst hypothesis that female ministers experience more stress than their male counterparts, an independent-samples t test was conducted since the nominal level independent variable was being analyzed with the interval level dependent variable. Counter to the stated research hypothesis and the available literature concerning potential gender diﬀerences, this particular test was not found to be signiﬁcant, t (28) = .54, p > .05. As was shown in Table 3, female clergy scored (M = 190.17, SD = 29.84) only slightly higher than the male clergy (M = 185.38, SD = 19.01). Correlational analysis of the ﬁve subscales (Table 4) showed that general stress and image/identity stress were statistically signiﬁcant for all respondents. The coping skills subscale was statistically correlated with the other four subscales with moderate to high strength. This may indicate that the more coping mechanisms a minister acquires and applies the less stress he or she will endure. Although there were only 30 respondents, the researcher wanted to examine potential gender diﬀerences (Table 5). This examination may be limited due to the small subsample size. Analyzing the data separately, the image/identity stress correlation with general stress continued to exist for both genders, but the coping skills subscale for the male clergy decreased in signiﬁcance whereas for the female ministers it increased. As seen in Table 5, gender issue stress had a much higher correlation among the clergywomen than it did for their male counterparts. Table 4. Correlational Analysis of Five Subscales of ROS

General Image/Identity Occupational Gender Issues

Image

Occupational

Gender Issues

.702**

.201 .246

.198 .235 .150

Note: ** Correlation is signiﬁcant at the 0.01 level (2-tailed). * Correlation is signiﬁcant at the 0.05 level (2-tailed).

Coping skills .577** 466** .403* .425*

gender differences in stress

131

Table 5. Correlational Analysis of Five Subscales of ROS by Gender

General Image/Identity Occupational Gender Issues Coping Skills

General

Image

–

.822** .080 – .439 .088 – –.267 .143 .230 .466*

.584** .319 –.540* .409

Occupational

Gender Issues

Coping skills

.695** .658** .123 – –.068

.698** .686** .341 .686** –

Note: Female scores are in the upper right portion; male scores are in the lower left. ** Correlation is signiﬁcant at the 0.01 level (2-tailed). * Correlation is signiﬁcant at the 0.05 level (2-tailed).

The researcher sought to examine stress and the amount of involvement in clergy support group. The results of the independent-sample t test showed that those ministers who are involved with a support group scored signiﬁcantly lower (M = 181.93, SD = 21.58) than those who do not (M = 200.36, SD = 25.46), t (28) = –2.03, p < .05. While signiﬁcance was reported for clergy involvement in support groups in general at the 95% conﬁdence level, when the ﬁle was divided according to gender, statistical signiﬁcance was no longer sustained (Table 6) although it became evident that membership in a support group approaches signiﬁcance much more among female clergy. There was more than a 30-point diﬀerence in mean ROS scores between those women ministers who belonged to support groups and those who did not. Table 6. Stress and Support Group Membership According to Gender Mean

SD

Female clergy in support group (n = 8)

178.25

24.81

Female clergy not in support group (n = 5)

209.25

29.24

Male clergy in support group (n = 13)

184.19

20.07

Male clergy not in support group (n = 4)

189.24

17.06

t

–2.05

–.453

Sig.

.065

.657

132

stephen joseph fichter

The third hypothesis that stated there would be an inverse correlation between years in ministry (M = 21.57, SD = 10.29) and overall stress (M = 187.46, SD = 23.95) was found to be statistically signiﬁcant (r = –.345, p < .05). The strength of this correlation is moderate. When the data were divided by gender, no correlation was found between the clergywomen’s number of years in ministry and their overall stress score (r = –.399, p > .05) nor was it found among the male clergy (r = –.316, p > .05). The fourth hypothesis examined whether there was a correlation between the clergy’s rating of perceived aﬃrmation from parishioners (M = 3.77, SD = 1.14) and overall stress scores. Results showed that there was a statistically signiﬁcant high correlation (r = –.551, p < .01). When the data were examined by gender, a high negative correlation was found to exist for the female ministers between perceived aﬃrmations and overall stress (r = –.787, p < .01), while for the male ministers there was no signiﬁcant correlation between these two same variables (r = –.170, p > .05). The ﬁfth hypothesis examined whether there was an inverse correlation between time spent in daily personal prayer (M = 37.69, SD = 26.99) and overall stress. Results showed no signiﬁcant relationships for the combined group (r = .260, p = NS), nor for the women (r = .330, p = NS) nor for the men (r = .245, p = NS) when analyzed separately. Discussion Contrary to what was stated in previous research concerning the higher levels of stress found among female clergy, the speciﬁc ﬁndings of this project did not support such a hypothesis as seen in the very small diﬀerential between mean ROS scores for female and male clergy. One possible explanation as to why the female ministers in this particular study did not experience much more overall stress could be due to an increased acceptance of contemporary clergywomen by society in general and by their churches not experienced by those ordained in the early 1970s. As the trailblazing generation of clergywomen forged the way for much larger subsequent generations, it seems stress levels have diminished. It would be prudent to recall the results from Lehman’s study (1987) concerning the generally positive attitude of lay church members toward women clergy.

gender differences in stress

133

Not only have parishioners come to feel more comfortable with female spiritual leaders, but perhaps clergywomen themselves have felt more at ease in their relatively new status in a profession dominated by men for millennia. That interpretation can also be linked to the third hypothesis that states that, as ministers gain more experience over the years, their stress levels diminish. Perhaps those clergywomen whose stress escalated over the course of time simply abandoned their ministry and embarked upon other journeys as was reported by Charlton (2000) and hence reduced the ranks of overstressed women ministers. Or perhaps the normal human developmental process by which both men and women tend to grow calmer with the passage of time is the key factor in the progressive diminishment of stress. It may be that, as a person gains more spiritual insights and broader perspectives over the course of time, a more placid (and hence less stressful) attitude towards life prevails. Of all the variables in this project, the one that was the most salient was the perceived aﬃrmation by parishioners. Those clergy who sensed more aﬃrmation reported less stress. In this particular case, the cause and eﬀect principle cannot be conﬁrmed in either direction. It is not known whether kind words spoken by parishioners after Sunday services lead to a greater sense of fulﬁllment in clergy and hence to lower stress levels or whether those ministers who live with less stress function better and hence are more often the object of praise from their parishioners. In either case, it is interesting to note that the clergywomen were more aﬀected by this factor than were their male counterparts. The study by Zikmund and Lummis (1998) mentioned in the literature review indicates a very similar pattern of correlation between the opinion of laypeople towards their clergy and the clergy’s sense of job satisfaction. Finally, and contrary to this researcher’s intuition, no correlation was found between the amount of time spent in private prayer on a daily basis and the overall ROS scores. Perhaps a larger sample size would have revealed something of signiﬁcance, but the current data did not validate such intuition. This speciﬁc topic could be yet another interesting angle for future research in the ﬁeld of clergy stress.

134

stephen joseph fichter Limitations and Implications

This study was limited due to its cross-sectional design and the very small number of respondents. Only 30 clergy from one wealthy suburban county in the state of New Jersey participated and hence it may be diﬃcult to generalize these results to clergy in less aﬄuent areas of the country. Perhaps the comfortable working conditions in this speciﬁc metropolitan milieu contributed to less overall stress in both female and male clergy. Another serious limitation which other researchers in the ﬁeld (Rayburn, Richmond & Rogers, 1982) have pointed out is the social desirability phenomenon which may cause clergy to report less stress than they actually experience. Notwithstanding the abovementioned serious limitations, there were several implications for clergy, their parishioners, and their denominational leaders. For the clergy, it was beneﬁcial for them to know that those who participate in support groups tend to experience less stress than those who do not. It seemed especially true for female ministers. Perhaps the ﬁrst step for a clergyperson who is feeling pressured in his or her ministry would be to ﬁnd out where and when the nearest support group meets. Another consideration that could encourage both male and female clergy during the more painful periods of life would be to know that there is a correlation between longevity in ministry and reduced stress. In regard to the perceived aﬃrmation hypothesis, which proved to be highly signiﬁcant for lower levels of stress among clergy, parishioners who learn about this correlation may be more ready to provide positive feedback at the end of a Sunday service or any other church activity. In particular, it would be helpful for laypeople to know that female ministers seem to be most positively aﬀected by this practice. It is possible that aﬃrmation leads to greater job satisfaction and hence less stress. These and many other related themes merit more attention by researchers in the future. Denominational leaders, many of whom are already attempting to establish mentoring programs for new clergy, may ﬁnd conﬁrmation in the results produced by this study as to the eﬃcacy and usefulness of clergy support groups. They may also provide education to congregations concerning patterns of aﬃrming pastoral leadership. Finally, the Pastoral Care Department at the Christian Health Care Center, under whose aegis this project was performed, may consider ways to assist both clergy and denominational leaders in their eﬀorts to provide clergy support.

gender differences in stress

135

References Bailey, R. W. & Bailey, M. (1984). Coping with stress in the minister’s home. Nashville, TN: Broadman. Birk, J. M., Rayburn, C. A. & Richmond, L. J. (2001). Religious Occupations and Stress Questionnaire (ROS): Instrument development. Counseling & Values, 45, 136–144. Blackmon, R. A., (1984). The hazards of inistry. Unpublished Ph.D. dissertation, Fuller Theological Seminary, Pasadena, California. Charlton, J. (1997). Clergywomen of the pioneer generation: A longitudinal study. Journal for the Scientiﬁc Study of Religion, 36, 599–613. —— (2000). Women and clergywomen. Sociology of Religion, 61, 419–424. Dewe, P. J. (1987). New Zealand ministers of religion: Identifying sources of stress and coping strategies. Work and Stress, 1, 351–363. Drakeford, J. (1981). The awesome power of the healing thought. Nashville, TN: Broadman. Faulkner, B. (1981a). Burnout in the ministry. Nashville, TN: Broadman. Faulkner, B. (1981b). Stress in the ministry. In B. Faulkner (ed.). Stress in the life of the minister (pp. 11–24). Nashville, TN: Broadman. Hatcher, S. W. & Underwood, J. R. (1990). Self-concept and stress: A study of Southern Baptist ministers. Counseling & Values, 34, 187–196. Jones, D. L. & Francis, L. J. (1992). Personality proﬁle of Methodist ministers in England. Psychological Reports, 70, 538. Konieczny, M. E. & Chaves, M. (2000). Resources, races, and female-headed congregations in the United States. Journal for the Scientiﬁc Study of Religion, 39, 261–271. Lehman, E. C., Jr. (1987). Research on lay church members’ attitudes toward women clergy: An assessment. Review of Religious Research, 28, 319–329. —— (2001). Clergy women’s world: Musings of a fox. Review of Religious Research, 43, 5–13. Lummis, A. T. & Nesbitt, P. D. (2000). Women clergy research and the sociology of religion. Sociology of Religion, 61, 443–453. Malony, H. N. (1988). Men and women in clergy: Stresses, strains, and resources. Pastoral Psychology, 36, 164–168. McDuﬀ, E. M. & Mueller, C. W. (1999). Social support and compensating diﬀerentials in the ministry: Gender diﬀerences in two Protestant denominations. Review of Religious Research, 40, 307–330. McDuﬀ, E. M. (2001). The gender paradox in work satisfaction and the Protestant clergy. Sociology of Religion, 62, 1–21. McSwain, L. & Treadwell, W. C. (1981). Conﬂict ministry in the church. Nashville, TN: Broadman. Mertler, C. A. and Vannatta, R. A. (2001). Advanced and multivariate statistical methods. Los Angeles, CA: Pryczak Publishing. Musson, D. J. (2001). Male and female Anglican clergy: Gender reversal on the 16FP5? Review of Religious Research, 43, 175–183. Perl, P. (2002). Gender and mainline Protestant pastors’ allocation of time to work tasks. Journal for the Scientiﬁc Study of Religion, 41, 169–178. Piper, J. L. (1995). Work stress among Lutheran clergywomen in the USA and Norway. Doctoral Thesis from Graduate School of the University of Minnesota. Rayburn, C. A. (1991). Counseling depressed female religious professionals: Nuns and clergywomen. Counseling & Values, 35, 136–148. Rayburn, C. A., Richmond, L. J. & Birk, J. (1996). Religious Occupations and Stress Questionnaire (ROS). Washington, D.C.: U.S. Copyright Oﬃce. Rayburn, C. A., Richmond, L. J. & Rogers, L. (1982). Women, men and religion: stress within sanctuary walls. Journal of Pastoral Counseling, 17, 75–83. —— (1983). Stress among religious leaders. Thought: Fordham University Quarterly Review, 58, 329–344.

136

stephen joseph fichter

—— (1988). Female rabbis, ministers, and priests and the religion and stress questionnaire. Journal of Pastoral Counseling, 23, 61–63. —— (1988). Female rabbis, Protestant clergywomen, and stress. Journal of Pastoral Counseling, 23, 64–66. Royle, M. H. (1987). Using bifocals to overcome blindspots: The impact of women on the military and ministry. Review of Religious Research, 28, 341–350. Ruppart, R. (1985). Psychological types and occupational choices among religious professionals: A psychosocial, historical perspective. Doctoral dissertation from New York University. Shehan, C. L., Cody, S. R. & Wiggins-Frame, M. (2002). Work-family conﬂict and personal happiness among clergy women: A qualitative study. Southern Sociological Society conference paper. Sullins, P. (2000). The stained glass ceiling: Career attainment for women clergy. Sociology of Religion, 61, 243–266. Zikmund, B. B. & Lummis, A. T. (1998). Women, men, and styles of leadership. Christian Century, 115, 478–485.

RELIGIOUS CHOICE AND THE MEANING OF CHURCH IN THE LIVES OF EVANGELICAL SEEKERS Richard W. Christopherson Abstract This study presents a description of the church life of American evangelicals that is centered in the lives of individual believers. A sample of evangelical Christians were asked to discuss their experiences in moving from church to church. The interview data reveal a general process by which people leave churches and seek new aﬃliations. The process involves (1) separation from their childhood faith and inherited religion; (2) the subsequent search for an “authentic” church; and (3) the accommodation strategies employed to ﬁt into the new church. The analysis centers on the tension between the evangelicals’ well developed sense of spiritual individualism and the institutionalized authority of congregations and denominations. Their decisions about church involve multiple “dimensions of religious identiﬁcation” (Hervieu-Léger, 1998, p. 218). Respondents interpret the meaning of their church lives in terms of their personal experiences with the sacred and their identiﬁcation with universal Christian values.

The religious lives of evangelical Christians are ﬁlled with choices about what to believe and how to behave. The authority for making those choices is drawn from two diﬀerent and potentially contentious sources. On one hand, there is the subjective authority grounded in an individual’s experience with the sacred. Guidance is provided by such other worldly directives as one’s “personal relationship with Christ,” “answers to prayer,” and “divine calling.” On the other hand, there is the objective authority drawn from the Bible, Christian history, and from the traditions of the evangelical movement. Such worldly guidance is acquired through socialization and is institutionalized in denominations, congregations, formal doctrines and creeds, and in the work and careers of professional clergy and other religious authorities.1 Evangelicalism is a complex, interdenominational Christian movement grounded in agreements about 1 The relationship between clergy and laity is explored in Christopherson (1994). Even in churches where the “priesthood of all believers” is a core doctrine, clergy still derive special authority from their professional skill and from their claim to divine calling. Particularly at times of crisis and joy, the clergy’s presence is taken for granted, and parishioners defer to their pastor for guidance and sacramental intervention.

Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

138

richard w. christopherson

certain core beliefs (Hunter, 1983; Marsden, 2000). There is institutional structure comprised of schools, evangelistic organizations, publishing houses, radio and television broadcasting, and social and political action groups (Smith, 1998). Marsden (2000) says that, “. . . the test of being a card-carrying evangelical is that of having a fairly strong transdenominational identity, whatever happens to be one’s denominational aﬃliation” (pp. 5–6). Within the world of evangelicalism there is an uneasy alliance between institutional authority and personal experience—between the power to choose that God has granted individuals, and the power to choose that is acquired through participation in religious institutions. Among the most basic choices facing evangelicals are questions about church. Where to go? With whom to aﬃliate? How deeply to commit? Decisions about churches hinge on the problematic relationship between an evangelicals’ highly developed sense of spiritual autonomy, and their dependence on a church community and corporate worship. American evangelicals are both profoundly individualistic—they testify to their “personal salvation”—and profoundly corporate—80% are in church services at least once a week, and 60% participate in additional church activities weekly (Smith, 1998). By listening to evangelicals talk about their experiences when they are moving from church to church, we learn something about what church means to them, and how they bring that understanding to bear on their relationship with organized religion. Such research provides an image of the church life of American evangelicals that is centered in the real lives of individual believers. Cultural Pluralism, Individualism and the Religious Market Observers of modern religion have long raised concerns about the potentially corrosive eﬀects of cultural pluralism and individualism on the ability of institutions to shape morality and guide human behavior. Recent studies by Bellah, et al. (1985), Hervieu-Léger (2000), Smith (1998), Roof (1999), and Wuthnow (1998), for example, all consider the impact of individual religious choice on persons and institutions. For some there is a concern about the future vitality of religion in a social situation where faith is increasingly personalized and links to tradition are weak. What happens when religion is detached from the “communities of memory” that serve to temper

evangelical seekers

139

utilitarian and expressive individualism? Bellah et al. (1985) are deeply concerned by the fact that the “American understanding of autonomy of the self places the burden of one’s own deepest self-deﬁnitions on one’s own individual choice” (p. 65). They point out that traditionally there has been a “certain priority of the religious community over the individual. . . . Relationship of the individual to God is ultimately personal, but it is mediated by the whole pattern of community life. There is a givenness about community and tradition. They are normally not a matter of individual choice” (p. 227). From this perspective, the prevalence of religious choice presents problems for the maintenance of faith and commitment, i.e. true belief is diﬃcult to hold onto in an increasingly individualized and rationalized social context. Becker (1998) sums up Bellah’s concern, along with Wuthnow’s (1998) and Roof and McKinney’s (1987), as: [A] metanarrative about the declining signiﬁcance of denominational attachment, increasing rates of member switching, the growth of an individual-expressive or personalistic style of commitment, and growing divisions between liberals and conservative. This metanarrative rhetorically links this institutional restructuring [of American religion] to various forms of decline—in community, in public religion, in social capital. (p. 19)

By contrast, “new paradigm” thinkers see a bright future for religion, not in spite of, but because of pluralism, individualism, and the freedom to choose for or against religious faith. For Stark and Finke (2000), “The most signiﬁcant single feature of a religious economy is the degree to which it is unregulated and therefore market driven” (p. 36). Warner (1993) contends that with so many churches to select from “the breakdown of ascriptive ties to religion can enhance, rather than reduce, the elemental nature that believers attribute to their experiences. . . . newfound religion is self-aﬃrming” (p. 1078). A free market of religious possibilities stimulates religious growth and personal commitment. These notions are the starting points for Roof ’s (1999) study of “baby boomers” seeking faith in the American “spiritual marketplace,” and for Smith’s (1998) examination of the strength and vitality of American evangelicalism. For example, Smith reasons: Moderns authenticate themselves through personal choice. Therefore, modern religious believers are capable of establishing stronger religious identities and commitments on the basis of individual choice than

140

richard w. christopherson through ascription. And we should expect this to be exceptionally true of American evangelicals, whose faith traditionally emphasizes making a personal ‘decision for Christ’ through an individual conversion experience. (p. 104)

Wuthnow (1998) looks at changes in American spirituality over the past 50 years and observes a “subtle reordering that has taken place in how Americans understand the sacred itself . . . a traditional spirituality of inhabiting sacred places has given way to a new spirituality of seeking” (p. 3). He concludes that in this situation: A person does not have an ascribed identity or attain an achieved identity but creates an identity by negotiating among a wide range of materials. Each person’s identity is thus understandable only through biography. The search that diﬀerentiates each individual is itself part of the distinct identity that person creates. (pp. 9–10)

Where new paradigm thinking celebrates the self-authenticating potential of individualized faith—religion freely chosen—in the end, Wuthnow is more cautious. In a situation where “growing numbers of Americans piece together their faith like a patchwork quilt” (p. 2), a spirituality of seeking, by itself, is “too ﬂuid to provide individuals with the social support they need or to encourage the stability and dedication required to grow spiritually and to mature in character” (p. 16). Even if individuals should choose to connect with institutional religion, their level of commitment is not likely to go beyond treating the church “as a supplier of goods and services” (p. 15). Wuthnow argues that real commitment requires something more: Spiritual practice takes this kind of seeking a step further, adding the vital element of sustained commitment, without which no life can have coherence. Practices may be messier than the commanders of largescale institutions would like, but they ultimately sustain these institutions by giving individuals the moral fortitude to participate in them without expecting to receive too much from them. If Americans’ practices were, indeed, better than their creeds, there would be much about which to be optimistic. (p. 198)

Hervieu-Léger (1998) also presents the image of modern individuals actively stitching together their religious lives from the symbolic resources available to them. In the world of high modernity, “Individuals are led to produce for themselves (if they produce it at all) their relationship to the believing lineage from which they take their identity” (p. 218, italics in the original). She argues in Religion as a

evangelical seekers

141

Chain of Memory (2000) that “post-traditional religion . . . defers recognition of a tradition’s power of generation to the eﬀectiveness of individual commitment” (p. 167), and it is up to us to ﬁnd a way to believe. In modernity, being religious is not so much knowing oneself begotten as willing oneself so to be. This fundamental reworking of the relationship with tradition which characterizes modern religious believing opens up theoretically limitless possibilities for inventing, patching together and playing with systems of meaning that are capable of ‘establishing tradition.’ (p. 167)

The broken links in Hervieu-Léger’s “chain of memory” mean that individuals must settle issues of religious belief and belonging on their own. The modern environment is rich with possibility; an unprecedented number of suppliers of religious goods compete for the attention and loyalty of institutionally unencumbered seekers. However, in this radically pluralistic culture of personal choice, guidance from the past is in short supply. Individuals are called on to piece together their own religious and spiritual identities. Exactly who they are is yet to be determined, and the resolution of their quest depends primarily on their own will and their ability to commit. The diverse subculture of American evangelicalism contains its own share of itinerant believers,2 seekers cut loose from family and faith community who now move about in the ﬂuid world of religious possibility.

2 Statistical patterns of mobility between religions and religious denominations have been explored at some length. Smith’s (1998, p. 48) data, for example, show that evangelicals do a better job of retaining their young people than do fundamentalists, mainline or liberal Protestants, or the “nonreligious.” In his sample, only 22% of those raised in evangelical families grew up to be something other than evangelical, and “not one respondent raised in an evangelical (or fundamentalist) family became non-religious (or Roman Catholic) as an adult.” Likewise the evidence presented by Bibby (1999) shows that relatively little switching is going on, and when people do move to another congregation they tend to stay close to their denomination of origin. Bibby’s (2003, p. 7) latest look at “circulating saints” indicates that most people in conservative Protestant churches stay in the religious group in which they were raised, and, in fact, the primary key to the continued growth of evangelical denominations is their success at holding on to their young people. Hadaway and Marler (1993, p. 111) also found that conservative Protestants are the least likely to switch, and “most Americans are stayers; they do not switch from one denominational family to another.” They point to individual factors useful in predicting stability or change in religious identiﬁcation—the most important relate to the family and socialization. The person most likely to switch denominations “is

142

richard w. christopherson

The present study looks closely at the experiences of evangelical seekers. They are conservative Christians committed to a life of faith, but marginal to the speciﬁc traditions and institutional structures of organized religion. The retrospective accounts of their journeys from church to church provide the data for this project. A working assumption is that we can understand something about the meaning of church by looking at instances where connections are weak, aﬃliations are changing, and where views of religious and spiritual reality are being reinterpreted. In such cases, the stories evangelical seekers tell about their eﬀorts to ﬁnd a church provide insight into the modern struggle to link belief with collective actions, private conviction with public commitment. Such research requires that we “listen to the voices of the people in the population who are actually making these [religious] choices” (Ammerman, 1997, p. 129). By listening carefully we can discover more than just what people are doing in the religious marketplace, we can also learn something about what church means to them and how religion works for people who choose to believe and must seek out a place to belong. Method Participants An availability/snowball sampling technique was used to select 21 individuals who had switched churches and were willing to be interviewed about their experiences.3 A notice describing the research and a calling for volunteers was placed in the employee bulletin of an evangelical Christian university. Eleven individuals who responded characterized by marginal church aﬃliation during childhood, heterogeneous aﬃliation patterns within families and marriages, and a lack of regular contact with relatives.” Finally, Wuthnow (1988, p. 89) uses data from the GSS to point out that educational attainment “plays a signiﬁcant role in determining (both) the likelihood that someone will have attended churches of diﬀerent denominations. . . . (And) of people actually switching denominations.” 3 In this study “switching churches” encompasses any movement from one congregation to another. Such moves may be across denominational lines, or they may be made within the boundaries of religious organizations. In some cases, such changes involved profound personal experience (being born-again, baptized or rebaptized), in other instances a switch was made simply to ﬁnd a more compatible congregation or clergy. Using Stark and Finke’s (2000:114) terminology, this is a sample of “reaﬃliates.”

evangelical seekers

143

to the notice were interviewed and were also asked to suggest additional respondents. This snowball technique yielded 10 more subjects including 4 from another Christian university, and one individual employed by a scholarly foundation (whose stated mission is to advance “the renewal of Christian scholarship”). The three institutions are all located in Southern California. The respondents’ employment status at these institutions, and the statement of faith required of employees, served as an operational measure of the respondents’ “evangelical” religious identiﬁcation. Ages ranged from late 20’s to 70 years of age; the modal age was mid-40’s. Eleven females and 10 males were interviewed. Eight individuals are university faculty members; 13 are employed in administrative and staﬀ positions. Nineteen of the respondents are Anglo, 1 is African-American, and 1 is Mexican-American. This is an educated sample, i.e. 12 have master’s or doctoral degrees, 8 have bachelor’s degrees, and only 1 did not complete college. Eleven were raised in evangelical Protestant families, 7 in mainline or liberal Protestant families, 1 in a strong Roman Catholic family, 1 in a nominal Catholic family, and 1 in a nominal Mormon family. Procedure Intensive, open-ended interviews of between 90 minutes and 3 hours were conducted in the winter and spring of 2003 using the interview techniques outlined in Loﬂand & Loﬂand (1995). An interview guide was developed, and respondents were asked to tell the story of their church life by focusing on transitions from one congregation to another. They were speciﬁcally asked to talk about their childhood experiences with church, their decisions to leave former churches, strategies for locating new churches, their current church situation, and to respond to questions about their religious and spiritual identity. Interviews were tape recorded and transcribed. The inductive analysis began concurrently with the interview process and continued after the interviewing had been completed. Following the analytical strategies discussed by Loﬂand & Loﬂand (1995) transcripts were read and reread, and a coding scheme was developed as a way of discerning patterns in the data. Of particular interest were patterns in individual’s identiﬁcation with local congregations. Concepts were discovered in the data by looking closely at the language used by respondents to describe and explain their experiences

144

richard w. christopherson

with faith and organized religion. The objective was to uncover the “meanings” that they attach to their church experience especially to the episodes of church switching. Meanings are the linguistic categories that make up the participants’ views of reality and which deﬁne their own and others’ actions. Other frequently used words for meanings are norms, understandings, interpretations, deﬁnitions of the situation, typiﬁcations, ideology, belief, frames, worldviews, perspectives and stereotypes. . . . Meanings are transbehavioral in the sense that they do more than describe behavior—they deﬁne, justify and otherwise interpret behavior as well. (Loﬂand & Loﬂand, 1995, p. 113)

Patterns in the respondents’ identiﬁcation with congregations were discovered, and the meanings of their experiences with organized religion explored. An overall analytical framework was developed for understanding and presenting the data. Results Accounts of confrontations with childhood faith and inherited religion are a central part of the interviews for this study. Respondents recalled experiences that marked their decision to “leave home” and strike out on a diﬀerent path.4 These accounts follow two general patterns (although the results are similar in both types of experience). First, those who were raised in non-evangelical families leave home when they are born-again. It is this event, “getting saved,” that separates them from their former religious identity and is the catalyst for conﬂict between them and their “unsaved” family and friends.

4 Loveland (2003, p. 154) has shown that by itself, “childhood socialization does not appear to produce lasting religious preferences.” His ﬁndings contradict Hadaway and Marler’s (1993) conclusion that the quantity and quality of religious socialization is predictive of switching behavior, and they modify Sherkat and Wilson’s (1995) notion of “embedded preferences.” Loveland argues that childhood religious preferences are fragile and holding on to them is dependent on “the individual’s ongoing relationship with religious institutions later in life,” on maintaining social ties with their family of origin, and on the “distinctiveness” of the religious group in which they were raised. Individuals who spend more time with their families, formally join a church as a young person, are regular lifetime participants in church, and develop preferences for unique religious goods (such as the Catholic Eucharist) are the least likely to switch.

evangelical seekers

145

Secondly, those raised in evangelical homes—let’s call them cradle evangelicals—also testify to being born-again, but for them accepting Christ does not mean conﬂict with their family. Getting saved is not deviant in an evangelical family; in fact, it means that a child is becoming the person they have been socialized to be. For this reason, accounts of leaving home provided by cradle evangelicals are not stories about ﬁnding God; they are stories about encountering doubts, experiencing family crises, or catching a sympathetic glimpse of something religiously and spiritually diﬀerent. Markers of their new identity as “seekers” include such things as: liberal education, marriage outside their faith community, death, divorce, and abuse. For both born-again and cradle evangelicals, the break with their past reveals an underlying conﬂict between the institutionalized authority of religious traditions and authority grounded in their personal experience and commitment. Born-again evangelicals For respondents raised in non-evangelical homes, the born-again experience initiates an ongoing process of religious and spiritual change. For example, a female respondent grew up in a nominal Presbyterian family and attended a Catholic women’s college because it was near her parents’ home. In college she identiﬁed herself as “an agnostic existentialist” because “that stance made more sense to me than anybody’s religion.” When she was in her early 20’s, married, and a college graduate with a master’s degree, she sought help for an addiction to prescription drugs. She selected a treatment center that happened to be run by evangelicals: While I was talking to the associate director he presented Christ to me and I became a Christian. Born again. On my own, I couldn’t get out of my drug problem, and he oﬀered me hope of healing. I thought, “I don’t know anything about the Bible,” but I walked out of there really diﬀerent. Healed of the drug problem. . . . I met my husband that afternoon after he was done at work, and we drove home together. I remembered this little phrase, and I wanted to tell my husband that I’d become a Christian. As he was driving down the road, I said, “I’m a born again Christian. I’ve accepted Christ.” He almost ran into the guardrail. I said, “It’s wonderful, I feel healthier than I’ve felt in years” . . . He called my dad and said, “She’s ﬂipped her lid.” He was worried about me psychologically. “She’s talking about being Christian, and I don’t know what to do.” I felt like I was on cloud nine. All the pills went down the toilet.

146

richard w. christopherson

Her newfound faith initiated a diﬃcult period in her marriage. She explained: “I really had changed and he hadn’t. He kept saying to me, ‘You’re not the girl I married,’ and he was absolutely right. He was still a radically secular person. It became very painful for both of us.” A man who grew up in the 1960s and 1970s was sent to Catholic school by his mother until the sixth grade at which point his formal, family-sponsored religious socialization ceased: I don’t think my father has ever been inside a church building. He stayed at home, but my mother attended for a few years, and she sent us to Catholic school. When she decided to withdraw us, we all went to local public schools for junior high and high school, and I don’t think we attended the Catholic Church once after that.

After high school he attended a local community college and then left for a university in Northern California: That afternoon when I arrived, there were some guys witnessing near the campus. They were from a [Christian] commune south of there called Lighthouse Ranch, and they led me to Christ after about three hours. . . . The conversation began when I asked them what time it was, and one of them pointed to me and said, ‘It’s time for you to get saved!’ . . . I called my folks—actually in my third week at Lighthouse Ranch—and I urged them to sell everything and come move up there with the utmost seriousness. And my father and mother felt that I’d just been inducted into some Jonestown cult. They said, “Just do us a favor and ﬁnish college.”

Finally, a woman raised in a devout Roman Catholic family talked about her born-again experience and her rejection of her parents’ faith: It took me until my senior year (of college) before I heard the Gospel in a way that made sense. It was a time of probably the most confusion—certainly spiritual confusion. I had always believed in a God and a Trinity, but the idea that God had any knowledge of who I am, or even desiring to, any of the interpersonal relationship kind of things, any idea of the atoning work, that was missing. Once I understood that, I thought immediately, “I am going to have to leave the Catholic Church.” I tied in leaving the Catholic Church with accepting a diﬀerent understanding of Christianity altogether.

When asked if she told her parents, she replied: I wrote my sister a long letter telling her what had happened. She promptly turned it over to my parents because she was convinced I’d

evangelical seekers

147

joined a cult. My parents were convinced I’d joined a cult. We had ten very diﬃcult years. I was asked to leave the house. There were repercussions. Not just, “Oh you’ve left the Catholic Church.” They went through a great soul searching as parents. “What did we do wrong that our Catholic daughter left the church?” Even today my dad will still hand me Catholic tracts every once in awhile and wink, “Thought you should read this.”

For those brought up in non-evangelical families, the born-again experience marks an exit from their prior religious life; it cuts them oﬀ from their past. (“I thought immediately, I am going to have to leave the Catholic Church.”) Their families are confused, perhaps hurt and angry. Of course, their “exit” is simultaneously an “entry” into a new social world. However, at this point it is much clearer what they are opting out of than what they are getting into. Bornagain individuals enter evangelicalism without beneﬁt of the inherited institutional attachments and understanding that characterize those who are cradle evangelicals. All of the born-again respondents for this study talked about the individuals, families, and organizations that were instrumental in shaping their new lives. They were given things to read, directed to churches and schools, prayed over, and taught new standards of behavior. Attempts were made to attach a profoundly personal experience to some sort of institutional framework. For example, the woman who met Christ in the drug treatment center talked about what happened immediately after her experience: He gave me half a dozen of those little Campus Crusade books with the verse to memorize and Bible readings and a Bible. . . . They were careful to direct me towards an Assemblies of God church in (nearby city), and said, “You may not like it but that’s the best we can do.” That Sunday I went, and the people were very kind to me so I started going to church there.

The young man who was “saved” when he arrived at the university campus accompanied his new spiritual mentors back to the Christian commune where his resocialization ensued. He stayed with them for a few weeks, participated in daily Bible studies, and weekly all-night prayer meetings. Subsequently he was involved in two evangelical churches in a nearby town. For a while he lived with a family that he met at one of the churches, and to this day his relationship with his adopted Christian family is “just as close” as with his real family. “They still call me son.”

148

richard w. christopherson

Cradle evangelicals Cradle evangelicals are socialized into a distinctly individualistic religious tradition. Every evangelical is expected to choose for him or herself, which way they will go. No matter how structured and intentional their religious upbringing, it’s understood that individuals will act autonomously, exercise free will, and make a choice. Children inherit the raw material of their faith, but as they grow up they are encouraged to transform that into authentic personal commitment. As cradle evangelicals make faith personal, their choices may begin a process of separation from family tradition. Here are three examples. A pastor’s son who grew up in the 1940s and 50s, talked about his sheltered childhood: The church was the center of our lives socially and in every other way. There was no life without church. So I was very loyal to the Free Methodists because I knew nothing else. I wouldn’t call it an intrinsic loyalty—and that’s why the transition from it was so painful— because I knew nothing else, but needed something else.

He graduated from a “very conservative” denominational college and later enrolled in a Methodist seminary intending to follow in his father’s footsteps into ministry. This decision began a process of disengagement from his family’s church: Seminary was a very positive experience for us, and that was an eye opener. When (my wife) and I found people who didn’t cross all the T’s and dot all the I’s, but had more passion for Christ and mission, it created disequilibria for us. The ﬁrst step of liberation—it was truly a spiritual feeling of liberation—came when we made an intentional decision to take our children, and we drove 55 miles to the capital of Kentucky and went to our ﬁrst movie together, “The Sound of Music.” When we walked out of that theater we were walking on air. It was truly a Maslovian peak experience. We felt so clean.

For this seminarian, a motion picture symbolically marked his entry into a larger world of Christian possibility and the end of his loyalty to the Free Methodists. His account of a “peak experience” is reminiscent of those given by born-again evangelicals. (“I felt I was on cloud nine.”) He hadn’t moved very far, but he was a diﬀerent kind of Protestant, one who had discovered deeper Wesleyan roots, and one who had been opened up to an intellectual and more worldaﬃrming faith. A woman raised in a Presbyterian minister’s family talked about her growing independence and the beginnings of a break with her father’s staunch Calvinism:

evangelical seekers

149

At the age of sixteen, after having theology and Christianity poured into me for years—and as I understood the scriptures then, I wanted to experience baptism. My dad and I went to war over this. So, out of respect for him I did not do this. Then, when my children were born my dad of course wanted to baptize them, but I asked him to dedicate them instead. It was under duress, but my dad did dedicate them.

As her children grew up, on their own they moved out of the Presbyterian Church in which she had raised them, and aﬃliated with a youth-oriented church in a nearby town. Eventually, she and her husband followed their children to the new church—a congregation that practices believers’ baptism by immersion. She explains what happened: As I studied, I simply had to step in faith; it was a spiritual experience. I was in church one night, they were performing baptisms, and one of the people who was speaking on behalf of a person being baptized, said, “If anyone out there is thinking about this and wondering, do it in obedience to Christ.” It didn’t take me two seconds to get out of my seat. Because we (her new church) believe in the priesthood of all believers, family or friends or whatever can perform that baptism. So my children stood and testiﬁed on my behalf and my husband performed my baptism. I’m a very egalitarian Christian female, and to allow my husband to perform that was just a huge thing for my life and a special presentation of God’s love. It was very special for us.

In regard to her father, she said: A month after I was baptized I ﬁnally got the nerve to tell my parents. My dad would not have wanted to come. No, that was a celebration for me. I think it would have concerned him deeply because he is so strongly ingrained in his belief. I wouldn’t have wanted to throw it in his face in any way. In my case it was an act of submission to do this, and submission as I’ve said is a new doctrine for me. I don’t go there very well.

A ﬁnal example involves a woman who grew up in the Salvation Army. Before she had ﬁnished high school, she was promoted to the rank of Corp Sergeant Major, and with this advancement she took her mother’s place in the church hierarchy. However, during this same time period the Corp Oﬃcer sexually molested her. This tragedy precipitated a series of crises that ultimately led to her decision to leave the Salvation Army: Nobody knew about it, and I couldn’t tell anybody because I was afraid, so I kept that little vicious secret for years. It was an ongoing

150

richard w. christopherson thing and I felt very trapped. I took the scripture that says, “There is no temptation taking you but such is common to man, but God is faithful and he will allow you a way of escape.” So, at that time my way of escape was to commit to the Salvation Army and become an oﬃcer. I thought I could get away from him. And I wanted to be— this mentality of my mother was so in me—I wanted to be a good person.

Some years later, still an oﬃcer in the Salvation Army, she was raped by the husband of a friend. When she informed her superior oﬃcer of the attack and the resulting pregnancy, he ordered her to move to a diﬀerent city, stay in one of the church’s homes for unwed mothers, and then give up the baby for adoption. So, okay, I’m a lieutenant and I do what I’m told. I found out that although they ordered me not to talk about what had happened, other people had talked. Also, they kept moving me every six months, and so at that point I decided to leave. It was lonesome, really lonely. And I am the Salvation Army. This was me. I grew up that. But now I had to leave, and it was like a divorce—very traumatic in my life. It took me until the 1990s before I could look at a Salvation Army truck driving down the street and not cry.

The radical break with her childhood church was also a break with her family and her self. Because she did not tell her parents about her pregnancy or her reasons for leaving the Salvation Army, “There was a constant rift between us.” She quit going to church and married within the year. She explained, “I told God that I didn’t feel like I was giving up on God—I hadn’t left him—I just felt like I’d left religion.” After her marriage she did return to church and has been involved in several congregations including Baptist, Presbyterian, and a series of independent and Pentecostal churches. Cradle evangelicals may abandon the faith traditions in which they were raised, but they do not necessarily lose faith in God. It is the institutionalized church that bothers them. In breaking with the past, they join with born-again evangelicals in searching for a community of believers where their personal faith experience can ﬁnd expression. They are not trying to conform to a diﬀerent tradition so much as they are working to create new ways of being religious. Theirs is a more ﬂexible approach to believing and belonging — institutionally tentative, but personally passionate. They continually evaluate present aﬃliations and imagine a corporate religious life that more closely mirrors their self-conception. In a “post-traditional” religious

evangelical seekers

151

world where truth belongs to the individual and institutional pronouncements are met with skepticism, born-again and cradle evangelicals join together as free agents scouring the landscape for symbolic resources and personal experiences that will validate their status as children of God. The Quest: Looking for a Home Evangelical seekers highlight two underlying concerns when they talk about their search for a church. First they are concerned with the “authenticity” of the church, and secondly with there own ability to ﬁt themselves into the congregation. The concern with authenticity can be understood as a subjective judgment about the spiritual integrity of the congregation and clergy. To be authentic, a church must be connected to the transcendent ideals of Christian belief and practice; it has to be a place where people “really” worship and “really” pray. In addition to authenticity, they need to successfully identify aspects of their selves in the congregation and institutional structure. In an eﬀort to ﬁt in, seekers develop creative strategies to accommodate their interests and experiences with those of the congregation. When a church seems to be inauthentic, and/or when an individual can’t ﬁt in, seekers look for a way to leave. Locating and connecting with a church is an active process undertaken by evangelical seekers who work to construct their religious identities from the symbolic materials available to them. Is this an authentic church? It’s much more likely that evangelical seekers are looking for an authentic church than for a Baptist, Methodist, or other denominational church, or even a church that is conveniently located. The concern for authenticity frames the evangelical seeker’s quest for a congregational home. A respondent explained that, “The core of the faith is the real commitment [of individuals], and no matter what the denomination or style of the service, people who feel strongest about the core of the faith are closer to each other than those who do not.” In this sense, the authenticity of congregations grows out of the pure hearts and hard work of individual members and clergy. It is personal authenticity that energizes the congregation as a whole. The discovery of an authentic church, and the ability of a seeker to recognize such a fellowship, validates the truth of his or her own

152

richard w. christopherson

spiritual quest. A respondent described his experience as a newly born-again Christian: “It was real church. This is very important to understanding me. I’ve had a hard time adjusting to anything since. . . . There were some dynamics there that were so organically true, you know, authentic. Shall we say non-institutional?” There is a strong sectarian quality in respondents’ concern for churches that are “non-institutional” and “authentic.” Bellah, et al. (1985) writes that the sect, by deﬁnition, is a fragile organization because it is “secondary to individuals and depends on their continued purity and constant eﬀort to maintain it” (p. 245). This kind of seeker brings the grace that God has bestowed directly on them into the church fellowship. The church does not purify its members; the congregational fellowship is authenticated by the lives and work of the faithful. It is up to individuals to locate and then to sustain such a church. Inevitably, in looking for a church that is authentic, seekers are confronted with the discrepancy between the raw experience at the center of their personal faith and the reality of religious life, as it has been symbolically preserved and organized in institutions. Thomas O’Dea (1983) points out that: . . . the religious institution is itself a dilemma. . . . In religious institutionalization the supraempirical is put together with empirical forms of embodiment; the supra-mundane is put together with mundane forms of expression; the relation to the beyond is put together with here and now social, ideational, ritual forms. (p. 57)

The opposite of an authentic church is precisely these “social, ideational, ritual forms.” Respondents used words such as “religious,” “empty ritual,” “cultural Christianity,” or “just playing church” to describe what they want to avoid in churches. The concern with preserving the purity of an individual’s core experience with the sacred is accompanied by a distrust of tradition and routine modes of worship. People leave churches because they feel predictable. Predictability is one sign that the congregation’s connection with authentic faith and practice has been compromised. For example, in evaluating an evangelical church that he and his wife had recently left, a man commented, “It started to seem manufactured to me. It just didn’t feel like a true church.” Another respondent was in a quandary about authenticity after the pastor preached a “terriﬁc sermon about prayer,” but when the service ended:

evangelical seekers

153

As he came down the aisle where I was standing he asked how I was. I said, “I’m in a hard place,” and I opened up to him brieﬂy that I was tired and pretty discouraged. Now [a previous pastor] would have all but anointed me right then. Here’s a guy who just preached a sermon on prayer and the importance of praying and he just walked oﬀ. I thought his spiritual radar screen is about as undeveloped as any I have ever seen. That starts to create dissonance when you ask, “That was a great sermon but is it backed up by a life?”

On the other hand, a family that has moved in and out of several churches now feels that they are in the right place: The main core is the genuineness of spiritual identity and quest. That’s what I love most. That’s what I want for my kids. You don’t feel religious there. You don’t feel that you’re asked to be and to think in a diﬀerent way than you would outside in regular life. It’s not institutional conformity—it’s Christian faith for everyday life. You come as you are. You ask the questions you ask. They’re not afraid of any questions.

The woman who was born-again during her visit to the drug treatment center eventually found something authentic in the celebration of the Catholic Mass. Her journey has taken her through the Assemblies of God, Presbyterian, Christian Missionary Alliance, and now to an active church life as a Roman Catholic. Her story about discovering Catholicism illustrates two qualities that are common among evangelical seekers, an abiding faith in personal experience with the sacred and a pragmatic individualism when making choices about religion. She was temporarily living in Europe while ﬁnishing her Ph.D.: I realized I was so alone. I was missing my husband; I had no structure, like a church or (my teaching job.) I didn’t have friends there, except the other American students, and they were dangerous! I thought, “I’m going to fall apart morally with these associates if I don’t do something and it’s going to have to be spiritual.” So, I started going to Mass at a little medieval chapel that was near where I ate lunch. I started taking the Eucharist like a Catholic and that was a real experience in itself. It wasn’t as powerful as the (born-again) experience, but I could walk in depressed or alone, take the Eucharist, walk out and everything was in order. Peace every single day.

In acting “like a Catholic” she was operating between the lines of institutionalized authority. She wasn’t a Catholic—wasn’t authorized to take communion—in fact she described her participation in the Eucharist as “faking it.” But her self-authorized experience with the

154

richard w. christopherson

daily Mass was powerful and authentic. It provided a transcendent moment. A cradle evangelical grew up in an immigrant, Pentecostal congregation, but has since aﬃliated with a variety of churches including Calvary Chapel, Presbyterian, Baptist, Christian Reformed, and Evangelical Free. He oﬀered a description of what a genuine church service is like: We walked through that door for the ﬁrst time, and [the pastor] welcomed everyone and opened up with a prayer that was a real prayer. Then we sang beautiful praise choruses. There was a praise band, but nobody up there was performing. They really were worshipping on piano and drums. There was a pastoral prayer—he prayed, really prayed and he prayed for ﬁve minutes. This was anything but perfunctory. There were guys here that reminded me of guys from [the church] where I had grown up. There were bums here that had just found Christ recently. And yet there was a guy from the university. I thought, ‘This is a lovely, very diverse congregation.’ We felt very at home there.

The subjective sense that one church is “real” and another is “just religion,” guides the search for a church home. An authentic seeker has the ability to recognize an authentic church in the faces of individuals in the congregation, in the power of the sacraments, in the prayers of the pastor, and the music that is played and sung. In eﬀect, the moral authority necessary to make the proper distinction, to choose the real over the fraudulent, lies within the individual believer. Personal faith, accumulated experience, and the purity of their commitment to Christ direct the evangelical seeker’s decisions about organized religion. Can I ﬁt in? A church must be genuine, but it will not be without ﬂaws. A respondent remarked that even if he found a perfect church, “the ﬁrst Sunday I showed up it would no longer be perfect.” Another said, “I am a big believer in ‘fallenness,’ and that applies to the church as well.” In an imperfect institutional environment, commitments are tentative and staying in a church requires accommodation and compromise. A recently retired couple summed up a common strategy: “We don’t have any loyalty to the tradition we’ve grown up with, so that isn’t really directing us toward a particular group. We’re trying to ﬁnd a church where we ﬁt in.” The best churches “feel right,”

evangelical seekers

155

but if seekers ﬁnd that ﬁtting in requires them to give up too much, they will leave. Identiﬁcation with churches is part of the process by which they construct and maintain their religious identities, and the constant possibility of exit necessitates an ongoing clariﬁcation of who they are and what church means to them. For example, a single woman who recently moved to a small independent congregation because she had begun to feel “invisible” in a much larger church, considers her chances of connecting with the new place: I do have relationships here, but Sunday mornings are still tremendously lonely experiences because everyone is running. It is sort of happy bedlam, but trying to hold an adult conversation with someone whose children are pulling on their pant leg, “Got to go, we’ve got a soccer game.” Church ends and ﬁve minutes later everyone is gone. This is my opportunity for fellowship. . . . (Also) I am a classically trained musician, but we have a rock band on Sunday morning and orange metal folding chairs. That has taken a long time to get used to. Then, we need to have at least 2 or 3 people who are single in the whole church body. So I don’t know. Leaving would involve a lot of prayer.

The Presbyterian minister’s daughter who followed her children to an independent, youth-oriented church, talked about her children and her own experience adapting to the new church: The youth group [at our former Presbyterian church] was not meeting my children’s needs as they grew in their faith. We had raised them from day one to be cutting-edge Christians. My youngest daughter said it the best. “At [our new church] it’s about God’s word and leading people to Christ and at [our former church] it is about the game that they’re playing in the youth group that night.”

When asked about her role at her new church, she replied: At this point we’re not really involved. We are kind of checking things out. My involvement is to pray daily for the pastor and his wife— kind of behind the scene. . . . Our background is Presbyterian and this is out of the Church of Christ. So this is diﬀerent doctrinally for us. Also, and this is a hard one for me as a female, I am used to being up front and being permitted to do whatever, but that doesn’t happen here. So, for this season, because I’m gaining so much more, for this season, it seems OK. But this is diﬀerent because I’ve always been so outspoken.

When one’s personal experience and convictions don’t match the contours of the new church, ﬁtting in requires creativity and ﬂexibility. So, “for this season” she is willing to suspend her Presbyterianism

156

richard w. christopherson

and her identity as an independent woman and make a tentative commitment to a new church. She is “not really involved” and is on the margins of both institutions. Asked to identify her self in terms of organized religion, this lifelong churchgoer, someone who never misses a Sunday, provided a common evangelical explanation: I’m absolutely not religious. I think of religious people as those who go through the motions. I have a deep, daily, on-going relationship with Jesus Christ—there is very little ritual in my life. Basically I walk with God and I speak with him all day long. Hopefully he infuses everything I do.

Her personal faith precedes and is independent of her involvement in the new church. She is personally secure in the authenticity of her relationship with Christ, but her membership in this local church is fragile and dependent on her ability to compromise and wait things out. A woman who separated from her family’s black Baptist church during her late teens embarked on a journey that has taken her through several Protestant churches, both black and white. At one point she tried a mostly white, evangelical church because it was near her home, and because she had friends who were members: I joined the choir, and I got involved in the new members class, which is kind of my pattern. Anyway, while I’m there, I noticed how conservative politically they were, and they were publishing those things in the church newspaper. I’d be in the choir room and hear some of the other members talking about political issues. They had no understanding—truly none—of racial issues. They didn’t seem to be bothered by not taking a stand on those things. I also discovered that they did not support women in the pulpit. So now I’ve got race, and I’ve got gender issues. I had to leave.

Because of her experiences at this church and others that she subsequently attended, she resolved to ﬁnd a church that was intentional in addressing questions about race and gender. Eventually she began attending an Episcopal parish where ﬁtting in presented a diﬀerent set of problems: One of the pastor’s ﬁrst question to me was, “How have you found your experience here as an African-American woman?” I said, “Why are you asking me that?” And he explained that their staﬀ goes through weekly training in terms of being open to people. I thought, “This is so aﬃrming and comforting to me that they are conscious of that and meeting and supporting me.”

evangelical seekers

157

[So this is a good church for you?] Spiritually I don’t know if I ﬁt. A lot of people are really into the liturgy, and I’m not into that. I’ve always been opposed to that kind of ritualism, so it has taken me some eﬀort. I think I went through the four-week orientation class twice, but I ﬁnd I’m not much into scripted prayer. I really have trouble with that. I’m used to having more of a spiritual move to say the prayer, and being moved spiritually in order to speak for the community. But I think my kids will be ﬁne there, and I don’t ﬁnd that they are going to be hurt there.

She has serious reservations about the heart of the Anglican tradition—the liturgy and the Book of Common Prayer—but she is willing to overlook her aversion to “ritualism” because this is a church where her identity as a woman and a person of color is aﬃrmed, and where racial diversity is accommodated. Finally, the man who was born-again upon arrival at his university campus also talked about the diﬃculty of ﬁtting his theology and politics into a conservative evangelical church: I’m evangelical in terms of orthodox faith, ecumenical in the I endorse a lot of the platforms of the World Council of which would be identiﬁed as liberal Christianity. But I’m eral Christian. I’m an orthodox Christian with many views be categorized as liberal or left of center.

best sense. Churches, not a libthat could

[How do your politics aﬀect your life at your current church?] It makes it diﬃcult, but like I said, it’s a good discipline. [At my current church] They’re still singing, “When Christ returns we’re going to ﬂy away.” Or they say, “I’m saved, they’re damned, if you don’t come to Christ you’ll be damned too.” They proof-text these pronouncements, and that irks me, but there’s so much that is good, and it’s important for me to stay rooted in that tradition. I need the tradition . . . to keep strength and direction and evangelical rootedness. It tethers me, a deep part of me.

Those roots are rediscovered at a Saturday night worship service that exudes “the feeling of the early Jesus movement.” Services are informal and led by “well known and very, very talented, gifted, and humble musicians.” Irksome objections to the social and political implications of his church’s theology are set aside. His solution is the reverse of the African-American woman’s—here the worship experience conﬁrms his personal experiences as a born-again Christian, but the church’s theology leaves something to be desired. Outside of church, he ﬁnds spiritual resources in a range of settings including

158

richard w. christopherson

a small group of Christian colleagues whom he meets with regularly for prayer and sharing. This ecumenical evangelical ﬁnds the relational and symbolic resources he requires to sustain faith even as he remains marginal to the collective life of his current congregation. Working out a resolution Many evangelical seekers express a sense of resignation about their choice of a church. They are troubled. It is clear to them that their own experience with the gospel and their commitment to Christ are too big to be contained within a particular congregation. Final loyalty is reserved for something larger and more mystical—the “authentic core” of Christian faith. However, they need the local church, and they feel called to corporate worship and fellowship. A woman, who has been involved in several churches since her born-again experience over 20 years ago, appeals to Biblical authority and expresses a common sense of obligation: Scripture really commands that we join with a body of believers. It’s clear that it is God’s design for people. Despite the list [of former churches] that I gave you, I really don’t hop around. If I truly have diﬃculties with a church, I really give it a lot of thought and prayer before I make a change. I’ll be in church the rest of my life and I will be as active in the church as God allows me to be.

Of course, the respondents for this study do have diﬃculties with churches, and they are far less apt to call themselves “Protestants,” “Presbyterians,” or even “evangelicals,” than they are to identify as “a follower of Jesus,” “just Christian,” or to say, “I have a relationship, not a religion.” They are committed to church going, and hold out hope of ﬁnding a congregational home, but their experience with a range of churches has left them tentative and even cynical about commitment to a congregation. Respondents commonly recalled times when their church lives were less complicated, and they felt accountable and securely bound to others. For example, the African-American woman who talked about struggling to ﬁt into an Episcopal parish reﬂected on what has been missing since she left home many years ago. She grew up in church with her “parents, grandparents, and all my aunts and uncles. Everybody was there, and we went every Sunday. . . . There is a lot of comfort where you know a large number of families and there is that common experience.” When her “saintly” grandmother died and her grandfather remarried, “It kind of divided things. Then after

evangelical seekers

159

that, as we were growing up, we just went our own way.” Now, after years searching for a church home in a wide range of conservative, liberal, black, and white churches, she is sometimes carried back to her family’s religious roots and memories of her own childhood faith: When I’ve had a crisis, a death or whatever, it has brought me back to the [African-American] church. I have relatives there, and there is something very comforting when I go. It is like a homecoming. I can’t quite describe it. Yet at the same time I know that it isn’t really me. It’s a hard place. One Sunday I stopped (at a store) to get something, and there was a family, husband, wife, and kids, and they were all dressed in white. They had just been to church, or they were on their way. I remember Sundays were like that, and there is sort of a longing and this feeling. . . . Of course, they were looking at me and probably thinking, “Girl you can go any old way you want!” There is something wonderful about being a part of something and belonging, but I’m not that person anymore.

The man who grew up in an immigrant, Pentecostal Church also remembers his childhood congregation as an extension of his own family. It was “a real community,” but a community whose fundamentalist roots he has outgrown. His story is about the discovery of a more inclusive Christian fellowship and an increasing willingness to ignore institutional boundaries, but it’s also a story tinged with the frustration of marginality and transience. He labeled his pattern of church attendance “sampling”—simultaneous involvement with multiple congregations. On a given Sunday, he and his wife attend an early service at a Reformed church where he appreciates the “excellent, solid preaching,” but where he has “honest theological diﬀerences,” and “it would feel claustrophobic to become a part of that.” Later in the morning they attend a growing evangelical church where they are still formally on the membership roll, although he sees signs that the church is becoming “an evangelical, yuppie social club.” Likewise his wife is “no longer comfortable” in this congregation. Then, on less frequent occasions, they also attend services at a liturgical church that they love “for the high holy days, the music and the aesthetics of the building.” When asked, “Are any of these places your church home?” he responded: I honestly don’t know what to do. I feel conﬂicted about this. “Shame on me!” Is this consumerism practiced to a point where it’s disgusting? Am I missing something? So I always want to keep this in perspective and say, “Father, we are blessed.” There are so many excellent

160

richard w. christopherson churches in [his town]. But a church home, a real family, those are things you don’t just put on like a suit.

For him there is a danger in losing sight of the authentic quest and the irreducible core of commitment that marks both the true church and the faithful seeker. There is a diﬀerence between religious tourism and a sojourner’s life of service and divine purpose. As a substitute for a single church home, “sampling” is a practice that brings him close to the line separating integrity and hypocrisy. If it is true that he has succumbed to the market and is simply consuming religious products, then the authenticity of his search is compromised, and his ability to discern a genuine church is in jeopardy. However, if his search is genuine, perhaps he is simply a “non-denominational” individual comfortable in a range of congregations, but at home in none. Finally, the born-again evangelical who discovered her aﬃnity for Roman Catholicism while studying in Europe, talked about an emergent recognition of her Catholic identity. She recalled an incident that marked a turning point when she “ﬁnally decided that I had become more Catholic than Protestant.” The impetus was a meeting with her academic advisor. After reading one of her papers he inquired, “You’re Catholic aren’t you?” His question surprised her even though this was the precise issue that she had been dealing with during her time in Europe. I said to him, “You know me, I’m a conservative Protestant, now why would you think I’m Catholic?” He said, “Your chapter is Catholic. You write like a Catholic. You’re very sensitive to the symbols. You really have a Catholic mentality that I didn’t expect because I know your background.” That really made me think. If somehow this is showing up in my intellectual work, there’s something spiritual and deep going on that’s out of my control. I decided that I’d better catch up with it, and felt that the way to catch up with it is through the organizational commitments that go along with Catholicism. So when I came back here I started going to [her local Catholic parish], and I talked to my husband about it and ﬁnally talked to my parents about it.

[So how do you see yourself now? Are you a Catholic?] The only term I’ll accept is Christian. I almost see myself on a spectrum—the born-again, Pentecostal, Presbyterian, Missionary Alliance experiences and now Catholicism. I don’t feel like I’ve turned against my previous experience. I think that has enhanced me, but what I

evangelical seekers

161

didn’t have in Protestant churches was the Eucharist and that is so key. I’m willing to do whatever I need to do to be a good Catholic for the sake of the nourishment and the renewing power of the Eucharist.

Her solution to the church problem is a novel one. She embraced a tradition outside the evangelical mainstream, and has been instructed, conﬁrmed and formally accepted into the Roman Catholic Church. Of course, like a true evangelical it is personal experience, “something deep and spiritual,” that directs her choice. Acting like “a good Catholic,” while remaining just “Christian,” is the path that ensures her continued spiritual vitality. She has settled in, is no longer switching or sampling, and now understands this surprising decision as a synthesis of previous religious identities and aﬃliations. Discussion Dimensions of Religious Identiﬁcation The data collected from this sample of educated evangelical seekers suggest that they make decisions about church in a broad context of religious prerogatives. It is not only a decision to attend or not attend, or about which church to select. There are multiple “dimensions of religious identiﬁcation” (Hervieu-Léger, 1998, p. 218), involving various choices and types of authority, that come together in certain combinations in the lives of these religious seekers. The meaning of their church participation and aﬃliation can be examined in light of the four dimensions of identiﬁcation that Hervieu-Leger labels: (1) “communal,” belonging to a particular local church and accepting the social and symbolic markers of membership in that congregation; (2) “ethical,” which refers to the universal aspects of the religious message—what people believe and value independent of their local aﬃliation or lack of aﬃliation; (3) “cultural,” the heritage of a given religious tradition including its symbolic legacy, way of life, institutional structure, etc.; and (4) “emotional,” or the “intense experience” of individuals that produces a sense of belonging which may or may not be connected to a particular religious community (Hervieu-Léger, 1998, pp. 218–222). Given this conceptual scheme, four aspects of the data deserve discussion. First, when born-again and cradle evangelicals break from their family’s religion, it is primarily the ascribed “cultural” identiﬁcation that is left behind. They don’t want to be cultural Christians. For

162

richard w. christopherson

them, this is “playing church” and/or “empty ritual,” and this is what they abandoned when they left their inherited religion behind. When evangelicals say, “I’m non-denominational,” “just a follower of Jesus,” or “I’m just Christian,” they are aﬃrming their independence from the cultural dimension of religious identiﬁcation. Second, personal experience with the sacred provides authority and guidance as seekers embark on their quest for a new spiritual home. In a sense, they cultivate and collect “emotional experiences” and use those gathered memories to reconﬁrm their connection to things that are transcendent. They talk about events such as their born-again encounters, sacramental epiphanies, and sense of God’s call on their lives—peak experiences that are interpreted as evidence of personal authenticity and link to the divine realm. Such experiences are drawn from both seekers’ private lives, and from corporate experiences such as congregational worship and prayer, baptism, and Holy Communion. Their quest is not the “completely private, idiosyncratic, self-serving spirituality” that Wuthnow (1998, p. 16) was concerned about, and that he discovered in portions of his broadbased sample of seekers. They are looking for a church, and church stories are a persistent narrative thread in accounts oﬀered by respondents for the present study. However ﬁckle and transient their aﬃliations prove to be, the history of congregational involvement recounted by evangelical seekers tempers their individualism and places boundaries around the territory to be explored. Wuthnow points out that “spiritual practices [as opposed to ‘spiritual seeking’] have a moral dimension, for they instruct people in how they should behave toward themselves and with each other (1998, p. 16). This morality is evident in the care with which these evangelicals move about in the world of churches, and the concern they have for authenticity in their spiritual lives. Experiences with the sacred conﬁrm God’s work in their lives, and are recalled as rites of passage marking the path they have traveled and providing meaning for what has transpired. Third, evangelical seekers, with or without a clear congregational or denominational identiﬁcation, hold tightly to the “ethical” message of the Christian gospel. They are “mere Christians”—true believers who claim a personal relationship with God and identify themselves within the mystical “body of Christ.” Such identiﬁcation is not dependent on institutional ties, but represents their hope for spiritual fellowship that transcends the worldly problems of congregational life. Such a stance imbues these seekers with a kind of

evangelical seekers

163

prophetic power. In the quest for an authentic congregation, their commitment to universal Christian values results in a claim to know better than others what choices they should make about churches. Fourth, evangelical seekers struggle with “communal” identiﬁcation, but refuse to give up on the idea of ﬁnding an authentic church. Seeking is not an end in itself—they want a spiritual home, but their rejection of cultural Christianity, dependence on direct experience with the sacred, and their abiding faith in the transcendent dimension of the Christian gospel combine to create a dilemma in respect to congregational ﬁdelity. As much as they want a church home, they don’t want to be just members of a congregation or just another adherent of a denominational group; such social markers of religious belonging are inadequate symbols of their true identity as Christians. The congregational lives of evangelical seekers are contingent and pragmatic, guided by a passion for spiritual authenticity and need for corporate worship, but tempered by the reality of broken fellowship and human fallibility. Churches provide a location for spiritual practice and sacramental grace, and churches help with the socialization of their children, but the desire of their heart is for something even more than this. Conclusion This study describes the transient church lives of a sample of educated, evangelical Christians. They live in the midst of an expansive religious market, and their tolerance of and experience with a range of churches works to blur institutional boundaries. Despite the certainty of their calling to an active congregational life, they can remain institutionally unidentiﬁed for long periods of time, free to migrate from congregation to congregation without much restriction from denominational or theological labels. Their experience with churches sheds light on the meaning of church in a pluralistic and individualized social environment, and provides insight into the ways that religion works for modern believers. This small study is suggestive of additional research on other subsets of modern religious believers. In terms of American Christians, there are a host of church stories yet to be recorded. A few examples: evangelicals who have found a home in mainline and liberal Protestant churches, in Roman Catholic parishes, or in Orthodox

164

richard w. christopherson

congregations; Catholics, Orthodox, mainline, and liberal Protestants who have settled into evangelical churches; church switchers who continue to believe, but who have abandoned their quest for a church home and are resigned to a non-congregational religious life; and seekers who decide to stay in congregations because they want to work to bring about institutional change. What we know about the evangelical seekers interviewed for the present study is that they do not identify with “cultural Christianity”— they have not been redeemed by virtue of inherited religion—and that they do identify with the universal/ethical and the emotional/ experiential dimensions of Christian faith. This combination of rejection of the cultural dimension and embrace of the universal and experiential dimensions of Christianity, contributes to the diﬃculties they have with churches. The worldly reality of congregational life pales in comparison to the transcendent ideals of the Christian gospel and the purity and power of personal spiritual experiences. Tensions arise between these contrasting perspectives, and the “authenticity” of congregations is compromised. Pastors may not be able to really pray; sacraments might lack evident power; individuals can lose their spiritual vitality; the joy of Christian fellowship may not develop; the church’s theology sometimes conﬂicts with the ancient Biblical ideals of peace and justice. In light of this three-cornered relationship between absolute Christian truth, personal experience with the sacred, and a congregation’s communal life, the liabilities of local churches are readily exposed. Because evangelical seekers depend on the authority of personal belief and individual experience, and because they lack deep commitment to particular cultural traditions, it is diﬃcult for them to generate the institutional loyalty that anchors believers in one place. Given a vast, anonymous, urban church market, seekers can continue, seemingly indeﬁnitely, in their quest for a church home. They seek without abandoning the conviction that God has called them to be in church. Their loyalty and their hope are anchored in a uniﬁed and transcendent body of believers, a communion of saints not visible to the naked eye. They experience this mystical fellowship whenever their personal faith and practice is most attuned, and the church of their choice is centered on the authentic Christian gospel.

evangelical seekers

165

References Ammerman, N. (1997). Religious choice and religious vitality: The market and beyond. In Lawrence Young (Ed.). Rational choice theory and religion: summary and assessment. New York: Routledge. Becker, P. E. (1998). Congregations in conﬂict: Cultural models of local religious life. Cambridge: Cambridge University Press. Bellah, R., Madsen, R., Sullivan, W., Swidler, A. & Tipton, S. (1985). Habits of the heart. Berkeley: University of California Press. Bibby, R. W. (1999). On boundaries, gates, and circulating saints: A longitudinal look at loyalty and loss. Review of Religious Research 41:149–164. —— (2003). The circulation of the saints: One ﬁnal look at how conservative churches grow. Paper presented at the Annual Meeting of the Paciﬁc Sociological Association, Pasadena, CA. Christopherson, R. (1994). Calling and career in Christian ministry. Review of Religious Research, 35:219–37. Hadaway, C. K. & Marler, P. L. (1993). All in the family: Religious mobility in America. Review of Religious Research, 35:97–116. Hervieu-Léger, D. (1998). The Transmission and formation of socioreligious identities in modernity: An analytical essay on the trajectories of identiﬁcation. International Sociology, 13:213–228. —— (2000). Religion as a chain of memory. New Brunswick, N.J.: Rutgers University Press. Loﬂand, J., & Loﬂand, L. (1995). Analyzing social settings: A guide to qualitative observation and analysis. Belmont, CA: Wadsworth Publishing Company. Loveland, M. T. (2003). Religious switching: Preference development, maintenance, and change. Journal for the Scientiﬁc Study of Religion, 42:147–57. O’Dea, T. F. & Aviad, J. O. (1983). The sociology of religion, (2nd ed.). Englewood Cliﬀs, N.J.: Prentice Hall. Roof, W. C. (1999) Spiritual marketplace: Baby boomers and the remaking of American religion. Princeton, N.J.: Princeton University Press. Roof, W. C. & McKinney, W. (1987). American mainline religion. New Brunswick, N.J.: Rutgers University Press. Sherkat, D. E., & Wilson, J. (1995). Preferences, constraints, and choices in religious markets: An examination of religious switching and apostasy. Social Forces, 73:993–1026. Smith, C. (1998). American evangelicalism: Embattled and thriving. Chicago: University of Chicago Press. Stark, R. & Finke, R. (2000). Acts of faith: Explaining the human side of religion. Berkeley: University of California Press. Warner, R. S. (1993). Work in progress toward a new paradigm for the sociological study of religion in the United States. American Journal of Sociology, 98:1044–1093. Wuthnow, R. (1998). After heaven: Spirituality in America since the 1950s. Berkeley: University of California Press. —— (1988). The restructuring of American religion: Society and faith since World War II. Princeton, N.J.: Princeton University Press.

SPIRITUALITY AND RELIGIOUS VARIABLES AS PREDICTORS OF WELL-BEING IN SEX OFFENDERS Brendan Geary, Joseph W. Ciarrocchi and N.J. Scheers* Abstract The purpose of this study was to examine empirically the contribution of spirituality and attendance at religious services to well-being, over and above demographic and personality factors, in a sample of male sex oﬀenders. The sample consisted of 195 men in outpatient sex oﬀender treatment programs or individual therapy. Consistent with previous research, the religious variables contributed added unique variance to cognitive well-being and positive aﬀect, but not to negative aﬀect. Spirituality predicted to the positive components of well-being, and attendance at religious services predicted independently to cognitive, but not emotional components of well-being. Spirituality was a more powerful predictor than religious practices. The implications of these ﬁndings for sex oﬀender treatment and recovery are discussed in the wider context of recent developments in the ﬁeld of positive psychology.

Recently there has been a growing interest in the role of religion and spirituality in physical and mental health. Researchers and therapists have begun to explore the role of religion and spirituality in recovery, and ways of involving this in treatment (Miller, 1999). There has been considerable research and writing in the area of addictions treatment, partly due to the inﬂuence of the Twelve-Step movement, which emphasizes the role of God or a Higher Power in recovery. Although in some parts of the world there has been a signiﬁcant decline in church going, this has not been true to the same extent in the United States of America. In the past 15 years, there has been a growing body of scientiﬁc literature exploring the relationship between religion, spirituality, and mental health. Seybold and Hill (2001, p. 21), in a review of recent research, write that, “the impact of religion and spirituality on physical and mental health . . . is largely beneﬁcial.”

* Author note: the authors would like to thank Greg Lehne, Ph.D., and Gerard McGlone, Ph.D., for comments and suggestions in the process of the research and writing of this article.

Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

168

geary, ciarrocchi and scheers

While some writers and researchers are beginning to look at the treatment of sex oﬀenders from a holistic perspective that involves spirituality and religion (Earle, Dillon and Jecman, 1998, Ellerby and Stonechild, 1998), there is almost no empirical research in this area. Many sex oﬀenders are church attendees, have a spiritual life, and try to involve religion in their personal path of recovery. Sardjian and Nobus (2003, p. 920), for example, write, “There must be a recognition and appreciation of the oﬀenders’ spiritual life, with treatment aiming to refocus it and harbor the adaptive beliefs implicit in the oﬀenders’ religion rather than deny or chastise the religion itself.” The positive psychology movement focuses on well-being and not just pathology, on resilience rather than failure, on resources rather than deﬁcits. The growth of interest in positive psychology has led to research on the factors that contribute to well-being. Steel and Ones (2002, p. 776) write that well-being can “largely be described by two components: aﬀect and satisfaction/happiness.” Watson (2000) demonstrates that well-being is associated with experiencing high positive aﬀect and low negative aﬀect, and that these two states are highly correlated with the personality dimensions of neuroticism and extraversion. He goes on to show that there are other factors that contribute to individual diﬀerences in well-being. Life satisfaction, or cognitive well-being, is a construct that relates to comparisons made by the individual with standards of the good life. DeNeve and Cooper (1998) conducted a meta-analysis of 137 personality traits and subjective well-being. They noted that most studies examined either demographic or personality factors, but few regression analyses incorporated both sets of variables. They conclude, “demographic variables and life events have a surprisingly small eﬀect on long-term satisfaction with life” (p. 219). They suggest that personality is an important correlate of subjective well-being, particularly over the long-term. While there is empirical evidence to suggest that demographic factors contribute little to well-being (Diener, 2000, Watson, 2000), it was decided to include age, education, current income, and relationship status in the study. Hanson (2002) has shown that recidivism rates are diﬀerentially related to age with diﬀerent types of oﬀenders, so it is necessary to include this as a variable with sex oﬀenders. Financial resources, education, and relationships, also contribute to well-being, and therefore should be controlled for as well.

spirituality, well-being, and sex offenders

169

There is general agreement that personality, as measured by the Big Five factors of neuroticism, extraversion, openness, agreeableness, and conscientiousness, is a signiﬁcant predictor of well-being, with neuroticism and extraversion taking the lion’s share of the variance (DeNeve and Cooper, 1998, Diener, Suh, Lucas & Smith, 1999, Watson, 2000). Watson (2000) and others (Myers & Diener, 1995, Diener, 2000) suggest that religion and spirituality can contribute unique and signiﬁcant variance to individual well-being. It is therefore worth investigating what, if anything, these factors contribute to the wellbeing of sex oﬀenders. Common sense indicates that the happy person is less likely to oﬀend, and research conﬁrms that negative aﬀect, especially when combined with anger or frustration, inclines oﬀenders to act out (Hanson, 2000). Thus, it is important to discover factors that contribute to well-being among sex oﬀenders. Both correlational and regression studies support the view that spirituality and religion are related to well-being in the general population (Gartner, Larson, and Allen, 1991, Mickley, Carsen and Soeken, 1995, Koenig, McCullough, and Larson, 2001). This study explores what contribution, if any, these factors contribute to well-being in sex oﬀenders. There are currently no empirically published articles examining the unique contribution of spirituality to well-being in this group. While there is sparse data available on the contribution of spirituality to well-being in sex oﬀenders, there is a growing literature supporting the view that spirituality and/or religion can contribute positively to people in recovery from addictions (Carroll, McGinley, and Mack, 2000). In a recent study of an outpatient substance abuse sample, Piedmont (in press) reports that spiritual transcendence at time one testing predicted coping ability, stress levels, and global symptoms eight weeks later, even when controlling for age, gender, mental status, negative aﬀect, and the ﬁve-factor personality domains. The present study seeks to determine whether such ﬁndings apply to sex oﬀenders. Carnes (1991) oﬀers the results of a survey of sex addicts. He writes that in the repair stage (between 18 and 36 months after beginning treatment), “For most, spirituality was the primary factor, the main basis for change” (p. 199). It is important to bear in mind that not all sex oﬀenders are sex addicts, and that these are the results of a survey and not a study. Nevertheless, this information does provide a better-than-anecdotal basis for the hypothesis that

170

geary, ciarrocchi and scheers

spirituality and religion may be important contributors to recovery for sex oﬀenders. The purpose of this study was to measure the relationship between spirituality and attendance at religious services, to positive aﬀect (PA), negative aﬀective (NA), and satisfaction with life (SWL) in sex oﬀenders in outpatient treatment programs, over and above selected demographic variables and personality as measured by the NEO FiveFactor Inventory (NEO-FFI). A hierarchical multiple regression was conducted to determine ﬁrst the unique contribution of spirituality and then of attendance at religious services on subjective well-being (SWB), over and above selected demographic variables and personality as measured by the NEO-FFI. Four models were developed for each of the three dependent variables. Demographic and situational data were gathered using a questionnaire designed for this study that included the following variables: age, education, income, relationship status, ethnicity, stress, health, prison experience, and length of time in treatment. Items on Private Religious Practices, Organizational Religiousness, and Religious Preference were selected from the Brief Multidimensional Measure of Religiousness/Spirituality: 1999 (Fetzer Institute, 1999). The item on attendance at religious services was used as an independent variable. Therapists were asked to provide information on the nature of each participant’s oﬀense/behavior, and/or the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV; American Psychiatric Association, 1994) Axis I and Axis II diagnoses. Method Participants The participants in the research were all male. Data were collected from 25 groups in nine treatment centers, and seven individual therapists between July, 2002—January, 2003. The participants in one group in a rural setting declined to participate (11 members), as did members of three groups in a major metropolitan city (16 members). Individual clients of a therapist in a suburban area also declined to participate. The groups who agreed to participate comprised 253 men. Thirteen men were either not present at the time of the survey or were not able to complete it (e.g., because of language limitations, reading diﬃculties, or absence from group for health or

spirituality, well-being, and sex offenders

171

other reasons). Members of groups in one treatment center were invited to complete the survey at home. Seventeen out of 35 men in the groups in this treatment center returned completed surveys. A total of 35 men in groups that cooperated with the study refused to participate and 18 of these came from one treatment center. Two hundred and ﬁve surveys were returned. Eighteen subjects were contacted through individual therapists. All others were contacted through their groups. Twenty-three participants completed the questionnaire outside of group time. The others (164) completed the questionnaire during their time in group therapy. Ten surveys were deleted due to incomplete data, leaving a total of 195 usable data sets. Materials The Satisfaction with Life Scale The Satisfaction With Life Scale (SWLS) is a ﬁve-item instrument that is a unidimensional narrow-band measure of the cognitive dimension of subjective well-being (Diener, Emmons, Larsen, Griﬃn, 1985). Respondents are asked to indicate their response to ﬁve statements ranging from (1) strongly disagree to (7) strongly agree. Pavot and Diener (1993) report six studies with coeﬃcient alphas between .79 and .89. Test-retest coeﬃcients between two weeks and two months ranged from .64 to .84. The mid-point on the scale is represented by a score of 20, but most groups report means between 23 and 28, which are in the range of slightly satisﬁed to satisﬁed. The PANAS Scales The Positive and Negative Aﬀect Schedules (PANAS) were designed by Watson, Clark and Tellegen (1988). The internal consistencies for the scales are high, with alphas ranging from .86 to .90 for PA, and from .84 to .87 for NA. The PANAS scales are reliable, relatively independent measures of positive and negative aﬀect that have demonstrated reliability with normal adults, inpatient psychiatric populations, as well as with undergraduates. The NEO-FFI The NEO Five-Factor Inventory (NEO-FFI), developed by Costa and McCrae (1992), is a short form of the NEO Personality Inventory

172

geary, ciarrocchi and scheers

(NEO-PI), developed by the same authors. It is a 60-item, self-report inventory. Items are scored on a ﬁve-point Likert scale from strongly disagree (1) to strongly agree (5). This shortened instrument provides scores for the ﬁve global domains of the NEO-PI without providing scores for the six facets in each domain. The ﬁve personality domains measured by the NEO are neuroticism (N), extraversion (E), openness (O), agreeableness (A), and conscientiousness (C). There are 12 items in the NEO-FFI to measure each of these domains. Costa and McCrae report coeﬃcient alphas of .92, .89, .87, .86, and .90 for N, E, O, A, and C respectively, for each of the domains, based on a sample of 1,539 men and women (1992). Since the NEO-FFI is a shorter inventory it is to be expected that some variance will be lost in exchange for brevity and convenience. Costa and McCrae (Costa and McCrae, 1992, p. 54) report that, on average, the NEOFFI scales “appear to account for about 85% as much variance in the convergent criteria as do factor scores.” The Faith Maturity Vertical Scale The Faith Maturity Vertical Scale (FMS) is a measure of spirituality that forms a subscale of the shorter form of the Faith Maturity Scale, which is the 11-item revision of Donohue’s 12-item short form of the full Faith Maturity Scale (Benson, Donohue, & Erickson, 1993). Benson et al. (1993) report a correlation of r = .94 between the original scale and the 12-item short form. This scale was used by Piedmont and Nelson (2001) in their psychometric evaluation of the FMS. The vertical subscale evaluates the individual’s relationship with God. They report an overall reliability of .92 for the short form of the FMS for a sample of 1,786 undergraduate students (1288 women and 496 men). Religious Practices Ten items comprise the Religious Practices Questionnaire (RPQ) selected from the Brief Multidimensional Measure of Religiousness/ Spirituality: 1999, (Fetzer Institute, 1999). The following unspeciﬁed reliabilities are reported for the items used in the RPQ: religious service attendance (r = .70), other public religious activities (r = .70), private prayer (r = .55), meditation (r = .51), Bible reading (r = .56), feel God is punishing me (r = .37), wonder if God has abandoned me (r = .37), carry beliefs to other areas of my life (r = .56).

spirituality, well-being, and sex offenders

173

Procedure The questionnaires were distributed to participants either in groups or by their therapists. After completing the questionnaires, each participant was asked to hand his envelope to the therapist who added a sheet indicating the participant’s oﬀense/behavior and/or DSM IV Axis I and Axis II diagnosis, without examining any questionnaires inside the envelope. Instructions were provided for each instrument. Results Demographic data The mean age of the total sample is 40.6. Almost three-quarters (73.3%) of the group were Caucasian (141), 35 were African-American (17.9%), 8 were Hispanic (4.1%), 3 were Asian-American (1.5%) and 6 indicated “other” (4.6%). Just over two-thirds (68.2%) have spent some time in jail. Slightly over a third (37.4%) were either married or in a committed relationship; 57 (29.2%) reported that they were married, and 16 (8.2%) indicated that they were in a committed relationship. The majority (62.6%) lived alone for various reasons; 72 (36.9%) reported that they were single, 48 (24.6%) reported that they were separated or divorced, and 1 (0.5%) reported that he was widowed. The group was evenly divided between college graduates and non-graduates. The annual income for the majority (80%) is less than $40,000, but over two-thirds are in full employment (69.7%). Therapists were asked to insert a form indicating the DSM-IV (1994) diagnosis for each participant. In the course of data collection, it was discovered that there was a lack of consistency among therapists in the area of DSM-IV diagnosis. Some therapists did not give diagnoses to clients, and, in situations where the client did not meet the criteria for a DSM-IV diagnosis, the approaches to diagnosis diﬀered. The diagnosis for pedophilia (302.2), for example, requires, “Over a period of at least six months, recurrent, intense sexually arousing fantasies, sexual urges or behaviors involving sexual activity with a prepubescent child or children” (APA, 2000, p. 572). In situations where the sex oﬀender had abused a child, but all criteria for the DSM-IV diagnosis were not present, the therapists sometimes used V61.21, sexual abuse of a child, or gave no diagnosis. In order to collect useful information on the nature of the oﬀenses,

174

geary, ciarrocchi and scheers

the diagnosis form was then altered to include a description of the oﬀense/behavior, as well as the diagnosis, if this was available. Forty-seven (24%) participants received the DSM-IV diagnosis of 302.2 pedophilia. The other 53 participants with child-related oﬀenses may have been involved in sexual behavior with a child that did not fulﬁll the DSM-IV criteria for pedophilia. Nineteen diagnosis forms indicated an intra-familial oﬀense. One therapist indicated that the participant was in treatment because of Internet-related child pornography oﬀenses. Ten participants (5.1%) were exhibitionists, 9 (4.6%) were rapists/sexual aggressives, 6 (3.1%) were voyeurs, 3 suffered from sexual addiction (1.5%), and one had the diagnosis of transvestism (302.3). No data were available for 66 participants (34%). Religious Practices Approximately one third of the sample report attendance at a religious service (34.4%) and read the Bible or religious literature (32.3%) once a week or more often. Almost half of the sample report saying prayers or grace before meals at home on a weekly basis or more often, and almost two thirds (59.5%) pray privately or meditate once a week or more often, but use of religious media on a regular basis is reported by less than a quarter (23.6%) of the sample. Conversely just over a quarter never attend religious services (26.2%), one third reports never reading the Bible or religious literature (33.5%), and more than one third never say prayers like grace before meals at home. However, just fewer than 20% report never praying privately or meditating. Over half the sample never make use of religious media (53.8%). Almost half of the sample reported that they made use of religion or spirituality to help in their recovery process “quite a bit” or “a great deal.” Just under 40% reported that they tried to carry their religious beliefs over into their daily lives “quite a bit” or “a great deal.” The majority (68.7%) reported that they did not feel that God was punishing them for their sins or lack of spirituality, and slightly more (72.8%) did not feel abandoned by God. Religious Aﬃliation Protestants comprised just over half the sample, 56.9% (111), with Catholics comprising 17.9% (35), and 1.5% (3) indicating Jewish. The remaining participants were almost evenly divided between

spirituality, well-being, and sex offenders

175

12.3% who chose “other” (24), and 11.3% who reported “none” (22). Reliabilities The reliabilities (Cronbach’s alpha) of the scales used in the study were all acceptable. The alphas for the scales used in the study were as follows: SWLS, r = .77, PA, r = .88, NA, r = .90, and FMS vertical, r = .93. These are all acceptably high and comparable with published data. Correlations Table 1 shows the pattern of correlations among the scales used in this study and the religious/spiritual variables. As was expected, the Faith Maturity Vertical Scale and attendance at religious services were highly correlated (r = .61, p < .01, 2-tailed). There was also a strong correlation between NA and neuroticism (r = .61, p < .01, 2-tailed), however, the other scales were not highly correlated. There were signiﬁcant correlations between the FMS vertical scale and all other scales apart from openness. Attendance at religious services correlated positively with satisfaction with life (r = .23, p < .01, 2-tailed) and negatively with neuroticism (r = –.19, p < .01, 2-tailed). We can conclude that the variables used in this study are related, but not redundant. (See Table 2 for means, range, and standard deviations). Table 1. Intercorrelations for well-being, personality, and religious variables Measure

1

1. Positive aﬀect — 2. Negative aﬀect –.09 3. Satisfaction with life .23** 4. Neuroticism –.29** 5. Extraversion .39** 6. Openness .16* 7. Agreeableness –.01 8. Conscientiousness .30** 9. Faith Maturity .29** 10. Attendance at .10 religious services

2

3

4

5

6

7

8

9

10

— –.43** — .60** –.39** –.14* .29** –.03 .17* –.24** .11 –.23** .18* –.20** .31** –.10 .22**

* p > .05 (2-tailed), ** p > .01 (2-tailed)

— –.28** –.01 –.35** –.37** –.28** –.19**

— .29** — .23** –.02 .28** –.18* .31** .13 .12 .02

— .34** — .15* .23** — .00 .07 .61** —

176

geary, ciarrocchi and scheers

Table 2. Mean, Standard Deviation, Range of Reported Scores and Range of Possible Scores of Predictor and Outcome Variables Variable

Mean

Positive aﬀect Negative aﬀect Satisfaction with life Neuroticism Extraversion Openness Agreeableness Conscientiousness Faith Maturity

33.3 23.4 16.7 22.9 27.9 26.1 31.2 33.0 31.1

SD 7.9 .3 6.7 8.2 5.9 6.3 5.7 6.7 8.0

Range Min

Max

13.7 10.0 14.0 3.0 10 15 13 10 8.0

50.0 50.0 34.0 43 42 48 45 47 56.0

Hypothesis Testing The hypotheses of this study were that spirituality and religious practices would show incremental validity in predicting to the well-being variables over and above demographic variables and personality factors. To determine whether these variables improve predictability, a series of hierarchical multiple regressions was performed with each facet constituting the criterion in separate models. For each model, demographic variables were entered in step 1, personality in step 2, and either spirituality or attendance at religious services in step 3. Partial F-tests were calculated at each step to determine whether a particular set of variables signiﬁcantly accounted for the explained portion of the variance. Model 1. Spirituality as a predictor of well-being As can be seen in Table 3, demographic variables contributed around 3% of the variance to SWL, NA, and PA and were not signiﬁcant. While the demographic variables as a group were not signiﬁcant, relationship status predicted positively to NA in all three models (see Table 5 for beta weights). Those in a relationship experienced higher levels of NA. Personality, taken as a whole, was a signiﬁcant predictor of all three variables, contributing 36% of the variance to NA, 23% to PA, and 19% to SWL. Neuroticism predicted negatively to SWL and PA and positively to NA, whereas extraversion, agreeableness,

spirituality, well-being, and sex offenders

177

and conscientiousness were signiﬁcant predictors of PA, but not of SWL or NA. Spirituality (Faith Maturity Vertical) was a signiﬁcant predictor of SWL and PA conﬁrming one portion of the hypotheses of this study. Faith Maturity Vertical contributed unique added variance over and above demographic variables and personality, predicting to SWL. It was not a signiﬁcant predictor of NA. Table 3. Hierarchical Regression Analysis Predicting Satisfaction With Life, Positive Aﬀect, and Negative Aﬀect, with Faith Maturity, and Attendance at Religious Services after Controlling for Demographic Variables and Personality Criterion Variable

Demos R2

Personality DR2

FMSV DR2

Attendance at Rel Services DR2

Satisfaction with life a Positive Aﬀectb Negative Aﬀecta

.03 .02 .03

.19*** .23*** .36***

.03** .02* .00

.00 .01 .00

Model 2. Attendance at religious services as a predictor of well-being Table 4 indicates that attendance at religious services contributed 2.0% unique variance to SWL but was not a signiﬁcant predictor of PA or NA. These results suggest that regular attendance at religious services contributes to cognitive well-being, but does not have a signiﬁcant impact on positive or negative emotional states. Table 4. Hierarchical Regression Analysis Predicting Satisfaction With Life, Positive Aﬀect, and Negative Aﬀect, with Attendance at Religious Services, and Faith Maturity Vertical after Controlling for Demographic Variables and Personality Criterion Variable

Demos R2

Satisfaction with lifea .03 Positive Aﬀectb .02 Negative Aﬀecta .03

Personality DR2

Attendance at Rel Services DR2

FMSV DR2

.19*** .30*** .36***

.02* .00 .00

.02* .02* .00

Note. Demographic variables are ﬁrst step in each analysis, and personality is second step. a N = 195 b N = 194 * p < .05, ** p < .01, *** p < .001

178

geary, ciarrocchi and scheers

Models 3 and 4. Exploratory hypotheses for spirituality and attendance at religious services To explore the predictability of spirituality and attendance at religious services relative to each other, two additional regression model analyses were conducted. In the ﬁrst (Table 3, Model 3) “attendance at religious services” was entered after “spirituality.” It adds nothing additional in predicting well-being. In the second (Table 4), “spirituality,” when entered after “attendance at religious services,” predicted to SWL and PA, but not to NA. Thus “attendance at religious services” predicts some aspects of well-being, but not over and above “spirituality.” “Spirituality,” however, does predict aspects of wellbeing over and above “attendance at religious services.” This suggests that “attendance at religious services” is mediated by “spirituality.” Table 5. Beta Weights for Demographic Variables, Personality Factors, and Faith Maturity predicting to Well-being variables Variable Demographic Age Education Relationship Status Current Income Personality Neuroticism Extraversion Openness Agreeableness Conscientiousness FMS Vertical

SWLac

NAac

PAbc

–.05 –.10 .91 .74

–.05 .61 2.92* –.61

–.08 1.47 –.54 –.70

–.18*** .06 .07 –.04 .02 .10**

.51*** .05 –.04 –.04 .01 –.02

–.10* .24*** .04 –.16** .19*** .08*

a

N = 195 N = 194 c Satisfaction with life, SWL, positive aﬀect, PA, negative aﬀect, NA. * p < .05, ** p < .01, *** p < .001 b

Discussion Among the demographic variables, only relationship status contributed signiﬁcantly to predicting well-being and, in this case, positively to negative aﬀect. In other words, sex oﬀenders in outpatient treatment programs who are in a relationship (most of whom (29.2%) were

spirituality, well-being, and sex offenders

179

married, as opposed to living in a committed relationship (8.2%)) appear to experience more negative aﬀect than those who are not in a relationship. This is an opposite ﬁnding to the usual relationship between marital-relationship status and well-being in the general population. In this clinical sample, the nature of the men’s problems may have aﬀected their marital/relationship satisfaction, and thus, indirectly, led to an increase in their levels of negative aﬀect. Regarding the other demographic variables, it is conceivable that the nature of their oﬀense, and the powerful inﬂuence of their socially unacceptable sexual desires and behavior, are such powerful factors that, apart from the important inﬂuence of personality, demographics have little impact. This makes the signiﬁcant contribution of spirituality and attendance at religious services in this clinical population all the more intriguing. Over and above these demographic variables, personality added to all three components of well-being. This conﬁrms the crucial importance of controlling for personality in well-being studies, and in investigations of religion and spirituality. Spirituality contributed positively to cognitive well-being and positive aﬀect over demographics and personality. Attendance at religious services predicted to cognitive well-being, but was not a signiﬁcant predictor of positive aﬀect or negative aﬀect. This suggests that regular attendance at religious services may contribute to a global sense of happiness with life in a sample of sex oﬀenders, while spirituality may contribute to both emotional and cognitive well-being. Two exploratory sets of hypotheses were examined. The ﬁrst related to the question of the possible contribution of attendance at religious services to well-being after accounting for the eﬀects of spirituality. For no aspect of well-being was this supported. The second related to the possible contribution of spirituality to well-being after accounting for the eﬀects of attendance at religious services. Spirituality contributed added explanatory variance over and above attendance at religious services and other variables when predicting to cognitive well-being and positive aﬀect, but not negative aﬀect. We can conclude that the hypotheses of this study received partial support. Spirituality predicted to the positive components of well-being, and attendance at religious services predicted independently to cognitive, but not emotional components of well-being. Spirituality was a more powerful predictor than attendance at religious services.

180

geary, ciarrocchi and scheers

The primary goal of this study was to investigate the contribution of spirituality to well-being in sex oﬀenders. Some treatment providers argue for integrated (Freeman-Longo, 2001) or blended (Ellerby et al., 2000) models. The results of this research lend empirical support to those who argue in favor of including a spiritual component in treatment (Miller, 1999). It also supports treatment providers and researchers who have seen the beneﬁcial results of attendance at religious services or personal spirituality in the lives of men in treatment. These results have implications for broader questions as well as for the psychology of religion. In this research, spirituality and attendance at religious services both contributed to SWL. Ellison, Gay, and Glass (1989, p. 107) wrote that measures that reﬂect cognitive dimensions of well-being “may be more sociologically interesting and meaningful because they have as their referents long-term states of being.” They found that both frequency of attendance and devotional intensity predicted cognitive well-being over and above other variables. Walsh (2001), in his study of pathological gamblers, found that spirituality, as measured by the Self-Transcendence Scale, contributed an additional 5% of the variance to Life Satisfaction. GroganHenderson (2002), in her study of breast cancer patients, found that the FMS vertical scale contributed 2.6% added variance over and above other predictors, which included dispositional optimism. These ﬁndings are robust in their consistency of replication. Attendance at religious services also contributed to SWL, although not as strongly as spirituality. Regarding frequency of attendance, Ellison, Gay, and Glass (1989, p. 114) wrote, “its eﬀect is positive and signiﬁcant in each case.” This is consistent with the conclusions reached by George, Ellison, and Larson (2002) regarding the beneﬁcial eﬀects of religious participation on psychosocial resources such as self-esteem, self-eﬃcacy, and mastery, which are related to satisfaction with life. Ellison (1991) reported that religious variables accounted for between 5% to 7% of the variance in well-being, whereas only 2% to 3% was related to aﬀective well-being. While both studies found higher variance accounted for in cognitive well-being than in aﬀective wellbeing, the results of the present study are not as strong as Ellison’s ﬁndings. However, it should be noted that Ellison did not control for the eﬀects of personality, and, as the present study indicates, this represents a major confound in religious research. The eﬀects of the religious variables are not as strong when, as in this study, control-

spirituality, well-being, and sex offenders

181

ling for the eﬀects of personality. Ellison concluded that religion has more inﬂuence on the cognitive dimension of well-being. This is also consistent with the results of the present study. To understand religion’s association with well-being, it is useful to see its relationship to meaning. Baumeister (1991) writes that meaning is a cognitive function, and that it can be used to regulate the self, including emotional states. He suggests (1991, p. 183), “religion is a powerful aid to regulating one’s emotional states, as in coping with misfortune and retaining faith in higher values.” Discovery of the oﬀense and involvement in the criminal justice system can result in considerable stress for sex oﬀenders and may lead to signiﬁcant negative feelings in the initial stages of treatment. If Baumeister is correct, then religion has the potential to alleviate the long-term negative impact of these events by providing a new meaning framework, which can lead to emotional self-regulation. This is particularly important for this population since dysphoric aﬀective states are often the trigger for abusive behavior. The hypotheses regarding negative aﬀect were not supported. This is consistent with Ellison’s (1991) thesis that religion may do little to decrease negative aﬀect. Park noted (2002, p. 14) that, “contrary to prediction, religion was not signiﬁcantly related to depressed mood.” This is supported by Walsh (2001), and Grogan-Henderson (2002), who did not ﬁnd any signiﬁcant relationship between spirituality and aﬀect balance. Ciarrocchi and Deneke (2004, p. 226) write, “the only relationship religion has with well-being is with positive components, and no relationship whatsoever with negative emotion.” On the other hand, a recent meta-analysis (Smith, McCullough, & Poll, 2003), found a small, but constant relationship between religiousness and depression. Two points about the review are noteworthy. First, none of the 200 plus studies cited controlled for personality (particularly neuroticism), and second, in studies that examined moderating variables, stress level was found signiﬁcant. This may mean that, although spirituality may not predict negative aﬀect over personality in general, it may have an impact for highly stressed groups. However, even this suggestion does not fully explain the present study’s lack of ﬁndings for negative aﬀect. One would presume that sex oﬀenders with considerable legal problems qualify as a highly stressed group. Only future work that controls for personality can determine more precisely the relationship between spirituality and religiousness with negative aﬀect.

182

geary, ciarrocchi and scheers

Spirituality, but not attendance at religious services, contributed to positive aﬀect in this sample of sex oﬀenders. The associations were weaker than with SWL, but they were signiﬁcant, after accounting for personality and the demographic variables. Maton and Pargament (1987, p. 198) suggest that religion can oﬀer ultimately hopeful beliefs and perspectives on life and life’s crises. They also suggest that a distinctive attribute of a religious world-view is that it can oﬀer both internal and external loci of control in individuals (1987, p. 198). They write (1987, p. 173): “A commitment to an ‘upright’ life style may have a direct preventative impact on members (of congregations) who are lacking in impulse control or enmeshed in a maladaptive subculture.” This is an important observation for a group that clearly has problems in the area of impulse control. Ciarrocchi, Piedmont, & Williams (2003) used an incremental validity model to measure the unique contribution of spirituality, as measured by the Faith Maturity Scale, to prosocial behavior, over and above personality. They found that personality was the major predictor for women, but that faith maturity was the major predictor for men, suggesting that among males a close relationship with God was a stronger predictor of prosocial behavior than personality. They report that the Faith Maturity Vertical Scale contributed 17% unique explanatory variance to prosocial behavior after controlling for the eﬀects of personality ( b = .43, p < .001). They write (2003, p. 72), “spirituality may play a critical role as a motivational force in prosocial behavior that personality alone does not guide.” Since the majority of sex oﬀenders are men, this could be an important ﬁnding. Emotions are related to how people construe personal meaning. Meaning, therefore, is crucial in the experience and interpretation of emotion. Fredrickson (2000; 2002) refers to Victor Frankl’s (1959) insight into the importance of meaning in people’s lives. Fredrickson (2002, p. 211) writes: “Finding positive meaning is the most reliable path to cultivating positive emotions.” Those who hold religious or spiritual beliefs are perhaps more likely to interpret both positive and negative experiences in a positive, or at least meaningful, light. She argues that people ﬁnd meaning by such routes as the cognitive process of reframing unpleasant experiences in a positive way, the pursuit and attainment of realistic goals, and ﬁnding positive meaning in the ordinary events of life (2002). These have been demonstrated to lead to well-being.

spirituality, well-being, and sex offenders

183

This research has suggested that spirituality and religion may contribute to increasing positive aﬀect and subjective well-being in sex oﬀenders. This makes sense in the light of Fredrickson’s model. It may be that the combination of the meaning-structure that religion oﬀers, with the social support, moral values, and prescribed ethical behaviors can, when supported by the positive experience of therapy, contribute to ongoing well-being in sex oﬀenders. Focusing on what is to be avoided (negative goals) is less eﬀective than oﬀering a vision of what can be achieved (Emmons, 1999). Ward’s (2002; Ward & Stewart, 2003a; 2003b) proposal that treatment providers promote whatever contributes to the well-being of oﬀenders resonates with those who propose a more holistic approach to treatment. A combination of being supportive of sex oﬀenders who are involved in religious practice and spirituality, and the promotion of well-being through therapeutic involvement, apparently oﬀers a way of supporting self-regulation and enhancing the motivation of oﬀenders to change their behavior and lifestyles. Conclusion Strengths The sample size for this study (N = 195) was relatively large. It made it possible to enter demographic, personality, and religious variables in the regressions without sacriﬁcing statistical power. The sample had a high response rate and participants came from a major urban area, suburban, and rural areas. The high response rate was no doubt due to the generosity of treatment providers in allowing the majority of respondents to complete the surveys during their group time. Slightly over half (18) of the 35 refusals came from one treatment center where the men were asked to complete the survey at home and return it later to the group therapist. There were no refusals in 11 groups, one refusal in 8 groups, and three refusals in 3 groups. Four men who refused indicated that they did not want to sign the informed consent form. Others may have refused for other reasons. While it is possible that the scores of those who refused may have biased the results, the low rate of refusals overall enables us to be reasonably satisﬁed that the results represent a cross section of male sex oﬀenders in outpatient treatment. The use of an incremental validity model allowed for a stringent empirical test of the contribution of spirituality and attendance at

184

geary, ciarrocchi and scheers

religious services to well-being in a sample of sex oﬀenders. The research made use of instruments that have good psychometric properties, and the results can be compared with other research ﬁndings in the areas of religion, well-being, personality, and sex oﬀender treatment. Limitations This study used a convenience sample that was mostly Christian and Caucasian and from only one area. It represents a cross-section of only one area of the United States. Attitudes and religious practices can change over time, and caution is required before generalizing these results to other parts of the United States or to other countries. A number of men in treatment use antiandrogenic and/or other psychotropic medications. The eﬀects of these medications on wellbeing were not measured in this study. The information on diagnosis and sexual oﬀense/behavior lacked the speciﬁcity necessary to make useful group comparisons. Future research should focus more clearly on the diﬀerent types of oﬀense behaviors and how these relate to personality, spirituality, and well-being. The study also failed to enquire about the respondents’ sexual history, sexual orientation, or history of sexual, emotional, and/or physical abuse. These variables could have signiﬁcant implications for personality, spirituality, and well-being. They were omitted from the study in order to make the survey manageable, and to prevent it being experienced as overly intrusive. It would have been helpful to have included a question about the religion in which the participants were raised as well as current religious preference, to see if some participants have chosen to change religious allegiance. Also, it was not possible to evaluate the eﬀects of discovery, arrest, and treatment on religious involvement. This could also be important and signiﬁcant in attempting to measure the relationship between spirituality and well-being in a sample of sex oﬀenders. Directions for Further Study There were signiﬁcant similarities between this study and others in the area of positive psychology, personality traits, spirituality, and well-being. Researchers in the area of religion, health, and personality may want to explore the sex oﬀender’s image of God and how

spirituality, well-being, and sex offenders

185

this relates to a range of religious and psychosocial outcomes (Ciarrocchi, Piedmont, & Williams, 2002). It was not possible to explore the kind of religion that these men favor, nor the kind of theological template they use in their understanding of themselves, their paraphilic disorder nor oﬀensive behavior, and the relationship of spiritual themes such as forgiveness, grace, reconciliation, hope, and gratitude operating in their lives and attitudes. The sample used for this study comprised men between 18 and 75 years of age in outpatient treatment programs that did not have a spiritual component as part of the treatment protocol. This study calls for comparisons with treatment programs that intentionally include spirituality as a focus of treatment. It also invites comparisons with incarcerated sex oﬀenders, non-oﬀenders in sexual addiction programs, and oﬀenders in programs in other parts of the United States and in other countries, particularly in countries and social environments where attendance at religious services is not as high as in the United States. These results, at a minimum, provide support for a deeper examination of the role spirituality and religious practices can play in the treatment and natural environments of sex oﬀenders. References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington DC: Author. Baumeister, R. (1991). Meanings of life. New York: Guilford. Benson, P. L., Donohue, M. J., & Erickson, J. A. (1993). The Faith Maturity Scale: Conceptualization, measurement, and empirical validation. Research in the Social Scientiﬁc Study of Religion, 5, 1–26. Carnes, P. (1991). Don’t call it love. New York: Bantam. Carroll, J. F. X., McGinley, J. J., & Mack, S. E. (2000). Exploring the expressed spiritual needs of drug-dependent males in modiﬁed, therapeutic community treatment. Alcoholism Treatment Quarterly, 18, (4) 79–92. Ciarrocchi, J., & Deneke, E. (2004). Happiness and the varieties of religious experience: Religious support, practices, and spirituality as predictors of well-being. Research in the Social Scientific Study of Religion, 15, 204–233. Ciarrocchi, J. W., Piedmont, R. L., & Williams, J. E. G. (2002). Image of God and personality as predictors of spirituality in men and women. Research in the Social Scientiﬁc Study of Religion, 13, 55–73. —— (2003). Love thy neighbor: Spirituality and personality as predictors of prosocial behavior in men and woman. Research in the Social Scientiﬁc Study of Religion, 14, 61–75. Costa, P. T., Jr. & McCrae, R. R. (1992). Revised NEO Personality Inventory and NEO Five Factor Inventory professional manual. Odessa, Florida: Psychological Assessment Resources.

186

geary, ciarrocchi and scheers

DeNeve, K. M., & Cooper, H. (1998) The happy personality: A meta-analysis of 137 personality traits and subjective well-being. Psychological Bulletin, 124, 2, 127–229. Diener, E. (2000). Subjective well-being: The science of happiness and a proposal for a national index. American Psychologist, 55, 34–43. Diener, E., Emmons, R. A., Larsen, R. L., & Griﬃn, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49, 71–75. Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125 (2), 276–302. Earle, R. H., Dillon, D., & Jecmen, D. (1998). Systemic approach to the treatment of sex oﬀenders. Sexual Addiction and Compulsivity, 5, 49–61. Ellerby, L., Bedard, J., & Chartrand, S. (2000). Holism, wellness, and spirituality: Moving from relapse prevention to healing. In D. R. Laws, S. M. Hudson & T. Ward (Eds.), Remaking relapse prevention with sex oﬀenders (pp. 427–452). Thousand Oaks, CA: Sage. Ellerby, L., & Stonechild, J. (1998). Blending the traditional with the contemporary in the treatment of aboriginal sexual oﬀenders: a Canadian experience. In W. L. Marshall, Y. M. Fernandez, S. M. Hudson, & Tony Ward (Eds.), Sourcebook of treatment programs for sexual oﬀenders (pp. 399–415). New York: Plenum. Ellison, C. G. (1991). Religious involvement and subjective well-being. Journal of Health and Social Behavior, 32, 80–99. Ellison, C. G., Gay, D. A., & Glass, T. A. (1989). Does religious commitment contribute to individual life satisfaction? Social Forces, 68, 1, 100 123. Emmons, R. A. (1999). The psychology of ultimate concerns. New York: Guilford. Fetzer Institute. (1999). Brief multidimensional measurement of religiousness/spirituality: 1999. Multidimensional measurement of religiousness/spirituality for use in health research: A report of the Fetzer Institute /National Institute on Aging Working Group. (85–88). Kalamazoo, MI: Author. Frankl, V. E. (1959). Man’s search for meaning. Boston: Beacon Press. Fredrickson, B. L. (2000). Cultivating positive emotions to optimize health and wellbeing. Prevention and Treatment, 3, Retrieved January 22nd, 2003, from the World Wide Web: http://www.journals.apa.org/prevention/volume3/toc-mar07–00.html. Fredrickson, B. L. (2002). How does religion beneﬁt health and well-being? Are positive emotions active ingredients? Psychological Inquiry, 13(3), 209–213. Freeman-Longo, R. (2001). Paths to wellness. Holyoke, MA: Neari Press. Gartner, J., Larson, D. B., & Allen, G. A. (1991). Religious commitment and mental health. Journal of Psychology and Theology, 19, 6–25. George, L. K., Ellison, C. G., & Larson, D. B. (2002). Explaining the relationships between religious involvement and health. Psychological Inquiry, 13, 190–200. Grogan-Henderson, K. (2002). Optimism and spirituality: Predictors of subjective well-being in breast cancer patients. Unpublished doctoral dissertation, Loyola College in Maryland. Hanson, R. K. (2000). Risk assessment. Oregon: ATSA. —— (2002). Recidivism and age: Follow-up data from 4,673 sexual oﬀenders. Journal of Interpersonal Violence, 17, 1046–1062. Koenig, H. G., McCullough, M. E., & Larson, D. B. (Eds.). (2001). Handbook of religion and health. New York: Oxford University. Maton, K. I., & Pargament, K. I. (1987). The roles of religion in prevention and promotion. Prevention in Human Services, 5, 161–205. Mickley, J. R., Carson, V., & Soeken, K. L. (1995). Religion and adult mental health: State of the science in nursing. Issues in Mental Health Nursing, 16, 345–360. Miller, W. R. (Ed.). (1999). Integrating spirituality into treatment. Washington DC: APA. Myers, D. G. & Diener, E. (1995). Who is happy? Psychological Science, 6, 10–19. Park, C. L. (2002, March 12t.). Religion as a meaning-making framework in coping with life stress. Paper presented at 7th Annual Roundtable on Religious Research, Columbia, MD.

spirituality, well-being, and sex offenders

187

Pavot, W. & Diener, E. (1993). Review of the satisfaction with life scale. Psychological Assessment, 5, 164–172. Piedmont, R. L. (in press). Spiritual transcendence as a predictor of psychological outcome from an outpatient substance abuse program. Psychology of Addictive Behaviors. Piedmont, R. L., & Nelson, R. (2001). A psychometric evaluation of the short form of the Faith Maturity Scale. Research in the Social Scientiﬁc Study of Religion, 12, 165–183. Sardjian, A., & Nobus, D. (2003). Cognitive distortions of religious professionals who sexually abuse children. Journal of Interpersonal Violence, 18, 905–923. Seybold, K. S., & Hill, P. C. (2001). The role of religion and spirituality in mental and physical health. Current Directions in Psychological Science, 10, 21–25. Smith, T. B., McCullough, M. E., & Poll, J. (2003). Religiousness and depression: Evidence for a main eﬀect and the moderating inﬂuence of stressful life events. Psychological Bulletin, 129, 614–636. Steel, P., & Ones, D. S. (2002). Personality and happiness: A national-level analysis. Journal of Personality and Social Psychology, 83, 767–781. Walsh, J. M. (2001). Spirituality and recovery from pathological gambling. Loyola College in Maryland. Unpublished doctoral dissertation. Ward, T. (2002). Good lives and the rehabilitation of oﬀenders: Promises and problems. Aggression and Violent Behavior, 7, 513–528. Ward, T., & Stewart, C. (2003). Criminogenic needs and human needs: A theoretical model. Psychology, Crime, & Law, 9, (2), 125–143. —— (2003). The treatment of sex oﬀenders: Risk management and good lives. Professional Psychology: Research and Practice, 34, (4), 353–360. Watson, D. (2000). Mood and temperament. New York: Guilford. Watson, D., Clark, L., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative aﬀect: The PANAS Scales. Journal of Personality and Social Psychology, 54, 1063–1070.

MAINLINE EVANGELICAL RENEWAL MOVEMENTS: A PRELIMINARY INQUIRY Sean F. Everton* Abstract In recent years, evangelical renewal movements have arisen within mainline Protestant denominations. While in the past, such sectarian movements have often split with their parent denominations, these groups express no such desire, but instead are adamant that they want to remain within their denominations and return them to the faith of their founders. Although it is clear that evangelical renewal movements draw their support from theological conservatives, it is unclear whether they come from mainline Protestant backgrounds or from another religious tradition such as evangelical Protestantism. Human capital theory suggests a higher probability that conservative mainline Protestants come from mainline Protestant backgrounds than from other religious backgrounds. Analysis of the 1996–1998 General Social Surveys provides support for this prediction. After discussing the results, a number of directions for future research are identiﬁed.

In recent years, evangelical renewal movements (ERMs) have arisen in mainline Protestant denominations. Most are well organized, growing rapidly, and exert considerable inﬂuence at the denominational level (Howell, 2003; McKinney & Finke, 2002). While many of these movements have come together under umbrella groups such as The Institute on Religion and Democracy and The Association for Church Renewal, and they all express extreme distaste with the direction that their respective denominations have taken, they show little interest in leaving and forming their own denominations. Most claim that they are committed to remaining within and renewing their own particular denominations. The American Anglican Council (1996), for example, considers itself “a network of individuals, parishes, and specialized ministries who aﬃrm biblical authority and Anglican orthodoxy within the Episcopal Church. In response to the Lord’s

* Author Note: I wish to thank Nancy Tuma, Larry Iannaccone for helpful comments on earlier drafts of this paper. They should not be held responsible for any of my conclusions. Correspondence concerning this article should be addressed to Sean F. Everton, Department of Sociology, Stanford University, Stanford, CA 94305, [email protected]. Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

190

sean f. everton

grace, we have committed ourselves to proclaiming the Good News to every person and working to reform and renew our Church” (italics added). Likewise, the United Methodist Confessing Movement states that “we are a movement within the United Methodist Church, inseparable from it, not apart from it. All of us intend to stay within it” (Confessing Movement, 2003). Currently, over 30 ERMs are active forces within mainline Protestant denominations; the major ones are summarized in Table 1. Table 1. Major Evangelical Renewal Movements within Mainline Protestant Denominations American Baptist Churches USA American Baptist Evangelicals Christian Church (Disciples of Christ) Disciples Renewal Episcopal Church American Anglican Council Episcopalians United Evangelical Lutheran Church of America Evangelical Lutheran Confessing Fellowship Word Alone Presbyterian Church (U. S. A.) Presbyterian Layman Presbyterians for Faith, Family & Ministry Presbyterians for Renewal Presbyterians Pro-Life United Church of Christ Biblical Witness Fellowship The Renewal Fellowship United Methodist Church The Confessing Movement Good News Lifewatch Renew Network Transforming Congregations

The rapid rise and increasing popularity of ERM’s raises the question as to who supports them. Some have suggested that they draw their primary support from individuals who grew up in religious traditions other than mainline Protestantism (e.g., Hinson, 1993). Others

mainline evangelical renewal movements

191

believe that clergy and other denominational authorities have done a poor job at educating members regarding the traditions of their faith (Howell, 2003). Currently, we do not have access to data that permits us to directly answer this question since there are no publicly available surveys of mainline ERMs. We can, however, indirectly assess it with available data by comparing the probability that mainline Protestant conservatives (MPCs) come from mainline Protestant backgrounds with the probability that they come from some other religious tradition (e.g., evangelical, Roman Catholic, etc.). This study draws on human capital theory, especially as it has been applied to religious behavior (Iannaccone, 1990; Myers, 2000). After a brief review of human (and religious) capital theory, the four hypotheses are derived and are then tested using multinomial logistic modeling analysis, the results of which are then analyzed and discussed. Several promising avenues for future research are identiﬁed in the conclusion. Iannaccone (1990, 1995) derives the concept of religious production and religious capital from Becker’s (1975, 1976) work on human production and human capital. According to human capital theory, people’s skills and knowledge play signiﬁcant roles in determining the quantity and quality of the goods and services they produce. Over time, people build up stocks of skills and knowledge (i.e., human capital) through education, experience, and specialized training. The more people invest in building up these stocks, the better returns they receive from their productive activities. Thus, people who invest time and money in obtaining a college degree will generally earn higher salaries than will those who do not. Similarly, people who spend more time on job-related activities will earn higher salaries than those who do not because the skills they acquire add to their stock of human capital. Becker and his students have applied human capital theory to what they call household production, by which they mean the various “commodities” that families produce, such as leisure, children, home-cooked meals, and aﬀection. While families do not produce these commodities for resale on the market, they do produce them by drawing on scarce resources such as purchased goods, household labor, and stocks of skills and knowledge. As such, how they produce these goods is seen as an economic decision that can be analyzed using economic models (Ierulli, Glaeser, & Tommasi, 1995). Thus, just as individuals can build up skills related to their jobs, they

192

sean f. everton

can also acquire skills related to the production of household commodities. While people generally acquire these skills through “on the job” training, they also acquire them through specialized training events such as cooking, parenting, and gardening classes. And the more that families invest in acquiring these skills, the better equipped they are to produce and experience familial bliss. Iannaccone (1990) argues that religious practice is one such productive process. “Purchased goods, household time, and human capital aﬀect a family’s capacity to ‘produce’ religious satisfaction just as they aﬀect its capacity to produce meals, health, oﬀspring, and recreational enjoyment” (Iannaccone, 1990, p. 299). In other words, the more people invest in acquiring skills and experience speciﬁc to their religious tradition (i.e., religious capital), the more religious satisfaction they are able to produce and experience. Add to this the fact that it is next to impossible to enjoy and derive meaning from religious services apart from any knowledge of and familiarity with the doctrines, rituals, and customs of their religious traditions, and it becomes clear that the quality of fellowship a congregation experiences depends largely on how much time and energy its members invest in their faith tradition and their relationships with one another. Furthermore, research indicates that the religious capital people acquire as children plays a strong role in determining their religious practices as adults. When people receive religious training as children, they tend to get it from their parents and the churches, synagogues, or temples their parents support. Thus, as they “mature and decide for themselves what beliefs they will accept and what church they will attend, they naturally gravitate to those of their parents” (Iannaccone, 1990:300). Even if they do switch religions or denominations, they will generally switch to traditions similar to the ones in which they were raised (Kluegel, 1980). Thus, it seems likely that mainline Protestant conservatives come from mainline Protestant backgrounds rather than from evangelical Protestant (or any other religious or nonreligious) backgrounds. This study will examine four hypotheses. They are: Hypothesis 1: Individuals from mainline Protestant backgrounds will have a higher probability than those from evangelical Protestant backgrounds of identifying with conservative mainline Protestantism than with evangelical Protestantism.

mainline evangelical renewal movements

193

Hypothesis 2: Individuals from mainline Protestant backgrounds will have a higher probability than those from Black Protestant backgrounds of identifying with conservative mainline Protestantism than with evangelical Protestantism. Hypothesis 3: Individuals from mainline Protestant backgrounds will have a higher probability than those from Roman Catholic backgrounds of identifying with conservative mainline Protestantism than with evangelical Protestantism. Hypothesis 4: Individuals from mainline Protestant backgrounds will have a higher probability than those from non-Christian backgrounds of identifying with conservative mainline Protestantism than with evangelical Protestantism. Method Sample To test these hypotheses, I analyze the National Opinion Research Center’s (NORC) 1996–1998 General Social Surveys (Davis & Smith, 1998). The General Social Surveys (GSS) are cross-sectional surveys of a sample of non-institutionalized individuals 18 years and older, residing in the United States. I use the 1996–1998 GSS because in 1996 the NORC began asking Protestant respondents whether they identiﬁed themselves as a fundamentalist, evangelical, mainline, or liberal Protestants. As I discuss in more detail later, I draw on this question to construct the polytomous dependent variable used in this analysis. To complicate matters somewhat, in 1994 the NORC switched to a biennial, dual-sample format, which means that each GSS consists of essentially two separate surveys (ballots A and B). Under this format, a large number of the questions appear on both ballots, but some appear on only one. This occurred with respect to the Protestant self-identiﬁcation question noted above. In 1996 it appeared on both ballots, but in 1998 it appeared only on ballot A. Consequently, I created a subset of the 1996–1998 GSSs by dropping ballot B from the dataset before analyzing the data. Because the two ballots are assigned randomly, this is not problematic.

194

sean f. everton Measures

Dependent Variable Because the GSS does not contain a variable that distinguishes conservative mainline Protestants from other mainline Protestants, I construct one through a series of steps by combining various objective and subjective measures of denominational aﬃliation and theological orientation. Using the series of questions employed by the GSS to sort respondents into 1 of over 150 possible denominational aﬃliations (Davis, Smith, & Marsden, 1999), I ﬁrst grouped respondents into the seven broad religious traditions—mainline Protestant, evangelical Protestant, Black Protestant, Roman Catholic, Jewish, Other and Nonreligious—identiﬁed by Steensland, et al. (2000). Next, I used the following two questions from the GSS to separate conservative mainline Protestants from other mainline Protestants: (1) Which of these statements comes closest to describing your feelings about the Bible? (a) The Bible is the actual word of God and is to be taken literally, word for word. (b) The Bible is the inspired word of God but not everything in it should be taken literally, word for word. (c) The Bible is an ancient book of fables, legends, history, and moral precepts recorded by men. (2) When it comes to your religious identity, would you say you are a fundamentalist, evangelical, mainline, or liberal Protestant or do none of these describe you? Respondents who said they belonged to a mainline Protestant denomination and identiﬁed themselves as biblical literalists, fundamentalists or evangelicals1 were coded as conservative mainline Protestants. Respondents who said they belonged to a mainline Protestant denomination but did not identify themselves as biblical literalists, fundamentalists or evangelicals were coded as other mainline Protestants. Third, I grouped respondents who identiﬁed themselves as Jewish, members of other faiths, or non-aﬃliated into a single category,

1 I identiﬁed biblical literalists as those, who, when asked to describe their feelings about the Bible, answered that “it is the actual word of God and is to be taken literally, word for word.” I identiﬁed fundamentalist and evangelicals as those who, when asked their religious identity, said they considered themselves to be fundamentalist or evangelicals.

mainline evangelical renewal movements

195

“Non-Christian.” Finally, I constructed a polytomous variable with the following categories: (a) mainline Protestant Conservative, (b) mainline Protestant Other, (c) evangelical Protestant, (d) Black Protestant, (e) Roman Catholic, and (f ) Non-Christian. Table 2 summarizes the construction of this variable. Table 2. Summary of Classiﬁcation Scheme for Dependent Variable Original Sevenfold Denominational GroupingSteensland et al. (2000) Mainline Protestant Mainline Protestant Evangelical Protestant Black Protestant Roman Catholic Jewish Other Faith Unaﬃliated

Biblical literalist, self-identiﬁedfundamentalist or evangelical?

Final Polytomous Dependent Variable

Yes No Either Either N/A N/A N/A N/A

Conservative Mainline Other Mainline Evangelical Protestant Black Protestant Roman Catholic Non-Christian Non-Christian Non-Christian

Independent Variables and Controls Because I argue that respondents’ background religious aﬃliation signiﬁcantly predicts their choice of faith tradition (if any) as adults, I created a series of dummy variables identifying respondents’ background religious aﬃliation at the age of 16 years. Using Steensland et al.’s (2000) classiﬁcation scheme, I ﬁrst grouped respondents by the religious tradition in which they were raised (background religious aﬃliation), and then created dummy variables for respondents from mainline Protestant, Black Protestant, Roman Catholic, and non-Christian ( Jewish, other, and non-religious) backgrounds, leaving respondents with evangelical Protestant backgrounds as the reference category. I control for several demographic variables: respondents’ age (in years, ranging from 18 to 89), education (years completed, ranging from 0 to 20), gender (female), race (Black), region of residence (South, West, Midwest, with Northeast as reference category) and occupational prestige score (ordinal scale, ranging from 0 to 86). I use occupational prestige instead of income because the GSS’s income variable groups respondents who earn $25,000 or more into a single category (approximately 63%), making the measure’s distribution highly problematic. Furthermore, nearly 12% of all cases have missing information on

196

sean f. everton

income. After analyzing the income variable’s missing cases, I determined they were not missing randomly with respect to several other variables used in this analysis—in particular, age, education, gender, and region of residence (south and west). By contrast, the occupational prestige variable is more normally distributed than is the income variable. Also, only 4.6% of its cases are missing, and they are missing non-randomly with respect to age, gender, race, region (west), and Black Protestant backgrounds. Taken together, these factors led me to select occupational prestige level over income2 as a control variable. Procedure To test my hypothesis, I used multinomial logistic regression analysis, estimating the regression coeﬃcients using the maximum likelihood method. A check of the multinomial logistic regression model’s requirement of independence from irrelevant alternatives (Hausman & McFadden, 1984) revealed no violation of this assumption. When the dependent variable is nominal (i.e., when it has no natural ordering), as it is in this case, it is more appropriate to use the multinomial logit model (MNLM) rather than estimating a series of independent binary logit regression models since it enforces the logical relationship between the parameters and uses the data more eﬃciently (Long, 1997). It is helpful to think of the MNLM as an extension of the binary logit model that simultaneously estimates binary logits for all possible comparisons among outcome categories. In fact, analysts have demonstrated that estimates from binary logits provide consistent estimates of the MNLM’s parameters (Begg & Gray, 1984). The estimates from each binary logit model (BLM) would be based on diﬀerent samples, however, while the MNLM is capable of simultaneously estimating all of the logits on the entire sample.

2 I estimated two multinomial logit regression models that included income as a covariate (not shown). Before doing so, I ﬁrst recoded income so that it ranged from $500 to $60,000 (in thousands). The ﬁrst model, I estimated included only income as a covariate. The second included only both income and prestige. In the ﬁrst model income was a signiﬁcant predictor of choosing other mainline or Black Protestantism over evangelical Protestantism. In the second it was a signiﬁcant predictor of choosing Black Protestantism or Roman Catholicism over evangelical Protestantism.

49.70*** (17.03) 14.20*** (2.73) 46.83*** (14.51) .57 .04+++ .19 .33 .28 .20 .81+++ .11+++ .01+++ .05+++ .03+++ 537

53.82*** (17.06) 12.96* (2.68) 42.93 (13.17) .63+ .10

.18 .46++ .26 .10+++

.70+++ .16++ .02+++ .09+++ .04+++ 200

Other Mainline Protestants

.02+++ .01+++ .93+++ .03+++ .01+++ 341

.14++ .55+++ .24 .07+++

.08+++ .51+++ .25 .16+++ .10+++ .75+++ .03+++ .07+++ .05+++ 1,029

44.21 (15.91) 12.37*** (2.86) 38.33*** (13.30) .65+++ .95+++

Black Protestants

45.73 (16.51) 12.92*** (2.82) 42.51* (13.00) .56 .05+++

Evangelical Protestants

Source: General Social Surveys, 1996–1998. Note: 1Reference category + p < .05 ++ p < .01 +++ p < .001—Chi-square test for independence (two-tailed) * p < .05 ** p < .01 *** p < .001—T-test for diﬀerence between means (two-tailed)

Female Black Region of Residence Northeast1 South Midwest West Background Religious Aﬃliation Mainline Protestant Evangelical Protestant1 Black Protestant Roman Catholic Non-Christian Final N

Occupational Prestige

Education (in years)

Sociodemographic Variables Age (in years)

Conservative Mainline Protestants

.04+++ .02+++ .01+++ .91+++ .02+++ 1,000

.33+++ .21+++ .25 .21

44.09* (16.44) 13.41 (2.78) 43.15 (13.52) .52 .04

Roman Catholics

.15+++ .12+++ .03+++ .24+++ .46+++ 823

.22 .24+++ .21++ .33+++

40.78*** (14.95) 14.12*** (3.01) 44.24* (14.58) .46+++ .08+++

Non-Christian

.22 .25 .10 .31 .12 3,930

.20 .35 .25 .20

45.10 (16.56) 13.43 (2.90) 43.28 (13.86) .54 .13

Total

Table 3. Means and Standard Deviations for Variables Used in the Analysis of the Eﬀects of background Religious Aﬃliation on the Probability of Adult Religious identiﬁcation

mainline evangelical renewal movements 197

198

sean f. everton Results and Discussion

Table 3 displays the means and standard deviations for the independent variables used in the analysis, separated by the various categories of the dependent variable. For continuous variables included in the analysis, I used t-tests to compute for statistical signiﬁcance between the mean of a particular group and the rest of the sample population, whereas for categorical variables I conducted a chi-square test for independence between a particular group and the rest of the sample population. With respect to the sociodemographic variables, conservative mainline Protestants are, on average, signiﬁcantly older than the rest of the population, with a mean age of 53.82 years old as compared to the population mean of 45.10 years. Other mainline Protestants are the next oldest group with a mean age of 49.81 years old, one that diﬀers signiﬁcantly from the rest of the population. The mean age of evangelical and Black Protestants do not diﬀer signiﬁcantly from the rest of the population mean, while the mean age of Roman Catholics and non-Christians are signiﬁcantly younger than Protestant counterparts. Turning toward education levels, Table 3 indicates that conservative mainline Protestants are, on average, less educated than other mainline Protestants, having completed just 12.96 years of education as compared to an average of 13.43 years for the total sample. Their mean education level is on par with that of evangelical Protestants, somewhat more than that of Black Protestants, and less than that of other mainline Protestants, Roman Catholics, and nonChristians.3 The mean occupational prestige scores of conservative mainline Protestants and Roman Catholics are somewhat lower than that of the population as a whole but do not diﬀer signiﬁcantly from everyone else. The mean occupational prestige scores of other mainline Protestants and non-Christians are signiﬁcantly higher than others in the group, while those of evangelical Protestants and Black Protestants are signiﬁcantly lower. Table 3 also indicates that a higher

3 As one of the reviewers pointed out, the lower level of education and of occupational prestige of conservative mainline Protestants may simply reﬂect cohort eﬀects and not as causes of or results from religious identiﬁcation.

mainline evangelical renewal movements

199

percentage of females than males identiﬁes themselves as conservative mainline Protestants or Black Protestants, and a higher percentage of males than females identiﬁes themselves as non-Christians. A higher percentage of females than males also identiﬁes themselves as other mainline Protestants, evangelical Protestants and Roman Catholics, but the diﬀerence is not statistically signiﬁcant. Not surprisingly, African-Americans overwhelmingly identify with Black Protestant denominations. Certain patterns emerge when looking at respondents’ region of residence. Conservative mainline Protestants are signiﬁcantly more likely to live in the South and less likely to live in the West than the rest of the population taken as a whole. By contrast, the Midwest is the only region in the country where other mainline Protestants are signiﬁcantly more likely to live than are others. Evangelical Protestants are similar to conservative mainline Protestants in that they are signiﬁcantly more likely to found in the South and less likely to be found in the West than are respondents identifying with other groups. However, they diﬀer from conservative mainline Protestants in that they are signiﬁcantly less likely to live in the Northeast. Black Protestants are a signiﬁcant presence in the South, but not so in either the Northeast or West. Roman Catholics are signiﬁcantly more likely than are others to live in the Northeast, but signiﬁcantly less likely to live in the South. Finally, consistent with previous research, respondents from other or no faith traditions are signiﬁcantly more likely than others to live in the West and signiﬁcantly less likely to live in other parts of the country (Stark & Bainbridge, 1985, 1987; Wuthnow, 1976). Most striking is the high percentage (70%) of conservative mainline Protestants who come from mainline backgrounds. Compared to the 16% who come from evangelical backgrounds, the 2% from Black Protestant backgrounds, the 9% from Roman Catholic backgrounds, and the 4% from other or no religious backgrounds, these results provide preliminary support for the argument that conservative mainline Protestants are far more likely to come from mainline Protestant backgrounds than from non-mainline backgrounds. Similar patterns are found for the other groups as well. Other mainline Protestants are more likely to come from mainline backgrounds; evangelical Protestants are more likely to come from evangelical backgrounds; Black Protestants are more likely to come from Black Protestant backgrounds; Roman Catholics are more likely to come

200

sean f. everton

from Roman Catholic backgrounds; and non-Christians are more likely to come from non-Christian backgrounds. Table 4 presents the ﬁt statistics for the three multinomial logistic models used in the analysis. Model 1 includes only sociodemographic variables (except region of residence) while Model 2 includes the sociodemographic variables from the ﬁrst model plus the regional dummy variables. Model 3 includes all of the variables included in Model 3 plus the background religious aﬃliation variables. As the model chi-square statistics indicate, all three models are statistically signiﬁcant at the p < .001 level. Furthermore, Model 2’s ﬁt signiﬁcantly (p < .001) improves upon Model 1’s ﬁt, while Model 3’s signiﬁcantly (p < .001) improves upon Model 2’s. Thus, Model 3 is my preferred model of adult religious identiﬁcation. Table 4. Fit Statistics for Multinomial Logistic Regression Models for Adult Religious Identiﬁcation (N = 3,930) Fit Statistics MODEL Chi-Square

Improvement Chi-Square

Degrees of Freedom

Model: Description Model 1: • Sociodemographic Variables

1,753.54***

Model 2: • Sociodemographic Variables • Region of Residence

2,155.84***

402.30***

40

Model 3: • Sociodemographic Variables • Region of Residence • Background Religious Aﬃliation

6,123.93***

3,968.09***

60

Source: General Social Surveys, 1996–1998. *** p. <. 001 (two-tailed)

25

mainline evangelical renewal movements

201

Table 5 presents the maximum likelihood estimates of Model 3. Given the six-category dependent variable, there are ﬁve separate sets of coeﬃcients: the log-odds of choosing to identify oneself as a conservative mainline Protestant rather than as an evangelical Protestant (column 1); the log-odds of choosing to identify oneself as an other mainline Protestant rather than as an evangelical Protestant (column 2); the log-odds of choosing to identify oneself as a Black Protestant rather than as an evangelical Protestant (column 3); the log-odds of choosing to identify oneself as a Roman Catholic rather than as an evangelical Protestant (column 4); and the log-odds of choosing to identify oneself as a member of another faith or no faith at all rather than as an evangelical Protestant (column 5). Because the primary focus here is whether conservative mainline Protestants come from mainline Protestant backgrounds rather than evangelical Protestantism, my discussion of the results concentrates on the coeﬃcients presented in the ﬁrst column. Table 5. Maximum Likelihood Estimates of Preferred Multinomial Logistic Regression Model Predicting Adult Religious Identiﬁcation (N = 3,030). Conservative Other Mainline vs. Mainline vs. Evangelical Evangelical Sociodemographic Variables Age (in years) .02*** Education –.01 (in years) Occupational –.00 Prestige Female .17 Black .56 Region of –.50 Residence1 South Midwest –.42 West –1.21*** Background Religious Aﬃliation2 Mainline 3.41*** Protestant Black Protestant .30 Roman Catholic 1.86*** Non-Christian 1.23** Intercept –3.62 Model Statistics Model 6,123.93***

.01* .09**

Black Protestant vs. Evangelical

–.01 .07

.01

.–02

.00 –.13 –.87***

.01 3.13*** .18

–.51* –.74**

Roman Catholic vs. Evangelical

.01 .05 –.01

NonChristian vs. Evangelical

–.01** .11*** –.01

–.35* –.19 –1.13***

–.52*** –.08 –1.01***

–.54 –.22

–.49* –.75**

–.63** –.10

3.86***

2.07**

2.76***

2.04***

.63 1.44*** 1.29**8 –3.85

5.25*** 3.01*** 1.76* –5.38

1.49* 6.20*** 2.83**8 –3.52

2.07*** 2.84*** 3.76*** –1.88

sean f. everton

202 Table 5 (cont.)

Conservative Other Mainline vs. Mainline vs. Evangelical Evangelical Chi-Square –2 Log Likelihood Degrees of Freedom

Black Protestant vs. Evangelical

Roman Catholic vs. Evangelical

NonChristian vs. Evangelical

–3,470.34 60

Source: General Social Survey, 1996–1998 Notes: 1Reference category us Northeast 2 Reference category is Evangelical Protestant * p < .05 ** p < .01 *** p <. 001 (two-tailed)

Table 6. Maximum Likelihood Estimates of Multinomial Logistic Regression Models Predicting Choice Between Other and Conservative Mainline Protestants (N = 3,930). Mainline Other vs. Mainline Conservative

Sociodemographic Variables Age (in years) Education (in Years) Occupational Prestige Female Black Region of Residence1 South Midwest West Background Religious Aﬃliation2 Mainline Protestant Evangelical Protestant Black Protestant Roman Catholic Non-Christian Intercept Model Statistics Model Chi-Square –2 Log Likelihood Degrees of Freedom

Model 1

Model 2

Model 3

Model 4

Model 5

–.01*

–.01*

–.01*

–.01*

–.01*

.11** .01 .16 –.70* –.37

.11** .01 .16 –.70* –.37

.11** .01 .16 –.70* –.37

.11** .01 .16 –.70* –.37

.11** .01 .16 –.70* –.37

–.09 .48

–.09 .48

–.09 .48

–.09 .48

–.09 .48

.45

.11 –.33

–.45 –.11 –.88** –.40 .21

.33 –.42 .05 –.23

–.76 –.28 .10

6,123.93*** –3,470.34 60

Source: General Social Survey, 1996–1998 Notes: 1Reference regional category is Northeast 2 Reference category Model diﬀer only in terms of reference category * p < .05 ** p < .01 *** p < .001 (two-tailed)

.88** .43 .76 .47 –.66

.40 –.05 .28 –.48 –.18

mainline evangelical renewal movements

203

Of the socio-demographic variables, age is the only signiﬁcant predictor of whether respondents identify with conservative mainline Protestantism or evangelical Protestantism. Everything else being equal, the odds of choosing conservative mainline Protestantism rather than evangelical Protestantism are multiplied by 1.02 (e4) for each additional year. That is, the odds that a 50-year-old individual is a conservative mainline Protestant rather than an evangelical Protestant is 1.22 [(e.020) 10 = e.20 = 1.22] times that of a 40-year-old individual. This result is consistent with those presented in Table 3, which showed that conservative mainline Protestants are, on average, older than evangelical Protestants. Looking at the other columns, one can see that age and education are signiﬁcant predictors of whether someone will identify with either other mainline Protestantism or nonChristian traditions than with evangelicalism. Both are positively associated with identifying oneself with other mainline Protestantism than with evangelicalism, while age is negatively associated and education is positively associated with identifying oneself as a nonChristian rather than as an evangelical Protestant. These results dovetail those presented in Table 3, which clearly indicate that other mainline Protestants tend to be older and better educated than evangelical Protestants while non-Christians tend to be younger and better educated. Interestingly, everything else being equal, women are signiﬁcantly less likely than men are to choose Roman Catholicism over evangelical Protestantism. Looking back at Table 3 we can see that while more women than men identify themselves as Roman Catholics, the percentage of women who do so (52%) is less than the percentage of women in the population as a whole (54%). Finally, but not surprisingly, women are less likely to identify themselves as non-Christian than evangelical Protestantism. A respondent’s region of residence plays a role in what faith tradition with which they choose to identify. Everything else being equal, Southerners are signiﬁcantly less likely than Northeasterners to choose conservative mainline Protestantism, other mainline Protestantism, Roman Catholicism or other faith traditions over evangelical Protestantism. They are not, however, signiﬁcantly less likely than Northeasterners to choose Black Protestantism over evangelical Protestantism. Living in the Midwest and West produces similar but not identical eﬀects to those from living in the South. Midwesterners are signiﬁcantly less likely than Northeasterners to choose non-conservative mainline Protestantism, Roman Catholicism, or other faith

204

sean f. everton

traditions over evangelical Protestantism. They are, however, no more likely than Northeasterners to choose either conservative mainline or Black Protestantism over evangelical Protestantism. People living in the West are signiﬁcantly less likely than those living in the Northeast to choose conservative mainline Protestantism, other mainline Protestantism or Roman Catholicism over evangelical Protestantism. Westerners, however, are no less likely than Northeasterners to choose Black Protestantism or other faith traditions over evangelicalism. Most importantly, Table 5’s results provide support for this article’s underlying argument that those who identify with conservative mainline Protestantism are more likely to come from mainline Protestant backgrounds than from evangelical Protestantism. Respondents from mainline backgrounds are estimated to be 30.4 (e3.41) times more likely than those with evangelical backgrounds to identify with conservative mainline Protestantism than evangelical Protestantism. In fact, respondents from mainline backgrounds are more likely than are those from evangelical backgrounds to choose any of the ﬁve non-evangelical faith traditions over evangelical Protestantism.4 Similarly, respondents with Roman Catholic or non-Christian backgrounds are more likely than are those from evangelical backgrounds to identify with any of the non-evangelical traditions over evangelical Protestantism. Taken together, these results suggest that mainline ERMs draw the bulk of their support from long-time mainline Protestants who are simply trying to hold on to the religious traditions and beliefs of their childhood. According to McKinney and Finke (2002), a primary rallying point of these movements is the perception that mainline churches have gone “soft” on certain types of sexual behavior, such as premarital or homosexual sex.5 All insist that such behavior is inconsistent with the teachings of historic Christianity and mainline Protestantism, and they are correct. Historically, apart from for a few isolated exceptions (see e.g., Boswell, 1981; Helminiak, 1994;

4

Respondents from mainline backgrounds are 47.5 times (e3.86) more likely to choose non-conservative mainline Protestantism, 8 times (e2.07) more likely to choose Black Protestantism, 16 times (e2.76) more likely to choose Roman Catholicism, and 8 times (e2.04) more likely to choose non-Christian faith traditions over evangelical Protestantism. 5 As one reviewer noted, liberalizing attitudes toward abortion among mainline Protestants has been a source of controversy as well.

mainline evangelical renewal movements

205

Spong, 1994), Christianity has maintained a rather strict sexual ethic, and it has only been in recent years that certain sectors within mainline Protestantism have begun to relax it. Thus, it is reasonable to argue that mainline ERM supporters are not interlopers from other traditions, but long-time mainline Protestants, who are seeking to return their denominations back to the traditions and beliefs of a previous generation. All of this is not to argue that mainline Protestant conservatives are as theologically conservative as their parents. Chances are they are not. Studies have shown that over the last 30 years even theological conservatives have become increasingly liberal on most issues (Hoﬀman & Miller, 1997). They just have not moved as far to the left as have their denominations. Indeed, it is this increasing gap that has helped give rise to mainline ERMs in the ﬁrst place. It would be inappropriate to conclude, however, that individuals raised in mainline Protestantism are more likely to choose one form of mainline Protestantism over the other. Table 6 presents the maximum likelihood estimates of choosing other non-conservative forms of mainline Protestantism over conservative mainline Protestantism. The models diﬀer from one another only in terms of the omitted background religious aﬃliation category. In the ﬁrst column, mainline Protestant is the omitted category, in the second it is evangelical Protestant, in the third, Black Protestant, the fourth, Roman Catholic, and the ﬁfth, Non-Christian. What is clear is that education is the only consistent predictor of whether respondents will choose non-conservative (other) mainline Protestantism over conservative mainline Protestantism.6 More years of education increases the likelihood of a respondent identifying with other mainline Protestantism than conservative mainline Protestantism. Background religious aﬃliation is only a signiﬁcant predictor when comparing respondents from mainline Protestant to those from Roman Catholic backgrounds. Respondents from mainline backgrounds are more likely than are those from Roman Catholic backgrounds to choose non-conservative mainline Protestantism over conservative mainline Protestantism.

6 The reverse, of course, is also true. If I had chosen to estimate MNLM models where the choice was conservative mainline Protestant over other mainline Protestant, the only diﬀerence in the table would be that the signs on the coeﬃcients would be reversed.

206

sean f. everton Conclusion

I began this article by asking whether the individuals who support contemporary mainline ERMs were themselves raised in mainline Protestant churches, or do they come from other religious traditions such as evangelical Protestantism. While I could not answer this question directly since there are no publicly available surveys of rankand-ﬁle ERM members, I uncovered indirect evidence by examining the religious roots of theological conservatives within mainline Protestantism, the group from which ERMs presumably draw the bulk of their support. I found that there is a higher probability that conservative mainline Protestants come from mainline rather than from other religious backgrounds, and in particular, evangelical backgrounds. Of course, the data make it diﬃcult to distinguish between the various groups with any exactness, which points to a limitation of this study and why it should only be seen as preliminary in nature. Because I grouped mainline fundamentalists, evangelicals, and biblical literalists together as conservative mainline Protestants, in the future researchers might want to explore the diﬀerences that exist between the groups and their relationship to ERMs. Smith and his colleagues (Smith, 2000; Smith, Emerson, Gallagher, Kennedy, & Sikkink, 1998) have persuasively argued that evangelicals are more likely than are fundamentalists to engage the world in any substantial way. Thus, it would be interesting to determine whether evangelicals are more likely than fundamentalists to actively participate in ERMs. To ascertain whether this is true, a survey of actual rankand-ﬁle ERMs would, of course, be needed, and such a survey would, in turn, permit direct testing of the argument explored here. Researchers may also want to consider exploring the backgrounds of ERM leaders and engage in in-depth qualitative studies of ERM values, hopes and actions. There remains, of course, the question as to whether ERMs will succeed in their quest. Although conservative Southern Baptists have successfully retaken control of the Southern Baptist Convention (Ammerman, 1990), Finke and Stark (1992) have expressed doubt as to whether this will have any long-term eﬀects. Furthermore, in the past, mainline Protestant movements similar to contemporary mainline ERMs have mobilized in protest against the liberalizing tendencies of their denominations with the stated purpose of reform-

mainline evangelical renewal movements

207

ing, and not leaving, their respective denominations.7 After years of failed eﬀorts, however, most gave up and broke away to start new denominations. Only time will tell whether mainline ERMs will succeed in recapturing the traditions and beliefs they inherited from their denominational founders. References American Anglican Council. (1996). American Anglican Council: A Narrative. Dallas, TX: Author. Ammerman, N. T. (1990). Baptist battles: Social change and religious conﬂict in the Southern Baptist Convention. New Brunswick, NJ: Rutgers University Press. Becker, G. S. (1975). Human capital: A theoretical and empirical analysis (2nd ed.). New York: Columbia University Press for the National Bureau of Economic Research. —— (1976). The economic approach to human behavior. Chicago: University of Chicago Press. Begg, C. B., & Gray, R. (1984). Calculation of polychotomous logistic regression parameters using individualized regressions. Biometrika, 71, 11–18. Boswell, J. (1981). Christianity, social tolerance, and homosexuality: Gay people in Western Europe from the beginning of the Christian era to the fourteenth century. Chicago: University of Chicago Press. Confessing Movement. (2003). Our confession. Retrieved March 29, 2004, from http:// www.confessingumc.org Davis, J. A., & Smith, T. W. (1998). General social surveys, 1972–1998. [machinereadable data ﬁle]. Principal Investigator, James A. Davis; Director and CoPrincipal Investigator, Tom W. Smith; Co-Principal Investigator, Peter V. Marsden, NORC ed. Chicago: National Opinion Research Center. Davis, J. A., Smith, T. W., & Marsden, P. V. (1999). General social surveys, 1972–1998 Cumulative Codebook. Chicago: National Opinion Research Center. Finke, R., & Stark, R. (1992). The churching of America, 1776–1990: Winners and losers in our religious economy. New Brunswick, NJ: Rutgers University Press. Hausman, J., & McFadden, D. (1984). Speciﬁcation tests for the multinomial logit model. Econometrica, 52(5), 1219–1240. Helminiak, D. A. (1994). What the bible really says about homosexuality. San Francisco: Alamo Square Press. Hinson, E. G. (1993). Will the Baptist tradition survive in the ABC/USA? Speech given at St. Paul’s United Methodist Church, San Jose, CA. Hoﬀman, J. P., & Miller, A. S. (1997). Social and political attitudes among religious groups: Convergence and divergence over time. Journal for the Scientiﬁc Study of Religion, 36(1), 52–70. Howell, L. (2003). United Methodism @ risk: A wake-up call. Kingston, NY: Information Project for United Methodists. Iannaccone, L. R. (1990). Religious participation: A human capital approach. Journal for the Scientiﬁc Study of Religion, 29(3), 297–314.

7 See, for example, Torbet’s (1973) account of the numerous protest movements that arose in protest against the liberalizing tendencies of the American (Northern) Baptist Churches.

208

sean f. everton

Iannaccone, L. R. (1995). Risk, rationality, and religious portfolios. Economic Inquiry, 38(2), 285–295. Ierulli, K., Glaeser, E. L., & Tommasi, M. (1995). Introduction. In M. Tommasi & K. Ierulli (Eds.), The new economics of human behavior. Cambridge, England: University of Cambridge Press. Kluegel, J. R. (1980). Denominational mobility: Current patterns and recent trends. Journal for the Scientiﬁc Study of Religion, 19(1), 16–25. Long, J. S. (1997). Regression models for categorical and limited dependent variables. Thousand Oaks, CA: Sage. McKinney, J., & Finke, R. (2002). Reviving the mainline: An overview of clergy support for Evangelical movements. Journal for the Scientiﬁc Study of Religion, 41(4), 773–785. Myers, S. M. (2000). The impact of religious involvement on migration. Social Forces, 79(2), 755–783. Smith, C. S. (2000). Christian America? What Evangelicals really want. Berkeley and Los Angeles: University of California Press. Smith, C. S., Emerson, M. O., Gallagher, S., Kennedy, P., & Sikkink, D. (1998). American Evangelicalism: Embattled and thriving. Chicago: University of Chicago Press. Spong, J. S. (1994). Living in sin. San Francisco: Harper. Stark, R., & Bainbridge, W. S. (1985). The future of religion. Berkeley: University of California Press. Stark, R., & Bainbridge, W. S. (1987). A theory of religion. Bern, Switzerland: Peter Lang. Steensland, B., Park, J. Z., Regnerus, M. D., Robinson, L. R., Wilcox, W. B., & Woodberry, R. D. (2000). The measure of American religion: Toward improving the state-of-the-art. Social Forces, 79(1), 291–318. Torbet, R. G. (1973). A history of the Baptists (3rd ed.). Valley Forge, PA: Judson Press. Wuthnow, R. (1976). The consciousness reformation. Berkeley: University of California Press.

HAPPINESS AND THE VARIETIES OF RELIGIOUS EXPERIENCE: RELIGIOUS SUPPORT, PRACTICES, AND SPIRITUALITY AS PREDICTORS OF WELL-BEING Joseph W. Ciarrocchi and Erin Deneke* Abstract The present study proposes a stringent test for religious variables by examining their power to predict subjective well-being (happiness) over and above nonreligious indicators. The total sample included 489 participants (57% women) with a mean age of 44. A series of multiple hierarchical regressions was performed with each wellbeing facet as the criterion in separate models. Personality explained a signiﬁcant amount of variance for negative emotion (21%), positive emotion (28%), and cognitive well-being (17%). Congregational support and public religious practices added no unique variance to any aspect of subjective well-being. Spirituality made a signiﬁcant, independent contribution in predicting positive emotion (5%) and cognitive well-being (1%), but none to negative emotion. These ﬁndings support the notion that spirituality makes a unique contribution in predicting subjective wellbeing, and its motivational power may result from its link to positive aﬀect and life satisfaction.

Although prominent psychologists (Freud, 1927/1961; Ellis, 1986) once maintained that religion psychologically harmed people, the ﬁrst wave of empirical research identiﬁed forms of religious motivation that diﬀerentiated positive from negative outcomes (Allport, 1950). Recent empirical research has revealed numerous positive associations between religion and psychological well-being (Pargament, 2002b). Currently, there is little dispute that a relationship exists between religion and well-being. A major discussion is underway, however, over how to interpret this relationship and what best accounts

* This research was supported in part by a Faculty Development Grant, Dean of the College of Arts and Sciences, Loyola College in Maryland. The authors thank Stephanie Marinelli for editing and manuscript preparation. Portions of this paper were presented as a poster at the 2004 Annual Convention of the American Psychological Association, Honolulu, Hawaii. Correspondence concerning this article should be addressed to Joseph W. Ciarrocchi, Pastoral Counseling Department, 8890 McGaw Road, Suite 380, Columbia, Maryland, 21045. E-mail: jwc@ loyola.edu Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

210

ciarrocchi and deneke

for it. One viewpoint sees religion as constituting something unique to human experience and happiness ( James, 1902/2002). A second interprets the eﬀects as the result of ‘mundane mediators’ ( Joiner, Perez, & Walker, 2002) that explain away religion’s unique eﬀects. We begin ﬁrst by reviewing these contrasting positions, and then move into a discussion of methodological strategies for sorting out the various claims. Contrasting Explanations If one mark of a ﬁeld’s maturity is serious discussion by renowned researchers, then a recent symposium (Baumeister, 2002) on religion and psychology indicates the ﬁeld is indeed growing up. Although the symposium represented divergent positions on how best to interpret the ﬁndings of psychology and religion, each article acknowledged the importance of the accumulated data to this point. The target articles and responses focused primarily on the eﬀects of religion, for good or ill, on psychological, social, and health outcomes. At the risk of oversimplifying, the articles drew several conclusions. First, religious practices and experiences have measurable eﬀects on important psychosocial and health outcomes. Second, these eﬀects seem to be independent of at least some plausible alternative mediators; e.g., social support, age, and health status. Third, these eﬀects may be culture-speciﬁc; i.e., in Protestant Christian populations. Fourth, the eﬀects are relatively small but statistically signiﬁcant over a wide range of outcomes, instruments, and age ranges. Finally, a major dispute centers on whether religious practices and beliefs contribute anything unique to these outcomes not better explained by important psychosocial variables. Some argued that religion is unique both in regard to its substance and function (Pargament, 2002a, b), while others argued that it may be a ‘mundane mediator’ ( Joiner, Perez, & Walker, 2002). Some went further stating, “. . . the active ingredients of religion in relationship to real-life outcomes are shared with many possible sources. . . . [I]t may be a better strategy to focus on the ingredients rather than on one particular source, even religion” (Funder, 2002, p. 213). The role that faith and spirituality play in human functioning has engaged philosophers, social scientists, and theologians and has led to many diﬀerent perspectives on their relationship. Some nineteenth-

religious varieties of happiness

211

century thinkers, such as Freud, Marx, and Nietzsche, viewed religion primarily as a means to manage the pain and suﬀering of life. This perspective views religion negatively in terms of its eventual impact on the individual and society. Whether viewed as the opiate of the people (Marx), an obsessional compulsive neurosis (Freud), or as the source of people’s justiﬁcation for living a resentful existence while awaiting eventual vindication in the afterlife (Nietzsche), religion functions as an illusion that at best buﬀers people from painful reality, and at worst prevents people from engaging the present world in an eﬀective, progressive manner. Other thinkers in that era saw more positive possibilities in religion. William James (1902/2002), although personally skeptical of organized religion, linked religious belief in some cases to adaptive psychological functioning. For example, in an age that lacked formal psychological or medical treatment for social ills such as alcoholism and drug abuse, James identiﬁed numerous cases in which religious conversion led to personal transformation and recovery. Indeed, these vivid stories led the founders of Alcoholics Anonymous to draw heavily from James’ depictions of successful abstainers, and to recommend attaining similar states to their co-members (Kurtz, 1979), thereby encouraging spirituality as an aid to recovery. James (1902/2002) might ﬁnd the current scientiﬁc discussion around psychology and religion bemusing. He certainly saw religion as a legitimate object of inquiry: “. . . to the psychologist the religious propensities of man must be at least as interesting as any other of the facts pertaining to his mental constitution” (p. 4). He went further; however, than many would today in that he believed religious phenomena diﬀered from other psychological phenomena. “The essence of religious experience, the thing by which we ﬁnally must judge them, must be that element or quality in them which we can meet nowhere else” (p. 52). “Like love, like wrath, like hope, ambition, jealousy, like every other instinctive eagerness and impulse, it [religion] adds to life an enchantment which is not rationally or logically deducible from anything else” (p. 55). James would not have disagreed with Pargament’s assertion (2002b) that religion is distinctive and unique. A hiatus of several decades occurred before scientiﬁc psychology seriously explored the contribution of religious belief to adaptive human functioning. Allport added considerable sophistication to research on the topic by identifying various ‘brands’ of religion. Some

212

ciarrocchi and deneke

forms of religious experience related to healthy functioning and some were related to negative social characteristics such as prejudice (1954; 1966). The contemporary scene, while not neglecting the downside of religious belief, has further explored the adaptive, functional nature of religious belief and practice. The largest body of evidence emerges from studies in the psychology of religion and coping. Pargament’s empirical work (1997), as well as his interpretive summaries of the literature (2002a, b), leave little room for doubt that religious belief and practices have a robust relationship with adaptive functioning. Explanatory Models With this history as backdrop, two current approaches in psychological theory and method have the potential to further advance psychology of religion. The ﬁrst is the positive psychology movement (Snyder & Lopez, 2002). This movement sees in psychology’s recent past a neglect of positive human functioning when compared to its emphasis on negative psychological states and human behavior. To counteract this neglect, considerable research is addressing positive aspects of human functioning and variables that describe more uplifting features of human experience; e.g., optimism, hope, altruism, etc. This movement’s critique of past clinical and experimental psychology applies to the psychology of religion as well. The overwhelming emphasis in empirical research in the psychology of religion relates to people coping with psychological stress, physical illness, or diﬃcult life circumstances (Pargament, 1997). Hedonic psychology—the study of happiness or subjective wellbeing—oﬀers an alternative frame of reference for the psychology of religion. A series of large surveys indicates that religious belief has a small, yet signiﬁcant measurable relationship to happiness even when controlling for relevant demographic variables (Argyle, 1999). Empirical evidence suggests that religious belief and spirituality is a fruitful area of inquiry for healthy functioning as well (Pargament, 1997). Fredrickson (2002) has proposed one model that could link positive psychology with religion. Her broaden-and-build model of positive emotions posits a mediating eﬀect of religious practices and beliefs on positive emotion. In this model, positive emotional experiences broaden the thought-action repertoire of individuals so that

religious varieties of happiness

213

their mind-sets expand, in contradistinction to the narrow thoughtaction tendencies associated with negative aﬀect. In dangerous situations, people focus narrowly on threat with an action-tendency to either escape or attack. Positive emotions, on the other hand, lead to more expansive cognitive patterns such as joy, contentment, interest, and gratitude. Experiencing these mind-sets, in turn, leads to people building enduring personal resources that support them in future coping situations. This model hypothesizes, therefore, that positive emotion mediates the beneﬁcial eﬀects of religion through the “most reliable path” (2002, p. 211) of meaning. In other words, religion provides meaning which, in turn, creates positive emotions leading to the development of personal capacities that then ensure health and well-being. Control theory (Carver & Scheier, 1998) and its correlative expectancy-value model of motivation, provide a second avenue for exploring the link between religion and subjective well-being. Control theory is one of a family of theories that views behavior as organized around goals. Goals help people regulate their behavior by motivating them to pursue desired values or avoid undesired ones. This process involves a feedback system that includes input (perception), the goal or reference value, and the ability to compare the two. If there is a discrepancy between input and goal, the person acts to reduce the discrepancy (behavior), either toward a desired goal or away from an undesired goal. Goals exist at four diﬀerent levels according to their degree of abstraction. At the highest level is the system concept or ideal self. Next are “be” goals that operate as principles that organize the self-ideal. These are followed by program or “doing” goals that move people toward fulﬁlling their principles. Finally, sequences or motor-control acts are the concrete steps (e.g., taking money out of a wallet) embedded in the larger program goal (giving to charity), which accomplishes loving my neighbor (principle), and means that I am a religious Jew (system concept or ideal self ). Expectations, along with the goal/values, round out the motivational end of control theory. The levels of conﬁdence or doubt people have in positive or negative outcomes predict whether people actually engage or disengage in goal pursuits. Emmons (1999) has extended control theory’s analysis of goals into the religious realm. Speciﬁc goal content and structures are related to subjective well-being. Those related to positive well-being included aﬃliation-intimacy, generativity, intrinsic goals, approach

214

ciarrocchi and deneke

goals, lower-level goals, goal commitment, goal-progress, and spiritual strivings (p. 47). Goals related to negative well-being included power goals, extrinsic goals, goal conﬂict, goal ambivalence, higherlevel goals, and interpersonal goal hindrance. Even a cursory examination of these ﬁndings would line up most aspects of traditional religion as supporting positive well-being, and several proscribed religious goals that are linked to negative well-being. In explaining the relationship between spiritual strivings and subjective well-being, Emmons focused on spirituality’s ability to provide personality integration, considered a major criterion for healthy functioning. He cites Allport (1950) who suggests that “. . . the healthy mind requires a hierarchical organization of sentiments. . . .” (p. 79) and that religion holds more promise than therapy as an integrative agent (Emmons, 1999, p. 123). What is the mechanism for this relationship? Emmons points to religion’s ability to provide meaning. Taken together, then, control theory and the model of positive emotions would predict that spirituality leads to subjective well-being through broad-based goals that motivate a variety of programs for engaging life and ﬁnding meaning. These goals and meaningful programs, in turn, generate conﬁdence and positive emotions. Benefit of Incremental Validity Model One major methodological obstacle at this point prevents enthusiastic endorsement of the relationship between spirituality and wellbeing —one that hampers the ﬁeld in establishing links between spirituality and other variables as well. This is the problem of whether spiritual/religious variables add anything to predicting outcomes above and beyond existing psychological variables. From a methodological standpoint do religious variables have incremental validity (Piedmont, 1999a)? Or, as Van Wicklin (1990) has asked, do they represent anything more than the ‘religiﬁcation’ of already existing psychosocial variables? An incremental validity model uses various statistical strategies to identify these potentially confounding mediators to determine whether the predictor variables of interest add additional predictive variance to outcomes. Without such statistical controls, multiple alternative explanations for the eﬀects of religiousness are cogent. Recent criticisms of research in the ﬁeld of religion and health, for example, claim that religious research has failed to control for

religious varieties of happiness

215

relevant variables that plausibly explain the reputed link between religion and physical health (Sloan, Bagiella, & Powell, 2001). A useful example of this failure is not controlling for personality in much religious research. It is well-known that personality accounts for a large amount of variance in predicting well-being (Diener & Lucas, 1999); yet one is hard-pressed to ﬁnd even a few studies that take this into consideration in surveys of religion and coping (Pargament, 1997). With a few exceptions (Ciarrocchi, Piedmont & Williams, 2003; Emmons, 1999; Francis & Katz, 2002; Piedmont, 1999b), religious research has ignored the incremental validity problem in terms of controlling for personality. This neglect is problematic and curious given the reported impact of personality. Religious research often controls for assorted demographic variables; yet, reviews of psychological well-being report that age, sex, marital status, income, race, and education, even when combined, account for only about 15% of the variance in subjective well-being (Argyle, 1999). Personality considerably exceeds that total amount (Diener & Lucas, 1999). An incremental validity model in religious research, particularly attending to the relationship between personality and criterion variables, therefore, is well suited to further our understanding of spirituality and religiousness’ connection with happiness. A further advantage in utilizing personality measures in religious research is their heuristic value for examining the nature of religious experience and behavior (McCrae, 1999). Correlations between ﬁvefactor scales and religiosity are often signiﬁcant but typically run in the low range. For example, intrinsic religiosity correlated with agreeableness and conscientiousness in a large sample (N = 1129) of Canadian undergraduates (Taylor & MacDonald, 1999), while extrinsic religiosity correlated negatively with openness. In a U. S. undergraduate sample, intrinsic religiosity correlated positively with agreeableness and conscientiousness as in the previous study but, contrary to that study, correlated negatively with openness (Piedmont, 1999a). Further, openness was unrelated to extrinsic religiosity. The inconsistent results may have occurred from using diﬀerent measures of religious orientation. Personality correlated signiﬁcantly with perceived closeness to God (Faith Maturity Scale) on all ﬁve factors for women but only on agreeableness for men (Ciarrocchi, Piedmont, & Williams, 2002) in an undergraduate sample (N = 989). These studies suggest that personality has a measurable connection to religious

216

ciarrocchi and deneke

attitudes and is diﬀerentially associated by gender. Although not typically identiﬁed by critics of religious research, the studies cited point to the potential mediating eﬀect of personality in religious research. Another major challenge for religion and well-being research is the complex nature of subjective well-being. The consensus in the research ﬁeld is that happiness, or subjective well-being, consists of three related, but independent, facets: negative emotion, positive emotion, and cognitive well-being (Kahneman, Diener & Schwartz, 1999). How religious beliefs and practices relate to these components is unknown. Do they relate to all three facets similarly, or do some forms of religious experience aﬀect individual components? The existing literature that diﬀerentiates spirituality’s relationship with negative versus positive dimensions of well-being is sparse. A seminal study by Watson and Clark (1993; Clark & Watson, 1999) found that religious beliefs were positively related to positive emotionality but unrelated to negative emotionality. Research that measures positive and negative emotions through balanced aﬀect methods has, in general, supported this pattern. A balanced aﬀect scale measures both positive and negative aﬀect, then subtracts negative aﬀect from positive emotion. Studies employing that technique in religious research report great variability. Studies using only the single balanced aﬀect score have found positive relationships between it and attendance at religious services (Graney, 1975), as well as with intense religious experiences (Hay & Morisy, 1978). A study looking at only positive and negative aﬀect (Yates, Chalmer, St. James, Follansbee, & McKegney, 1981) found that attendance at church, perceived closeness to God, and importance of church attendance correlated signiﬁcantly with positive but not negative aﬀect. Finally, two studies employing all three components of aﬀect balance found no relationship at all between any aspect of well-being and measures of either religious coping or attitudes toward Christianity (Francis, Jones, & Wilcox, 1997; Koenig, Siegler, & George, 1989). To clarify these relationships, Francis and Kaldor (2002) surveyed 989 Australian adults in a stratiﬁed community survey. Personal prayer, church attendance, and belief in God correlated positively with positive emotion but were independent of negative emotion, consistent with Watson and Clark’s (1993) earlier ﬁndings. Church attendance added no additional predictive power to positive emotion beyond prayer and belief in God.

religious varieties of happiness

217

Even more sparse are studies of subjective well-being and religion that control for personality. Religious attitudes predicted happiness, as measured by the Oxford Happiness Scale, in a sample of 298 Hebrew-speaking undergraduates over and above the personality variables of introversion and extroversion (Francis & Katz, 2002). This study employed an incremental validity model and found that religious beliefs contributed independently to predicting happiness. It did not examine cognitive well-being separately, however, rendering unclear the relationship between religious attitudes and this feature of subjective well-being. A series of recent studies controlling for personality with ﬁve-factor personality measures has tested speciﬁcally the relationship between spirituality and most facets of subjective well-being. Controlling for age, gender, marital status, and personality with Methodist ministers (N = 320) spirituality predicted psychological exhaustion and cognitive well-being (Golden, 2002; Golden, Piedmont, Ciarrocchi, & Rodgerson, 2004). In a sample of Maltese undergraduates (N = 312), spiritual transcendence predicted both positive aﬀect and cognitive well-being over and above personality but did not predict negative aﬀect (Galea, 2003). A similar pattern emerged in a sample of male sex oﬀenders (N = 194) with faith maturity predicting positive aﬀect and cognitive well-being but not negative aﬀect (Geary, 2003). Spiritual transcendence predicted cognitive well-being over personality but not either negative aﬀect or positive aﬀect (Walsh, 2001) in problem gamblers (N = 100). Finally, faith maturity predicted cognitive wellbeing over personality but was unrelated to negative emotionality in female breast cancer patients (Grogan-Henderson, 2002; GroganHenderson & Ciarrocchi, 2003). These results are notable in that, despite using a variety of instruments to measure both personality and well-being, the pattern of the outcomes is similar in most of the studies. Spirituality predicted the positive aspects of subjective well-being, but failed to predict negative emotionality, as ﬁrst noted by Watson and Clark (1993). Overall these results render Fredrickson’s (2002) theory plausible that religion’s major inﬂuence on well-being may be its ability to enhance positive emotions. On the other side, a substantial body of empirical work in the area of coping demonstrates a link between religious forms of coping and stress reduction (Pargament, 1997). The discrepancy between the coping literature and the studies cited above may occur for two

218

ciarrocchi and deneke

reasons. First, religion may have a diﬀerential impact when people are in crisis from when they are not. The coping literature typically examined people’s functioning under stressors such as bereavement or environmental hardships. Perhaps religious beliefs mediate anxiety in crisis situations and enhance positive emotions and well-being under ordinary life circumstances. Indeed, a meta-analysis of over 200 empirical studies found a consistent negative relationship between religiousness and depression, and that relationship was moderated by stress (Smith, McCullough, & Poll, 2003). Yet, despite the preponderance of evidence for the link between religious coping and depression, studies on religion and coping typically do not control for personality. In the absence of data to the contrary, the eﬀects of religious coping on reducing stress could be mediated by personality, in particular, the trait of negative emotionality. The psychology of religion, therefore, like its sister disciplines, has focused perhaps too narrowly on human distress and thus missed opportunities to understand religiousness’ relationship to healthy, adaptive human functioning. If that is the case, religion, far from being the opiate of the people, may take its place in humanistic psychological projects as one pathway for attaining “the farther reaches of human nature” (Maslow, 1971). The present study attempts to position itself within the general framework of well-being research by using standard outcome measures of subjective well-being and well-accepted personality constructs. Taking this approach should allow clearer interpretation of ﬁndings that relate religion and spirituality to subjective well-being. Second, in light of various critiques of religious research (Sloan, Bagiella, & Powell, 2001), this study will provide a stringent test of these relationships by controlling for personality, age, gender, and speciﬁcally religious support ( Joiner, Perez, & Walker, 2002). Although no single study can hope to control for all possible confounds, testing religion and spirituality against well-established mediators of well-being should shed light on whether the correlates between religion and well-being are unique. The study hypothesizes that spirituality, as measured by people’s perceived closeness to God, will predict subjective well-being uniquely over the aforementioned control variables. However, in light of recent research that controls for personality, we posit that spirituality will predict the positive, but not the negative, dimensions of happiness. To replicate previous ﬁndings (Francis & Kaldor, 2002), we predict,

religious varieties of happiness

219

further, that religious practices will have no added eﬀect on subjective well-being. Method Participants Graduate students in a psychology of religion class solicited volunteers from their acquaintances over a three-year period. There were 209 men (42.7%) with a mean age of 43.69 years (range: 18 to 84 years) and 280 women (57.3%) with a mean age of 43.64 (range: 17 to 87). The sample was 77% Caucasian, 10% African-Americans, 8% Asian/Paciﬁc Islanders, 2% Hispanic, and 3% other. Religious aﬃliation was 48% Protestant, 36% Catholic, 2% Jewish, 5% not speciﬁed, 2% other and 7% atheist/agnostic. Measures Well-Being Measures In keeping with the multidimensional understanding of subjective well-being, outcome measures tapped positive and negative aﬀect as well as cognitive well-being. The positive and negative aﬀect scales were two 6-item scales, one for each dimension, whose items were derived from various established aﬀect scales (Mroczek & Kolarz, 1998). Each scale arranges items on a 5-point Likert scale from ‘none of the time’ to ‘all of the time.’ Instructions included the stem question “During the past 30 days, how much of the time did you feel . . . ?” Sample items include, nervous, hopeless (negative aﬀect), and cheerful, full of life (positive aﬀect). The 30-day time frame paints a recent aﬀect context and avoids temporary mood ﬂuctuations in measures of current feeling. Alpha reliabilities were .73 and .86 for negative and positive aﬀect respectively. The Satisfaction-With-Life-Scale is an accepted measure of cognitive well-being, widely used in happiness research (Diener, Emmons, & Larsen, 1985). The scale consists of ﬁve items arranged in a 7point Likert fashion from strongly disagree to strongly agree. Sample items include, “I am satisﬁed with my life,” and “In most ways my life is close to my ideal.” Alpha reliability for the current study is .82.

220

ciarrocchi and deneke

Personality Measure We measured personality according to the ﬁve-factor model, a standard taxonomy that captures a broad array of human characteristics (McCrae & Costa, 1999). The traditional ﬁve factors include neuroticism (tendency to experience negative emotion), extraversion (tendency toward sociability and positive experience of emotions), openness to experience (degree to which one is open to diﬀerent types of experience), agreeableness (tendency to be cooperative and oriented toward others), and conscientiousness (degree to which one sets goals and adheres to responsible behavioral patterns). The Bipolar Adjective Scale measured the ﬁve factors in the study. This 80-item scale measures responses on a 7-point Likert-type scale (McCrae & Costa, 1985, 1987). Scores for each dimension are obtained by summing responses for each domain of the ﬁve factors. Sample items include, ‘imaginative-down-to-earth’, and, ‘aﬀectionate-reserved’. Spirituality and Religiousness Measures Consistent with the understanding that the religious experience encompasses various dimensions, participants completed the 40-item Multidimensional Scale of Religion and Spirituality (MSRS, Idler, et al., 2003). Developed by a combined working group from the National Institute on Aging and Fetzer Institute, the MSRS covers eleven domains of religious experience, such as religious aﬃliation, public and private religious experiences, forgiveness, and religious coping. Following its development by the working group, the MSRS was included in the 1998 General Social Survey (N = 1445) and had reliability estimates ranging from moderate to good (Idler et al., 2003). Although participants in the current study ﬁlled out the entire scale, the current study focuses on four domains in the MSRS: public religious activities (2 items), congregation beneﬁts and problems (4 items combined to measure a single domain of religious support), and daily spiritual experiences (6 items). This last scale comprises the domain of spirituality in this study and refers to people’s perceived closeness to and desired relationship with God. Sample items for each of the scales include: ‘How often do you go to religious services?’(public religious activities), ‘If you are ill, how much would the people in your congregation help you out?’ (religious support), and ‘I feel God’s presence’ (daily spiritual experiences). Despite the small number of items, internal consistency was moderate to good. Alpha reliability was .72 for religious support (con-

religious varieties of happiness

221

gregation beneﬁts = .93, congregation problems = .71), .79 for organized religious activities, and .91 for daily spiritual experiences. These reliability coeﬃcients are highly similar to those obtained in the national probability sample (Idler et al., 2003, p. 352). Results Descriptive Statistics Table 1 presents the means, range, and standard deviations for predictor and outcome variables. The ﬁve-factor T-scores are all between 45 and 55. T-scores below 45 and above 55 are considered distinctive, suggesting that scores in this study resemble those of the general population found in previous research (McCrae & Costa, 1985, 1987). Independent t-tests were run to analyze diﬀerences in gender and few emerged. Women scored signiﬁcantly higher in measures on neuroticism ( p < 0.03) and openness ( p < 0.01) and negative religious support ( p < 0.05). Table 2 indicates the intercorrelations for the study variables. Inspection of this table illustrates several points. First, consistent with previous research, neuroticism and extraversion correlate with all well-being indices. Correlations, however, are in the small to lowmoderate range suggesting that temperament and well-being are not redundant concepts. Second, except for negative religious support, the spiritual and religiousness variables correlate in the small range with well-being indices. The eﬀect sizes for spirituality and well-being ranges from 3% to 13%, while the eﬀect size for organizational religion and religious support range from less than 1% up to 2%. Third, signiﬁcant intercorrelations among the spirituality and religiousness variables range from .10 to .62, with three correlations in the small and four in the moderate range. Religiousness and spirituality appear to be non-overlapping constructs in the Multidimensional Measurement of Religiousness/Spirituality (Idler et al., 2003) at this level of analysis. The strength of these relationships replicates ﬁndings in the general social survey (Idler et al., 2003). Finally, multiple signiﬁcant relationships exist between personality and the spirituality/religiousness variables. The larger correlations consistently occurred for spirituality rather than religiousness. These relationships among personality, spirituality, and religiousness as well as spirituality and religiousness’ relationships with well-being, conﬁrm the necessity of sorting out the possible mediating role of personality.

ciarrocchi and deneke

222

Table 1. Means and Standard Deviations for Personality, Spirituality, Religiousness, and Well-being Variable

M

SD

N

Personality: Neuroticism Extraversion Openness Agreeableness Conscientiousness

47.01 50.44 52.17 53.16 50.44

9.21 9.57 9.76 9.18 10.73

471 471 471 471 471

Fetzer subscales: Daily Spiritual Experiences Positive Religious Support Negative Religious Support Organizational Religiousness

25.20 6.02 3.20 7.65

6.98 2.04 1.20 1.54

489 489 483 489

Well-being: Positive Aﬀect Negative Aﬀect Satisfaction with Life

21.25 10.94 25.13

3.62 3.46 6.09

489 489 489

Table 2. Intercorrelations for Personality, Spirituality, Religiousness, and Well-being 1 1. Neuro. 2. Extrav. 3. Open. 4. Agree. 5. Consc. 6. Spirit. 7. PRS 8. NRS 9. Organ. Rel. 10. Pos. Aﬀect 11. Neg. Aﬀect 12. SWL

–0.23 –0.24 –0.38 –0.46 –0.25 –0.13 0.02 –0.11 –0.44 0.47 –0.37

2

0.36 0.45 0.31 0.31 0.20 0.02 0.12 0.38 –0.22 0.22

3

4

0.86 0.14 0.07 0.05 0.11 0.04 0.09 –0.07 0.05

0.35 0.37 0.24 0.06 0.18 0.33 –0.21 0.24

5

0.18 0.10 0.03 0.71 0.35 –0.29 0.33

6

0.50 0.19 0.56 0.36 –0.17 0.21

7

8

9

10

11

0.26 0.62 0.40 0.14 –0.04 0.14 –0.09 0.05 –0.09 –0.55 0.05 –0.05 0.04 0.57 –0.39

Note: Neuro. = Neuroticism; Extrav. = Extraversion; Open. = Openness; Agree. = Agreeableness; Consc. = Conscientiousness; Spirit. = Spirituality; PRS = Positive Religious Support; NRS = Negative Religious Support; Organ. Rel. = Organizational Religiousness; Pos. Aﬀect = Positive Aﬀect; Neg. Aﬀect = Negative Aﬀect; SWL = Satisfaction with Life. p < .05 in boldface

religious varieties of happiness

223

Tests of main hypothesis A series of multiple regression analyses tested the main hypotheses that spiritual experience predicts positive but not negative emotions and well-being. To determine whether spirituality improves predictability of these well-being facets over psychosocial, personality, and religious support, a series of hierarchical regressions was performed with each facet constituting the criterion in separate models. For each model, age and gender were entered on step 1, personality on step 2, religious support on step 3, and ﬁnally spiritual experience on step 4. Partial F-tests were calculated at each step to determine whether a particular set of variables signiﬁcantly increased the explained proportion of the variance. As expected, Tables 3–5 indicate that personality explains a signiﬁcant amount of the variance for negative emotion (21%), positive emotion (28%), and cognitive well-being (17%) when controlling for age and gender. The beta weights in Table 6 illustrate that neuroticism and extraversion account for most of the variance for the three well-being scales, with conscientiousness contributing only to cognitive well-being. That same table indicates gender had no eﬀect on any well-being facet and that increasing age was negatively related to emotional distress—a ﬁnding consistent with the well-being literature (Argyle, 1999). Contrary to the position of those seeing social mediators as the basis for religious beliefs’ eﬀect on well-being ( Joiner, Perez, & Walker, 2002), speciﬁcally religious support added nothing over and above personality in predicting subjective wellbeing on any well-being facet. Table 3. Hierarchical Regression for Negative Emotion Hierarchical Step

Predictor Variable

R2

ΔR2

F Change

1 2 3 4

Age, Gender Personality Religious Support Spirituality

.06 .27 .28 .28

.06 .21 .01 .00

13.73* 24.59* 2.75 .03

Note: N = 427 * p < .001

ciarrocchi and deneke

224

Table 4. Hierarchical Regression Analysis for Positive Emotion Hierarchical Step

Predictor Variable

R2

ΔR2

F Change

1 2 3 4

Age, Gender Personality Religious Support Spirituality

.01 .29 .30 .34

.01 .28 .01 .05

2.58 33.19* 3.88 29.14*

Note: N = 428 * p < .001

Table 5. Hierarchical Regression Analysis for Cognitive Well-Being Hierarchical Step

Predictor Variable

R2

ΔR2

F Change

1 2 3 4

Age, Gender Personality Religious Support Spirituality

.00 .18 .18 .19

.00 .17 .00 .01

.74 17.66** .73 4.77*

Note. N = 427 * p < .05; ** p < .001

Table 6. Beta Weights for Predictor Variables of Happiness According to Entry Order Predictor Variable Gender Age Neuroticism Extraversion Openness Agreeableness Conscientiousness Religious Support Religious Practices Spirituality

Negative Emotion

Positive Emotion

Cognitive Well-Being

.08 –.16** .42*** –.13* –.02 .06 –.03 –.07 –.03 –.01

.02 –.03 –.29*** .24*** –.04 –.00 .08 .01 .04 .25***

.01 –.06 –.25*** .08 –.04 .02 .17*** .01 –.04 .11*

Note: N = 427 * p < .05; ** p < .01; *** p < .001

religious varieties of happiness

225

These outcomes support the main hypotheses that no relationship exists between spirituality and negative emotion (Clark & Watson, 1999; Watson & Clark, 1993) over and above personality. Spirituality, however, did make a signiﬁcant independent contribution over personality in predicting positive emotion (5%) and cognitive well-being (1%). Religious practices, as measured by frequency in attendance at religious services and other organized religious activities, when analyzed in the same manner, contributed nothing over the demographic, personality, or religious support variables (R2 change for negative aﬀect = .001; positive aﬀect = .001; and cognitive wellbeing = .001, all nonsigniﬁcant). Table 6 indicates the beta weights are nonsigniﬁcant as well. Discussion Overall Findings Spirituality, deﬁned as perceived closeness to God, adds an element to well-being not accounted for by age, gender, personality, or the social support provided in the religious setting. Spiritual writers, as far back as Augustine, have maintained spirituality was essential to human happiness (Charry, 1997). Further, Augustine deﬁned spirituality as increasing one’s desire for God—a dimension highly similar to questions in the Daily Spiritual Experiences scale used in this study (“I desire to be closer or in union with God”). The role spirituality plays in its relationship with well-being varies according to well-being’s components. Spirituality contributes to positive emotion and cognitive well-being but not negative emotion. This replicates previous work that did not control for personality (Watson & Clark, 1993; Clark & Watson, 1999), as well as work that did (Francis & Kaldor, 2002; Piedmont, 2001). Even though there is a large body of empirical literature that supports religion’s association with coping with negative events (Pargament, 1997), these associations may well be accounted for by failure to control for personality or may be moderated by stressful events. If these results continue to be replicated, Fredrickson’s model (2002) that links religion’s eﬀect on well-being via positive emotions may provide an adequate heuristic conceptualization. Future research should explore the mediation model she described that posits religion’s

226

ciarrocchi and deneke

inﬂuence on positive emotion, resulting in development of personal resources that enhance well-being. These ﬁndings, therefore, are consistent with the positive psychology movement and provide a diﬀerent direction for psychology and religion. Research has discredited a number of assertions about religion put forward by Freud. Repeated studies, for example, demonstrated that one’s image of God is not a projection of an ideal fatherﬁgure (Piedmont, Ciarrocchi, & Williams, 2002), but is more closely related to one’s preferred parent or oneself. The current study discredits another key Freudian notion that religion is related to neurosis. Our ﬁndings suggest that the only relationship spirituality has to well-being is with its positive components, and it had no relationship with negative emotion. Of course, other dimensions of religious coping (e.g. belief in a punishing God) may be related to psychological distress (Pargament, 1997). Unlike spiritual experience, institutional religious practices do not contribute to well-being. Again, this ﬂies in the face of a large body of empirical literature that links attendance at religious events to adaptive functioning (Pargament, 1997). Once more, however, these studies have rarely controlled for personality or the purely religious support that such gatherings provide. The study’s weakness is its cross-sectional methodology that does not permit causal analysis. Well-being may lead people to look for spiritual meaning in life or some third variable may mediate the relationship between the two. As research fails to support explanations based on demographic variables, personality, and religious support, the search for mediators might look to those that tap higher-order constructs consistent with a positive psychology emphasis. Likely candidates could include optimism, hope, meaning, uncertainty reduction, and sense of coherence. An incremental validity model, therefore, is essential for understanding the measurable aspects of religious experience (Piedmont, 1999b; Rodgerson & Piedmont, 1998). It rules in fruitful areas of exploration (spiritual experiences) and challenges others (religious support or institutional religious practices). It contributes to a better understanding of the speciﬁcally psychological components to religious experience. These results suggest that religious researchers should put forth ﬁndings most cautiously when they lack credible controls for personality.

religious varieties of happiness

227

These ﬁndings contribute to answering the question that framed this study. Is religion unique? Spiritual experience, deﬁned as perceived closeness to or desire for God, predicts over and above demographic variables and personality to outcomes such as purpose in life, subjective well-being, and prosocial behavior across international samples, with observer ratings, in clinical and nonclinical settings using a wide array of instruments (Ciarrocchi, Piedmont, & Williams, 2003; Galea, 2003; Geary, 2003; Golden et al.; in press; GroganHenderson, 2002; Piedmont, 2001; Walsh, 2001). The cumulative eﬀect of these ﬁndings should lead to the next appropriate question, “How unique is religion in predicting happiness?” The answer to that appears to be twofold. First, from a purely statistical standpoint, the response is “Not much.” In the better-controlled studies, spiritual experience contributes anywhere from one to ﬁve percent of the variance when well-being is the outcome. Staying within an empirical framework, however, even small amounts of variance can be quite meaningful. An early meta-analysis reported that religion’s eﬀect on well-being was about the same as psychotherapy’s (Bergin, 1983) and this eﬀect has remained in later reviews (Hackney & Sanders, 2003). In the well-being studies cited above, religion’s eﬀect on well-being is greater than that of socioeconomic status. Or, in David Myers’ phrasing (2000), one’s faith is equally important to happiness as one’s funds and friends. To put this in perspective, a substantial portion of governmental budgets have the singular focus of improving citizens’ ﬁnancial status. From this vantage point, the cost-beneﬁt ratio to human happiness for voluntary faith communities in comparison to expensive governmental eﬀorts deﬁes quantifying. Positive Psychology Models What, then, is a productive model for understanding the unique nature of religion and spiritual experience as it relates to well-being? This report tentatively proposes that control theory and its expectancyvalue motivational model provide a fruitful framework for clarifying spiritual and religious phenomena. Control theory not only points to a link between religion and well-being as reviewed above, it also suggests the mechanism by which the link occurs. As noted, goal integration is one path to personality integration. Religion has the capacity to “sanctify” many domains of life, making secular activities

228

ciarrocchi and deneke

sacred (Pargament & Mahoney, 2002). Sanctiﬁcation can transform marriage into a sacrament, ecological awareness into stewardship, and work into a vocation. The end result from control theory’s viewpoint is that religious individuals can incorporate many, if not all, lower-level pursuits into an integrated system with higher-level ones. Religions enhance subjective well-being not only because the content of their goals fosters it (e.g., intrinsic versus extrinsic goals), but also because they promote fully integrated goal systems via the psychological process of sanctiﬁcation. An expectancy-value model points to other potentially fruitful areas of investigation as well. First, in what ways does religiousness form an integrated goal system? Do religious systems vary in their capacity to integrate goals? In what ways do religious systems create value conﬂict, e.g., contradictory dogmas, clashes with the dominant culture, etc.? Control theory provides an alternative model for understanding religiousness in a manner consistent with the positive psychology movement. Rather than viewing religiousness as inherently pathological, it can function as an important source of behavioral regulation. Religions help people self-regulate in multiple ways. They provide distinct reference values, instigate attentional processes to these values via self-reﬂection or prayer, prescribe virtuous behaviors, and enhance self-monitoring through social support within faith communities. Baumeister and his colleagues (Baumeister, Heatherton, & Tice, 1994) have identiﬁed self-regulation failure as the cause of much personal and societal distress in modern life. They have used the term ‘transcendence failure’ to describe one major mechanism that leads to loss of self-control. In transcendence failure people focus exclusively on the lower-level meaning of a behavior (the chocolate cake will satisfy hunger), but ignore or are unaware of its higher-order meaning (sugar products exacerbate diabetes). Although one need not believe in religious transcendence to have self-control, religions are eﬀective resources for promoting self-regulatory transcendence, as witnessed by the consistent negative correlation between religious attendance and health-impairing behaviors such as substance abuse (Gorsuch, 1995). Religion is also an avenue for inﬂuencing expectancies, a central concept in control and other social cognitive theories. Expectancies predict behavioral engagement or disengagement. Control theory

religious varieties of happiness

229

states that people’s expectancies can have many sources and includes religion as a potential inﬂuence (Carver & Scheier, 2002). Diﬀerent religious points of view could function to enhance self-conﬁdence or self-doubt. Which viewpoints lead to which type of expectancy could shed light on paths to important psychosocial and health outcomes, such as religion’s capacity to regenerate a sense of optimism and hope in people suﬀering trauma or feeling hopeless (Murphy et al., 2000). Control theory could illuminate debates within the psychology of religion such as the reputed distinction between being religious versus being spiritual (Zinnbauer, Pargament, & Scott, 1999). One obvious insight is that these terms refer to diﬀerent goal levels. Deﬁnitions of spirituality tend to emphasize higher-level principles or system concepts, while deﬁnitions of religion emphasize goals at the program level. This does not negate the fact that ‘spiritual’ people engage in practices at the program level or that ‘religious’ persons have belief systems at the level of principles. Indeed, correlations between the two are usually in the moderate to large range (e.g. .50; Mapa, Ciarrocchi, & Deneke, 2004). Finally, control theory can address how to understand individual diﬀerences in religious behavior. Evidence points to the similar function that higher-level goals and traits play in human behavior (Carver & Scheier, 1998). Models that view religiousness as a distinct personality disposition (Piedmont, 1999a) ﬁnd support in expectancyvalue theory. Religiousness in this view is a higher order trait (value) with multiple lower-order programs and sequences geared toward fulﬁlling a self-ideal. Just as there are individual diﬀerences in the degree to which people prefer living at higher versus lower levels of goal pursuits (Carver & Scheier, 1998), people diﬀer over their preferences for a higher-level orientation toward their ultimate concerns (Emmons, 1999). Similarly, Fredrickson’s (2002) broaden-and-build theory of positive emotion, with its suggestion that it mediates religion’s eﬀect on well-being, harmonizes with control theory. If, as some suggest (Carver & Scheier, 2002), religion is a potential source of optimism, religion could motivate people to engage in forms of coping conducive to adaptive living (Carver & Scheier, 1998). These coping responses to religious worldviews, in turn, may be responsible for the positive emotions linked to “improved health and well-being” (Fredrickson, 2002, p. 211).

230

ciarrocchi and deneke

Perhaps one way to summarize the symposium (Baumeister, 2002) described at the outset is to challenge both sides in the “Is-religionunique?” debate. The challenge for social scientists who wish to maintain religion’s uniqueness is to adhere to methodologies, as described in this article, that lend empirical support to their claims. The challenge for those who question its uniqueness is to attend again to James’ (1902/2002) rebuttal of biological reductionism. His main argument is that only distinctly religious emotions were “solemn” (p. 86). By this he meant that the mediators postulated in his day could neither account for religion’s capacity to guide grand projects, nor stretch people to sacriﬁce current gains for delayed rewards. James’ complaint is similar to the one positive psychology makes, that the dominant psychological models have a diﬃcult time accounting for what is most solemn in human behavior and experience. If scientiﬁc psychology is to contribute meaningfully to the concerns of public life, it requires a fuller understanding of the mechanisms behind religious motivation. Otherwise it must resort to explanations that pale in light of religion’s power to enrich wellbeing or provide a cause to kill for. This study suggests that we now have scientiﬁc tools for learning more about religious and spiritual mindsets. Positive emotion theory, expectancy-value motivational models, and incremental validity research strategies are proving more fruitful guides for understanding religion and spirituality’s correlates with well-being than eﬀorts to ﬁnd mediators that eliminate those relationships. References Allport, G. W. (1950). The individual and his religion: A psychological interpretation. New York: Macmillian. —— (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley. —— (1966). The religious context of prejudice. Journal for the Scientiﬁc Study of Religion, 5, 447–457. Argyle, M. (1999). Causes and correlates of happiness. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: the foundations of hedonic psychology (pp. 353–373). New York: Russell Sage Foundation. Baumeister, R. F. (2002). Religion and psychology: Introduction to the special issue. Psychological Inquiry, 13, 165–167. Baumeister, R. F., Heatherton, T. F., & Tice, D. M. (1994). Losing control: How and why people fail at self-regulation. San Diego, CA: Academic Press. Bergin, A. E. (1983). Religiosity and mental health: A critical reevaluation and metaanalysis. Professional Psychology: Research and Practice, 14, 170–184. Carver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. Cambridge: Cambridge University Press.

religious varieties of happiness

231

—— (2002). The hopeful optimist. Psychological Inquiry, 13, 288–290. Charry, E. T. (1997). By the renewing of your minds: The pastoral function of Christian doctrine. Oxford: Oxford University Press. Ciarrocchi, J. W., & Deneke, E. (2003, November). Optimism and spirituality as predictors of happiness: Nothing but or more than personality? Poster session presented at the annual convention of the Association for the Advancement of Behavior Therapy, Boston, MA. Ciarrocchi, J. W., Piedmont, R. L., & Williams, J. E. G. (2002). Image of God and personality as predictors of spirituality in men and women. Research in the Social Scientiﬁc Study of Religion, 13, 55–73. —— (2003). Love thy neighbor: Spirituality and personality as predictors of prosocial behavior in men and women. Research in the Social Scientiﬁc Study of Religion, 14, 61–75. Clark, L. A., & Watson, D. (1999). Temperament: A new paradigm for trait psychology. In L. A. Pervin & O. P. John (Eds.), Handbook of personality: Theory and research (2nd ed., pp. 399–423). New York: Guilford Press. Diener, E., Emmons, R. A., & Larsen, R. J. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49(1), 71–75. Diener, E., & Lucas, R. E. (1999). Personality and subjective well-being. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: The foundations of hedonic psychology (pp. 213–229). New York: Russell Sage Foundation. Ellis, A. (1986). The case against religion. Austin, Texas: American Atheist Press. Emmons, R. A. (1999). The psychology of ultimate concerns: Motivation and spirituality in personality. New York: Guilford Press. Francis, L. J., & Kaldor, P. (2002). The relationship between psychological wellbeing and Christian faith and practice in an Australian population sample. Journal for the Scientiﬁc Study of Religion, 41, 179–184. Francis, L. J., & Katz, Y. J. (2002). Religiosity and happiness: A study among Israeli female undergraduates. Research in the Social Scientiﬁc Study of Religion, 13, 75–86. Francis, L. J., Jones, S. H., & Wilcox, C. (1997). Religiosity and dimensions of psychological well-being among 16–19 year olds. Journal of Christian Education, 40(1), 15–20. Fredrickson, B. (2002). How does religion beneﬁt health and well-being? Are positive emotions active ingredients? Psychological Inquiry, 13, 209–212. Freud, S. (1961). The future of an illusion ( J. Strachey, Trans.). New York: Norton. (Original work published 1927). Funder, D. (2002). Why study religion? Psychological Inquiry, 13, 213–214. Galea, M. (2003). The impact of child abuse on the psycho-spiritual status, religious behavior, and family dynamics of Maltese college students. Unpublished doctoral dissertation, Loyola College in Maryland. Geary, B. (2003). The contribution of spirituality to well-being in sex oﬀenders. Unpublished doctoral dissertation, Loyola College in Maryland. Golden, J. L. (2002). Spirituality as a predictor of burnout among United Methodist clergy: An incremental validity study. Unpublished doctoral dissertation, Loyola College in Maryland. Golden, J., Piedmont, R. L., Ciarrocchi, J. W., Rodgerson, T. (2004). Spirituality and Burnout: An incremental validity study. Journal of Psychology and Theology, 32, 115–125. Gorsuch, R. L. (1995). Religious aspects of substance abuse and recovery. Journal of Social Issues, 51, 65–84. Graney, M. J. (1975). Happiness and social participation in aging. Journal of Gerontology, 30, 701–706. Grogan-Henderson, K. (2002). Optimism and spirituality: Predictors of subjective wellbeing in breast cancer patients. Unpublished doctoral dissertation, Loyola College in Maryland.

232

ciarrocchi and deneke

Grogan-Henderson, K. & Ciarrocchi, J. W. (2003, April). Optimism and Spirituality as predictors of subjective well-being in breast cancer patients. Poster session presented at the Sixth World Congress of Psycho-Oncology, Banﬀ, Alberta, Canada. Hackney, C. H., & Sanders, G. S. (2003). Religiosity and mental health: A metaanalysis of recent studies. Journal for the Scientiﬁc Study of Religion, 42, 43–55. Hay, D., & Morisy, A. (1978). Reports of ecstatic, paranormal or religious experience in Great Britain and the United States: A comparison of trends. Journal for the Scientiﬁc Study of Religion, 17, 255–268. Idler, E. I., Musick, M. A., Ellison, C. G., George, L. K., Krause, N., Ory, M. G., Pargament, K. I., Powell, L. H., Underwood, L. G., & Williams, D. R. (2003). Measuring multiple dimensions of religion and spirituality for health research: Conceptual background and ﬁndings from the 1998 General Social Survey. Research on Aging, 25, 327–365. James, W. (2002). The varieties of religious experience: A study in human nature. New York: Modern Library. (Original work published in 1902). Joiner, Jr., T. E., Perez, M., & Walker, R. L. (2002). Playing devil’s advocate: Why not conclude that the relation of religiosity to mental health reduces to mundane mediators? Psychological Inquiry, 13, 214–216. Kahneman, D., Diener, E., & Schwartz, N. (Eds.). (1999). Well-being: The foundations of hedonic psychology. New York: Russell Sage Foundation. Koenig, H. G., Siegler, I. C., & George, L. K. (1989). Religious and non-religious coping: Impact on adaptation in later life. Journal of Religion and Aging, 5, 73–94. Kurtz, E. (1979). Not God: A history of Alcoholics Anonymous. Center City, Minnesota: Hazelden. Mapa, M., Ciarrocchi, J. W., & Deneke, E. (2004, April). Are spirituality and religiousness distinct from each other and from personality? Poster session presented at the annual meeting of the Eastern Psychological Association, Washington, D.C. Maslow, A. (1971). The farther reaches of human nature. New York: Viking. McCrae, R. R. (1999). Mainstream personality psychology and the study of religion. Journal of Personality, 67, 1209–1218. McCrae, R. R., & Costa, P. T., Jr. (1985). Updating Norman’s “Adequate Taxononomy”: Intelligence and personality dimensions in natural language and in questionnaires. Journal of Personality and Social Psychology, 49, 710–721. McCrae, R. R., & Costa, P. T., Jr. (1987). Validation of the ﬁve-factor model of personality across instruments and observers. Journal of Personality and Social Psychology, 52, 81–90. McCrae, R. R., & Costa, P. T., Jr. (1999). A ﬁve-factor theory of personality. In L. A. Pervin & O. P. John (Eds.), Handbook of personality (pp. 139–153). New York: Guilford Press. Mroczek, D. K., & Kolarz, C. M. (1998). The eﬀect of age on positive and negative aﬀect: A developmental perspective on happiness. Journal of Personality & Social Psychology, 75(5), 1333–1349. Murphy, P. E., Ciarrocchi, J. W., Piedmont, R. L., Cheston, S., Peyrot, M., & Fitchett, G. (2000). The relation of religious belief and practices, depression, and hopelessness in persons with clinical depression. Journal of Consulting and Clinical Psychology, 68, 1102–1106. Myers, D. (2000). The funds, friends, and faith of happy people. American Psychologist, 55, 56–67. Pargament, K. I. (2002a). Is religion nothing but . . .? Explaining religion versus explaining religion away. Psychological Inquiry, 13, 239–244. —— (2002b). The bitter and the sweet: An evaluation of the costs and beneﬁts of religiousness. Psychological Inquiry, 13, 168–181. —— (1997). The psychology of religion and coping: Theory, research, practice. New York: Guilford.

religious varieties of happiness

233

Pargament, K. I., & Mahoney, A. (2002). Spirituality: Discovering and conserving the sacred. In C. R. Snyder & S. J. Lopez (Eds.), The handbook of positive psychology (pp. 646–659). New York: Oxford University Press. Piedmont, R. L. (1999a). Does spirituality represent the sixth factor of personality? Spiritual transcendence and the ﬁve-factor model. Journal of Personality, 67, 985–1013. —— (1999b). Strategies for using the ﬁve-factor model of personality in religious research. Journal of Psychology and Theology, 27, 338–350. —— (2001). Spiritual transcendence and the scientiﬁc study of spirituality. Journal of Rehabilitation, 67, 4–14. Piedmont, R. L., Ciarrocchi, J. W., & Williams, J. E. G. (2002). A components analysis of one’s image of God. Research in the Social Scientiﬁc Study of Religion, 13, 109–123. Rodgerson, T. B., & Piedmont, R. L. (1998). Assessing the incremental validity of the Religious Problem-Solving Scale in the prediction of clergy burnout. Journal for the Scientiﬁc Study of Religion, 37, 517–527. Sloan, R. P., Bagiella, E., & Powell, T. (2001). Without a prayer: Methodological problems, ethical challenges, and misrepresentations in the study of religion, spirituality, and medicine. In T. G. Plante & A. C. Sherman (Eds.), Faith and health: Psychological perspectives (pp. 339–354). New York: Guilford Press. Smith, T. B., McCullough, M. E., & Poll, J. (2003). Religiousness and depression: Evidence for a main eﬀect and the moderating inﬂuence of stressful life events. Psychological Bulletin, 129, 614–636. Snyder, C. R., & Lopez, S. J. (Eds.). (2002). Handbook of positive psychology. Oxford: Oxford University Press. Taylor, A., & MacDonald, D. A. (1999). Religion and the ﬁve factor model of personality: An exploratory investigation using a Canadian university sample. Personality and Individual Diﬀerences, 27, 1243–1259. Van Wicklin, J. F. (1990). Conceiving and measuring ways of being religious. Journal of Psychology and Christianity, 9, 208–219. Walsh, J. M. (2001). Spirituality and recovery from pathological gambling. Unpublished doctoral dissertation, Loyola College in Maryland. Watson, D., & Clark, L. A. (1993). Behavioral disinhibition versus constraint; A dispositional perspective. In D. M. Wegner & J. W. Pennebaker (Eds.), Handbook of mental control (pp. 506–527). Upper Saddle River, NJ: Prentice Hall. Yates, J. W., Chalmer, B. J., St. James, P., Follansbee, M., & McKegney, F. P. (1981). Religion in patients with advanced cancer. Medical and Paediatric Oncology, 9, 121–128. Zinnbauer, B. J., Pargament, K. I., & Scott, A. B. (1999). The emerging meanings of religiousness and spirituality: Problems and prospects. Journal of Personality, 67, 889–919.

TEMPTATION BIAS: SEEING ONESELF AS BETTER ABLE THAN OTHERS TO RESIST TEMPTATION Sheila Garos, James K. Beggan and Annette Kluck Abstract Temptation bias is an individual’s belief that he or she is better able than others to resist temptation. We hypothesized that temptation bias is moderated by religious commitment and obsessive sexual thought as measured by the Religious Commitment Inventory—10 and the Garos Sexual Behavior Index, respectively. Female college student participants reported their expected level of temptation in twelve scenarios that involved both sexual and non-sexual content. High religious commitment and high obsessive sexual thought were associated with greater temptation bias. We interpret results in terms of the motivation to maintain self-control.

Temptation plays a prominent role in the Bible, beginning with Satan’s temptation of Eve in the Garden of Eden. As described in the Bible, temptation is the desire to do something forbidden by God. Although contemporary psychological conceptualizations of temptation can be derived from opposing theoretical orientations, they can still be understood in light of Eve’s experience. In social learning theory, Bandura (1977, 1982) deﬁned temptation as responses to environmental or internal stimuli. Psychodynamic approaches view temptation as giving in to drives, compulsions, and obsessions. Temptation is an important component of numerous models of behavior change, relapse prevention, and self-regulation (Hodgson, 1989). Scholars describe resisting temptation using diﬀerent metaphors. In his seminal text on psychology, William James (1890/1950) deﬁned it as a form of exercise that improved with practice. In contrast, Baumeister (2002) viewed resisting temptation as an eﬀortful process that fails when defensive resources are depleted. Freitas, Liberman, and Higgins (2002) saw resisting temptation as the adoption of a self-guide where desired states are expressed as hopes or aspirations. We distinguish between the act of resisting temptation and beliefs about the ability to resist temptation. We intend to assess whether people will display a temptation bias by reporting they are better than others at resisting temptation. As such, temptation bias represents a speciﬁc type of self-serving judgment. We also sought to determine Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

236

garos, beggan, and kluck

whether temptation bias is correlated with religious commitment and obsessive thought. Resisting Temptation Existing research has considered factors that inﬂuence an individual’s ability to resist temptation, i.e., to engage in self-control. In a study of ministers’ self-reports regarding yielding to sexual temptation, Balswick and Thoburn (1991) identiﬁed several factors that helped ministers control sexually related behavior. Facilitating control factors included a sex-positive marital relationship with their wives, prayer, and spiritual discipline. Threats to self-control included low self-esteem and an achievement-oriented, Type A personality. Studies of children (e.g., Coady & Sawyer, 1986) have shown that the ability to resist temptation is related to stages of moral development, with more advanced children being better able to engage in self-control. Gender eﬀects for resisting temptation are mixed. Some studies (e.g., Coady & Sawyer, 1986) indicate no eﬀect for gender. Walsh (1967) found that females are better at resisting temptation than males, whereas Eisen (1972) reported a positive correlation between self-esteem and resistance to temptation for boys but not for girls. Sexual Temptation Temptation exists whenever an individual experiences the desire to perform an action that he or she believes should not be performed. Reasons for avoiding tempting behavior range from spiritual risks to health or safety risks. Given our population base of young, unmarried adults, we expected that sexually-related impulses, and the possible need to control those impulses, are particularly salient to our participants. We included several forms of sexual temptation. The desire for sex can create a strong approach-avoidance conﬂict. The high level of physical pleasure associated with sexual activities makes people want to seek out sexual opportunities. There is also a beneﬁt associated with resisting the sexual impulse. Many religions, such as those incorporating Judeo-Christian doctrines, caution against the act of sex outside of marriage because it is seen as a sin. By abstaining,

temptation bias

237

individuals avoid the possibility of contracting sexually transmitted diseases. Concerns about AIDS have highlighted the conﬂict between the desire for sexual activity and the advantages of sexual restraint. Abstinence also prevents unwanted pregnancy. In addition to the temptation for sex, we also considered sexually related acts that we suspected would be viewed as deviant by our participants. Speciﬁcally, we included items related to anal sex and viewing pornography. Cognitions About Temptation One important, but previously unexamined, aspect of resisting temptation is the individual’s appraisal of his or her ability to resist. Temptation includes a meta-temptation component that involves understanding how people deal with the awareness that they may encounter temptation. Hodgson (1989) notes that individuals can fail to retrieve and use information relevant to resisting temptation. “If I eat this, I’ll gain weight,” will help resist temptation only if it is applied during the tempting situation. One line of research consistent with meta-temptation was developed by Trope and Fishbach (2000) in an analysis of counteractive self-control. They noted that people develop and use a variety of techniques to ensure their self-control in future situations. For example, they may employ contingencies like receiving a reward only after completing a diﬃcult task. Trope and Fischbach noted that, “People may learn that certain temptations are diﬃcult to overcome on-line and that it is necessary to bolster the value of their long-term goals in advance and irreversibly precommit themselves to achieving these goals” (p. 504). In other words, they develop thoughts about temptation. In a study of romantic couples, Johnson and Rusbult (1989) found that more committed romantic partners were more likely to devalue potential alternatives, a strategy that helps committed individuals remain in a relationship. Individuals often have to decide whether to enter into a situation where self-control is necessary. Consider, for example, a married woman asked out for an after-hours drink by a colleague she ﬁnds attractive. She might think the invitation is harmless or suspect that he has romantic intentions. She might agree to have a drink but believe that if he does make an advance, she will be able to resist.

238

garos, beggan, and kluck

When the time comes, however, she may discover she lacks the willpower to do so. She has overestimated her ability to resist temptation. Alternatively, she may have been accurate in her self-appraisal and fully able to resist, despite feeling ﬂattered by his attention. In other words, to understand temptation, we need to consider people’s beliefs about their abilities to resist temptation. On the Possibility of a Temptation Bias It makes sense that people should be able to predict their own behavior in future situations. After all, people have a great deal of experience with themselves. Prediction would be even easier with clear guidelines for behavior. For married couples or individuals involved in an exclusive relationship, the question of whether to engage in sexual relations outside the couple is readily answered (typically, in the negative). However, evidence indicates people’s judgments often contradict what would be expected by logic or their own values (see Kleinmuntz, 1990 for a review). Moreover, judgments tend to be self-serving in direction (see Taylor & Brown, 1988 for a review). For example, people see themselves as more obedient than others to biblical commandments (Rowatt, Ottenbreit, Nesselroade, & Cunningham, 2002). Self-serving biases provide a means of protecting one’s ego and self-esteem (Forsyth, 1980; Holland, Meertens, & Van Vugt, 2002; Krahe, 1984; Robins & Beer, 2001). We refer to the tendency for an individual to judge the self as better than others at resisting temptation as temptation bias. Although it is appealing to deﬁne temptation bias more simply as a tendency to overestimate one’s ability to resist temptation, this latter deﬁnition has a serious limitation. It is not easy to compare an objective measure of an individual’s ability to resist temptation with the individual’s own judgment. As a result, we deﬁned temptation bias as a judgment made in relation to others. With a suﬃciently large group, on average people should see themselves as about average in their ability to resist temptation. A deviation from average would suggest that as a whole individuals’ judgments are biased.

temptation bias

239

The Motivation for Control The motivation for control has been viewed as a fundamental human drive (Connell & Wellborn, 1991; Deci & Ryan, 1985; Harter, 1978; Skinner, 1991, 1995; White, 1959) that asserts itself early in life (Heckhausen & Schultz, 1995). It can include a desire to possess selfcontrol; i.e., to refrain from making a choice that is the most attractive in terms of immediate consequences (Kuhl & Goschke, 1994; Logue, 1988). We view temptation bias in terms of ego protection (cf. Taylor & Brown, 1988) with regard to maintaining a view of oneself as in control. Thus, we were interested in determining whether temptation bias is greater for individuals with a greater need to maintain self-control. We focused on two factors related to an individual’s need for self-control: degree of religious commitment and the presence of intrusive, sexually related thoughts. Religious Commitment and the Motivation to Resist Temptation Religion provides one source of rules or prescriptions for behavior. Religiosity can be deﬁned as an acceptance of a set of beliefs about the origin of the world and the interrelations of objects in that world (Hill et al., 1998; Pargament, 1986; Zinnbauer et al., 1997; Zinnbauer, Pargament, & Scott, 1999). Typically, the focus of the beliefs is on supernatural or spiritual explanations rather than interpretations based on the methods of science such as experimentation and logical proof. Religiosity is a multidimensional phenomenon (Thompson & Remmes, 2002). In an early study, Clark (1958) surveyed scholars interested in the scientiﬁc study of religion and reported six distinct conceptualizations of religion, such as church-related activity and customs or a person’s relationship with God. Batson, Schoenrade, and Ventis (1993) suggested that religiosity is a “quest” with a focus on ﬁnding answers to existential questions. Rowatt and Kirkpatrick (2002) viewed religiosity in terms of attachment to God, in a manner consistent with Bowlby’s (1982) attachment theory. Religiosity can also be viewed in terms of orthodoxy and fundamentalism (Rowatt et al., 2002). The complexity of religion as topic of psychological study and as an aspect of human behavior is demonstrated by the

240

garos, beggan, and kluck

large number of psychometric scales designed to measure diﬀerent aspects of religiosity (see Hill & Hood, 1999). Religiosity has been studied in terms of its eﬀects on behavior and the expression of values. Schwartz and Huismans (1995) found a positive correlation between religious beliefs and traditional values (e.g., humility, respect for tradition), and a negative correlation between religious beliefs and hedonistic values (e.g., pleasure and sensuous gratiﬁcation). In contrast, other researchers (e.g., Gunnoe & Moore, 2002) deﬁned religiosity in terms of engaging in religious behaviors. We viewed religiosity in terms of religious commitment. In keeping with Worthington (1988), we deﬁned religious commitment as adherence to religious beliefs, values, and practices as applied to daily living. In the development of a religious commitment inventory, Worthington et al. (2003) considered interpersonal or intrapersonal religious commitment. Interpersonal religious commitment refers to observable behaviors involved in the expression and practice of organized religion (e.g., “I enjoy working in the activities of my religious organization”). In contrast, intrapersonal religious commitment refers to internalized spiritual beliefs (e.g., “It is important to me to spend periods of time in private religious thought and reﬂection”). One reason to expect that high religious commitment is associated with an increased temptation bias is the desire to preserve one’s self-image as a religious person. Burris and Jackson (2000) found that individuals with a religious orientation saw themselves as more religious when they received feedback that threatened their view of themselves as religious persons. Rowatt et al. (2000) similarly found that participants rated themselves to be more obedient to biblical injunctions than they rated others (the “holier-than-thou” eﬀect). Thus, we predicted a positive correlation between level of religious commitment and temptation bias. Obsessive Thoughts and Self-Control Obsessive thoughts can be deﬁned as a set of repetitive ideas, feelings, or impulses that interfere with a person’s normal mental function and are not subject to conscious control (Clark & Purdon, 1993). Clinical literature (e.g., Lee & Kwon, 2003) suggests obsessive thoughts are ego-dystonic, that is, anxiety-provoking, undesirable, disturbing

temptation bias

241

to the individual, and often experienced as intrusive or absurd. People hold unrealistically optimistic beliefs about their abilities to control their thoughts and behavior (Clark & Purdon, 1993; 1995; Freeston, Ladouceur, Gagnon, & Thibodeau, 1993; Purdon, 1999; Rachman, 1997; Salkovskis, 1985; Salkovskis & Forrester, 2002). Evidence (e.g., Wegner, 1994), however, suggests that people are unable to exert conscious control over their thoughts. In fact, the more they try to control a thought, the more the unwanted thought intrudes on behavior. As such, obsessive thoughts represent a threat to one’s desire for self-control. Thus, we assumed that people who are unable to control their thoughts might be highly motivated to believe that they could, as a means of maintaining a belief in their own abilities. We predicted that the tendency to display a temptation bias is positively correlated with the tendency to report obsessive thoughts. Current Study Temptation bias is the tendency to see oneself as better able to resist temptation than others. We predicted that people would display a temptation bias. In addition, we predicted that temptation bias would be positively correlated with religious commitment and the desire to maintain control of one’s thoughts. We chose to limit our study to women for four reasons. On a pragmatic level, the pool from which we drew our subjects was heavily skewed toward women. Thus, it was easier to recruit women to participate. In addition, a number of researchers (e.g., Mol, 1985; Suziedaliz & Potvin, 1981; Walter & Davie, 1998) have demonstrated that women are more religious than men. As a result, we anticipated that women might be more likely to display temptation bias. Third, it seems as if society encourages a more rigid adherence to sexual restraint for women than for men. Thus, we chose to conduct our initial test of temptation bias with a population we thought more likely to display an eﬀect. Finally, a number of the scenarios in which we were interested (e.g., breast implants) were more applicable to women.

garos, beggan, and kluck

242

Method Participants One hundred seventy six female college students participated in this study. Participants were enrolled in the introductory psychology course at a large, state university in southwest Texas. A large number of students take this course which contains a mainstream overview of subjects relevant to the ﬁeld of psychology. The course requires students to participate in experiments as part of their educational experience. Although students have ﬂexibility in choosing experiments, mandatory participation increases the types of students who take part. The sign-up sheets (intentionally) contained very little background information about the experiment. The experiment was described as a study in social judgment. Because there was no forewarning of the sexual nature of some questions, it is unlikely that factors such as the participants’ views of sexuality inﬂuenced their willingness to participate. Table 1 contains demographic information for respondents. Participants were predominantly single and Caucasian. All but 15% endorsed an aﬃliation with a Judeo-Christian religion (e.g., Baptist, Methodist, Catholic). Participants ranged in age from 17 to 31, M = 18.54, SD = 2.33. The mean age of participants’ ﬁrst sexual experience was 16 (SD = 1.93). All but one participant reported being in a sexually active relationship that included intercourse. We were surprised by the high level of sexual experience of participants. On a scale of 1 (not at all religious) to 7 (extremely religious), participants rated themselves to be moderately religious (M = 4.43; SD 1.36); all participants indicated that they engaged in prayer. Frequency of attendance at religious services is reported in Table 2. Table 1. Demographic Information of Participants (N = 176) Variable Age Ethnicity Caucasian African American Hispanic Asian

18.54 (n)

SD = 2.33) (%)

128 7 28 2

72.7 4.0 15.9 1.1

temptation bias

243

Table 1 (cont.) Variable Other Missing Marital Status Single Married Divorced Lives w/signiﬁcant other Missing Sexual Orientation Heterosexual Bisexual Missing Year in School Freshman Sophomore Junior Senior Religious Aﬃliation Catholic Baptist Methodist Church of Christ Lutheran Presbyterian Protestant Eastern Jewish Other None Missing

9 2

5.1 1.1

166 4 1 4 1

94.3 2.3 0.6 2.3 .6

171 2 3

97.2 1.1 1.7

143 18 12 2

81.3 10.2 .8 1.1

35 58 22 10 9 3 4 1 1 24 7 1

19.9 33.0 12.5 5.7 5.1 1.7 2.3 .6 .6 13.6 4.0 .6

Table 2. Temptation Scenarios: Mean Scores Sex Related Scenarios Flash breasts during Spring Break Go out on a date while in a relationship Have intercourse before marriage Watch a pornographic ﬁlm to please your boyfriend Try anal sex to please your boyfriend Have unprotected sex “Hook-up” with a guy and have sex

Mean

SD

2.63 3.48 4.15

1.77 1.91 1.99

3.08 2.61 3.16 2.11

1.65 1.84 2.13 1.65

244

garos, beggan, and kluck

Table 2 (cont.) Sex Related Scenarios Non-Sex Related Scenarios Copy answers to a ﬁnal exam Leave a restaurant without paying Try drugs at a party with friends Go out with someone of diﬀerent religion Get breast implants

Mean

SD

3.86 3.87 1.73 4.28 3.11

1.84 1.99 1.40 1.68 2.15

Note: Temptation bias scores could vary from 1 to 7, with a low number indicating greater bias.

Measures The Garos Sexual Behavior Index (Garos & Stock, 1998a; 1998b) is a four factor, 70-item scale developed to assess the cognitive, aﬀective, and behavioral dimensions of sexuality. The measure is constructed along a 5-point Likert-type scale with responses ranging from 1 (strongly agree) to 5 (strongly disagree). The GSBI is reported to have good internal consistency, factorial validity, and reliability (see Garos & Stock, 1998a for psychometric data). The Discordance Subscale of the GSBI measures the extent to which a person feels shame, conﬂict, or concern about his or her sexual behavior, sexual interests, and sexuality in general (e.g., I feel uncomfortable about my sexual practices). Items on this scale were scored such that higher scores indicated greater discordance. The Permissiveness Subscale provides information about the nature of an individual’s sexual values (e.g., Birth control should be readily available to young people). Higher subscale scores indicate a more unconventional, permissive, or liberal disposition toward sex and sexuality. Lower scores reﬂect a more conservative or traditional sexual ethic. The Sexual Obsession Subscale assesses the degree of preoccupation an individual has with sex and sexual stimuli (e.g., It seems like sex is always on my mind). This scale is scored such that lower scores reﬂect greater levels of preoccupation with sex. Finally, the Sexual Stimulation subscale indicates the degree of discomfort one has with feelings of sexual arousal and exposure to sexually arousing stimuli (e.g., I feel uncomfortable when a relationship becomes sexual). Lower scores on this subscale represent less comfort.

temptation bias

245

The Religious Commitment Inventory—10 (Worthington et al., 2003) is a 10–item, 2–factor scale designed to assess an individual’s degree of religious commitment (e.g., I make ﬁnancial contributions to my religious organization). Possible response choices include: (1) not at all true of me, (2) somewhat true of me, (3) moderately true of me, (4) mostly true of me, and (5) totally true of me. The Intra personal Religious Commitment subscale concerns personal spiritual beliefs. The Interpersonal Religious Commitment subscale refers to observable behaviors of a religious nature. The two subscales have a reported internal consistency reliability of .92 and .87, respectively. The overall Religious Commitment Inventory score is computed by summing across item scores. Internal consistency reliabilities have been reported to be .93. Test-retest reliabilities have a reported range from .83 to .86. The normative mean for a general sample of adults in the United States was 26 with a SD of 12. A score of 38 or higher indicates high religious commitment. The temptation bias measure was constructed for the present study. Each scenario was a dilemma in which the respondent had to choose whether to engage in a behavior despite her reservations about doing so (each scenario is listed in Appendix A). We developed items on the basis of face valid issues for the age and background of our participants. We divided the items into sex- and non-sex-related items, as shown in the Appendix and Table 3. Using a scale ranging from 1 (much easier for me to resist than others) to 7 (much harder for me to resist than others), respondents estimated how much better than others they would be at resisting temptation. A score of 4 indicated that the participant rated his or her ability to resist temptation as equal to others. The identity of the “others” was not speciﬁed. As such, there is ambiguity about with whom participants were comparing themselves. We address this point in the discussion section. Design We used a correlational design to examine temptation bias as a function of religious commitment and obsessive sexual thought. Responses were given anonymously. Religious commitment was assessed using the Religious Commitment Inventory (RCI; Worthington et al., 2003). Obsessive sexual thought was measured using the four subscales of the Garos Sexual Behavior Index (Garos & Stock, 1998a): Discordance,

garos, beggan, and kluck

246

Sexual Obsession, Permissiveness, and Sexual Stimulation. The dependent variable was participants’ judgment about their ability, in comparison to others, to resist temptation in 12 diﬀerent scenarios. Table 3. Correlations Between the Religious Commitment Scale, Subscales of the Garos Sexual Behavior Index, and the Sex-related and Non-sex-related subscales of the Temptation measure

T (Sex) T (Non) RAS SO D P SS

T ( Sex)

T ( Non)

RAS

SO

D

P

1.00 .317** –.424** .561** .083 .334** .446**

1.00 –.223** .108 –.073 .100 .230**

1.00 –.403** .048 –.421** –.325**

1.00 .146 .386** .662**

1.00 –.183* .087

1.00 .397**

Note: T(Sex) = Sex-related temptation items; T(Non) = Non-sex related temptation items. Lower scores indicate greater temptation bias. RAS = Religious Commitment Scale. Higher scores indicate greater commitment. SO = Sexual Obsession; D = Discordance; P = Permissiveness; SS = Sexual Stimulation. These items were scaled such that high scores on SO and SS means low levels of obsession and stimulation, respectively. High scores on D and P indicate high levels of discordance and permissiveness. ** p <.001; * p < .05.

Results Descriptives The mean religious commitment score was 28.53 (SD = 10.50), suggesting the participants showed a moderate degree of religious commitment. We computed mean scores for the seven temptation bias items related to sex (M = 3.03; SD = .98) and the ﬁve items not related to sex (M = 3.37; SD = .96). A paired sample t-test showed that participants displayed a signiﬁcantly greater temptation bias for sex-related items in comparison to non-sex related items, t(173) = 3.99, p < .001. Means for each temptation bias item are presented in Table 3. Table 4 contains correlations among the four GSBI subscales, the religious commitment scale, and the two temptation bias scores. As Table 4 indicates, religious commitment is correlated with the display of temptation bias for both sexually related and non-sexually related items. There was a stronger correlation for religious commitment and sexually related temptation bias compared to non-

temptation bias

247

sexually related temptation bias. We also obtained signiﬁcant correlations with temptation bias and certain subscales of the GSBI. The least correlated subscale was discordance, which did not signiﬁcantly correlate with temptation bias for either sex or non-sex items. Signiﬁcant moderate correlations were obtained between the Sexual Obsession, Permissiveness, and Sexual Stimulation subscales of the GSBI with temptation bias for sex-related items. Only one GSBI subscale (Sexual Stimulation) was correlated with temptation bias for non-sex items. Table 4. Summary of Hierarchical Regression Analysis for Predicting Temptation Bias for Sexually-related Items from scores on the Religious Commitment Scale and GSBI Subscale Variable Step 1 RAS Step 2 RAS Sexual Obsession

SE B

b

–.43

.07

–.45**

–.23 .07

.07 .02

.24* .39**

B

Note: Rsq = .20 for Step 1 (p < .001); D Rsq = .20 for Step 2 (p < .001). RAS = Religious Commitment Scale. Sexual Obsession = subscale of the Garos Sexual Behavior Index. ** p < .001; * p < .05.

The correlations among the GSBI subscales are similar to those reported by Garos and Stock (1998a). The highest correlation was obtained between Sexual Obsession and Sexual Stimulation (r = .662). This is understandable given that an individual bothered by obsessive sexual thoughts might ﬁnd arousal a potential threat. The lowest correlation among the GSBI subscales was obtained between the Sexual Stimulation and Discordance subscales (r = –.087). This makes sense, given that individuals with a great degree of conﬂict about their sexual behaviors and desires might be uncomfortable with sexual feelings. Self-Enhancement with Regard to Judgments about Resisting Temptation As expected, we obtained a strong self-serving bias for beliefs about resisting temptation with regard to sexual situations. The temptation bias for sex-related items (M = 3.03, SD = .98) signiﬁcantly diﬀered from 4, i.e., the midpoint of the scale, t(174) = 13.09, p < .001,

248

garos, beggan, and kluck

indicating participants believed they were better at resisting temptation than others. In addition, we also obtained a temptation bias for the non-sex items (M = 3.37; SD = .96), t(173) = 8.68, p < .001. Participants believed they would be better than others at resisting temptation in non-sexual situations. Table 2 contains scores for each item. Of the twelve items, nine showed a temptation bias. Two items (having sex before marriage and copying test answers) did not show evidence of temptation bias. One item (dating someone of a diﬀerent religion) showed a reverse eﬀect, such that participants thought they were less able than others to resist the temptation to date someone of a diﬀerent religion. Religious Commitment and Judgments about Resisting Temptation We predicted that the tendency to display a temptation bias is positively correlated with religious commitment. We conducted a hierarchical regression analysis that predicted temptation bias for sexually related items from scores on religious commitment in the ﬁrst step. As expected, religious commitment accounted for a signiﬁcant percentage of variance in temptation bias, R2 = .20, F(1,159) = 39.69, p < .001. The sign of the regression coeﬃcient (B = –.23) indicated that greater religious commitment was associated with more bias. We performed a similar analysis on the non-sex items. Once again, religious commitment accounted for a signiﬁcant percentage of variance in temptation bias, R2 = .04, F(1,158) = 6.41, p < .05. The sign of the regression coeﬃcient (B = –.18) indicated that greater religious commitment was associated with more temptation bias. Although the variance accounted for was signiﬁcant, it is important to recognize that religious commitment only accounted for 4% of the variability in temptation bias with regard to non-sex items. In contrast, 20% of the variability in temptation bias with regard to sex items was accounted for by religious commitment. Thus, religious commitment was a much better predictor of temptation bias in sexual, rather than non-sexual domains.

temptation bias

249

Table 5. Summary of Hierarchical Regression Analysis for Predicting Temptation Bias for Non-sexually Related Items from scores on the Religious Commitment Scale and GSBI Subscales B

SE B

∃

RAS

–.18

.07

–.20*

RAS Sexual Stimulation

–.15 .06

.08 .03

–.16 .27*

Variable Step 1 Step 2

Note: Rsq = .04 for Step 1 (p < .05); D Rsq = .05 for Step 2 (ns). RAS = Religious Commitment Scale. Sexual Stimulation = subscale of the Garos Sexual Behavior Index. ** p < .001; * p < .05.

Table 6. Correlations Between Interpersonal and Intrapersonal Religious Commitment Subscale Scores and Temptation Bias Scenarios Sex Related Scenarios

Interpersonal Religious Commitment

Flash breasts during Spring Break Go out on a date while in a relationship Have intercourse before marriage Watch a porn ﬁlm to please your boyfriend Try anal sex to please your boyfriend Have unprotected sex “Hook-up” with a guy and have sex Non-Sex Related Scenarios Copy answers to a ﬁnal exam Leave a restaurant without paying Try drugs at a party with friends Go out with someone of diﬀerent religion Get breast implants Note: * p < .05; ** p < .001

Intrapersonal Religious Commitment

–.356**

–.376**

–.281** –.381**

–.242** –.306**

–.108

–.092

.031 –.258**

.113 –.285**

–.215**

–.213**

–.134 .022 –.219** –.130

–.093 .023 –.248** –.194*

–.149*

–.084

250

garos, beggan, and kluck Obsessive Thoughts and Judgments about Resisting Temptation

To test the eﬀects of obsessive thoughts on temptation bias independent of religious commitment, in a second step in the regression analysis, we entered the four subscales of the Garos Sexual Behavior Index. The four subscales produced a signiﬁcant increase in the variance accounted for DR2 = .20), F(4, 154) = 12.68, p < .001. The only GSBI subscale that had a signiﬁcant regression coeﬃcient was the Sexual Obsession subscale (B = .07), t(155) = 4.20, p < .001. The sign of the regression coeﬃcient indicated that a low score on the Sexual Obsession subscale of the GSBI (indicating greater degrees of sexual obsession) was associated with greater temptation bias. For the regression analysis of non-sex items, the addition of the four GSBI subscales produced a marginally signiﬁcant increase in the variance accounted for in temptation bias (increase in R2 = .05), F(4, 154) = 2.01, p = .10. The only subscale with a signiﬁcant regression coeﬃcient was Sexual Stimulation (B = .06), t(154) = 2.47, p < .05. The positive sign of the coeﬃcient indicated that participants with greater discomfort with sexual arousal reported a greater belief that they could resist temptation better than others. Subscales of the Religious Commitment Inventory (RCI) Worthington et al. (2003) noted that the RCI is composed of two distinct but correlated factors: interpersonal and intrapersonal religious commitment. We found that the two subscales were signiﬁcantly correlated (r = .80, p < .001). Although we did not make separate predictions for these two aspects of religious commitment, as an exploratory analysis, we examined the correlations between the two religious commitment subscales and the twelve temptation bias scenarios. These results are presented in Table 6. We also recomputed regression equations using the interpersonal religious commitment subscale and the intrapersonal religious commitment subscale as separate predictors. Interpersonal religious commitment was a signiﬁcant (B = –.30), t(172) = 2.98, p < .01, but intrapersonal religious commitment was not (B = –.10), t(172) = 1.01, ns. For non-sex items, neither interpersonal religious commitment (B = –.13), t(172) = 1.25, ns nor intrapersonal religious commitment predicted temptation bias (B = –.07), t(172) = .67, ns. Thus, it appears that the signiﬁcant relationship between religious commitment as measured by a unidimensional index is mostly due to the items that involve interpersonal religious commitment.

temptation bias

251

The distinction that Worthington et al. (2003) makes between interand intrapersonal religious commitment is relevant to our examination of temptation bias. Neither of the Religious Commitment Inventory subscales was a signiﬁcant predictor of non-sex related items. We thought that religious commitment would be related to both sexrelated and non-sex related forms of temptation bias, given that religious teachings contain proscriptions against a broad range of behaviors that are both sexual and non-sexual in nature (e.g., stealing, cheating). When the two subscales were combined, they did predict temptation bias for non-sex items, albeit, to a small degree. This apparent discrepancy between the unidimensional and two-factor approach is likely due to the high correlation between the two subscales (r = .80). The variance accounted for by the regression equation using both subscales included the shared variance between them. However, the regression coeﬃcients for each subscale, account for only unique variance of each variable. Alternatively, because the non-sex-related items tap a range of possible areas of temptation, they may be conceptually less focused, thus creating greater error variance. General Discussion We found strong support for temptation bias, i.e., individuals’ tendencies to see themselves as better able than others to resist temptation. Religious commitment accounted for 20% of the variability in temptation bias for sex-related behaviors. This ﬁnding is consistent with the idea that religious proscriptions have an important inﬂuence on how people try to live their lives. Speciﬁcally, the JudeoChristian background of most of our participants emphasizes sexual restraint in premarital relationships. An individual with a high degree of religious commitment is invested in a self-view as a moral person. It is likely that this moral self-image includes a belief about the importance of restraining behavior, i.e., the importance of maintaining self-control. Because our data are correlational in nature, it is not possible to make causal statements. With this caution in mind, we describe some possible, albeit tentative, explanations for why temptation bias may exist. Although we have presented our research in terms of the motivation for self-control, other characterizations are plausible. Strong religious commitment could promote temptation bias because those with strong religious beliefs may have an easier time resolving conﬂict

252

garos, beggan, and kluck

as they have readily accessible guides for behavior. A body of evidence (Busemeyer, 1985; Fox & Cooper, 1997; Lee & Kim, 2002; Stayman & Batra, 1991) suggests readily accessible information facilitates decision making. A diﬀerent, but conceptually related, explanation is that individuals high in religious commitment readily access religiousbased guidelines for behavior. Frequency of prior access could play a role in their ability to resist temptation. A goal of future research is to better understand underlying processes related to temptation bias. We have presented temptation bias in terms of resisting negative behaviors. It is important to recognize that sensitive religious people, especially Christians, put a value on performing good acts as well as restraining bad ones. An important topic for possible future research is temptation bias with regard to performing good deeds. Temptation bias may be extended to include “sins of omission” in addition to those of commission. We predict that people are motivated to see themselves as less willing than others to fail to perform good, Christian acts. Obsessive sexual thoughts also accounted for a signiﬁcant proportion of variability in the display of temptation bias for sex-related behaviors. Individuals who reported a high level of sexual preoccupation reported being better able than others to resist temptation with regard to sexual domains. This ﬁnding can be interpreted in terms of the threat that obsessive sexual thoughts pose to the integrity of the self. Obsessive thoughts are often a precursor to engaging in behavior that an individual ﬁnds unappealing, regretful, or shameful. Maintaining a belief that one is better able than others to resist temptation provides a way to safeguard against the potential harms caused by acting on unwanted sexual thoughts. The eﬀects of religious commitment and obsessive sexual thoughts on participants’ responses to temptation scenarios that did not involve sexual domains were somewhat diﬀerent. Although religious commitment was a signiﬁcant predictor, it accounted for only 4% of the variance in temptation bias. This ﬁnding is surprising because we expected that religious commitment would account for a greater proportion of the variance even for non-sexually related domains such as modesty and honesty. One explanation for this low percentage of variance accounted for may be that behaviors involving honesty and cheating are seen as being fairly low on a continuum of severity of immorality. In contrast, sexual transgressions may be automatically

temptation bias

253

perceived as serious. Therefore, it is possible that we would have obtained a stronger eﬀect for religious commitment for non-sex items deemed more serious (e.g., stealing large sums of money, lying on the witness stand in a criminal trial). Further research is required to test this hypothesis. One important discrepancy in our results concerns the eﬀects of the Sexual Obsession and Sexual Stimulation subscales of the GSBI. For sex related items, Sexual Obsession was the only signiﬁcant predictor of temptation bias. This signiﬁcant relationship is consistent with the idea that a preoccupation with sexual matters might lead to unwanted or harmful sexual behaviors. As expected, Sexual Obsession was not a signiﬁcant predictor of non-sex items. Surprisingly, however, Sexual Stimulation was a predictor of temptation bias for non-sex items, such that participants who reported greater discomfort with sexual stimulation were more likely to display temptation bias. One, albeit, post hoc interpretation for this ﬁnding is that the Sexual Stimulation items may have measured participants’ willingness to engage in risky or “naughty” behaviors, such as swimming in the nude or going without underwear. Such behaviors are risky because they carry the possibility that certain (private) body parts will be seen by others. The ﬁnding that religious commitment did not predict non-sex related items is consistent with this explanation in that the non-sex items may have been interpreted as “daring” (e.g., cheating on a test risks getting caught by the instructor) rather than immoral. The failure to accurately predict one’s future behavior, with regard to resisting temptation, can have a number of psychological implications. Perhaps the most beneﬁcial one is that the overestimation of one’s ability to resist temptation, like other self-serving biases, can help maintain high self-esteem (Taylor & Brown, 1988). In other instances, the eﬀects of unrealistic optimism about the ability to resist temptation can lead to personal hardship. Individuals with social phobia tend to avoid situations where “real” dangers may not exist. Our results speak to an opposite problem; that is, people may underestimate the danger inherent in future situations where self-control is required to oﬀset harmful consequences (e.g., the classic statement, “Well, I’ll just have one”). An individual’s belief that he or she is better able than others to resist temptation might be a self-serving bias or it might be an accurate self-appraisal. Some of our participants are probably better than

254

garos, beggan, and kluck

average at resisting temptation. Others are probably worse than average. In addition, people may vary in their beliefs about their ability to resist temptation across areas. One person might be very good at resisting eating candy but give in to the desire to drink alcohol. In the long run, however, a representative sample should report being, on average, as good as others at resisting temptation. Our participants, however, displayed a strong tendency to report being better than others at resisting temptation. Future research should compare an individual’s actual ability to resist temptation with his or her beliefs about being able to do so. It would be quite intriguing to discover that people report being best at resisting temptations that they are actually poor at resisting, a prediction that would be consistent with the self-serving motivational perspective. In addition, temptation bias can help in understanding experiences of frustration when individuals are attempting behavior change. To the extent that people overestimate their ability to resist temptation, a failure in selfcontrol, such as “falling oﬀ the wagon,” is that much more unexpected. We have approached temptation bias from the point of view that resisting temptation is positive. Under certain conditions, there may be problems with resisting temptation. One especially clear case involves people with eating disorders, where resisting the temptation (in this case to eat) becomes so well developed that people endanger their health. In a like manner, failing to resist the temptation to quit exercising can lead to sports-related injury. This study has several limitations. One is the religious background of our participants. Although virtually all participants reported a Judeo-Christian aﬃliation, temptation bias could be found across religions, though the content area of that bias might diﬀer. As noted by Schwartz and Huismans (1995), despite diﬀerences in speciﬁc religious practices or beliefs, the importance of the relationship between religious commitment and values should generalize across various religions. A second limitation is that all participants were college students. This similarity in background may have reduced the variability in their values, beliefs, and past experiences. A test of temptation bias in ethnically diverse samples, across diﬀerent socio-economic backgrounds, and with diﬀerent ages is desirable. Third, it is important to note that only females participated in the study. We chose to examine women because available evidence (e.g., Mol, 1985; Suziedaliz & Potvin, 1981; Walter & Davie, 1998) sug-

temptation bias

255

gests that women are more religious than men. In addition, women’s sexual behavior is more circumscribed than that of men (Tanenbaum, 1999). As such, we expected women to self-report a greater tendency in comparison to others to resist temptation. Future research should attempt to generalize our results to men. A fourth limitation is that we have chosen to interpret our ﬁndings from a motivational framework; i.e., the desire to see oneself in a favorable light—speciﬁcally, as having self-control. However, as with most other psychological phenomenon, it is possible to interpret what appears to be motivationally based behavior from a cognitive point of view. Temptation bias might originate from a failure to retrieve information. For example, people may be less able to recall their bad behavior than their good behavior, due to a reluctance to think about bad behavior. Alternatively, it is important to remember that our measurement of temptation bias was based on participants’ self evaluation in relation to others. As such, there may have been ambiguity with regard to how participant deﬁned “others.” A person high in religious commitment may have referred to “others” as having low religious commitment. As a result, their display of temptation bias might have reﬂected the characteristics attributed to the referent, rather than a tendency to see oneself as better than the “other.” Future studies should be conducted to assess temptation bias using other methods. Our conclusions are based on the temptation scenarios that we generated. The scenarios have face validity as possible areas of temptation for young adults who report moderate to strong levels of religious commitment. It would be fruitful to examine other forms of temptation or develop a psychometrically sound temptation bias measure. Some of our sexually related items (e.g., try anal sex) might be seen as less transgressive within a committed relationship. Although temptation bias could result from numerous sources and exist in many domains, by formally recognizing it, we hope to contribute to a better understanding of how people strive to control their behavior. References Balswick, J., & Thoburn, J. (1991). How ministers deal with sexual temptation. Pastoral Psychology, 39, 277–286. Bandura, A. (1977). Self-eﬃcacy: Towards a unifying theory of behavior change. Psychological Review, 84, 191–215.

256

garos, beggan, and kluck

Bandura, A. (1982). Self-eﬃcacy: Mechanism in human agency. American Psychologist, 37, 122–147. Batson, C. D., Schoenrade, P. A., & Ventis, W. L. (1993). Religion and the individual: A social-psychological perspective. New York: Oxford University Press. Baumeister, R. F. (2002). Yielding to temptation: Self-control failure, impulsive purchasing, and consumer behavior. Journal of Consumer Research, 28, 670–676. Bowlby, J. (1982). Attachment and loss, Vol. 1: Attachment, 2nd ed. New York: Basic Books. Burris, C. T., & Jackson, L. M. (2000). Social identity and true believer: Responses to threatened self-stereotypes among the intrinsically religious. British Journal of Social Psychology, 39, 257–278. Busemeyer, J. R. (1985). Decision making under uncertainty: A comparison of simple scalability, ﬁxed-sample, and sequential-sampling models. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 538–564. Clark, D. A., & Purdon, C. (1993). New perspectives for a cognitive theory of obsessions. Australian Psychologist, 28, 161–167. Clark, D. A., & Purdon, C. (1995). The assessment of unwanted intrusive thoughts: A review and critique of the literature. Behavior Research and Therapy, 33, 967–976. Clark, W. H. (1958). How do social scientists deﬁne religion? The Journal of Social Psychology, 47, 143–147. Coady, H. & Sawyer, D. (1986). Moral judgment, sex, and level of temptation as determinants of resistance to temptation. The Journal of Psychology, 120, 177–181. Connell, J. P., & Wellborn, J. G. (1991). Competence, autonomy and relatedness: A motivational analysis of self-system processes. In M. Gunnar & L. A. Sroufe (Eds.). Minnesota Symposium on Child Psychology, (Vol. 23, pp. 43–77). Chicago: University of Chicago Press. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum Press. Eisen, M. (1972). Characteristic self-esteem, sex and resistance to temptation. Journal of Personality and Social Psychology, 24, 68–72. Forsyth, D. R. (1980). The function of attributions. Social Psychology Quarterly, 43, 184–189. Fox, J., & Cooper, R. (1997). Cognitive processing and knowledge representation in decision making under uncertainty. In R. W. Scholz & A. C. Zimmer (Eds.), Qualitative aspects of decision making (pp. 83–106). Lengerich: Pabst Science Publishers. Freeston, M. H., Ladouceur, R., Gagnon, F., & Thibodeau, N. (1993). Beliefs about obsessional thoughts. Journal of Psychopathology and Behavioral Assessment, 15, 1–21. Freitas, A. L., Liberman, N., & Higgins, E. T. (2002). Regulatory ﬁt and resisting temptation during goal pursuit. Journal of Experimental Social Psychology, 38, 291–298. Garos, S., and Stock, W. A. (1998a). Measuring disorders of sexual frequency and control: The Garos Sexual Behavior Index. Sexual Addiction and Compulsivity: The Journal of Treatment and Prevention, 5, 159–177. —— (1998b). Investigating the Discriminant Validity and diﬀerentiating Capability of the Garos Sexual Behavior Index. Sexual Addiction and Compulsivity: The Journal of Treatment and Prevention, 5, 251–267. Gunnoe, M. L., & Moore, K. A. (2002). Predictors of religiosity among youth aged 17–22: A longitudinal study of the national survey of children. Journal for the Scientiﬁc Study of Religion, 41, 613–622. Harter, S. (1978). Eﬀectance motivation reconsidered: Toward a developmental model. Human Development, 21, 89–97. Heckhausen, J., & Schultz, R. (1995). A life-span theory of control. Psychological Review, 102, 284–304. Hill, P. C. & Hood, R. W. (1999). Measures of religiosity. Birmingham, AL.: Religious Education Press.

temptation bias

257

Hill, P. C., Pargament, K. I., Swyers, J. P., Hill, R. L., McCullough, M. E., Hood, R. W., & Baumeister, R. F. (1998). In D. B. Larson, J. P. Swyers, & M. E. McCullough (Eds.), Scientiﬁc research on spirituality and health: A consensus report (pp. 14–30). Rockville, Maryland: National Institute for Healthcare Research. Hodgson, R. J. (1989). Resisting temptation: A psychological analysis. British Journal of Addiction, 84, 251–257. Holland, R. W., Meertens, R. M., & Van Vugt, M. U. (2002). Dissonance on the road: Self-esteem as a moderator in internal and external self-justiﬁcation strategies. Personality and Social Psychology Bulletin, 28, 1713–1724. James, W. (1890/1950). Principles of psychology (Vol. 1). New York: Dover. (Original work published 1890). Johnson, D. J., Rusbult, C. E. (1989). Resisting temptation: Devaluation of alternative partners as a means of maintaining commitment in close relationships. Journal of Personality and Social Psychology, 57, 967–980. Kleinmuntz, B. (1990). Why we still use our heads instead of formulas: Toward an integrative approach. Psychological Bulletin, 107, 296–310. Krahe, B. (1984). The “self-serving bias” in attribution research: Theoretical foundations and empirical ﬁndings. Psychologische Rundschau, 35, 79–97. Kuhl, J. & Goschke, T. (1994). A theory of action control: Mental subsystems, modes of control, and volitional conﬂict-resolution strategies. In J. Kuhl & J. Beckmann (Eds.) Volition and personality: Action versus state orientation (pp. 93–124). Seattle, WA: Hogrefe & Guber. Lee, H. J., & Kwon, S. M. (2003). Two diﬀerent types of obsession: Autogenous obsessions and reactive obsessions. Behaviour Research and Therapy, 41, 11–29. Lee, J. K., & Kim, J. K. (2002). A case-based reasoning approach for building a decision model. Expert Systems: International Journal of Knowledge Engineering and Neural Networks, 19, 123–135. Logue, A. W. (1988). Research on self-control. An integrating framework. Behavioral and Brain Sciences, 11, 665–709. Mol, H. (1985). The faith of Australians. Sydney: George, Allen, and Unwin. Pargament, K. I. (1986). The psychology of religion: A clinical/community psychology perspective. Journal of Psychology and Christianity, 5, 68–72. Purdon, C. (1999). Thought suppression and psychopathology. Behaviour research and Therapy, 37, 1029–1054. Rachman, S. (1997). A cognitive theory of obsessions. Behaviour Research and Therapy, 35, 793–802. Robins, R. W., & Beer, J. S. (2001). Positive illusions about the self: Short-term beneﬁts and long-term costs. Journal of Personality and Social Psychology, 80, 340– 352. Rowatt, W. C., & Kirkpatrick. (2002). Two dimensions of attachment to God and their relation to aﬀect, religiosity, and personality constructs. Journal for the Scientiﬁc Study of Religion, 41, 637–651. Rowatt, W. C., Ottenbreit, A., Nesselroade, K. P. Jr., & Cunningham, P. A. (2002). On being holier-than-thou or humbler-than-thee: A social-psychological perspective on religiousness and humility. Journal for the Scientiﬁc Study of Religion, 41, 227–237. Salkovskis, P. M. (1985). Obsessional-compulsive problems: A cognitive-behavioral analysis. Behaviour Research and Therapy, 23, 571–583. Salkovskis, P. M., & Forrester, E. (2002). Responsibility. In R. O. Frost & G. Steketee (Eds.), Cognitive approaches to obsessions and compulsions: Theory, assessment, and treatment (pp. 45–61). Amsterdam: Pergamon/Elsevier Science, Inc. Schwartz, S. H., & Huismans, S. (1995). Value priorities and religiosity in four western religions. Social Psychology Quarterly, 58, 88–107. Skinner, E. A. (1991). Development and perceived control: A dynamic model of

258

garos, beggan, and kluck

action in context. In M. Gunnar & L. A. Sroufe (Eds.), Minnesota symposium on child psychology (Vol. 23, pp. 167–216). Chicago: University of Chicago Press. Skinner, E. A. (1995). Perceived control, motivation, and coping. Thousand Oaks, CA: Sage. Stayman, D. M., & Batra, R. (1991). Encoding and retrieval of ad aﬀect in memory. Journal of Marketing Research, 28, 232–239. Suziedelis, A. & Potvin, R. H. (1981). Sex diﬀerences in factors aﬀecting religiousness among Catholic adolescents. Journal for the Scientiﬁc Study of Religion, 20, 38–50. Tanenbaum, L. (1999). Slut! Growing up with a bad reputation. New York: Perennial. Taylor, S. E., & Brown, J. (1988). Illusion and well-being: A social-psychological perspective on mental health. Psychological Bulletin, 103, 193–210. Thompson, E. H., & Remmes, K. R. (2002). Does masculinity thwart being religious? An examination of older men’s religiousness. Journal for the Scientiﬁc Study of Religion, 41, 521–532. Trope, Y., & Fishbach, A. (2000). Counteractive self-control in overcoming temptation. Journal of Personality and Social Psychology, 79, 493–506. Walsh, R. P. (1967). Sex, age, and temptation. Psychological Reports, 21, 625–629. Walter, T. & Davie, G. (1998). The religiosity of women in the modern West. British Journal of Sociology, 49, 640–660. Wegner, D. M. (1994). Ironic processes of mental control. Psychological Review, 101, 34–52. White, R. W. (1959). Motivation reconsidered: The concept of competence. Psychological Review, 66, 297–333. Worthington, E. L. (1988). Understanding the values of religious clients: A model and its application to counseling. Journal of Counseling Psychology, 35, 166–74. Worthington, E. L., Wade, N. G., Hight, T. L., Ripley, J. S., McCullough, M. E., Berry, J. T., Berry, J. W., Schnitt, M. M., Bursley, K. H., & O’Conner, L. (2003). The Religious Commitment Inventory—10: Development, reﬁnement, and validation of a brief scale for research and counseling. Journal of Counseling Psychology, 50, 84–96. Zinnbauer, B. J., Pargament, K. I., Cole, B., Rye, M. S., Butter, E. M., Belavich, T. G., Hipp, K. M., Scott, A. B., & Kadar, J. L. (1997). Religion and spirituality: Unfuzzying the fuzzy. Journal for the Scientiﬁc Study of Religion, 36, 549–564. Zinnbauer, B. J., Pargament, K. I., & Scott, A. B. (1999). The emerging meaning of religiousness and spirituality: Problems and prospects. Journal of Personality, 67, 889–919.

Appendix: Temptation Bias Measure Directions: Below are a variety of situations that people sometimes ﬁnd themselves in. Though you may not have actually been faced with these situations, please try to imagine how you would react if the situation DID apply to you. Read each scenario. After you read a scenario, answer the question below it by circling the number that best represents your response. Remember, answer the question AS IF the situation applied to you. You had a very religious upbringing. You identify yourself as “religious” and regularly attend services and participate in church-related activities. You have always assumed you would marry someone of the same faith. In math class you meet a guy that you begin dating and like very much. One day you discover that he has very diﬀerent religious beliefs than you do. In comparison to others, how easily could you resist the temptation to continue your relationship, despite this diﬀerence in beliefs?

temptation bias 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

259 5

6

7 Much harder for me to resist than others

It is Spring Break. You and your friends decide to go to New Orleans and attend Mardis Gras. At this festivity it is common for men to encourage women to ﬂash their breasts. This usually happens during the evening festivities. Men throw strings of beads and cheer at those women who do lift up their shirts. In comparison to others, how hard would it be to resist the temptation to lift up your shirt and show your breasts? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You are at a party with a group of people you know. A few people are using a drug that you have heard about but never tried. In comparison to others, how hard would it be to resist the temptation to try the drug? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You are in a committed relationship. There is a guy in one of your classes that you think is really good-looking. Recently you have found out he has broken up with his girlfriend. One night you run into him while shopping at the mall. You strike up a conversation and it is evident he is attracted to you. He tells you that he would like to take you on a date this Friday night. In comparison to others, how hard would it be to resist the temptation to go out with him? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You ﬁnished eating a large lunch in a café. The waiter brings your ticket and you notice that you were only charged for a Coke. In comparison to others, how easily could you resist the temptation to leave without paying for the rest of your lunch? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You and your ﬁancé are planning to get married in the next six months. The two of you have engaged in kissing and heavy petting. You can tell that your ﬁancé really wants to have intercourse. Though deep down you do too, you are still unsure of what to do. In comparison to others, how hard would it be to resist the temptation to go ahead and have intercourse?

260 1 2 Much easier for for me to resist than others

garos, beggan, and kluck 3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You have been with your boyfriend for over a year. The two of you have a satisfying sexual relationship. You know that your boyfriend owns some pornographic movies. You have seen a movie like this once before and didn’t really like it. One evening your boyfriend tells you that he would really like to watch a movie while the two of you are being sexual. You want to please him, but aren’t sure about the idea. In comparison to others, how easy would it be to resist the temptation to comply with your boyfriend’s request? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You do well in school. You’re grade point average is usually above a 3.5. You discover that one of your classmates has a copy of the answers to the ﬁnal exam in your physics class and oﬀers you a copy. In comparison to others, how hard would it be to resist the temptation to take the copy? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You and your boyfriend have been together for over a year. The two of you have a satisfying sexual relationship which includes having intercourse. On a few occasions the two of you have experimented with anal stimulation, which you have both found to be highly arousing and enjoyable. From time to time your boyfriend has mentioned an interest in trying anal sex. One night while you are having intercourse he tells you he would really like to try it. In comparison to others, how hard would it be to resist the temptation to try having anal sex? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You and your boyfriend have been sexual in the past. One evening the two of you are becoming sexual. You are not taking birth control and neither you or your boyfriend have any form of protection with you. In comparison to others, how easy would it be to resist the temptation to have sex anyway? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

Overall, you are pleased with your appearance. You like your body and ﬁgure but for some time have thought you would be even more attractive and appealing if you had larger breasts. You have been wearing breast-enhancing bras (e.g., the water bra, “push-up” bras) but ﬁnd them unsatisfying because they aren’t the “real

temptation bias

261

thing.” You recently inherited a sum of money that could cover the cost of the surgery. In comparison to others, how easy would it be to resist the temptation to spend the money to have your breasts enlarged? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

You are out at a club. A guy approaches you and asks you to dance. You ﬁnd him to be very attractive and agree. The two of you end up dancing most of the evening. In comparison to others, how hard would it be to resist the temptation to “hook-up” and have sex with him that night? 1 2 Much easier for for me to resist than others

3

4 Equally able as others to resist

5

6

7 Much harder for me to resist than others

THE LOGOPLEX AS A PARADIGM FOR UNDERSTANDING SPIRITUAL TRANSCENDENCE Ralph L. Piedmont * Abstract Spiritual transcendence represents a trait-based motivational construct hypothesized to underlie individuals’ eﬀorts at creating a personal sense of meaning. Although a signiﬁcant predictor of spiritual behaviors and aspirations, spiritual transcendence alone cannot be considered a complete explanation for all types of personal meaning. Working out of the writings of Victor Frankl, the Logoplex is presented as a theoretical framework for better understanding spiritual orientations within a broader framework of personal meaning creation. Transpersonalism vs. Materialism and Relationalism vs. Intentionalism are the orthogonal dimensions proposed to deﬁne the Logoplex. In order to test this model, a sample of 185 undergraduate volunteers completed several measures deemed to be appropriate markers of these two underlying dimensions. A factor analysis provided empirical support for the proposed model. The practical implications of the Logoplex for understanding people were discussed.

Over the past several years I have been involved in basic taxonometric research on the construct of spirituality. The basic goal was to determine whether spirituality represented some unique, individualdiﬀerences variable or was merely the blend of several existing dimensions. This is a crucial question for religious researchers because of the conceptual ramiﬁcations of the answer. If spirituality is a unique dimension of psychological functioning, then constructs which tap into this domain represent new psychological phenomena that hold the potential for expanding our understanding of people and the goals they pursue. Empirically, spirituality could then improve our ability to predict important psychosocial outcomes. If, instead,

* Author Notes: Request reprints from the author at the Department of Pastoral Counseling, Loyola College in Maryland, 8890 McGaw Road, Suite 380, Columbia, MD 21045. E-mail at [email protected]. Support for this research was provided by the Institute for Religious and Psychological Research. I would like to thank Dr. Joseph Williams for his assistance in collecting the data used in this study and Dr. Joseph Ciarrocchi for his helpful comments on an earlier draft of this paper. Special thanks are extended to Rose Piedmont, Dr. William Sneck, Lynn Davies, Peter Rogers, and three anonymous reviewers for their eﬀorts at both proofreading the manuscript and providing interesting insights. Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

264

ralph l. piedmont

spirituality merely represented the admixture of one or more existing constructs, then spirituality would simply reﬂect an interesting way individuals molded internal motivations toward particular types of goals. In order to accomplish this task, a motivational/trait approach to spirituality was taken. Such an approach views spirituality as a nonspeciﬁc aﬀective force that drives, directs, and selects behaviors. As an intrinsic source of motivation, spirituality would be a relatively stable construct over time and would impel individuals toward identiﬁable goals (Emmons, 1999). Spirituality would operate in ways consistent with other motivational traits, such as extraversion, power, aﬃliation, and conscientiousness. This approach is certainly diﬀerent from how spirituality is usually viewed, such as a cognitive schema (Beit-Hallahmi, 1989; Worthington, Kurusu, McCullough, & Sandage, 1996), a way of being (Elkins, 1988), or way of understanding (Wong, 1998). It is certainly very diﬀerent from theological understandings, in some of which spirituality is construed as an end product of spiritual maturity, in others as living under a relationship or commitment to God, and in still others as a process of spiritual growth and development (e.g., Conn, 1985; Downey, 1997). The advantages of taking a trait approach were fourfold. First, it made available a comprehensive taxonomy of individual diﬀerence variables, the Five-Factor Model of Personality (FFM), as a conceptual and empirical reference point for evaluating the spiritual constructs. Second, the psychometrics for creating trait measures are well developed, as are criteria for determining reliability and validity. Third, the use of a trait model enabled a more ready integration of the resulting spiritual constructs into the broader research agenda of the social sciences. Finally, the trait approach necessitated the development of a construct that would be nondenominational. Many current spirituality scales reﬂect primarily a mainline Protestant theological orientation which compromises their generalizability to non-Christian groups. With these advantages in mind, spirituality was deﬁned as an individual’s eﬀorts to construe a broad sense of personal meaning within the light of his/her own mortality. Knowing that we are going to die, each of us needs to construct some sense of purpose and meaning for the life being led. Why am I here? What purpose does my life serve? Why should I do the things I do? These are important questions that each of us needs to answer and our responses to them

the logoplex

265

set the tempo, tone, and direction for our lives. Managing our sense of mortality is an innate task for our species (e.g., Greenberg, Solomon, & Pyszczynski, 1997), and ideally, these answers help to pull together the many disparate threads of existence into a more meaningful coherence that gives us the will to live productively. However, there are many diﬀerent types of answers to these existential questions, with some responses providing more emotional support and psychological resilience than others. One way of categorizing responses to these existential questions is along the time frame people use to understand their lives which I refer to as their “event horizon.” For example, some may perceive their lives within the immediate context they inhabit, responding to the speciﬁc needs and demands of the here and now, a relatively short event horizon. Others may view their lives as part of a speciﬁc generation or cohort, and personal meaning develops in relation to how these people view their commitments to others in their generation and those that follow them a more moderate event horizon. Finally, still others may view their lives as part of some eternal ontological pathway that involves responsibilities toward others in both the “here and now” and in the “next life,” a long event horizon. The broader the event horizon one uses to create meaning, the more stability, resilience, and personal satisfaction that is experienced. I believe that constructing meaning is a basic motivational aspect of personality; we all need to ﬁnd some sense of purpose for our lives. According to Frankl (1997), without meaning our lives deteriorate into despair and emotional maladjustment; with it, especially a long range or ultimate meaning, our lives receive depth and emotional fulﬁllment. Because spirituality is one way in which ultimate meaning is created, it is not surprising that some see it as a central organizing aspect of personality (Allport, 1950; Batson, Schoenrade, & Vensu, 1992). For me, underlying people’s eﬀorts at addressing and answering the existential questions of meaning, purpose, and relevance is a motivation I have labeled “spiritual transcendence.” Spiritual transcendence represents a universal human capacity to stand outside of one’s own immediate existence and to view life from a broader, more integrative whole. To varying degrees, we begin to realize that there is a larger meaning and purpose to life. We are single threads in a larger tapestry of existence; although small and limited, each thread makes a unique contribution to the overall texture and imprint of the woven piece.

266

ralph l. piedmont

In developing the Spiritual Transcendence Scale (STS), a consortium of theological experts from diverse faith traditions, including Buddhism, Hinduism, Quakerism, Lutheranism, Catholicism, and Judaism was assembled. This focus group was used to identify qualities of spirituality that were common to all of these faiths. The resulting items were then analyzed within the context of the FFM and it was determined that spirituality represented a unique, distinct individual-diﬀerences construct (see Piedmont, 1999a). Initial validation of the instrument showed: (1) that its factor structure (a single overall domain with three correlated facets, labeled Connectedness, Prayer Fulﬁllment, and Universality) was replicable over several samples; (2) that scores on the STS converged over both self and observer ratings; and, (3) that scores on the STS predicted important psychosocial outcomes (e.g., prosocial behavior, stress experience, interpersonal style, meaning in life) even after the predictive eﬀects of the ﬁve major personality domains were removed (Piedmont, 1999a; 2001). Thus, the STS has good incremental validity over established personality constructs. Research has also shown that the dimensions of the STS were reliable and valid across both cultural and religious contexts, including Catholics in the Philippines (Piedmont, 2004) and Hindus, Muslims, and Christians in India (Piedmont & Leach, 2002). This series of studies provided initial support for the belief that spirituality can be construed as a broad source of intrinsic motivation that is independent of existing personality constructs. It also supports the contention that spiritual transcendence represents a universal aspect of human experience. The next question to emerge from this research was how to ﬁt spiritual transcendence (ST) into a broader conceptual framework. As an independent construct, ST represents a motivational impetus to seek spiritual experiences, but it certainly should not be seen as capturing all that is involved in the process of creating personal meaning. ST provides a motivational foundation to our eﬀorts to self-transcend (Frankl, 1997), but the form, content, and direction of these strivings is no doubt moderated by other aspects of our personality. Thus one needs to consider these other qualities if a full understanding of the meaning-making experience is to be garnered. To this end, the Logoplex (literally, meaning network) has been developed. The term was selected to acknowledge its conceptual indebtedness to Frankl’s logotherapeutic model. The Logoplex is a circumplex model that provides a framework for understanding the diﬀerent

the logoplex

267

types of meanings that individuals develop for construing their lives. The model also outlines the role ST plays in the formation of these orientations. In addition, the model appreciates the inherent multidimensionality of meaning formation, including not only ST but aspects of the personality dimensions of agreeableness and extraversion. This model, which is presented below, identiﬁes two broad domains that appear to this author to underlie the diﬀerent meaning orientations individuals can develop: Transpersonalism versus Materialism and Relationalism versus Intentionalism. The Logoplex A circumplex represents a circular ordering of traits around two independent dimensions. These traits represent varying amounts of the two deﬁning constructs. Scales that are closer together in space on the circumplex are more related than scales that are farther apart. Scales that are 90 degrees apart are independent of one another, while those 180 degrees apart are negatively related. What the model illustrates is that there are multiple ways in which individuals create meaning, and that there exists certain relationships among these various outcomes. As will be seen, the most salient aspect of meaning will be the event horizon upon which it is constructed. It is the assumption of this model that the broader the frame of reference, the richer and more durable the resulting meaning (i.e., a sense of self that remains stable and satisﬁed in the face of existential challenges or hardships). Spirituality per se represents one way an individual can create meaning. Because spirituality involves some relationship with an ultimate, usually eternal, being or reality, its event horizon is the broadest and therefore is considered to provide the highest level of meaning organization. The ﬁrst major dimension of the Logoplex is Transpersonalism versus Materialism. Transpersonalism represents Frankl’s (1959) motivational construct of a “will to meaning.” It is the eﬀort to create a personal sense of meaning and purpose. There are two components to creating meaning. First, there is the recognition that “. . . life is a time gestalt, and as such becomes something whole only after the life has been completed” (Frankl, 1966, pp. 99–100). This perspective enables a view of one’s life as an emerging entity that ﬁlls a need in nature. Second, as Frankl (1966) stated, “. . . being human

268

ralph l. piedmont

is directed to something other than itself ” (p. 102). Thus, Transpersonalism allows an individual to construct meaning that transcends his or her immediate sense of time and self. In contrast, Materialism maintains a focus on the immediate and the concrete. The materialist is primarily concerned with that which can be possessed and/or consumed in the here and now. Meaning is found only in terms of mechanical, tangible operations that follow prescribed paths; there are limits in vision, purpose, and meaning. The materialist maintains a very strong self focus. The second major dimension of the Logoplex is the Relationalism versus Intentionalism dimension. The Relational orientation reﬂects a desire for involvement in, communion with, and connection to larger social groups. It is a personal response to care for larger organizations or institutions and to see that they are aﬃrmed. What Adler referred to as Gemeinschaftsgefühl, or social interest, very much embodies this orientation. In contrast, the Intentional orientation reﬂects a much more self-centered, interpersonally cynical position. Intentionalism reﬂects a desire to put the interests of the self above any other considerations; one’s own personal needs are seen as being paramount. Another way to conceive of this domain is as an individual versus group focus. These two domains comprise the Logoplex, which is presented in Figure 1. The circumplex marks oﬀ four distinct quadrants that represent the various combinations of the two deﬁning constructs. The solipsism quadrant reﬂects the Material-Intentional combination. Individuals in this quadrant are hypothesized to be very much selffocused and self-involved. Their worlds are comprised of their range of immediate experiences. Everything is about themselves. They tend to focus only on their own needs, feelings, desires, and aspirations, a very self-contained perspective. They distrust the motivations of others and tend to avoid getting involved with people. Because the event horizon for creating meaning is the most restricted, individuals in this category are prone to experience high levels of mental distress because they are easily cut oﬀ from the nurturing experiences of both other people and institutions. Their lives lack broad perspective and therefore stressful events can easily upset their lives. Secular humanism represents the Material-Relational orientation. Individuals in this quadrant are hypothesized to be very much concerned with their own immediate needs and concentrate on realities that are experienced through their senses; however, they do have a

the logoplex

269

Optimal Levels of Human Functioning TRANSPERSONALISM Spiritual Transcendence Scale Prayer Fulﬁllment

Universality

Asceticism

Connectedness

Spritualism

INTENTIONALISM

RELATIONALISM Solipsism

Secular Humanism

MATERIALISM Minimal Levels of Well-Being

Figure 1. Proposed Logoplex Model.

broader orientation that includes larger social organizations. Individuals who are committed to social institutions or to the larger social good fall into this category. They have found a place for themselves in a larger group or society and are able to ﬁnd personal satisfaction in these groups. However, as far as personal meaning goes, these people do not ﬁnd value beyond their social institutions. They carry with them a sense of civic-mindedness and social responsibility; their service to these organizations helps to maintain a sense of purpose and contributes to their on-going survival. Individuals in this quadrant may have a highly developed social ethic, and may recognize their responsibilities to not only those in their own life cohort, but to those who come after them. Their experiences represent mental health as traditionally deﬁned, the absence of any debilitating symptoms. In fact, this lower half of the circumplex represents minimum levels of well-being, and marks out the area traditionally studied by psychology. The upper half of the circumplex contains qualities of the individual that are at the center of more optimal levels of human

270

ralph l. piedmont

functioning. The two quadrants here are perhaps the least well researched in the social sciences but have the most to contribute to expanding our understanding of people and the goals they pursue. The asceticism quadrant reﬂects a Transpersonal-Intentional orientation to ﬁnding meaning. Such individuals are assumed to be quite concerned about developing a broad sense of personal meaning within a transcendent context; there is the recognition that one is a member of a larger community of believers that transects both the material and spiritual worlds. However, these individuals remain detached from involvement in larger communities and organizations. The prototypes for this quadrant are the Desert Fathers and Mothers, cloistered monks and nuns, or hermits. People in this category experience very intimate relations with a larger, Ultimate Reality, but this relationship with their God is a very personal and singular one. Although there is a broad concern and care for all of humanity, and perhaps a desire to witness this sacredness to larger communities, they create very self-contained, and perhaps insulated, social worlds. Finally, the Transpersonal-Relational quadrant deﬁnes spiritualism as most traditionally construed. These are individuals who are hypothesized to have a broad sense of transcendent meaning and develop and/or express this understanding within a deﬁned community. The community is part of their larger meaning. Unlike the ascetics, connections with others, groups, or institutions, become salient aspects of how they construe the value of their lives. Unlike the secular humanist, there is a realization that their involvement with others arises from strong teleological convictions. Individuals such as Mother Theresa of Calcutta or Mahatma Gandhi ideally represent this category. Like the ascetic, these individuals bear witness to larger, transcendent realities, but unlike the ascetic they are enmeshed in a ministry that directly bonded themselves daily to the larger secular world. I hypothesize this quadrant as representing the most durable level of personal fulﬁllment and meaning. The triangle shown in Figure 1 represents that substantive aspect of the circumplical “pie” that is captured by my STS facet scales, their conceptual range of convenience. Clearly, ST does not represent the totality of one’s spiritual orientation. Certainly other variables are needed in order to better nuance how one’s eﬀorts at ﬁnding meaning are going to be expressed (other such factors may include personality dimensions, such as agreeableness and extraversion, and religion-speciﬁc beliefs). Although the general scores on the

the logoplex

271

STS can give a sense of how a person may be oriented toward developing a sense of meaning in his or her life, without consideration of the other dimension (Relational versus Intentional), one cannot determine the speciﬁc form this transcendent motivation will take. This model provides a framework for understanding a person’s orientation to the world and speculates on the potential diﬃculties he or she is likely to experience. For example, the solipsist is believed to have a proclivity toward high levels of psychological distress. Secular humanists may risk becoming cynics, seeing that social institutions do not provide the needed solutions, or perhaps that people may not be worthy of the civic structures they have created. The ascetic may experience crises of meaning in which his or her sense of purpose may melt in the heat of personal examination. Those in the spiritual quadrant may face crises of relevancy; such individuals may question whether they as single agents can really have the impact they intend. These individuals may never lose sight of their purpose and its value, but they may become disheartened with their own abilities to bring about desired ends. Testing the Model Although there is heuristic value to this model (e.g., it is simple and direct), there is a need to determine whether this proposed structure can hold up empirically. Do these dimensions emerge in the form articulated here? In order to address this question, a study was conducted to determine whether marker scales of these dimensions, when factor analyzed, would produce this two-dimensional circular ordering. In order to accomplish this, measures were selected that putatively reﬂected various aspects of these relevant dimensions. Given that there are no scales explicitly designed to capture the dimensions of Transpersonalism and Relationalism, markers needed to be identiﬁed. The STS scales served as markers of the Transpersonal dimension. It was hypothesized that the three facets of Connectedness, Prayer Fulﬁllment, and Universality would align in the circumplex as predicted by the model. The gregariousness scale from the NEO Personality Inventory-Revised (NEO PI-R; Costa & McCrae, 1992) was used to reﬂect the Relational orientation because high scores on these scales reﬂect a desire to seek many social contacts and tend to

272

ralph l. piedmont

readily establish interpersonal relationships. The modesty scale of the NEO PI-R was selected as an indicator of the Intentional aspect of this domain; low scores on modesty reﬂect a narcissistic, self-centered, and interpersonally distant orientation toward others. As of yet, no scale has been identiﬁed to capture the Material aspect. Several additional scales were also included. First were measures of self-actualization and personal meaning in life. These scales were hypothesized to load in the Spiritual quadrant because they reﬂect active attempts on the part of an individual to ﬁnd personal relevance within a cooperative, relational framework. A measure of meaning that captured more secular interests was also included. This scale ( Just World Scale), like the earlier ones, evaluated the degree to which individuals ﬁnd a certain meaning and regularity in life. Unlike the other scales, this one centered on a more pragmatic, secular orientation (e.g., people get what they deserve in life). The Just World Scale was hypothesized to fall in the secular humanist quadrant. Finally, a peer rating of overall aﬀective well-being was also included in order to avoid a singular reliance on self-reports. It was believed those with poor well-being (higher negative aﬀect) should fall in the solipsistic quadrant. Support for the proposed circumplex model would be found if a factor analysis of these scales yielded a two factor solution which, when plotted, would position the scales in the sectors anticipated. Method Participants Participants consisted of 141 women and 44 men who were undergraduate students at a Midwestern university. All volunteered to participate and received course credit for their involvement. The average age was 19 (range 17 to 37) and 43% were Catholic, 56% were Christian, and 1% were Jewish. These participants were requested to ask two individuals who had known them for at least three months to rate them on several scales. One hundred and seventy-three returned two peer-evaluations and 12 returned one. Of the 358 raters, 128 were men and 220 were women and 10 did not indicate a gender. Overall, raters knew their targets quite well, with an average length of acquaintance of eight years (range: 3 months to 23 years). Raters also were asked to indicate how well they knew the partici-

the logoplex

273

pants on a 1 (Not really that well; casual acquaintance) to 7 (Know each other very well, close friends) Likert scale. A mean rating of 6.2 (SD = .77) was obtained, indicating that the peer raters believed themselves to be very well acquainted with the participants. Measures Spiritual Transcendence Scale (STS) Developed by Piedmont (1999a), this 24-item scale is responded to on a 1 (Strongly Disagree) to 5 (Strongly Agree) Likert-type scale. There are three subscales: Universality, a belief in the unitive nature of life; prayer fulﬁllment, feelings of joy and contentment that result from personal encounters with a transcendent reality; and Connectedness, a belief that one is part of a larger human orchestra in which one’s contribution is indispensable in creating life’s continuing harmony. Six of the items are reverse scored in order to control for acquiescence eﬀects. Alpha reliabilities for these facets are .83, .87, and .64 for Universality, Prayer Fulﬁllment, and Connectedness, respectively. These scales have shown signiﬁcant cross-observer validity (Mean r = .43) as well as correlating with a range of important psychosocial criteria, such as attitudes toward sexuality, perceived social support, vulnerability to stress, and prosocial behavior (Piedmont, 1999a; 2001). These scales have also been shown to be independent of personality as measured by the FFM and to demonstrate incremental validity in predicting outcomes over these established personality domains (see Piedmont, 1999a). Finally, these scales have been shown to capture aspects of spirituality that are cross-culturally relevant. Piedmont and Leach (2002) found the STS scales to be reliable and valid in a sample of Indian Hindus, Muslims, and Christians. Bradburn Aﬀect Balance Scale Developed by Bradburn (1969), this 20-item true-false scale captures the dimensions of positive and negative aﬀect (PAS and NAS, respectively). Studies have shown the PAS and NAS scales to represent independent dimensions (Bradburn, 1969; Costa & McCrae, 1980). Scores on these scales have been shown to correlate with ratings of global happiness (Lowenthal, Thurner, & Chiriboga, 1975) and well-being (Costa & McCrae, 1984). This scale was completed by the peer evaluators. An overall measure of aﬀect balance was computed by subtracting the Positive Aﬀect Scale from the Negative

274

ralph l. piedmont

Aﬀect Scale. Thus, high scores on this dimension indicate an overall level of negative aﬀect. The inter-rater correlation for scores on this scale was r (187) = .27, p < .01. NEO Personality Inventory-Revised (NEO PI-R) Developed by Costa and McCrae (1992), this 240-item questionnaire was developed through rational and factor analytic methods to measure the domains of the FFM. For each factor, there are six facet scales that are designed to capture more speciﬁc traits. Items are answered on a ﬁve-point scale ranging from 1 (strongly disagree) to 5 (strongly agree), and scales are balanced to control for the eﬀects of acquiescence. Normative internal consistency estimates for the selfreport version for adults range from .59 to .92 (Costa & McCrae, 1992). Scales have shown evidence of convergent and discriminant validity across instruments, methods, and observers and have been related to a number of life outcomes, including frequency of somatic complaints, ability to cope with stress, burnout, and occupational success (Costa & McCrae, 1989; Piedmont, 1993). Only the facet scales of modesty and gregariousness were used in these analyses. The modesty scale was negatively reﬂected for the analyses reported here. The Purpose in Life Test Developed by Crumbaugh (1968) this 20-item scale measures a person’s “will to meaning” as construed by Victor Frankl. Responses are given on a 7-point Likert-type scale, the poles of which vary according to the question. Guttmann (1996) reviews the research literature on this scale which provides basic validity data, documenting the scale’s ability to capture the degree to which an individual has developed a personal sense of meaning in life. Self-Actualization Scale Created by Jones and Crandall (1986), this scale provides a measure of Maslow’s highest level of development. The 15 items are responded to on a 1 (disagree) to 4 (agree) Likert-type scale. Jones and Crandall (1986) have found that high scores on this scale were associated with individuals being extraverted, rational in their thoughts and behaviors, and inner directed.

the logoplex

275

Just World Scale. Developed by Rubin and Peplau (1975) this 20-item scale is based on the “Just World Theory” that individuals get what they deserve in life and consequently deserve what they get. This philosophical orientation to life takes a more utilitarian view of justice, one with little sympathy and compassion for the ill-fate of others. Items are responded to on a 0 (Strongly Disagree) to 5 (Strongly Agree) Likert-type scale. Scores on this scale correlated positively with authoritarianism, belief in an active God, and an internal locus of control. Procedure Participants completed all of the self-report materials in groups of 10 to 35. The order of scales was randomized to control for any order eﬀects. Raters completed their materials independently. When completed, raters placed their materials in an envelope provided by the experimenter, sealed and mailed them back to the experimenter. Scores on the rater version were obtained by averaging scores across the two raters. These aggregated scores were used in all analyses. Results The intercorrelation matrix of the 9 variables used in the factor analysis is presented in Table 1 along with descriptive statistics and alpha reliabilities. Means and SDs are comparable to those found in other student samples. As can be seen, there are numerous signiﬁcant associations among the variables. It should be noted that the NEO PI-R scale modesty was negatively reﬂected so that high scores would represent the low end of this dimension (e.g., narcissism and selfcenteredness). Scores were then subjected to a principal components analysis and two factors were extracted (a scree plot indicated that only two factors should be extracted). These factors accounted for 41% of the total variance. These factors were orthogonally rotated and the factor plot of scales around the two dimensions is presented in Figure 2. All items had loadings of .30 or above on at least one of the extracted factors (with the exception of the observer rating of negative aﬀect, which loaded –.21 on the Materialism/Transpersonal dimension).

ralph l. piedmont

276

Table 1. Intercorrelation among scales used in factor analysis

1. 2. 3. 4. 5. 6. 7. 8. 9.

Gregariousness Modestya Univers Fulﬁll Connected Negative Aﬀect JWS Self-Act PILT

Mean SD Alpha

1.

2.

3.

4.

1.0 –.32 .05 .10 .11 –.05 .17 .13 .32

1.0 –.10 .01 –.27 .14 –.17 –.08 –.25

1.0 .47 .45 –.16 –.14 .13 .12

1.0 .22 –.06 .00 .13 .23

5.

6.

7.

8.

9.

1.0 –.07 1.0 –.06 –.11 .03 –.06 –.01 –.23

1.0 –.05 .19

1.0 .37

1.0

51.33 50.18 33.91 28.01 21.90 2.44 50.88 103.59 43.86 10.62 12.87 5.03 5.89 3.46 1.23 8.38 12.30 4.67 .71 .67 .82 .82 .58 .60 .52 .54 .84

N = 185. r’s ≥ |.14| are signiﬁcant at the p < .05 level, two-tailed and r’s ≥ |.18| are signiﬁcant at the p < .01 level, two-tailed. a This scale was negatively reﬂected. Note. Univers–Universality; Fulﬁll—Prayer Fulﬁllment; Connected—Connectedness; Negative aﬀect is based on peer ratings; JWS—Just World Scale; Self-Act—Self Actualization Scale; PILT—Purpose in Life Test.

As can be seen in Figure 2, the STS scales clearly deﬁne the transpersonal pole, although Prayer Fulﬁllment and Connectedness fell in the quadrants opposite to those predicted, and universality appears to deﬁne the transpersonalism pole. The NEO PI-R gregariousness scale deﬁnes the positive pole and the (low) modesty scale deﬁnes the negative pole of the Relational versus Intentional dimension, respectively. The Purpose in Life Test and the Self-Actualization Scale both fall in the Spiritualism quadrant, as predicted. These scales capture eﬀorts of individuals to ﬁnd a sense of connection, fulﬁllment, and purpose in their lives that draws on a larger sense of community and involvement. This is in contrast to the Just World Scale which loads in the Secular Humanist quadrant. High scores on this scale reﬂect a sense of meaning that stresses more of an individualistic, self-focused perspective. High scores here do not reﬂect connections with larger realities. Rather, high scores indicate a coming to terms with life as it is encountered. Finally, ratings on the Negative Aﬀect Balance Scale fall in the Solipsistic quadrant, supporting the hypothesis that individuals who maintain a very self-oriented focus are not able to construct a personal sense of meaning that can withstand the fragmenting forces of reality. As a result, individuals in

the logoplex 1.0

277

universality

Material-Transpersonal Dimension

connectedness fulﬁllment .5 selfact

pilttot gregariousness

modesty 0.0

rnegaﬀ jwstot

–.5

–1.0 –1.0

–.5

0.0

.5

1.0

Intentional-Relational Dimension Figure 2. Plotted factor analysis results for Logoplex. Note. gregariousness-NEO PI-R Gregariousness facet; jwstot-Just World Scale; rnegaﬀ-rating of negative aﬀect balance; selfact-Self Actualization Score; pilttot-Purpose in Life Test score; universality-STS Universality facet; connectedness-STS Connectedness facet; fulﬁllment-STS Prayer Fulﬁllment facet scale; modesty-NEO PI-R Modesty scale, negatively reﬂected.

this sector are easily overwhelmed and lose their own sense of direction, purpose, and well-being. Although these loadings do not show a perfect circular ordering (which is a result of their lack of precision in capturing the putative latent domains of the circumplex), in general they do load in the quadrants anticipated. Discussion Overall, these ﬁndings provide some initial support for the Logoplex. Despite only using scales designed as markers of the underlying dimensions of the model, the proposed two-factor structure was obtained. Future research now has a point of departure for further

278

ralph l. piedmont

reﬁning the model and testing its underlying hypotheses. The Logoplex can also provide a framework for developing measures of spirituality that capture diﬀerent aspects of the spiritual experience. This model may form the basis for a taxonomy of spiritual constructs opening the door for religious researchers to better understand their measures and how those measures should relate to each other (Piedmont, 1999b). Psychological Implications of the Circumplex Model It was hypothesized that the Spiritualism quadrant represented the highest level of meaning making. The integration of a transpersonal orientation with a focus on people and groups provides an opportunity for people to become completely absorbed in the human experience. As a result, individuals can create a sense of personal meaning that is durable and can withstand the pressures of extreme stressors and life shifts. This quadrant calls the individual to reach outside of him or herself, to recognize the many larger needs of the community, and to respond in a caring and committed way. The individual becomes “one” with a larger reality and ﬁnds ultimate fulﬁllment. Support for this hypothesis was found in the data, where the measures of psychological maturity and meaning were found to load in this quadrant, along with the Prayer Fulﬁllment and Universality facets from the Spiritual Transcendence Scale. However, the liability for the Spiritualist is the “crisis of relevancy.” Those high on spiritualism may come to believe that they do not have the necessary skills, abilities, or talents to meet the social needs they encounter. As a result, they may question their own sense of meaning and mission. They may develop what Maslow (1971) has referred to as meta-pathologies, which include feelings of senselessness, nihilism, cessation of striving, and a loss of feeling needed. If the Spiritualism quadrant represents the optimal orientation, then its opposite quadrant, Solipsistic, represents the least eﬀective orientation. Individuals in this condition experience high levels of psychological distress, mostly due to their inability to construct a useful, reassuring sense of personal meaning that can endure the many contradictory and fragmenting forces endemic to society (Allport, 1950). Unlike the self-enhancing virtues of the meaningful and purposeful life characteristic of those in the Spiritualism quadrant, the Solipsists experience “existential disappointment” (Tillich, 1951) or “noögenic neurosis” (Frankl, 1959) as a result of their eﬀorts to extol their own

the logoplex

279

egoistic impulses. Their glass house of meaning easily shatters and results in despair and inertia. As Emmons (1999) noted in his analysis of personal strivings, “. . . individualistic, self-oriented spiritual strivings outside of faith-based communities may not be associated with adaptive psychological, physical and interpersonal outcomes” (p. 111). That the Negative Aﬀect Scale loaded in this quadrant provides some support for this conceptualization. In short, the solipsistic quadrant represents psycho-spiritual immaturity so movement out of this quadrant, in either direction, results in higher levels of development. There are several forces which may operate to keep an individual in the solipsistic quadrant. First, Frankl (1966) has noted that a continual focus on reductionistic thinking fosters an empty existentialism. According to Frankl, a strong life is built upon strong ideas. Without such large, overarching values life collapses into the atomistic world of basic impulses and immediate gratiﬁcation. I hypothesized that another factor is narcissistic injury experienced at the hands of important emotional care takers, such as parents. Individuals hurt by life at a young age will tend to retreat into themselves and ﬁnd recourse in their own eﬀorts and strivings. An emotional rebuﬀ from a powerful transferential object will sever the person’s emotional links to others as he or she matures. Thus, the outer world of people and groups becomes too threatening and is perceived as being unable to provide the consistent positive regard and validation the child and subsequent adult requires. The secular humanism quadrant reﬂects individuals who are able to make reasonably robust emotional connections with their environments. Involvements with civic and social organizations makes these individuals vibrant, involved citizens committed to bringing about speciﬁc social ends. That the Just World Scale loaded in this dimension supported these descriptions. However, because these individuals do not have a teleological vision of the world that both celebrates their own humanity and aﬃrms their own dignity and worth, the social ends pursued can be quite varied and chillingly pragmatic. What separates this quadrant from spiritualism above it, is the immediacy of focus. The practical realities of tangible, social structures are the center of attention, possibly resulting in the development of very parochial attitudes and allegiances. Yet, individuals in this quadrant can also express a deep sense of nobility and selﬂess dedication. Patriots who, as Abraham Lincoln described in his Gettysburg Address, gave their last full measure of devotion to their country

280

ralph l. piedmont

certainly represent the best of this category. Individuals in this category can have an event horizon for construing meaning that spans generations. As such, they may aspire to create a personal legacy to reach that distance through time. Then there are the political ideologues and fanatics who work to sculpt people to ﬁt their ideal civic organization. The Fascists and Communists are those who ruthlessly pursue the creation of a durable social order. Asceticism is the opposite pole of secular humanism and is distinguished from the latter by its belief in the wholeness of nature and its movement toward ultimate realities that transcend ordinary material existence. These groups are similar in that both may have ideologies that give a direction and purpose to existence. But what separates the ascetic is the larger event horizon that deﬁnes his or her ideological agenda. The cosmology of the aesthetic concerns itself with a reality that emerges once the material experience passes away. Meaning for the ascetic begins where secular humanism ends. Another important diﬀerence between these two groups is the more solitary or detached existence of the ascetic. He or she does not share the same absorption in larger social groups that characterizes both the secular humanist and the spiritualist. The ascetic feels him or herself in the world but not of it. Further, the prototypical ascetic evidences a single minded eﬀort to ﬁnd an individual connection with the divine or ultimate reality. Nonetheless, the transcendent, eschatological nature of the ascetic imbues him or her with a broader, more integrated and durable sense of self. Spiritual transcendence is hypothesized to increase with age. Age brings with it the growing inevitability of death and thus the pressing need for the individual to construct a sense of meaning and purpose. This change is usually expressed in concerns about creating a legacy, mentoring, and ﬁnding personal closure. These tasks coincide with Erikson’s (1950) notions of generativity and ego integrity, the ﬁnal two psychosocial stages of life. Individuals can only come to see the inevitability of their own lives and their place in the greater scheme of life by adopting a transcendent orientation. Piedmont (1999a) demonstrated a signiﬁcant age eﬀect on his connectedness and universality facet scales. Therefore the natural movement of growth over time is away from the self-involved impulses of the secular humanist and solipsist toward the broader, transcendent goals of the spiritualist and ascetic. However, when growth is either delayed or preempted the Logoplex can provide a conceptual paradigm for

the logoplex

281

constructing intervention strategies. For example, techniques aimed at developing a sense of mindfulness, gratitude, and self-acceptance could be used to help individuals move toward understanding their lives from a spiritual perspective. The dimensions of Transpersonalism and Relationalism are the guideposts for signaling directions for growth. From a clinical perspective, the Logoplex can help counselors identify the ways in which clients are constructing their sense of personal meaning. As the data from this study indicated, being solipsistic, ascetic, spiritual, or humanistic carries with it particular styles of interacting with the world, along with their own potential weaknesses. But more than just providing insight into clients, the Logoplex can also serve as a focus for making interventions aimed at promoting a more inclusive meaningmaking style. Limitations and Conclusions This study provides some very encouraging results supporting the empirical utility of the Logoplex. However, there are some issues that need to be considered in both evaluating these data and in designing future studies. First, the scales that were included for analysis were not ideal markers for each pole of the circumplex. Instead, possible markers were selected. Although mostly useful, it needs to be pointed out that the included scales did not organize themselves in the type of circular arrangement that a circumplex demands. Perhaps if scales were used that better reﬂected the dynamics outlined by the circumplex, a better pattern of results may have been found. Second, for some scales rather low estimates of internal reliability may have also compromised the analyses. Third, only the Connectedness facet scale loaded within the Asceticism quadrant (and this was not theoretically expected), thus little is learned about the motivations that characterize individuals here. Similarly, the loading of negative aﬀect in the Solipsism quadrant was not extreme enough to really deﬁne individuals here as well. Thus, more information is needed for testing hypotheses relating to these two dimensions. Finally, undergraduates may not be an ideal sample for this type of study. Young adults may have had too few spiritual experiences to be representative of all four quadrants. Without the inclusion of a full age range of adults, there may be a real restriction of range in scores for several scales (e.g., self-actualization scale).

282

ralph l. piedmont

The next steps in this process would be to develop scales that speciﬁcally assess various aspects of the circumplex. There are speciﬁc statistical criteria that can be employed for developing and testing circumplex models (e.g., Browne, 1992). It should also be recognized that although spiritualism is considered to represent the most durable meaning structure, this does not mean one has to believe in “God” to get there. What is the most salient aspect of the Logoplex is the event horizon within which a person places his/her meaning. The broader the event horizon (e.g., moving from the immediacy of one’s needs, to the recognition of others across generations, to seeing life as an eternal pathway), the more satisfying, genuine, and lasting the meaning one makes. However, not being in the spiritual or ascetic categories does not diminish the reality that such individuals can create a sense of personal meaning that is psychologically adaptive. Nor should this paper be construed as implying that satisfying and fulﬁlling lives cannot be found among those in the Secular Humanist, or even Solipsist, quadrants. The purpose of the Logoplex is not to prescribe what meaning and values a person should have. Rather, its goal is to heighten our awareness of the kinds of meaning one can have. Ultimately, as Frankl (1997) noted, it is up to the individual him or herself to determine what is of value and what is not of value; of what is good and what is bad. Nonetheless, the Logoplex provides a useful paradigm for understanding a wide variety of meaning orientations. The two dimensions of Transpersonalism versus Materialism and Relationalism versus Intentionalism deﬁne a conceptual space for appreciating the types of meaning individuals create for their lives and for understanding the goals people seek. The data reviewed here provided some empirical support for this conceptualization. This circumplex can serve as a beginning eﬀort in the development of a comprehensive taxonomy of meaningrelated constructs. The need for such a classiﬁcatory schema becomes critical given the plethora of psycho-theological constructs already in existence. The construct validity of many of these instruments is not well established (Gorsuch, 1988; Hall, Tisdale, & Brokaw, 1994). Locating these scales within the Logoplex can accomplish two goals. First, it can help identify the aspects of spirituality reﬂected in the content of these scales, and second it can identify areas of uniqueness and redundancy in measurement among the instruments (see Piedmont, 1999b). Developing spiritual concerns or growth entails a consideration of issues that are not traditionally entertained in main-

the logoplex

283

line psychotherapeutic circles and thus go beyond traditional notions of well-being and mental health. The Logoplex provides an outline of the relevant motivations that need to be considered in conceptualizing these types of issues. References Allport, G. W. (1950). The individual and his religion. New York: Macmillan. Batson, C. D., Schoenrade, P., & Vensu, W. L. (1992). Religion and the individual: A socio-psychological perspective. New York: Oxford University Press. Beit-Hallahmi, B. (1989). Prolegomena to the psychological study of religion. Lewisburg, PA: Bucknell University Press. Bradburn, N. M. (1969). The structure of psychological well-being. Chicago, IL: Aldine. Browne, M. W. (1992). Circumplex models for correlation matrices. Psychometrika, 57, 469–497. Conn, J. W. (1985). Spirituality and personal maturity. In R. J. Wicks, R. D. Parsons, & D. Capps (Eds.), Clinical handbook of pastoral counseling Vol. 1 (pp. 37–57). Mahwah, NJ: Paulist Press. Costa, P. T., Jr., & McCrae, R. R. (1980). Inﬂuence of extraversion and neuroticism on subjective well-being: Happy and unhappy people. Journal of Personality and Social Psychology, 38, 668–678. Costa, P. T., Jr., & McCrae, R. R. (1984). Personality as a lifelong determinant of well-being. In C. Z. Malatesta & C. E. Izard (Eds.), Aﬀective processes in adult development and aging. (pp. 267–283). Beverly Hills, CA: Sage. Costa, P. T., Jr., & McCrae, R. R. (1992). Revised NEO Personality Inventory: Professional manual. Odessa, FL: Psychological Assessment Resources. Crumbaugh, J. (1968). Purpose-in-life test. Journal of Individual Psychology, 24, 74–81. Downey, M. (1997). Understanding Christian spirituality. Mahwah, NJ: Paulist Press. Elkins, D. N. (1988). Towards a humanistic-phenomenological spirituality: Deﬁnition, description, and measurement. Journal of Humanistic Psychology, 28, 5–18. Emmons, R. A. (1999). The psychology of ultimate concerns: Motivation and spirituality in personality. New York: Guilford Press. Erikson, E. H. (1950). Childhood and society (2nd ed.). New York: Norton. Frankl, V. E. (1959). From death-camp to existentialism. Boston, MA: Beacon Press. —— (1966). Self-transcendence as a human phenomenon. Journal of Humanistic Psychology, 6, 97–106. —— (1997). Man’s search for ultimate meaning. New York: Insight Books. Gorsuch, R. L. (1988). Psychology of religion. Annual Review of Psychology, 39, 201–221. Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror Management Theory of self-esteem and cultural world views: Empirical assessments and conceptual reﬁnements. In M. P. Zanna (Ed.), Advances in experimental social psychology (pp. 61–139). San Diego, CA: Academic Press. Guttmann, D. (1996). Logotherapy for the helping professional. New York: Springer Publishing. Hall, T. W., Tisdale, T. C., & Brokaw, B. F. (1994). Assessment of religious dimensions in Christian clients: A review of selected instruments for research and clinical use. Journal of Psychology and Theology, 22, 395–421. Jones, A., & Crandall, R. (1986). Validation of a short index of self-actualization. Personality and Social Psychology Bulletin, 12, 63–73. Lowenthal, M. F., Thurner, M., & Chiriboga, D. (1975). Four stages of life. San Francisco, CA: Jossey-Bass.

284

ralph l. piedmont

Maslow, A. H. (1971). The farther reaches of human nature. New York: Viking. Piedmont, R. L. (1993). A longitudinal analysis of burnout in the health care setting: The role of personal dispositions. Journal of Personality Assessment, 61, 457–473. —— (1999a). Does spirituality represent the sixth factor of personality: Spiritual Transcendence and the ﬁve-factor model. Journal of Personality, 67, 985–1013. —— (1999b). Strategies for using the ﬁve-factor model of personality in religious research. Journal of Psychology and Theology, 27, 338–350. —— (2001). Spiritual Transcendence and the scientiﬁc study of spirituality. Journal of Rehabilitation, 67(1), 4–14. —— (2004). Doing cross-cultural research with spirituality: The Spiritual Transcendence Scale in the Philippines. Paper presented at the 2nd Annual Mid Winter Conference on Research in Spirituality and Religion. Columbia, MD. Piedmont, R. L., & Leach, M. M. (2002). Cross-cultural generalizability of the Spiritual Transcendence Scale in India: Spirituality as a universal aspect of human experience. American Behavioral Scientist, 45, 1886–1899. Rubin, Z., & Peplau, L. A. (1975). Who believes in a just world? Journal of Social Issues, 31, 65–89. Tillich, P. (1951). Systematic theology I. Chicago, IL: University of Chicago Press. Wong, P. T. P. (1998). Meaning centered counseling. In P. T. T. Wong & P. S. Fry (Eds.), The human quest for meaning (pp. 395–435). Mahwah, NJ: Erlbaum. Worthington, E. L., Kurusu, T. A., McCullough, M., & Sandage, S. J. (1996). Empirical research on religion and psychotherapeutic processes and outcomes: A 10-year review and research prospectus. Psychological Bulletin, 119, 448–487.

AUTHORS’ BIOGRAPHIES Amy Adamczyk is a doctoral student in sociology at the Pennsylvania State University where she also works as a research associate with the American Religion Data Archive (www.thearda.com). She has published in the Journal of Historical Sociology and has a forthcoming article with John Wybraniec and Roger Finke in the Journal of Church and State. Currently, she is writing her dissertation, which examines the contextual inﬂuences of interrelated moral communities on adolescents’ deviant behaviors. John Armstrong is a Catholic priest and currently the administrator at St. Christopher’s Cathedral in Canberra, Australia. He has recently completed a doctorate at the Australian National University in Canberra in 2001. His research examined changing patterns of church attendance in Australia between 1966 and 1996. He received a master’s in Pastoral Ministry and Spiritual Direction in 2003 at the Australian Catholic University in Canberra. He is involved in providing chaplaincy at primary and secondary school levels. He also has the responsibility of being the director of clergy life and ministry for the Archdiocese of Canberra and Goulburn. He has a particular interest in looking at changing patterns of relationships and their aﬀect on the spiritual lives of people he works with. James Beggan received his BA in psychology from the State University of New York at Buﬀalo. He received his PhD in psychology from the University of California, Santa Barbara. He is currently an associate professor of sociology at the University of Louisville in Kentucky. His research involves the study of the representation of gender and sexual stereotypes in mass media. He teaches courses on the sociology of the self. He also teaches statistics at the undergraduate and graduate levels. John J. Cecero is an assistant professor of psychology at Fordham University, where he teaches undergraduate and graduate courses in the clinical psychology PhD program. A Jesuit priest and clinical psychologist, he is interested in the role of spirituality and mental Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

286

authors’ biographies

health. He is the director of the Center for Spirituality and Mental Health at Fordham, and maintains a part-time clinical practice at the Cognitive Therapy Center of New York. He recently published a book on the integration of spirituality and cognitive therapy techniques, Praying through our lifetraps: A psycho-spiritual path to freedom (2002). Richard Christopherson is a professor of sociology at Azusa Paciﬁc University. His teaching and research interests include the sociology of religion, occupations and professions, and cultural studies. He has published articles and book chapters on topics such as photography and photographers, liberal arts education, and Christian ministry. Joseph W. Ciarrocchi, Ph.D., is professor and chairperson of the Graduate Programs in Pastoral Counseling, Loyola College in Maryland. He is a clinical psychologist whose research interests are in the integration of spirituality and psychology. His clinical areas are addictive behavior and the anxiety disorders. His most recent book is Counseling Problem Gamblers: A Self-regulation Manual for Individual and Family Therapy (2002), Academic Press. Erin Deneke is a doctoral student in the pastoral counseling program at Loyola College in Maryland. Her research interests are in the area of the integration of psychology and spirituality. Sean Everton is currently a Ph.D. candidate in sociology at Stanford University. He is also a part-time instructor at Santa Clara University. In addition to the sociology of religion, his interests include economic sociology, social network analysis and political sociology. His dissertation research is focusing on the role of organization prestige within the venture capital industry. Stephen Joseph Fichter plans to pursue a Ph.D. in the sociology of religion at Rutgers University. He earned his MSW from Fordham and studied spirituality, philosophy, and theology in Europe for 12 years in preparation for his priestly ordination. He currently serves at Saint Gabriel’s Parish in Saddle River, New Jersey while teaching philosophy and religion at Bergen Community College. He is the grandnephew of Joseph Fichter, Jesuit pioneer in the sociology of religion. stephen_ﬁ[email protected]

authors’ biographies

287

Leslie J. Francis is professor of practical theology at the University of Wales, Bangor, United Kingdom. He received his PhD and Scd from the University of Cambridge, and his DD from the University of Oxford. He is an Anglican priest and fellow of the British Psychological Society. His recent books include: Gone but not forgotten: Church leaving and returning (1998), The Long diaconate: Women deacons and the delayed journey to priesthood (1999), Exploring Luke’s Gospel: Personality type and scripture (2000), The values debate: Listening to the pupils (2001). Psychological perspectives on prayer: A reader (2001), Exploring Mark’s Gospel: Personality type and scripture (2002), The naked parish priest: What priests really think they are doing (2003). Sheila Garos received her Ph.D. in counseling psychology from Arizona State University. She is currently an assistant professor of psychology at Texas Tech University in Lubbock Texas. Her research focus has been the development of the Garos Sexual Behavior Index and socio-sexual behavior. She teaches courses on ethics, sexuality, group counseling and psychopharmacology. She also maintains a private practice specializing in the treatment of sexual disorders, dysfunctions, addictive behaviors, and marital therapy. Brendan Geary, Ph.D., is a Marist brother, and a recent graduate of the pastoral counseling program at Loyola College in Maryland. He was recently a therapist at the National Institute for the Study, Prevention, and Treatment of Sexual Trauma (Baltimore, Maryland) where he worked with victims and perpetrators of sexual abuse. Chris J. Jackson is a senior lecturer in psychology at the University of Queensland. He did his degrees, masters and PhD, in the UK before moving to Australia in 2000. He is a member of the Australian and British Psychological Societies and is registered in Australia and the UK. His research interests concern the measurement and structure of personality and how it can be applied to the prediction of real-world behaviour. At present, he is most interested in the biosocial approach to personality and laterality of personality. Chris understands the problems organizations face in selection and appraisal and has much experience of consultancy in Australia and the UK. Susan H. Jones is research associate of the Welsh National Centre for Religious Education, University of Wales, Bangor, United Kingdom.

288

authors’ biographies

She received her PhD from University of Wales, Bangor. She has written a number of articles in the ﬁeld of the psychology of religion, with special interest in clergy personality. She is an Anglican priest, Director of the Ministry Course at Bangor, and team vicar within the parish of Bangor. Annette Kluck is a doctoral student in the counseling psychology program at Texas Tech University where she received her MA in psychology. Her research interests include eating disorders and sociosexual behavior. Rachel MacNair graduated from Earlham College, a Quaker school, with a degree in peace and conﬂict studies in 1978. After many years of working in activist concerns, she went to the University of Missouri in Kansas City to get her PhD in psychology and sociology. Her books include: Perpetration-induced traumatic stress: The psychological consequences of killing (2002), The psychology of peace: An introduction (2003), both published by Praeger, and Gaining mind of peace: Why violence happens and how to stop it (Xlibris, 2003), which explains psychology to secondary school students. David O. Moberg is sociology professor emeritus at Marquette University and has been the co-editor of Research in the Social Scientiﬁc Study of Religion since its beginning. He is a member of the American Sociological Association, Association for the Sociology of Religion, Gerontological Society of American, International Sociological Association and its research committee on the sociology of religion, Religious Research Association, Society for the Scientiﬁc Study of Religion, and other professional societies, in several of which he has held signiﬁcant leadership positions. The most recent of his hundreds of publications are “Assessing and Measuring Spirituality: Confronting Dilemmas of Universal and Particular Evaluative Criteria,” Journal of Adult Development, 9(1):47–60, 2002; “Forum: Religion and Spirituality,” Social Compass, 49(1):133–138, 2002, and Aging and Spirituality: Spiritual Dimensions of Aging Theory, Research, Practice, and Policy (Haworth Press, 2001). Ralph L. Piedmont is professor of pastoral counseling at Loyola College in Maryland as well as the director of doctoral research. He is co-editor of Research in the Social Scientiﬁc Study of Religion as well as

authors’ biographies

289

a consulting editor for Measurement and Evaluation in Counseling and Development and the Journal of Personality Assessment. His research interests include the Five-Factor Model of Personality and its relationship to spiritual phenomena, and the impact of spiritual transcendence on mental and physical health outcomes. Christian T. Racine received his BA from Boston College in 1999. His senior thesis examined relationships between attachment theory and interpersonal dominance behavior. He is beginning his ﬁfth year in Fordham University’s doctoral clinical psychology program. He is currently analysing data for his dissertation examining the moderating eﬀects of religious coping and will begin a clinical internship at Long Island Jewish Medical Center in the fall of 2004. Christopher F. J. Rutledge is research associate at the Welsh National Centre for Religious Education, University of Wales, Bangor, United Kingdom. He received his PhD from University of Wales, Lampeter. He has written a number of articles on burnout among the clergy, clergy personality, and clergy roles. He is an Anglican priest, canon, and prebendary of Salisbury Cathedral and vicar of Talbot Green. N.J. Scheers, Ph.D., is director of the Oﬃce of Planning and Evaluation and a senior statistician at the U.S. Consumer Product Safety Commission. Her research interests include identifying risk factors for sudden infant death, and she has reported and published her work in outlets that include international conferences and various pediatric journals. She is an adjunct faculty member at Loyola College, Maryland, serves as coordinator of Ph.D. dissertations, and provides methodological guidance to students.

MANUSCRIPT REVIEWERS

In addition to the preceding authors, many scholars and scientists have made signiﬁcant contributions to the publication of this volume. They have anonymously screened the initial versions of these and other submitted manuscripts for methodological rigor and scientiﬁc signiﬁcance. They also have provided the authors with invaluable suggestions for improving their papers prior to the authors making their ﬁnal revisions. We, as editors, and the respective authors appreciate their expert services. Not only have they improved the quality of the research reported here, but they have contributed to the quality of the ﬁeld of the social scientiﬁc study of religion. Joseph W. Ciarrocchi Karel Dobbelaere Gabriel S. Dy-Liacco Leslie J. Francis Kevin Gillespie Jonathon Golden Janet M. Goodman

Loek Halman Mark Leach Nichole Murray-Swank Robert Nelson Thorleif Pettersson Thomas E. Rodgerson Martin Sherman

Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2005

INDEX OF NAMES Abu-Hilal, M. M., 75, 89 Acock, A. C., 52, 68 Adamczyk, A., 51, 285 Addams, Jane, 18, 34 Adler, 268 Albert, M., 100, 114 Albott, 104 Allen, G. A., 169, 186 Allen, H. E., 72, 93 Allie, S. M., 79, 89 Allport, G. W., 209, 214, 230, 265, 283 Altman, B., 39 American Anglican Council, 189, 207 American Psychiatric Association, 18, 29, 37, 104, 113, 170, 185 Ammerman, N. T., 142, 165, 206, 207 Anagnostopoulos, F., 78, 92 Anderson, K. G., 5, 16 Archer, R. P., 5, 13 Argyle, M., 212, 215, 223, 230 Armstrong, J. M., 41, 45–46, 48–49, 285 Aronowitz, S., 51, 68 Ashforth, B. E., 78, 86, 91 Aviad, J. O., 165 Axinn, W. G., 52, 69 Bachelor, P., 78, 90 Bagiella, E., 215, 218, 233 Bahr, H., 98, 113 Bailey, M., 124, 135 Bailey, R. W., 124, 135 Bainbridge, W. S., 199, 208 Bakker, A. G., 75, 77, 89, 93 Ball, S., 95–96, 101, 113 Balswick, J., 236, 255 Bandura, A., 29, 37, 235, 255–256 Banks, S. R., 86, 89 Barbanelli, C., 29, 37 Barnes, G., 113 Barnett, C. W., 86, 91 Baron, R. M., 108, 113 Barrett, P., 2–3, 5–6, 13 Bartkowski, J. P., 53, 69 Bartz, C., 86, 89 Batra, R., 252, 258

Batson, C. D., 239, 256, 265, 283 Baum, B. H., 72, 89 Baumeister, R. F., 181, 185, 210, 228, 230, 235, 256–257 Beck, A. T., 96, 113 Becker, G. S., 191, 207 Becker, P. E., 139, 165 Beckmann, J., 257 Beckwith, B. P., 66, 68 Bedard, J., 186 Beemsterboer, J., 72, 89 Beer, J. S., 238, 257 Begg, C. B., 196, 207 Beggan, J. K., 235, 285 Beit-Hallahmi, B., 264, 283 Belavich, T. G., 100, 113, 258 Belcastro, P. A., 74, 78–89 Bellah, R., 138–139, 152, 165 Bengtson, V. L., 52, 54, 68–69 Benson, P. L., 172, 185 Bergin, A. E., 227, 230 Bernstein, D. P., 114 Berry, J. T., 258 Berry, J. W., 258 Berwick, K. R., 76, 89 Bevis, C., 52, 69 Bibby, R. W., 141n, 165 Bijur, P., 99, 113 Birk, J. M., 117, 123, 127, 135 Black, C. D., 75, 91 Blackmon, R. A., 124, 135 Blass, T., 30, 37 Blazer, D., 113 Bolton, B., 79, 89 Boswell, J., 204, 207 Bowlby, J., 239, 256 Bowman, P., 99, 114 Boyd, H., 39 Bradburn, N. M., 273, 283 Bradford Wilcox, W., 41, 49 Bremner, J. D., 31, 37 Brennan, W., 29, 38 Brenninkmeyer, V., 77, 89 Brief, A. P., 76, 91 Britton, P., 77, 90 Brokaw, B. F., 282–283 Bronen, R. A., 37 Brookings, J. B., 79, 89

Research in the Social Scientiﬁc Study of Religion, Volume 15 © Koninklijke Brill NV, Leiden, 2004

294

index of names

Brooks, C., 42, 49 Brouwer, J., 77, 89 Brown, C. E., 79, 89 Brown, G., 96, 103, 112–113, 115 Brown, J., 238, 253, 258 Brown, L. B., 5, 14, 65, 69, 81, 90 Browne, M. W., 282–283 Burke, R. J., 75–76, 89, 91, 93 Burns, R. C., 72, 89 Burris, C. T., 240, 256 Bursley, K. H., 258 Busemeyer, J. R., 252, 256 Butter, E. M., 258 Buunk, B. P., 75, 77, 89 Byrne, B. M., 78–79, 89 Cahoon, A. R., 78, 93 Cain, 104 Calhoun, G., 114 Campbell, L., 114 Capel, S. A., 75, 88–89 Caplowitz, D., 51, 65, 68 Capps, D., 283 Caprara, G. V., 29, 37 Carine, B. E., 95, 101, 113 Carlson, 104 Carnes, P., 169, 185 Carroll, J. F. X., 169, 185 Carson, V., 169, 186 Carter, J. D., 79, 93 Carver, C. S., 213, 229, 230 Castles, I., 49 Cecero, J. J., 95–96, 101, 113–114, 285 Ceslowitz, S. B., 75, 89 Chalmer, B. J., 216, 233 Charlton, J., 118, 120–121, 135 Charney, D. S., 39 Charry, E. T., 225, 231 Chartrand, S., 186 Chaves, M., 121, 135 Chen, C. Y., 78, 90 Cheston, S., 232 Chiriboga, D. E., 273, 283 Christie, D. J., 38 Christopherson, R. W., 137, 165, 286 Ciarrocchi, J. W., 167, 181–182, 185, 209, 215, 217, 226–227, 229, 231–233, 263n, 286 Clark, D. A., 240–241, 256 Clark, L. A., 171, 187, 216–217, 225, 231, 233 Clark, W. H., 239, 256 Clydesdale, T. T., 58, 69

Coady, H, 236, 256 Cody, S. R., 125, 136 Cohen, H., 99, 113 Colarusso, C. A., 38 Cole, B., 258 Confessing Movement, 207 Conn, J. W., 264, 283 Connell, J. P., 239, 256 Conway, L. G., 30, 38 Cook, B. B., 86, 89 Cooley, E., 75, 93 Cooper, H., 168–169, 186 Cooper, R., 252, 256 Corcoran, K. J., 79, 89 Cordes, C. L., 72, 78, 89 Coristine, M., 101, 103, 114 Costa, P. T., Jr., 87, 89, 172, 185, 219, 220–221, 232, 271, 273–274, 283 Cotter, T., 13, 15 Cox, M. K., 78, 93 Crandall, R., 274, 283 Cronbach, L. J., 7, 13, 75, 81, 90, 175 Crowe, P., 100, 114 Crumbaugh, J., 274, 283 Cullman, E., 100, 114 Cunningham, P. A., 238, 257 Cunningham, V. J., 38 Curtin, L., 95–96, 101, 113 D’Andrea, J., 95–96, 113 D’Aquili, E., 30, 39 Dagg, P., 101, 103, 114 Dagher, A., 38 Daley, M. R., 72, 90 Darnell, A., 51, 53, 69 Dateline NBC, 26, 38 Davie, G., 241, 254, 258 Davies, L., 263n Davis, J. A., 55n, 69, 193–194, 207 De Jonge, J., 78, 90 De Rijk, A. E., 78, 90 DeBruyn, E. J., 5, 16 Decdouvelaere, F., 95, 113 Deci, E. L., 239, 256 Demerath, N. J., III, 52, 69 Demerouti, E., 77, 89 Deneke, E., 209, 229, 231–232, 286 DeNeve, K. M., 168–169, 186 Derecho, D. V., 39 Deutch, A. Y., 39 DeVaus, D., 58, 69

index of names Dewe, P. J., 117, 123, 135 Diener, E., 169, 171, 186–187, 215, 216, 219, 230–231 Dillon, D., 168, 186 DiPasquale, R., 113 Dolan, N., 72, 76, 90 Donnenwerth, G., 42, 49 Donohue, M. J., 172, 185 Doohan, H., 72, 90 Dougherty, T. W., 72, 78, 89 Douvan, E., 99, 114 Downey, M., 264, 283 Drakeford, J., 124, 135 Dubow, E., 100, 114 Dunham, C. C., 52, 54, 69 Dunn, J., 101, 103, 114 Durup, J., 77, 92 Earle, R. H., 168, 186 Ellerby, L., 168, 180, 186 Edgar, D. E., 49 Einsiedel, A. A., 72, 90 Eisen, M., 236, 256 Elkins, D. E., 5, 13 Elkins, D. N., 264, 283 Ellenhorn, L., 21, 39 Ellis, A., 209, 231 Ellison, C. G., 53, 64, 69, 180–181, 186, 232 Emerson, M. O., 206, 208 Emmons, R. A., 171, 183, 186, 213–215, 219, 229, 231, 264, 278, 283 Erickson, J. A., 172, 185 Erikson, E. H., 52, 69, 280, 283 Everton, S. F., 189, 286 Eysenck, H. J., 1–3, 5–6, 10, 13, 80, 81, 87, 90 Eysenck, M. W., 1, 13, 80, 90 Eysenck, S. B. G., 2, 5, 13, 81, 87, 90 Fairbank, J. A., 38 Farber, B. A., 76, 90 Faulkner, B., 124, 135 Feigelman, W., 51, 54, 64–65, 69 Felson, J., 51n Fernandez, Y. M., 186 Ferrari, M. M., 39 Fetzer Institute, 170, 172, 186 Fichter, J., 117n, 286 Fichter, S. J., 117, 286 Figley, C. R., 39 Fiksenbaum, L., 75, 91

295

Finke, R., 65, 70, 139, 142n, 165, 189, 204, 206–208, 285 Firth, H., 77, 90 Fishbach, A., 237, 258 Fitchett, G., 232 Fletcher-Marsh, W., 2, 15 Floderus, B., 5, 13 Foa, E. B., 23–24, 38 Follansbee, M., 216, 233 Forde, L., 13, 15 Forrester, E., 241, 257 Forsyth, D. R., 238, 256 Fowler, J. W., 58, 69 Fox, J., 252, 256 Francis, L. J., 1–5, 14–15, 71, 80–81, 87–88, 90, 119, 135, 215–218, 225, 231, 287 Frankl, V. E., 182, 186, 263, 265, 266–267, 274, 278–279, 282–283 Fredrickson, B. L., 182–183, 186, 212, 217, 225, 229, 231 Freeman-Longo, R., 180, 186 Freeston, M. H., 241, 256 Freitas, A. L., 235, 256 Freud, S., 52, 69, 209, 211, 226, 231 Freudenberger, H. J., 71, 90 Frey-Wouters, E., 21, 38 Friedman, M. J., 39 Friedman, S., 99, 113 Frost, R. O., 257 Fry, P. S., 284 Funder, D., 210, 231 Funk, R. B., 67, 69 Furnham, A., 13, 15 Gackiere-Eraldi, D., 95, 113 Gagnon, F., 241, 256 Galea, M., 217, 227, 231 Gallagher, S., 206, 208 Gallops, M. S., 21, 38–39 Gandhi, M., 270 Garﬁnkel, I., 41, 49 Garos, S., 235, 244–245, 247, 256, 287 Gartner, J., 169, 186 Gay, D. A., 180, 186 Geary, B., 167, 217, 227, 231, 287 Gentilini, J. M., 79, 90 George, L. K., 180, 186, 216, 232 Gianelli, D. M., 27, 38 Gill, J. J., 71, 90 Glaeser, E. L., 191, 208 Glaser, B., 114 Glass, D. C., 77, 78, 90 Glass, J., 52, 54, 69

296

index of names

Glass, T. A., 180, 186 Glock, C. Y., 52, 70 Glover, H., 24, 26–27, 38 Godley, S. H., 72, 76, 93 Gold, R. S., 74, 78, 89 Gold, Y., 74, 78, 90 Golden, J. L., 217, 227, 231 Golembiewski, R. T., 77, 78, 91 Goli, V., 113 Gorman, B. S., 51, 54, 64–65, 69 Gorsuch, R. L., 228, 231, 282–283 Goschke, T., 239, 257 Gose, K. F., 78, 92 Gough, H., 104, 113 Graney, M. J., 216, 231 Gray, R., 196, 207 Graziani, P., 95, 113 Green, B. L., 21, 38 Green, D. E., 78–79, 91, 93 Green, T. G., 74, 91 Greenberg, J., 265, 283 Greenberg, M., 39 Greenglass, E. R., 75–76, 91, 93 Greenwald, E., 39 Greenwood, L., 113 Griﬃn, S., 171, 186 Grogan-Henderson, K., 180–181, 186, 217, 227, 231, 232 Grosse-Holtforth, M., 99, 113 Grossman, D., 21, 32–33, 36, 38 Gunn, R. N., 38 Gunnar, M., 256, 258 Gunnoe, M. L., 240, 256 Gupchup, G. V., 75, 91 Gurin, G., 99, 114 Guttmann, D., 274, 283 Haas, A. P., 21, 38 Hackman, J. R., 76, 91 Hackney, C. H., 227, 232 Hadaway, C. K., 141n, 144n, 165 Hafer, M., 72, 76, 93 Hair, J., 1, 15 Haley, J. M., 2, 15 Haley, S. E., 25, 38 Hall, D. T., 76, 91 Hall, F. L., 21, 38 Hall, H. V., 21, 38 Hall, T. W., 282–283 Hammond, P. E., 53, 65, 67, 69 Hanson, R. K., 168, 186 Harris, A., 95–96, 101, 113 Harter, S., 239, 256 Harvey, C., 98, 113

Hatcher, S. W., 117, 124, 135 Hausman, J., 196, 207 Hautekeete, M., 95, 113 Hay, D., 216, 232 Hayes, B. C., 52, 65, 69 Hays, L. C., 78, 89 Heath, W. C., 59, 70 Heatherton, T. F., 228, 230 Heckhausen, J., 239, 256 Heilbrun, A., Jr., 104, 113 Helminiak, D. A., 204, 207 Hendin, H., 21, 38 Hermalin, J., 114 Hervieu-Léger, D., 137–138, 140–141, 161, 165 Hess, R., 114 Heyman, J., 117n Hickey, N., 24, 38 Higgins, E. T., 235, 256 Hight, T. L., 258 Hill, P. C., 167, 187, 239–240, 256–257 Hill, R. L., 257 Hinson, E. G., 190, 207 Hipp, K. M., 258 Hodge, J. E., 34, 38 Hodgson, R. J., 235, 237, 257 Hoﬀman, J. P., 205, 207 Holiday-Goodman, M., 75, 91 Holland, R. W., 238, 257 Hollin, C. R., 38 Hood, R. W., 240, 256–257 Hoogduin, K., 75, 93 Hough, R. L., 38 Howell, L., 189, 191, 207 Hubbard, C., 24, 38 Hudson, S. M., 186 Huebner, E. S., 76, 91 Huismans, S., 240, 254, 257 Hunsberger, B., 65, 69 Hunter, 138 Hunter, J. D., 53, 65, 67, 69 Iannaccone, L. R., 189n, 191–192, 207, 208 Idler, E. I., 220–221, 232 Ierulli, K., 191, 208 Inskeep, K., 42, 49 Ishler, K., 100, 114 Iwancki, E. F., 74, 79, 91 Izard, C. E., 283 Jackson, C. J., 1, 3–6, 10, 13–15, 287

index of names Jackson, J., 99, 114 Jackson, L. M., 240, 256 Jackson, R. A., 86, 91 Jackson, S. E., 72–78, 80, 91–92 James, W., 210–211, 230, 232, 235, 257 Janssen, P. M. P., 77, 89 Jecmen, D., 168, 186 John, O. P., 231–232 Johnson, D. J., 237, 257 Johnson, E. C., 18, 38 Joiner, T. E., Jr., 101, 103, 114, 210, 218, 223, 232 Jones, A., 274, 283 Jones, D. L., 2, 15, 119, 135 Jones, J. W., 90 Jones, S. H., 1–4, 14, 216, 231, 287 Jones, T., 38 Jordan, 101 Jordan, B. K., 38 Kadar, J. L., 258 Kahneman, D., 216, 230–232 Kaldor, P., 216, 218, 225, 231 Kaplan, H. B., 114 Katz, Y. J., 215, 217, 231 Kay, W. K., 1, 2, 14–15 Kellstedt, L., 58, 67, 69 Kennedy, P., 206, 208 Kenny, D. A., 108, 113 Key, C. C., 78, 93 Keyes, C., 41, 49 Kilfedder, C. J., 86, 91 Kim, J. K., 252, 257 Kirkpatrick, 239, 257 Kladler, A., 75, 93 Klausner, L. H., 74, 91 Kleinmuntz, B., 238, 257 Kline, P., 7, 12, 15 Klosko, J., 115 Kluck, A., 235, 288 Kluegel, J. R., 192, 208 Koenig, H. G., 99, 100, 105, 113–114, 169, 186, 216, 232 Koepp, M. J., 32, 38 Koeske, G. F., 79, 91 Koeske, R. D., 79, 91 Kolarz, C. M., 219, 232 Konieczny, M. E., 121, 135 Krahe, B., 238, 257 Krause, N., 232 Krishnan, K., 113 Kruppa, I., 24, 38 Krystal, J., 39

297

Kudler, H. S., 113 Kuhl, J., 239, 257 Kulka, R. A., 21, 38, 99, 114 Kurtz, E., 211, 232 Kurusu, T. A., 264, 284 Kwon, S., M., 240, 257 Ladouceur, R., 241, 256 Lankshear, D. W., 1, 14 Larsen, R. J., 219, 231 Larsen, R. L., 171, 186 Larson, D. B., 169, 180, 186, 257 Laufer, R. S., 21, 38, 39 Lawler, E. E., 76, 91 Lawrence, A. D., 38 Laws, D. R., 186 Leach, M. M., 266, 273, 284 LeBlanc, A. E., 78, 90 LeCroy, C. W., 76 91 Lee, C. W., 101, 103, 114 Lee, H. J. 240, 257 Lee, J. K., 252, 257 Lee, R. T., 78, 86, 91 Lehman, E. C., Jr., 118, 122, 132, 135 Lehne, G., 167n Lehrer, E. L., 51, 65, 69 Leiter, M. P., 73, 76–77, 86, 91–92 Lemkau, J. P., 76, 78, 93 Leong, G. B., 39 Leung, N., 101, 114 Leventman, S., 39 Levy, P., 5, 15 Liberman, N., 235, 256 Lifrak, S., 99, 113 Lifton, R. J., 26, 38 Likert, R., 80, 92 Lincoln, A., 279 Littler, K., 1, 14 Lively, B. T., 75, 91 Loﬂand, J., 143–144, 165 Loﬂand, L., 143–144, 165 Logue, A. W., 239, 257 Long, J. S., 196, 208 Longpre, R. E., 39 Lopez, S. J., 212, 233 Louden, S. H., 1–2, 14–15 Loveland, M. T., 144n, 165 Lowenthal, M. F., 273, 283 Lubin, B., 102–104, 114–115 Lucas, R. E., 169, 186, 215, 231 Lummis, A. T., 122, 133, 135–136 MacDonald, D. A., 215, 233 Mack, S. E., 169, 185

298

index of names

Macklin, M. L., 39 MacNair, R. M., 17, 21–22, 29, 36, 38, 288 Madsen, R., 165 Mahoney, A., 228, 233 Malatesta, C. Z., 283 Maloney, J. P., 86, 89 Malony, H. N., 117, 124, 135 Manlove, E. E., 87–88, 92 Mapa, M., 229, 232 Marinelli, S., 209 Marler, P. L., 141n, 144n, 165 Marmar, C. R., 38 Marsden, G. M., 138 Marsden, P. V., 55n, 69, 194, 207 Marshall, W. L., 186 Marx, K., 211 Maslach, D., 72–78, 80, 86, 92 Maslow, A. H., 218, 232, 278, 284 Maton, K. I. , 114, 182, 186 McAllister, I., 58, 69 McCarthy, D. M., 5, 16 McCrae, R. R., 87, 89, 172, 185, 215, 219, 221–232, 271, 273–274, 283 McCullough, M. E., 99, 114, 169, 186–187, 218, 233, 257–258, 264, 284 McDuﬀ, E. M., 118–119, 135 McEvoy, A., 79, 89 McFadden, D., 196, 207 McGinley, J. J., 169, 185 McGlone, G., 167n McIntec, J., 77, 90 McKegney, F. P., 216, 233 McKeown, P., 77, 90 McKinney, J., 189, 204, 208 McKinney, W., 52, 55, 58, 65, 70, 139, 165 McKnight, J. D., 77–78, 90 McMurran, M., 38 McSwain, L., 124, 135 Meador, K., 113 Meadow, K. P., 72, 92 Meadows, E. A., 23–24, 38 Meertens, R. M., 238, 257 Meier, S. T., 76, 92 Menaghan, E., 114 Mendel, W. M., 71–72, 92 Mertler, C. A., 127, 135 Michael, W. B., 78, 90 Mickley, J. R., 169, 186 Milewski, G. B., 114 Milgram, S., 29–30, 38

Miller, A. S., 205, 207 Miller, W. R., 167, 180, 186 Moberg, D. O., 288 Mol, H., 241, 254, 257 Molloy, G. N., 78, 92 Monos, D., 78, 92 Moore, K. A., 240, 256 Moore, R. J., 76, 93 Morgan, C. A., 32, 39 Morisy, A., 216, 232 Mroczek, D. K., 219, 232 Mueller, C. W., 118–119, 135 Munzenrider, R., 78, 91 Murphy, J. E., 86, 91 Murphy, P. E., 229, 232 Murphy-Hackett, E. S., 72, 92 Musick, M. A., 232 Musson, D. J., 119, 135 Myers, D. G., 169, 186, 227, 232 Myers, S. M., 52, 69, 191, 208 Nadelson, T. N., 32–33, 39 Nagy, S., 79, 92 National Bureau of Economics Research, 53, 67, 69 Neiderhoﬀer, A., 71, 92 Neighbors, H., 99, 114 Nelson, R., 172, 187 Neminoﬀ, R. A., 38 Nesbitt, P. D., 122, 135 Nesselroade, K. P. Jr., 238, 257 Newberg, A., 30, 39 Nielsen, J., 99–100, 113 Nietzsche, 211 Nobus, D., 168, 187 Nunnaly, N. C., 75, 92 O’Conner, L., 258 O’Dea, T. F., 152, 165 O’Driscoll, M., 78, 93 Okagaki, L., 52, 69 Oldham, G. R., 76, 91 Olson, 125 Ones, D. S., 168, 187 Ory, M. G., 232 Ottenbreit, A., 238, 257 Ozorak, E. W., 65, 69 Papadatou, D., 78, 92 Pargament, K. I., 95, 98–101, 105, 111–115, 182, 186, 209–212, 217, 225–226, 228–229, 232–233, 239, 257, 258 Park, C. L., 181, 186

index of names Park, J. Z., 208 Parsons, R. D., 283 Pastoreloli, C, 29, 37 Pathak, A., 99, 113 Pavot, W., 187 Pearce, L. D., 52, 69 Pearson, P. R., 2, 14 Pelsma, D. M., 78, 92 Pennebaker, J. W., 233 Peplau, L. A., 275, 284 Perez, L., 105, 114 Perez, M., 210, 218, 223, 232 Perl, P., 120, 135 Pervin, L. A., 231–232 Peterson, L., 42, 49 Petrocelli, J., 114 Peyrot, M., 232 Philipchalk, R., 5, 14, 81, 90 Piedmont, R. L., 79, 87, 93, 98, 114, 169, 172, 182, 185, 187, 214–215, 217, 225–227, 229, 231–233, 263, 266, 273–274, 278, 280, 282, 284, 288 Pieper, C., 99, 113 Pierce, C. M. B., 78, 92 Pines, A., 72, 77, 92 Piper, J. L., 118, 124–125, 135 Pitman, R. K., 24, 39 Pittelkow, Y., 52, 65, 69 Plante, T. G., 233 Plato, 23 Poire, R. E., 39 Poll, J., 99, 114, 181, 187, 218, 233 Pontefract, 101 Pope John Paul II, 118 Potvin, R. H., 241, 254, 258 Powell, L. H., 232 Powell, T., 215, 218, 233 Power, K. G., 86, 91 Powers, S., 78, 92 Pretorius, T. B., 75, 92 Price, L., 86, 92 Purdon, C., 240–241, 256–257 Purdy, R. R., 76, 78, 93 Pyszczynski, T., 265, 283 Rachman, S., 241, 257 Racine, C. T., 95, 289 Raﬀerty, J. P., 76, 78, 93 Randall, P., 37 Rank, M. R., 76, 91 Rause, V., 30, 39 Rayburn, C. A., 117, 122–125, 127, 134–135

299

Reed, M., 98, 114 Regnerus, M. D., 208 Remmes, K. R., 239, 258 Richmond, L. J., 117, 123–125, 127, 134–135 Riggar, T. F., 72, 76, 93 Rinck, C. M., 104, 115 Ripley, J. S., 258 Robbins, M., 1–3, 14–15 Roberg, R. R., 72, 93 Robins, R. W., 238, 257 Robinson, L. R., 208 Rodger, R., 1–2, 14, 80, 87, 90 Rodgerson, T. E., 79, 87, 93, 217, 226, 231, 233 Roe, K. M., 36, 39 Rogers, L., 124, 125, 134–135 Rogers, P., 263n Roland, B., 78, 92 Roof, W. C., 52, 55, 58, 65, 67, 69, 98, 114, 138–139, 165 Roozen, D. A., 51, 65, 70 Ross, N. R., 72, 92 Roth, R. A., 78, 90 Rouiller, R., 100, 114 Rowatt, W. C., 238–240, 257 Rowney, J. I. A., 78, 93 Royle, M. H., 118, 120, 136 Royster, B., 100, 114 Rubin, Z., 275, 284 Rudisill, J. R., 76, 78, 93 Ruppart, R., 119, 136 Rusbult, C. E., 237, 257 Rusinek, S., 95, 113 Rutledge, C., 1–2, 14–15, 71, 79–80, 88, 90, 93, 289 Ryan, R. M., 239, 256 Ryan, W., 72, 93 Ryba, K., 78, 93 Rye, M. S., 258 Salameh, K. M., 75, 89 Salkovskis, P. M., 241, 257 Sandage, S. J., 264, 284 Sanders, G. S., 227, 232 Sardjian, A., 168, 187 Sarros, A. M., 78, 93 Sarros, J. C., 78, 93 Savicki, V., 75, 93 Sawyer, D., 236, 256 Schaap, 75, 93 Schaufeli, W. B., 73, 75, 77–78, 86, 89, 90, 92–93 Scheers, N. J., 167, 289

300

index of names

Scheier, M. F., 213, 229–230 Scherer, R. F., 78, 93 Schlenger, W. E., 38 Schmidt, N. B., 96, 101, 103, 114 Schnitt, M. M., 258 Schoenrade, P. A., 239, 256, 265, 283 Scholte, R. J. H., 5, 16 Scholz, R. W., 256 Schuler, R. S., 91 Schultz, R., 239, 256 Schwab, R. L., 79, 91 Schwartz, N., 216, 230–232 Schwartz, S. H., 240, 254, 257 Scott, A. B., 98, 115, 229, 233, 239, 258 Scott, T. M., 37 Segal, M. L., 53, 69, 70 Seibyl, J. P., 37 Seybold, K. S., 167, 187 Shatan, C., 25, 39 Shehan, C. L., 125, 136 Shelp, F., 113 Sherkat, D. E., 51, 53–54, 69–70, 98, 114, 144n, 165 Sherman, A. C., 233 Sherrow, F., 51, 65, 68 Shorto, R., 98, 114 Sibert, T., 113 Siegler, I. C., 216, 232 Siganga, W. W., 75, 91 Sikkink, D., 206, 208 Silva, J. A., 28–29, 39 Skinner, E. A., 239, 257–258 Sloan, R. P., 215, 233 Smidt, C., 58, 67, 69 Smith, C., 99, 113 Smith, C. S., 138–139, 141n, 165, 206, 208 Smith, G. T., 5, 16 Smith, H. L., 169, 186 Smith, T. B., 181, 187, 218, 233 Smith, T., 99, 114 Smith, T. W., 55n, 56, 69–70, 193, 194, 207 Sneck, W., 263n Snyder, C. R., 212, 233 Socrates, 23 Soeken, K. L., 169, 186 Soliday, J., 75, 93, 104 Solomon, S., 265, 283 Solursh, L., 32–33, 39 Southwick, S. M., 32, 37, 39 Spangler, E., 78, 93

Spence, S. H., 86, 92 Spong, J. S., 205, 208 Sroufe, L. A., 256, 258 St. James, P., 216, 233 Stajich, G. V., 86, 91 Stanik, P., 100, 114 Stark, R., 52, 65, 70, 139, 142n, 165, 199, 206–208 Stayman, D. M., 252, 258 Stearns, G. M., 76, 93 Steel, P., 168, 187 Steensland, B., 194, 208 Steer, R., 96, 113 Steketee, G., 257 Stewart, 125 Stewart, C., 183, 187 Stickney, F. A., 78, 93 Stock, W. A., 244–245, 247, 256 Stonechild, J., 168, 186 Strachey, J., 231 Strayer, R., 21, 39 Such-Baer, M., 36, 39 Suedfeld, P., 30, 38 Suh, E. M., 169, 186 Sullins, P., 121, 136 Sullins, P. D., 58, 65, 70 Sullivan, W., 165 Suziedeliz, A., 241, 254, 258 Swidler, A., 165 Swyers, J. P., 257 Tanenbaum, L., 255, 258 Taylor, A. J., 78–79, 91, 215, 233 Taylor, G., 101, 103, 114 Taylor, S. E., 238–239, 253, 258 Telch, M. J., 101, 103, 114 Tellegen, A., 171, 187 Tetlock, P. E., 30, 38 Theresa, Mother, 270 Thibodeau, N., 241, 256 Thoburn, J., 236, 255 Thomas, G. J., 101, 114 Thomas, T. H., 1–2, 15 Thompson, E. H., 239, 258 Thurner, M., 273, 283 Tice, D. M., 228, 230 Tillich, P., 278, 284 Tipton, S., 165 Tirrell, C. A., 5, 13 Tisdale, T. C., 282–283 Tollefson, N., 78, 92 Tommasi, M., 191, 208 Torbet, R. G., 207n, 208

index of names Treadwell, W. C., 124, 135 Trope, Y., 237, 258 Tully, H. A., 72, 90 Tuma, N., 189n Turner, J., 76, 91 Tutu, D. M., 25, 35, 39 Udry, E. M., 75, 93 Underwood, J. R., 117, 124, 135 Underwood, L. G., 232 Valdimarsdottir, H., 78, 90 Van der Hulst, R., 77 89 Van der Kolk, 32, 39 Van Dierendonck, D., 75, 93 Van Hook, L., 99, 113 Van Vugt, M. U., 238, 257 Van Whitlock, R., 103–104, 114 Van Wicklin, J. F., 214, 233 Van Yperen, N. W., 77, 89 Vannatta, R. A., 127, 135 Varacalli, J. A., 51, 54, 64–65, 69 Vealey, R. S., 75, 93 Ventis, W. L., 239, 256, 265, 283 Veroﬀ, J., 99, 114 Wade, D. C., 75, 93 Wade, N. G., 258 Wagner, R. V., 38 Walker, R. L., 210, 223, 232 Walkey, F. H., 78–79, 91 Wallace, P. A., 72, 93 Waller, G., 101, 114 Wallis, C., 98, 114 Wallston, K., 99, 113 Walsh, J. A., 77, 93 Walsh, J. M., 180–181, 187, 217, 227, 233 Walsh, R. P., 236, 258 Walter, T., 241, 254, 258 Ward, T., 183, 186–187 Warner, J., 79, 93 Warner, R. S., 139, 165 Waters, M. A., 59, 70 Watson, D., 168–169, 171, 187, 216–217, 225, 231, 233 Watson, J. K., 59, 70 Wegner, D. M., 233, 241, 258

301

Weinstock, R., 39 Weishaar, M., 95, 115 Welburn, K., 101, 103, 114 Wellborn, J. G., 239, 256 Wells, T. J., 86, 91 Wheaton, B., 100, 115 White, R. W., 239, 258 Whitehead, A., 78, 93 Wicks, R. J., 283 Wiggins-Frame, M., 125, 136 Wigington, H., 78, 92 Wikler, N., 32, 39 Wilcox, C., 216, 231 Wilcox, W. B., 208 Williams, D. R., 232 Williams, J. E. G., 182, 185, 215, 226–227, 231, 233 Williams, J., 263n Willits, F. K., 67, 69 Wilson, 125 Wilson, G. G., 3, 5–6, 10, 13 Wilson, J., 53–54, 64, 70, 144n, 165 Winter, D. D., 38 Wolpin, J., 76, 93 Wong,, P. T. P., 264, 284 Woodberry, R. D., 208 Worthington, E. L., 240, 245, 250–251, 258, 264, 284 Wright, C. R., 78, 90 Wuthnow, R., 51, 70, 138–140, 142n, 162, 165, 199, 208 Wybraniec, J., 285 Yager, T., 21, 39 Yates, J. W., 216, 233 Ybema, J. F., 75, 89 Yehuda, R., 32, 39 Young, J. E., 95–96, 101, 103, 106–107, 110, 112, 114–115 Young, L., 165 Zanna, M. P., 283 Zastrow, C., 72, 93 Zikmund, B. B., 122, 133, 136 Zimmer, A. C., 256 Zimmerman, V., 75, 93 Zinnbauer, B. J., 98, 115, 229, 233, 239, 258 Zuckerman, M., 102–104, 115

SUBJECT INDEX Abortion, 27, 36, 204n Accomplishment, personal, 73–74, 76, 78, 81–87 Addiction, 28–29, 32–34, 145, 169 Aﬀect, 103–104, 216–217, 219, 222, 273–274, 276–277 Aﬃrmation, 132–134 African Americans, 199 Age comparisons, 43–44, 58–59, 82–84, 86–87, 197, 203, 280 Agreeableness, 169, 172, 175–176, 178, 215, 219, 222, 224, 267, 270 Alcoholics Anonymous, 211 American Anglican Council, 189 American Baptist Churches USA, 190 American Baptist Evangelicals, 190 American Methodist Episcopal (AME) Zion, 121 Anglicans, 1–16, 43–44, 71–93, 189 Anxiety, 4, 103–104, 218 Apartheid, 25, 35 Apostasy, 51, 54, 57, 62–63, 65, 68 Asceticism, 269, 270, 280–282 Association for Church Renewal, 189 Attachment theory, 239 Attendance, religious, 170, 172, 175, 177–180, 216, 226 Australian, 41–50 measurement of, 49–50 Australia, 41–50 Authenticity, search for, 151–154, 156, 158, 163–164 Authoritarianism, 275 Baptists, 44, 121, 190, 207n Southern, 124, 206 Balanced aﬀect, 216 Battle fatigue, 17 Beck Depression Inventory, 96 Bible, attitudes toward, 194 Biblical Witness Fellowship, 190 Bipolar Adjective Scale, 219 Black Protestants, 197–199, 201–205 Born-again evangelicals, 144–147 Bradburn Aﬀect Balance Scale, 273–274 Brief Religious Coping Questionnaire, 105

Buddhism, 44 Buddhist meditation, 30 Burnout, 71–93, 274 Cancer patients, 180, 217 Capital theory, 191 (see Human capital theory) Catholics, 43–44, 118, 122, 125, 144n, 153, 160–161, 197–205, 266 Childhood, 95–115, 144n Choice, religious, 137–165 Christian Church (Disciples of Christ), 190 Christian Reformed Church, 118 Christians, 266, 273 Church attendance, see Attendance 41–50 (in Australia) Measurement of, 49–50 Church of England, 119 See Anglicans Churches of Christ, 44–45 Circumplex, 267–270, 272, 277–282 Clergy, 1–16, 217 Burnout, 71–93 Criticisms of, 85–86 Job satisfaction, 122, 125, 133 Male/female comparisons, 3–4, 117–136 Stress, 117–136 Clinical assessment, 5 Cohabitation, 41, 42–46, 48–49 Commitment, religious, 240 Community, religious, 138–139 Connectedness, 273, 276–277, 280–281 Conscience, 36 Conscientiousness, 169, 172, 175–177, 178, 215, 219, 222–224 Conservative religion, 53, 179–208 Control theory, 213–214, 227–229 Conversion, 211 Coping, religious, 95–115, 169, 181, 212, 217–218, 225–226, 229 Counseling, 23–25 Cradle evangelicals, 145, 148–151 Cultural Christianity, 152, 161–164 Daily spiritual experiences, 220, 225 Death penalty, 36

subject index Denominations, classiﬁcation of, 56 Depersonalization, 72–74, 76, 78, 81–87 Depression, 76–77, 96, 99–100, 104, 121, 125, 181, 218 Desert Fathers & Mothers, 270 Disaﬃliation, see Apostasy Disciples of Christ, 119, 190 Disciples Renewal, 190 Dreams, 26–27 Dysphoria, 95–115 Early maladaptive schemas, 95–115 Education and religious switching, 51–70 Education comparisons, 197–198, 203, 205 Ego Integrity, 280 Ego protection, 239 Emotions, 213–214, 216, 223–226, 229–230 Emotional exhaustion, 72–74, 76, 78, 81–87 Episcopal Church, 53, 121, 189, 190 Episcopalians United, 190 Ethics, 17, 24, 36–37 Evangelical Lutheran Church of America, 118, 121, 190 Evangelical Lutheran Confessing Fellowship, 190 Evangelicals, 53–54, 121, 194, 198–199, 201–206 Renewal movements, 189–208 Seekers, 137–165 Eve, 235 Existentialism, 279 Expectations, 213, 228–229 Extraversion/Extroversion, 2–3, 7–10, 12, 71, 81–84, 87–88, 168–169, 172, 175–176, 178, 217, 219, 221–224, 267, 270 Extroversion, 2–3, 7–10, 12, 71, 81–84, 87–88, 217 Extrinsic religiosity, 215 Eysenck Personality Proﬁler, 1–16 Eysenck personality tests, 4–5, 81–82 Faith, 210 Faith communities, 227 Faith Maturity Scale, 172, 175, 177, 182, 215 Family, 191 Fetzer Institute, 220 Five Factor Model of Personality

303

(FFM), 264, 266, 273 Flashbacks, 20, 24, 28 Forgiveness, 35 Fundamentalists, 56, 60–62, 67–68, 194, 206 Gamblers, 180, 217 Garos Sexual Behavior Index (GSBI), 235, 244–250, 253 Gemeinschaftsgefühl, 268 Gender (sex) comparisons, 3–4, 45–46, 117–136, 199, 203, 216, 221, 236 General Social Survey (GSS), 52, 55, 189, 195, 220 Generativity, 280 Goals, 213–214, 227–228 God, 23, 111, 167, 172, 174, 215–216, 225, 264, 270, 282 belief in, 66, 98, 275 closeness to, 218, 220, 227 image of, 184, 226 Good News, 190 Gregariousness scale, 271, 274, 276–277 Guilt, 4, 22, 24, 26–27 Hallucination, 18 Happiness, 168–179, 209–233, 273 Health, 167 religion and, 210, 214–215 Hedonic psychology, 212 Hinduism, Hindus, 44, 266, 273 Human capital theory, 189, 191–192 Humanistic psychology, 218 Hypocrisy, 160 Identiﬁcation, religious, 161–163 Income, 195–196 Incremental validity model, 214–219, 226, 230 India, 266 Individualism, 138–142, 148, 279 Institute on Religion and Democracy, 189 Institutionalization dilemma, 152 Intentionalism, 268–269, 272, 276–277, 282 Intrinsic religiosity, 215 Introversion, 87, 217 Islam, 43–44 Job satisfaction, 76, 122 Judaism, 44

304

subject index

Judgments, 238 Just World Scale, 272, 275–277, 279 Justice, 35 Killing, 17–39 Liberals, religious, 56, 61, 64 Lie scale, 82–83 Life satisfaction, 168, 175–177, 180 Lifewatch, 190 Logoplex, the, 262–284 Lutherans, 44–45, 120, 122, 124–125 Evangelical L. Church of America, 118, 121, 190 Missouri Synod, 67 Marital status and religion, 41–50 Maslach Burnout Inventory, 71–93 Materialism, 267–268, 275, 277, 282 Maturity, 278 Meaning, 144, 181–182, 192, 213–214, 264–272, 274, 278–280, 282 Mental health, 167–185, 269 Meta-pathologies, 278 Modernity, 140–141 Modesty scale, 272, 274–277 Mood, 100–101, 110 Moral development, 236 Mortality, 264–265 Motivation, 30, 182, 213, 230, 239–240, 241, 254–255, 264–267, 271, 283 Multidimensional Scale of Religion & Spirituality, 220 Multiple Aﬀect Adjective Check List, 103 Muslims, 266, 273 Mystical experience, 30–31 National Institute on Aging, 220 National Opinion Research Center (NORC), 55, 195 National Vietnam Veterans Readjustment Study, 21–22 Natural law, 23 NEO Five-Factor Inventory, 171–172 NEO Personality Inventory-Revised, 271, 274, 276–277 Neurosis, 226 Neuroticism, 2–3, 7–10, 12, 71, 81–84, 87, 168–169, 175–176, 178, 219, 221–224 New Zealand, 123 Non-Christians, 197–203

Nuns, 30, 122, 125 Nurses, 77–78 Obsessive thoughts, 240–241, 245–246, 250 Occupational prestige, 195–198 Openness, 169, 172, 175–176, 178, 215, 219, 222, 224 Orthodox religion, 43–44 Oxford Happiness Scale, 217 PANAS, 171 Parental inﬂuence, 51–70, 95–115, 191–192 Pastoral care, 23–25 Peak experiences, 148, 162 Pedophilia, 173–174 Pentecostalism, 43–45 Perpetration-induced traumatic stress, 17–39 Personal accomplishment, see Accomplishment Personality, 215–220 Of clergy, 1–16, 80 Personality assessment, 1–16 Philippines, 266 Pluralism, cultural, 138–142, 163 Police, 77–78 Positive and Negative Aﬀect Schedules, 171 Positive psychology movement, 168, 212, 226–230 Posttraumatic stress disorder, 17–39 Symptoms, 18–21, 28, 33–34 Prayer, 132, 133, 174 Belief in, 98 Prayer Fulﬁllment, 273, 276–278 Prayers, Franciscan nuns’, 30 Pre-marital sex, 42, 204 See also: Cohabitation Presbyterian Church (U.S.A.), 190 Presbyterian Layman, 190 Presbyterians, 43–44, 118, 121, 125, 190 Presbyterians for Faith, Family & Ministry, 190 Presbyterians for Renewal, 190 Presbyterians Pro-Life, 190 Priesthood of believers, 137n Protestant conservatives, 179–208 Protestants, 55–68 Mainline, 189–208 Psychology, religion and, 209–212, 218, 226

subject index Psychotherapy, 227 Psychoticism, 3, 7–10, 12, 71, 81–84 Public policy, 36–37 Purpose, 264–265, 267, 269 Purpose in Life Test, 274, 276–277 Quaker paciﬁsm, 34 Rabbis, 125 Race, 59 Recidivism, 168 Reductionism, 230, 279 Reformed churches, 121 Regional comparisons, 197, 199, 202, 203–204 Relationalism, 268–269, 271, 276, 281, 282 Religion, conceptualizations of, 239–240 Religiosity, 239–240 Religious attendance, 41–50 See also: Attendance Religious commitment, 240, 246–253 1, 251, 252–253 Religious Commitment Inventory, 235, 245, 250–251 Religious experience varieties, 209–233 Religious market, 138–142 Religious Practices Questionnaire, 172 Religious switching, 51–70, 137–165 Renew Network, 190 Renewal Fellowship, The, 190 Salvation Army, 44–45 Sanctiﬁcation, 228 Satisfaction With Life Scale, 171, 219 Secular Humanism, 268–269, 271–272, 276, 279–280 Secularization, 65 Self-actualization, 272 Self-Actualization Scale, 274, 276–277 Self boundaries, 26–27 Self-control, 228, 236–237, 239, 241, 251, 253, 255 Self-esteem, 4, 236, 238, 253 Self-image, 240, 251 Self-transcendence, 27 Seventh Day Adventists, 44–45 Sex oﬀenders, 167–187 Sexual ethics, 204–205 Sexual temptation, 235–261 Shell shock, 17 Sin, 236, 252

305

Social status, 52–53, 64 Social workers, 79 Socialization, 52, 66, 142–143n, 144n, 148 Solipsism, 268–269, 271–272, 276, 278–279, 280–282 South Africa, 25, 35 Spirit, human, 22–23, 37 Spiritual experience, 30 Spiritual health, 35 Spiritual marketplace, 139 Spiritual quest, 151–161 Spiritual transcendence, 169, 217, 263–284 Spiritual Transcendence Scale (STS), 266, 270–271, 273, 278 Spiritualism, 269, 270–271, 276, 278, 279–282 Spirituality, 98, 140, 162, 167–187, 209–233, 263–264, 266–267, 278 And religion, 229 Sports, 33 Stress, 100–101, 117–136, 217–218, 274 Suﬀering, 22–23 Support groups, 131, 134 Switching, religious, 51–70, 192 Sympathy, 25, 34–35 Temptation bias, 235–261 Theological conservatives, 205 Theoneurology, 30–34 Therapeutic counseling, 23–25 Transcendence failure, 228 Transdenominationalism, 138 Transforming Congregations, 190 Transpersonalism, 267–268, 271, 275–277, 281–282 Twelve-Step movement, 167 United Church of Christ, 119, 121, 190 United Methodist Confessing Movement, 190 United Methodists, 120–122, 125, 190 Uniting Church, 43, 44 Universality, 273, 276–278, 280 Vatican, 22 Veterans, 21–22, 24, 26, 28, 32–34, 36 Victimhood, 17, 26, 72 Violence, 22, 28 Prevention of, 27–30

306 Warfare, 36–37 Well-being, 273 Well-being predictors, 167–187, 209–233 Religion and, 290–233 Women, 241

subject index Ordination of, 118 Women clergy, 117–136 Workaholics, 71 Word Alone, 190 Young’s Parenting Questionnaire, 103