'cll » .. '9' Li xv. L.. . _ I . 1.. “4..., i333 . . V , A . . . h\ . . . ,, . . ... ..V . . l, . ...-..., nil: . . . . . . 2 . . . . . 7 . ‘ ‘ C . . . . u L ."rflllello A... J.’Vfl‘. ‘ L... . ‘ . .. V . ..T ... . ., , h . , . . . . . { , . .. . . . . {43!}. (”Ml-u. (vimllfur . ‘ . . I . .. y . .. . . x .. . ..1 .o .VU:’.I I) ‘Mvtt! : ... r . ‘ . . Nannilr. unwary-1. ......unr \ 51L. kw. . a: 3 a L l Ink claw L. .. ; .. ..I IE 0.0. 57:75,? it .1 g}; y .1 -... v.3. ....I 35?...“ 1.7. 1, _ uhx..v ;: ... _ It- . fink 11:... .A ‘57:}! .i4 ‘ . 03‘: t n»: (aux)... E.Lv L333 1. , ‘..I .n l runway. “11... . L; . . .vruflfiniéé... .. _ V. .Méqunmm L .. _ (Cz’f K. . , v 1 ’3"): J v... I... . ‘ ,. .... .. ... . .. “it?! , . , 1:3." ,..:L,L.....z..uw. ..-. arid. .. nf .)_v Luncuicnu. .IEHLLV..¢42 I‘- r . .. ..LI‘ ‘2 4“ |l .vw . . I. \. .u ‘9 . . . ..v «$5.3. W) 044.- I .i . IILJaitf n). . .V. 00’ (Cw-J‘v‘aln 0“: an! . .I - .LVI 2.3L: yin» . ‘L' Al‘sfluwug‘lz‘... #339“? I}. u L L. 1... ,v. tub-Vt .. sis-...... I .Jth. ,I .- L. ...!I |Anrb.|o.\ « l...‘ (.529. (L.- ... .:.v‘.l.I.-).. .7. Cull... .. .. t. i u .e . .y}.-..nL.z..\€ v n . N. . .3. I. . .I. l «QR-ins KL .‘ (v! . .‘(tciv .. -..) Liv...:1\p_ 1‘. II .blnih. .1 «u \c ktL‘l t‘ . y .\\I . IJIIIO. D \u.‘ . (a .I‘. 0"} . . If.” a. . 1. .IL, ‘.’ c . ‘v:"$'A-Lu>orl t . . . ,Q ..rloflh. ..Jld- .{uot . ‘ ..\. -.e ‘1 )- .5... L . 35.5 IL» rottwii...n.: . Farina". A»: n. 2 . 'V .P'.. C ’1." u. . . . . 21:3 ., Fig.1?! ' {5.1!}; . . . . 71“. . ‘ . . ‘ y c .l a 1 I L . Iv. Ft 0.. . VI. A , . .. ;u.w..,.{....s,c...‘...y.h . . . . . .. LP 3?... 19.33.?!35. (I ..... tabrirfl. ..cn ”v.0! LA . . . . y . b“|‘.uflall '4.- . .. ‘ {‘ll‘lg-fi . , . . ‘2. . . ,. . v F 130.11... 010...: flu" I0 '15.]; u 0‘ . . . A n , u. '0 ...;ti . .30.”): .543... 5.1. L, in.“ , . .1- a . t. tn 1! c. ‘ . .1 ‘ . l . .. ...“.I‘ . ...: ..fét... . .. II I.) , V ,. , . t L5 a u : Narnia: m.“ ..1 ¢ -4, 6.73.. . . 3 u 1 .. u ....ufimvhanfiutn; .. ...» .52 ..r..m..@n..n.vu.uu...! v in: up: w ). hon-[fin]: Jun" , r .. k 12.1%,- . ...v{.... . t...$J .UJ A..:).»f.-Orkr.. \wl. "Y..N..c.‘x 11-. -. JHLQ..‘.L..\. .12....trfk. at}... ...». 2‘ L... 3;“... ..vl ...? . b . -... .... e -.-. Ll ‘ 11.3.4. ' 1‘431V. 1. y: ., xo. 0 ‘3‘...- NW“ "D! . up . . .. . c , . . , 3 .v ‘ 2b: . v .. . 1 .f ‘ . . . I , . I 1 ‘ y . . 11...... L at? ,, are»? “Q I j . 3 r z .gfim: Emu... a: an»??? E 49%;; 3%“ $31} I.” . _ _ ...: ; ..3 “Kb; q: t K -. 5.10.3. « -.4. : $§mmfim ‘ , c...i)c.i:h.$ .51.}.le“3: 9331.; (mph). , -.. Eff?! 1.311.: ,Iglucmru - . 30w!...».% I . .n . .. .. L t o I . Y 3 . . E. I .JFI‘ AI' 1 LI’ ! ‘l’ L. I . II! E p ’ I.‘ ’ I i l I 'bl ' I WWWWWWW WWWWWWWWWWWWWWWWW WW WWW WWW WW WWI 2 2; x9 5’ 9,1, 3129300 W‘ Uwrifiu’v W MiGhECflmn {T511}; University W m. This is to certify that the dissertation entitled RELIGIOSITY, IDENTITY DEVELOPMENT, AND HEALTH OUTCOMES IN A LATE ADOLESCENT SAMPLE presented by Larry Antosz has been accepted towards fulfillment of the requirements for ID”) degreein gyc/ffll-déy // JMA Major professor Date 2% 9/; /;9?; MS U is an Affirman'w Action/Equal Opportunity lnsliturion 042771 RELIGIOSITY. IDENTITY DEVELOPMENT. AND HEALTH OUTCOMES IN A LATE ADOLESCENT SAMPLE By Larry Antosz A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1989 ‘> ‘8 5"? U3 (go ABSTRACT RELIGIOSITY. IDENTITY DEVELOPMENT. AND HEALTH OUTCOMES IN A LATE ADOLESCENT SAMPLE BY Larry Antosz The current study (N=440) attempted to replicate and extend the findings of a previous study by this author on the role of religion in coping with stress in a late adolescent sample. That study suggested that religion. particularly personal religious beliefs and prayer. may help late adolescents cope with the stresses associated with their developmental period by selectively influencing their perceptions of minor daily events. The current research investigated whether this relationship between religion and the perception of minor daily events was in turn related to physical and mental health outcomes. Several hypotheses were made about the indirect and direct relationship of specific religious variables with health measures. Correlational and path analyses failed to support many of the hypotheses. Only some small positive. as well as negative. direct links between religion and health outcomes were found. In general. the results for this sample of late adolescents were consistent with the findings in the literature for adult samples that religion has a positive but small relationship to measures of well-being. Additional analyses uncovered some information about the relationship of religion to general identity development as well as pointing to some of the components of religion that seem to be particularly salient for this age group. Based on this sample. it appears that there are important gender differences in the structure and function of religion. For this sample. religion seemed to be closely associated with the Foreclosure identity status for males. while it related to the Achievement status for females. For both males and females of this age group. the personal meaning that is associated with religious belief and commitment appears to be the crucial element in religion. In particular for females. the social aspects of religious involvement seem to be important. Finally. this study provided further psychometric support for the religiosity measure developed by this author in a previous study. Copyright by LARRY ANTOSZ 1990 To Rose. my wife and best friend. for your love and support that helped make a dream come alive. ACKNOWLEDGMENTS I would like to thank Dr. Elaine Donelson for her three years of generous support for me and my research in this area and for allowing me the freedom to follow ideas that were important to me. Drs. Mary Ann Reinhart. Larry Messe. and Bob Zucker for their support. suggestions. and cooperation that contributed to making this an exciting and rewarding project. vi TABLE OF CONTENTS LIST OF TABLES ......................................... Introduction ...... ..... . ..... ... ............... ........ Previous Research by the Author ..... ..... . ......... Scale development .............................. Results ............................. ........... Develpmental approach (beliefs) .. ...... ........ Personal Religious Beliefs .............. ........ ... Attributional Style ....... ......... ................ Hassles and Uplifts ............. ............ . ...... Suggestions for Further Research . .............. .... The Current Study ...................................... Cognitive Appraisal ................................ Direct Effects of Religion ......................... Personal Meaning ........ . ......................... . Social Support ....... ..................... ......... Ego Identity ....... ................................ Gender Differences ................................ . Measures of Religion ............................... Method .......... . ........................... '..... ..... . Subjects .. ...... .. ................................. Measures ... ........................................ Religious Involvement Survey ............ ....... Hassles and Uplifts ................. ........... Psychological symptoms and well-being: MHI ..... Physical symptoms: CHIPS ......... ..... ......... Ego identity: EOMEIS-z ......................... Purpose-in-Life: PIL .......... ....... .......... Social support .............. ......... .......... Demographics .......... ..... ......... ...... ..... Procedure ............. ....... ...................... Results ..................... ......... . ............. .... Psychometric Properties of the Measures ............ Religious Involvement Survey (RIS) ............. Gender Differences in the R15 .... ............ .. DeJong Religiosity Scale . ......... ... .......... vii ix 1 3 4 6 8 9 14 16 19 22 22 23 24 25 29 30 31 37 37 39 39 41 42 44 4s 47 48 49 49 51 51 51 53 TABLE OF CONTENTS LIST OF TABLES ........ ......... ...... ....... ....... .... Introduction .................... ............... . ....... Previous Research by the Author ....... ............. Scale development .... ...... .. ........... . ...... Results 0.00..OOOOOOOOOOOOOOOOOOOOOOO 000000 .0000 Develpmental approach (beliefs) . ....... ........ Personal Religious Beliefs ........... .............. Attributional Style ...... . ..... ... ................. Hassles and Uplifts .............. .................. Suggestions for Further Research ............... .... The Current Study . .................. . .................. Cognitive Appraisal ........ . ....................... Direct Effects of Religion ......................... Personal Meaning ..... . ............................. Social Support ... ..... . .......... .. ......... ....... Ego Identity ......... . ............................. Gender Differences ..... ... ......................... Measures of Religion ............................... Method .......... . ..... .... .................. '..... ...... Subjects . .......................................... Measures ........................ ......... . ......... Religious Involvement Survey ............ ....... Hassles and Uplifts ......... ................... Psychological symptoms and well-being: MHI ..... Physical symptoms: CHIPS ......... ..... . ........ E‘O identity: EOMEIs-z OOOOOOOOOOOOOOOOOOOOOOOOO Purpose-in-Life: PIL ........................... SOCi a1 support 0 O O O O O O O O O O O O O ...... O O O O O O O O O O O O O DamO‘raphics O O O O O O O 0 O O ..... O O O O O O O O O O O O O O O O O O O 0 Procedure 0 O O O O O O O O O O O O O O O ........ O ...... O O O O O O O O O O O Resu1ts ......OOOOOOOO0.000000000000000000000.00.0000... Psychometric Properties of the Measures ............ Religious Involvement Survey (RIS) . ..... ....... Gender Differences in the R15 .... .............. DeJong Religiosity Scale .. ....... .. ..... . ...... vii 1 3 4 6 8 9 14 16 19 22 22 23 24 25 29 30 31 37 37 39 39 41 42 44 4s 47 48 49 49 51 51 51 53 59 Other religious measures ............... ....... . 62 Purpose-in-Life (PIL) . ......................... 63 Ego identity status (EOMEIS-Z) ... ........... ... 64 Mental Health Index (MHI) .. ......... .... ...... . 66 Hassles and Uplifts ................ ....... ..... 68 Medical indices .................... ............ 72 Gender Differences ........................ ..... .... 73 Hypotheses Testing and Related Analyses ............ 74 Religion directly related to outcomes for females .................................. 79 Religion directly related to outcomes for males 0.0.0.0..00.00.000.000.00.00.00.000. 80 Identity and religion ..... ..................... 81 Religion indirectly related to outcomes for females ......OOOOOOOOOIOOOOOOOOOO ........ O. 86 Religion indirectly related to outcomes for males ...................................... 92 Further Analyses ..... ...... ................ ....... . 94 Partial correlations for females ......... ...... 94 Partial correlations for males ... ..... . ........ 101 Regression Analyses .......... ................ .. 103 Discussion ............................. . ............... 110 The Sample .. .................................... ... 110 Hypotheses . ............ . ........................... 112 Identity Status ...... ........ ..................... . 115 Components of Religious Involvement ............... . 119 Other Aspects of Religion ........................ .. 123 Other Methodological Issues ....................... . 125 The RIS ..... ..... .................. . ..... .......... 128 Summary ........................................... . 130 APPENDICES ........... . ...... .. ....... .. ..... ..... ...... ’133 APPENDIX A: Religious Involvement Survey (RIS) ..... 133 APPENDIX B: Personal Background Questionnaire ...... 137 LIST OF REFERENCES ........................... . ....... .. 139 viii Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table 10: 11: 12: 13: 14: 15: LIST OF TABLES A Summary of the Major Hypotheses .... ........ 36 R15 Subscales and Reliabilities .... ..... ..... 55 R15 Subscale Intercorrelations ............... 57 I Tests by Gender Groups for RIS Subscales ... 58 Correlations Between R15 and DeJong Subscales ........................ ..... ..... 61 Intercorrelations Among the Identity Scales .. 65 Current Sample and Normative Data on the EOMEIS-z .......... .... .............. 66 Current Sample and Normative Data on the Hassles and Uplifts .......... . ..... . 70 I Tests by Gender Groups for Other Major Variables ..... ......... ... ................. 75 Intercorrelations Among the Religious and Outcome Variables .......................... 77 Intercorrelations Among the Religious and Ego Identity Variables ................. ........ 82 Frequencies of Identity Status Groups by Gender ...... .... ............. .............. 85 Results of Analysis of Variance of the R15 Variables by Identity Status Groups ........ 87 Zero-order and Partial Correlations Between the R15 and Health Variables Controlling for Hassles and Uplifts .................... 95 Zero-order and Partial Correlations Between the R15 and Health Variables Controlling for Purpose-in-Life ........................ 100 ix Table 16: Correlations of Health Habits With Religious and Health Variables ............. 105 Table 17: Stepwise Regression Analyses for the Five Health Variables ............. ........ . ..... 106 Introduction A Gallup survey of religion in America (1987) found that ninety-one percent of the respondents stated a religious preference. with sixty-nine percent of the sample claiming membership in a church or synagogue. With regard to the personal salience of religion. fifty-five percent ranked religion as very important in their lives. In a similar national poll (Gallup. 1981). almost one-third of the respondents (31%) considered their religious beliefs to be the most important aspect of their lives. Although religion is an integral part of the American culture. it is currently not a popular topic of psychological research (Byrnes. 1984). This was not always the case. At the end of the last century and through the 1920's. the psychology of religion was a much-discussed interest area and the topic of numerous empirical investigations (Ames. 1910: Coe. 1910: Leuba. 1912; Starbuck. 1899: and James. 1961). Without trying to unravel the historical causes for the decline in religious research. it is sufficient to observe that religion has not only lost its previous status in mainstream psychology. but it has also taken on a negative connotation due to its association with racial prejudice (Allport & Kramer. 1946: Allport. 1966) and psychopathology (Bergin. 1983: Sevensky. 1984). Religious beliefs by psychotherapy clients are frequently treated as defensive maneuvers by their therapists (Strunk, 1979) and the social control dimensions of religion are emphasized in professional and educational literature while the social support aspects of religion receive less exposure and study (D'Antonio et al.. 1982). Although current research has failed to replicate previous negative associations with religion. only weak links between religion and positive life outcomes have been found. and many of these have focused on global measures of life satisfaction and subjective well-being (Glik. 1986: Peterson & Roy. 1985: Witter. Stock. Okun. a Haring. 1985: Duke 5 Johnson. 1984). A concise description of the ambiguous state of affairs in religious research can be found in the tentative conclusion by Bergin. Masters. and Richards (1987) that "religiousness is not necessarily indicative of emotional disturbance." (p. 197) One line of research investigating more specific outcomes associated with religion follows the observation of Strunk (1979) that for some people. religion takes on an "important dimension that becomes an authentic coping process" (p. 194). Lindenthal. Myers. Pepper. and Stern (1970) found that mentally impaired individuals and normal controls turned to prayer with equal frequency in response to less controllable events such as catastrophes and health problems. Hay & Morisey (1985) found that religious interpretations of experiences were especially prevalent in crisis situations for their English countryside sample. Others have also found aspects of religion as mediating in individual coping processes with particularly stressful events (Zimmerman. 1984: Sanderson & Crawley. 1982: Zuk. 1959). However. the utility of religion for coping with life stressors has not received unequivocal support. For example. McLure & Loden (1982) found that while time spent on religious activities was related to overall life satisfaction. it was not related to perceived life stress. leading to the suggestion that the subjects did not use their religion to deal with everyday problems. Even assuming the tentative empirical link between religion and coping with stressful events. it is not clear how religion impacts the coping process or what particular religious variables might be involved. For instance. it has been suggested that religiosity may contribute to an individual's sense of meaning and purpose in life (Peterson a Roy. 1985) or provide a base of social support (Glik. 1986). These speculations on the precise functioning of religion as a factor in the coping process are complicated by the lack of consensus about the salient dimensions of religiosity. Research over the past twenty years have identified from two (Allport & Ross. 1967) to twenty-one (King & Hunt. 1975) different factors or dimensions of religiosity. Conversely. it has also been argued that religiosity is basically a unidimensional phenomenon (Clayton & Gladden. 1974). Ezezieee Reeeereh he the Aether My previous research (Antosz. 1988. 1989). which will be discussed. was an initial attempt to identify particular aspects of religion that may be involved in the daily coping processes of some individuals as well as to determine to what extent developmental differences may effect any relationship between religion and coping. To this latter end. the results from a late adolescent and an adult sample were compared. Specifically. it was hypothesized that religiosity functions in the overall coping process by reducing the subjective perception of certain life events as being stressful. This study attempted to isolate patterns of religious variables that would discriminate subjects who differentially perceive similar situations as stress- producing. The dependent measure was the Daily Hassles and Uplifts Scales (Kanner. Coyne. Schaeffer. & Lazarus. 1981). On the Hassles Scale. a subject can endorse any of 117 daily events as stressful and occurring within the last month. The Uplifts Scale describes 135 similar daily events that may have brought some measure of joy or satisfaction to the subject. Thus it was specifically predicted that aspects of religiosity would be negatively correlated with the number of Hassles endorsed. While there was no firm basis for a hypothesis for the relationship between religion and the frequency of Uplifts endorsed. a positive correlation was expected. §§§l§ dexeleemento The salient components of religiosity for both an adult and a late adolescent sample were investigated. To meaure the different aspects of religion for these groups. a review of the literature was undertaken to find an appropriate measurement instrument. Despite uncovering a variety of religious measures. none of them fulfilled the following criteria required for this study: (a) a brief and psychometrically sound instrument. (b) capable of measuring specific components of religiosity. (c) not specific to a particular religious denomination or Christianity in general. (d) and able to discriminate between personal and institutional religious beliefs and practices. As a result. the Religious Involvement Survey (RIS) was constructed for this study. The RIS was generally patterned after the measure used by Cornwall. Albrecht. Cunningham. and Pitcher (1985) which tapped the institutional and personal modes of religiosity along cognitive. behavioral. and affective dimensions in a Mormon sample. Due to the many significant differences in responses between the adolescent and adult samples in my study. each group was analyzed individually. While explanatory factor analysis resulted in generally uninterpretable scales. it did suggest patterns of relationships. This led to a series of confirmatory factor analyses combining conceptual and statistical parameters. and resulting in five very similar scales for both the adult and late adolescent samples. These scales consisted of measures of Personal Religious Beliefs. Personal Prayer. Church Worship. Church Beliefs. and Church Non-worship Activites. Each scale had adequate internal consistency (alphas >.80) for each sample. Despite the moderate to high intercorrelations between the scales (a range of .39 to .75). they differentially related to the other independent variables as well as to the dependent measures. the frequency scores on both the Daily Hassles and Uplifts scales. The pattern of intercorrelations of the religious scales with the other variables was different for each of the samples. Beeelte- With regard to the dependent measures of perceived Hassles and Uplifts. the late adolescent sample showed some positive correlations between the Personal Religious Beliefs scale and frequency scores on both the Hassles (g: .16. p<.05) and Uplifts (3:.25, p<.001) scales. These correlations were unaffected by controlling for the effects of measures of self-esteem and attributional style. This is particularly interesting since the measure of attributional style included attributions of control for various events to God as well as to Self. Powerful Others. and Chance. Meanwhile some of the other religion scales shared considerable variance with God-attribution scores. and thus small but significant zero-order correlations of these religious variables with the Hassles and Uplifts scales disappeared when the locus of control variables were statistically controlled. Further regression analyses of the relationship between religion and the other independent variables with frequency scores on the Hassles and Uplifts scales indicated that the Personal Religious Beliefs scale was the single most important predictor of Uplift scores. Looking at the best-fit regression models utilizing the magnitude of the multiple R as well as the relative size of the Standard Error of the Estimate (SEE). the Hassles scores were best predicted (3:.28. g squared=.06. §E§=20.92) by a combination of self-esteem scores (§§;§=—.18). Personal Religious Beliefs (beta=.16). gender (bgt§=.15). and self- attributions (bgt§=.12). Despite various regression procedures and combinations of variables. the Personal Religious Beliefs scale was the only variable to significantly predict frequency scores on the Uplifts scale (3:.27. g squared=.07. §§E=22.36). The relationship between religious beliefs to perceived daily Hassles and Uplifts was somewhat different for the adult sample. There were no significant zero-order correlations between any religous variables and the number of Hassles endorsed. With attributional style held constant. a small but significant negative correlation between Hassles and Personal Religious Beliefs was uncovered. Subsequent regression analyses indicated that the number of Hassles endorsed was best predicted by scores on the self-esteem scale and Chance-attributions (3:.46. g squared=.l9. §§§=18.59). The number of daily Uplifts was predicted by Church Beliefs and sex (8:.27. g squared=.05. §§§=26.28). Thus it appears that certain aspects of religion are related to the perception of daily minor events for the late adolescent sample but not for the adult sample. These results in general do not support the initial hypothesis that religion aids in the coping process by reducing the subjective perception of certain life events as stressful. Nonetheless. they are consistent with a developmental approach to religious beliefs and behaviors. Elkind (1970) and Fowler (1981) integrate the growth of religious thinking and identity with concurrent events in the cognitive and psychosocial development of the individual. Both Elkind and Fowler agree that the onset of formal operational thinking in adolesence contributes to the development of religious thinking in the individual. The adolescent's ability to reflect on and question childhood religious beliefs and practices can be utilized to find solutions for current conflicts and in the process. forge a more complex and personal religious identity. This search by adolescents for a personal faith more consistent with their overall outlook on life is why Hurlock (1973) contends that adolescent attitudes and beliefs rather than their religious practices are indicative of their current interest in religion. This would explain why surveys of high school and college students like those by Nordin (1972) and Conger (1973) reveal a discrepancy between reports of church attendance and the importance of religion to the adolescents. Fowler (1981) contends that progress toward religious identity. like other aspects of the individual's movement toward general psychosocial identity achievement. is related to an increase in cognitive structures. social interactions and experiences with arising conflicts. If strong personal religious beliefs are assumed to be some marker of progress in the achievement of religious identity. it should be expected that religiosity functions like other accumulated psychosocial strengths in supporting positive adaptational outcomes for the individual. As argued by Kanner et al. (1981) and Weinberger et al. (1987). the Hassles and Uplifts Scale serves as an effective predictor of current and subsequent psychological problems and physical symptoms. If the above assumptions about religious identity formation are correct. then the strong association between Personal Religious Beliefs and frequency of reported Uplifts by the late adolescent sample would be expected. As conceived by Kanner et al.. the daily Uplifts are positive experiences that can possibly serve "as emotional buffers against stress disorders." (1981. p. 6) Thus the data from the Antosz (1989) sample suggests that personal religious beliefs could play a role for some late adolescents in buffering them from the stress associated with their ongoing attempts to cope with developmental tasks. However. the less robust but still significant relationship between Personal Religious Beliefs and frequency of Hassles reported does pose an apparent problem to this conceptualization. 32259251 321181995 §§li§£§ Some solution to this dilemma can be found by looking f , ggggg , 10 beyond the mere frequecy scores of the Hassles and Uplifts scales and analyzing the specific items endorsed. comparing the responses of those scoring high on the Personal Religious Beliefs scale with responses of those scoring at the low end. The highest and lowest thirds of the sample were utilized in lieu of a median split in order to more sharply highlight differences between groups. Regardless of the comparison method. there is a larger percentage of females in the high Personal Religious Belief groups. However. a Chi-square analyses of these differences are non- significant. with less discrepancy between the percentages of men and women in groups occurring in the top and bottom thirds of the Personal Religious Belief scores. While there does not appear to be a clear pattern of inter-group differences on the Hassles Scale. some of the items more frequently endorsed by the group high in Personal Religious Beliefs tend to conceptually cluster together.' For instance. those higher in Personal Religious Beliefs express more concerns over living arrangements (e.g.. "home maintenance (inside)" and "neighborhood deterioration"). Other items refer to financial concerns such as "financial security." "financial responsibility for someone who doesn't live with you." and "cutting down on electricity. water. etc." Other Hassles for this group seem to revolve around issues associated with personal reflection such as "the meaning of life." "being exploited." "inner conflicts.’ "regrets over past decisions." and "getting ahead." As a 11 whole. the Hassles more frequently endorsed by those with stronger Personal Religious Beliefs are not atypical of the issues confronting late adolescents attempting to individuate from their families and face the realities of an adult world. The differences between the high and low Personal Religious Beliefs groups are more striking on the Uplifts Scale. There are statistically significant differences on items pertaining to health. with the high Personal Religious Beliefs group more frequently endorsing items like "staying or getting into good physical shape." "getting enough sleep." "feeling healthy." "relaxing." and “having enough (personal) energy." It would also appear that this group finds successful. active coping as a source of positive feelings (e.g.. "meeting a challenge." "confronting someone or something." "making decisions." "past decisions panning out." "resolving conflicts over what to do." "thinking about the future." "resolving inner conflicts." "being efficient." "capitalizing on an unexpected opportunity." and "meeting your responsibilities"). Additionally this group of adolescents is more likely to find their daily perks in altruistic acts ("giving a compliment." "doing volunteer work"). and mastery-related behaviors ("using skills well at work." "practicing your hobby." and "fixing/repairing something (besides at your job)"). Some of the specific Hassles reported more frequently by the late adolescents high in Personal Religious Beliefs are 12 also related to a source of gratification. For instance. these adolescents tend to report being "concerned about the meaning of life" and yet they more frequently endorsed "life being meaningful" as a daily uplift. Similarly. "home maintenance (inside)" is a frequent Hassle for these adolescents. yet there is the bonus of the "home (inside) pleasing to you." "Financial security" is a source of concern for this group. so "getting unexpected money" more frequently becomes an unanticipated joy. The last point suggests the possibility that the expectations of this particular group of adolescents. namely those high in Personal Religious Beliefs are less idealistic and more realistic. For example. a young person actively in the process of separating from his or her family might have more opportunity to experience the financial insecurities of self-support. Although this group’s personal religious beliefs may provide them with a set of ideals to strive for. they seem to better appreciate real-world inequities and thus are pleasantly surprised by "finding no prejudice or discrimination when you expect it." Finally the largest discrepancies on the Uplifts items between the high and low groups on the Personal Religious Beliefs scale are on "praying" and "meditating." This is not unexpected since there is a very high correlation between the Personal Prayer and the Personal Religious Belief scales. In fact. combining these two scales into a "personally religious" scale and entering it into the 13 regression equation for predicting the number of uplifts endorsed. results in a slightly larger multiple 3 than just using the score on the Personal Religious Belief scale. In summary. then. the group of adolescents who espouse stronger Personal Religious Beliefs seem to experience more daily Uplifting events than their counterparts scoring lower on this particular religous variable. However. for this high personal belief group. most positive experiences do not just "happen" to them: rather they actively effect positive events. In support of this. one of the two uplift items that the less religious group more frequently endorsed was "being visited. phoned. or sent a letter." which would seem to be a fairly passive source of gratification. (The only other item in the uplifts scale endorsed significantly more frequently in this group is "sex.") Overall. it appears that the high Personal Religious Belief group of late adolescents is actively working on tasks developmentally appropriate for this age group and finding some joy and gratification in doing so. However. given the instructions for filling out the Uplifts scale ("circle the events that made you feel good 12 the 2552 mggtg"). no conclusion can be made as to whether or not the the less personally religious group is also engaged in similar tasks. Even if this group of late adolescents is developmentally similar to the high belief group. it would appear that they experience little gratification in dealing with the issues of this age period as indicated by endorsing 14 fewer of these developmentally appropriate events as Uplifting. The high Personal Religious Belief group's active involvement would seem to provide considerable benefits in the form of personal successes and satisfactions but not without some cost in the increased number of stressors they experience in confronting their important developmental tasks. Attributienel Stxlee Examining the attributional styles of this late adolescent group. particularly in relationship to Personal Religious Beliefs. the results are as expected. The low Personal Religious Belief group is significantly higher (gz4.23. p<.001) in Person (self) attributions. and significantly lower (i=9.36. p<.001) in God attributions than the high Personal Religious Belief group. The high belief group is somewhat. though not significantly (t=1.70. p<.09) lower on Chance attributions and about the same in Powerful Other attributions as the low belief group. Breaking down the scores on the locus of control variables of the entire late adolescent sample into high. medium. and low. the high belief group scored low-high-medium-medium on the Person. God. Chance. and Other variables respectively. By contrast. the attributional profile for the low belief group is high-low-medium-medium. To view the locus of control variables from a more traditional perspective (Internal-External). the scores for the God. Chance. and Powerful Other attributions can be summed up to give a 15 general External score. Comparing the high and low religious belief groups according to this dichotomy. the high belief group is significantly lower on the Internal dimension (t:-4.23. p<.001) and significantly higher (i=4.47. p<.001) on the External locus of control dimension than the low belief group. These findings are consistent with previous studies of religious commitment and causal attribution (Gorsuch & Smith. 1983: Ritzema. 1979). However. given the usually favorable traits and outcomes associated with individuals with a high Internal and low External attributions. the differences between the high and low belief groups on the Hassles and Uplifts scales would seem paradoxical. A posssible resolution for this dilemma can be found in studies looking at religion and attributions. Gorsuch and Smith (1983) emphasize the distinction between "cause" and "responsibility" for an event (p. 349). Utilizing written vignettes varying on severity of outcome. Gorsuch and Smith found that the more religious subjects attributed greater effort to the actors in stories with extreme outcomes even though they did not simulataneously attribute more responsibility for the outcomes to the actors. Gorsuch and Smith concluded that in events with severe outcomes. the more religious subjects may view a person as the direct cause of an outcome by virtue of intention and individual effort while still attributing ultimate responsibility for the outcome to God. Such a formulation would be consistent 16 with the notion of God being responsible for an outcome when "a person functions as God's 'agent'" (Gorsuch a Smith. 1983. p. 349). The particular locus of control measure used in the study by Antosz (1987) includes extreme (getting "into a car accident") and ultimate ("...what will happen in my life") outcomes and no distinction is made between cause and responsibility. Looking at God attributions in relationship to problem- solving. Pargament et al. (1988) concluded that "the content of 'God control' is a multi-dimensional one" that can relationship with God. (p. 102) Each of these perceived personal relationships with God are differentially related to measures of personal competence. At least in respect to their sample. Pargament and his collegues found that the collaborative style (working with God) was more frequently endorsed and was associated with higher personal competency scores than the deferring (manipulation by God) style. fleeelee see flelifie While these recent data point to a definite association between some religious variables and one stress indicator (the Daily Hassles & Uplifts Scale). a functional relationship between these religious variables and effective coping with stress has not been demonstrated. Although there is some initial support for the utility of the Hassles & Uplifts Scale as a predictor of psychological (Kanner et al.. 1981) and physical symptoms (Weinberger et al.. 1987: 17 Kanner et al.. 1981: Monroe. 1983: Zarski. 1984: Holahan et al.. 1984). the case for a causal link between major or minor life events and reported symptoms is hardly a consensual matter (e.g.. Grant et al.. 1987). Even allowing for a hypothetical causal relationships between the daily Hassles and current and subsequent symptomology. it is still not clear how strong personal religious beliefs would be related to an individual's coping system. Some hints in this area are provided by certain Uplift items that were more frequently endorsed by the group with high Personal Religious Belief scores. For instance. "praying" & "meditating" were two Uplifts items more often endorsed by the high religious belief group. It is possible that certain religious behaviors like these can serve an ameliorative role with experienced stress by dampening its cumulative impact on the individual (Holahan a Moos. 1987; Aldwin & Revenson. 1987). The two items on the Personal Religious Belief scale that most strongly correlate with the frequency score on the Uplifts scale are "I try to carry religion over into all my dealings in life" and "I do not think about my personal religious beliefs very often" (reversed scored). Another item on the religious questionnaire that did not load on any of the five subscales but that correlated highly with the number of Uplifts endorsed was "as a rule. I do not share my personal religious beliefs with others" (reversed scored). It is possible that religious beliefs play some preventive 18 function in the cognitive appraisal of potentially stressful events (Pearlin et al.. 1981: Folkman a Lazarus. 1980). The group with high Personal Religious Belief scores more frequently endorsed the Uplifts item. "life being meaningful." Again. while these are only correlations. it has been suggested by other researchers that personal faith can be a salient framework for providing direction in the daily affairs of some individuals (Zika & Chamberlain. 1987: Ben-Sira. 1985: Hadaway & Roof. 1978). It is interesting to note that while for the high Personal Religious Belief group none of the items on the Church Beliefs scale are related to the Uplifts Scale. two items on the religious survey with very strong correlations with the frequency score of the Uplifts scale pertain to this group's affiliation with an institutional religion ("the church is not a very important part of my life". which is reverse scored. and "I believe my church is the true religion"). This would suggest that in addition to their personal faith. members of this particular group have also made a strong commitment to an organized religious body. These descriptions of a strong commitment to and involvement with religious beliefs and institutions as well as the satisfaction gained from confronting developmentally appropriate tasks appear similar to the components of commitment. control. and challenge that characterize the notion of "hardiness" in coping with stress as conceptualized by Kobasa. Maddi. & Courington (1981). 19 §esaee£iee§ £2: Eertbe: Eeeeereb The study just described (Antosz. 1988: 1989) suggests some interesting relationships among one aspect of religiosity and the psyhosocial development and coping processes of late adolescents. The data identify a link between Personal Religious Beliefs and the cognitive appraisal of daily events. However. there is an inherent paradox between the item analysis and overall scoring of the Hassles and Uplifts Scale. which represents the dependent variable. Based on the descriptive analysis presented. it seems that those late adolescents reporting stronger Personal Religious Beliefs appear to be dealing more effectively with developmentally appropriate tasks. As a result. it is tempting to predict that this group would thus score better on measures of adaptational outcomes like physical and mental health. However. as indicated. the group high in Personal Religious Beliefs endorsed significantly more items on the Hassles as well as the Uplifts subscales than their less religious counterparts. According to previous research with this instrument. frequency scores on the Hassles subscale have been positively related to physical (weinberger et al.. 1987: Zarski. 1984: DeLongis et al.. 1982) and psychological (Holahan et al.. 1984: Monroe. 1983: Kanner et al.. 1981) symptoms and negative well-being (Zika & Chamberlain. 1987). Additionally for females. higher scores on the Uplifts subscale have also been related to increases in negative 20 affects and psychological symptoms (Kanner et al.. 1981). Thus in light of these other findings with the Hassles and Uplifts Scale. an alternate prediction might be that Personal Religious Beliefs are positively related to poor adaptational outcomes. A closer look at the data may lessen the appeal of this latter prediction. For instance. while both Hassles and Uplifts scores were significantly related to Personal Religious Beliefs in the late adolescent sample. the relationship of this scale with Uplifts was stronger than with Hassles. As previously summarized. the regression analysis predicting Hassles frequency scores included Personal Religious Belief. sex. self-esteem. and self- attributions. However. the variable of Personal Religous Beliefs was the sole predictor of Uplifts scores. Given the previously described item analysis of the Hassles and Uplifts. it is possible that the elevated Hassle scores of the group high in Personal Religious Belief is reflecting some of the stress associated with this group's increased involvement in the developmentally appropriate tasks of late adolesence. However. as previously noted regarding this group. they also appear to be experiencing more joy and satisfaction in dealing with these developmental tasks than their less religious counterparts. A possible relationship between developmental issues and scores on the Hassles and Uplifts Scale is also supported by the fact that the pattern of correlations 21 between Personal Religious Beliefs and Hassles and Uplifts for the late adolescent sample was not found in the adult sample (Antosz. 1989). Additionally. the positive relationship between Uplift scores and negative affects and psychological symptoms reported by Kanner et al. (1981) was based on a middle-aged sample. The results of my previous research seem to suggest that religion may be related to the coping process in a particular developmental group. specifically late adolescents. However. to gather further support for this notion. another study with a similar sample of late adolescents will need to not only replicate the relationship between religion and the perception of minor daily events but also go one step further and examine the relationship between the perception of stress and health outcomes. In addition. much of the previous speculative discussion about the relationship between religion and relative progress on developmental tasks calls for a more direct measure of developmental status to substantiate these ideas. Finally. while some initial relationship between religion and the perception of minor daily events has been identified. the specific components and processes by which religion effects the coping mechanisms of individuals need to be further elucidated and tested. The Current Study In this study the relationship between religiosity and adaptative outcomes is more directly investigated utilizing measures of psychological and physical health. Speculations stemming from the previous data about the specific coping process associated with religiosity are tested. Additionally. the relationship of religiosity with other aspects of ego identity development in late adolescents are . also examined in more detail. Finally. more psychometric data on the Religious Involvement Scale (RIS) developed for the previous study are gathered to determine the overall utility of the R15 as a research tool. After examining religiosity and adaptational outcomes. Peterson and Roy (1985) concluded that the nature of the relationship between well-being and religiosity may vary depending upon the aspect of well-being and the dimension of religiosity under consideration. Other support for this notion comes from the previous study (Antosz. 1988) which found significant relationships between two of five religous variables (Personal Prayer and Personal Religious Beliefs) and the cognitive appraisal of positive daily events as measured by the Uplifts subscale. With a sample of college undergraduates. Cohen and Hoberman (1983) found that on a measure of depressive symptoms. positive life events acted as a buffer against the stressful impact of negative life 22 23 events. Cohen and Hoberman found a weaker buffering effect for postive life events with physical symptoms. Considering that others have also postulated a stress-buffering role for positive life events (Lazarus.Kanner. & Folkman. 1980: Reich & Zautra. 1981). it is predicted that the religious variables of personal prayer and personal beliefs will have an indirect effect on Psychological Well-being and physical symptoms through the cognitive appraisal of daily life events. More specifically. it is expected that these two religious variables will be directly related to the number of positive life events reported which in turn will be positively related to a measure of Psychological Well-being (Hypoth. b1) and inversely related to measures of psychological distress and physical symptoms (Hypoth. b2). (These and the following hypotheses will be summarized later.) 21129.91 Eiieeie 9.12 Beliaiee In addition to the indirect effects of these religious variables. it is also predicted that they will have direct effects on the health outcome measures. Specifically. the Personal Prayer variable seems to tap a positive. affective component of religiosity. As a result. high scores on this religious subscale should be directly related to higher scores on the Psychological Well-being scale which measures positive affect as well as other variables (Hypoth. a1). The items on the Personal Religious Beliefs subscale. on the other hand. overtly refer to cognitive and behavioral 24 components of religiosity. From the previous analysis of this subscale's relationship to certain items on the Uplfits subscale. it was noted that individuals scoring high on this variable endorsed more health-related behaviors like good sleep and exercise habits and less use of drugs. Because of this connection between certain religious beliefs and health promoting behaviors. it is predicted that scores on the personal religious beliefs variable will have a direct and inverse relationship with the number of physical symptoms reported (Hypoth. a2). In the previous analysis of the data from the Antosz (1988) study. some speculation was made about the possible functional relationship between strong personal religious beliefs and an individual's general coping system. The constructs of personal meaning and social support are examined in the current study as possible explanatory variables in this relationship between religiosity and coping. Eereeeel Meeeies One of the more frequently mentioned functions of religion is the sense of purpose and direction that it provides for the individual (Peterson 5 Roy. 1985: Hadaway & Roof. 1978: Grossman. 1975: Greeley. 1972: Allport. 1950). Davidson (1972) reviews the suggestions of other writers regarding religion's ability to provide individuals "with meaningful explanations (e.g.. God's will) for events and conditions (e.g.. death) which may be difficult to explain 25 at the natural level alone." (p. 66) In addition to explaining events. Spilka. Shaver. and Kirkpatrick (1985) posit that religious belief systems can "satisfy the individual's need or desire to predict and control events" (p. 8). The lack of purpose and meaning in life have been implicated in maladaptive outcomes like drug addiction (Newcomb & Harlow. 1986) and alcoholism (Jacobson. Ritter. & 1 Mueller. 1977). Thus. some treatment programs for these chemical dependencies have specifically focused on the a construct of life meaning (Gruner. 1984: Crumbaugh. 1981). As a result. it is predicted that in the current study. personal religious belief will be positively associated with self-reports of meaning and purpose in life which in turn will be positively associated with psychological well-being (Hypoth. b3) and inversely related to psychological distress (Hypoth. b4). If religion does function in the coping process by instilling meaning and purpose in the lives of individuals. then the direct relationship between psychological health and personal meaning should diminish when the effects of religion are partialled out. §9§iél §HREQEI Some of the effects of social support are similar to those resulting from having a sense of purpose and meaning in life. In a survey study. Klinger (1977) found that less than 502 of those polled cited religious faith or occupational success as an important source of meaning to them. However almost all the respondents mentioned 26 relationships with others. the feeling of being loved and wanted. as a major factor in making their lives meaningful. In studying the religious needs of college students. Pargament. Enchemendia. Johnson. and McGath (1984) found that regardless of the level of religious involvement. a significant proportion of students cited belonging to a group ”which makes one feel accepted and loved" as an important factor in selecting a religious group to join (p. 277). Bruhn and Philips (1987) identify personal faith and and the anticipation of positive events (hope) as two closely related themes that extend throughout the life span and help shape an individual's perception about giving and receiving social support. From this developmental point of view. religiosity may effect how social support behaviors are learned and utilized in other life areas. Thus it would appear that the social support function is an important part of religion. suggesting that religious variables other than personal belief and personal prayer may be related to the coping process. There is already support for the positive relationship between attendance at religious services and involvement in religous activities with psychological well-being (Bruhn & Philips. 1987: Witter et al.. 1985: Bergin. 1983: McClure & Loden. 1982). In a health survey of a large undergraduate sample. Comstock and Slome (1973) unexpectedly found statistical differences in the reporting of four health-related problems between female students with no stated religous affiliation 27 and those claiming any religious affiliation. Specifically. the no-affiliation group reported greater frequencies of bad drug experiences. contraceptive needs. upset stomachs. and respiratory problems. While the social control function of religious beliefs may account for the inter-group differences in the first two problem areas (Studer & Thornton. 1987: Gruner. 1984). it is not clear which religious variables. if any. are associated with the last two somatic problems. Since Comstock and Slome (1973) did not find the same patterns of results for male students. it is possible that there are gender differences in the effect and importance of the social support aspects of religiosity. Gender differences in the pattern of intercorrelations among the religious variables in the Antosz (1988) study suggest that the social aspects of religious affiliation are more important for females than males. For instance. the variable of Church Worship. which taps both private and social aspects of religion. correlated more highly with Personal Prayer for males while it was more closely associated with Church Nonworship Activities for females. In a non-religious context. Walker and Greene (1987) found that social support operated differently for male and female adolescents. For the males. peer social support served a buffering effect in relation to negative life events and psychopyhsiological symptoms. Such an interaction was not found for females. It appeared that 28 while high levels of social support did not protect females from higher frequencies of stressful life events. low levels of peer support were associated with high symptom levels regardless of the frequency of negative life events. These results are consistent with a main effects interpretation of the relationship between peer support and symptoms (Cohen a Wills. 1985). In their review of the literature on social support. Cohen 5 Wills (1985) also concluded that various support functions are differentially effective in buffering stress for males and females. For instance. studies by Husaini. Neff. Newbrough. & Moore (1982) and Henderson (1981) found that confidant support acted as a stress buffer for females but not for males. Meanwhile. Henderson. Byrne. Duncan- Jones. Scott. & Adock (1980) found a buffering effect of companionship activities for males but not for females. Although females may be more prone to become socially involved in organized religion. it would seem possible that both males and females who are socially integrated into a religious group could benefit from the various social support functions available. Thus it is predicted that scores on the Church Nonworship Activities variable and on a short measure of social interaction within a religious group will be positively correlated with Psychological Well-being scores (Hypoth. a3) and negatively correlated with Psychological Distress scores (Hypoth a4). Yet it is expected that the positive effects of this relationship 29 between church involvement and Psychological Well-being will be seen for a higher proportion of the female subjects. Given the theoretical speculation about the links between social support and physical illness (Krantz. Grunberg. & Baum. 1985: Jemmott & Locke. 1984: Cohen & McKay. 1984: Gore. 1981). it is predicted that the same measures of social involvement in religion will be negatively associated with physical symptoms (Hypoth. a5). It is possible that this latter relationship will not be as strong for females. Walker a Greene (1987) suggest that peer support may not effectively discourage adolescent females from somatization in response to increased levels of stress. Ese Identity One of the tentative conclusions drawn from the previous study by this author (Antosz. 1988) as previously discussed was that those students reporting a high level of personal religious beliefs appeared to be more actively engaged in behaviors associated with the developmental tasks of late adolesence. Since this conclusion was based on individual items endorsed on the Uplifts subscale. it is not clear whether this particular group was actually confronting these developmental issue more frequently or whether they simply derived more satisfaction or joy from engaging in these activities. Parker (1985) suggests that personal religious development may parallel the general ego identity development of individuals. He further states that overall ego identity “can be and often is expressed in the 30 development of religious values." (p. 47) Using Fowler's (1981) stages of religious development and Marcia's (1966) identity statuses. Mischey (1981) found a strong relationship between the stages of religious and ego identity development. However. even Parker (1985) admits that identity formation in adolescents does not always occur simultaneously in all domains. a premise that receives support from others (Thorbecke & Grotevant. 1982: Coleman. 1974. 1978). QQBQEI Qi££§E§BE§§ This last notion about the unevenness of ego identity development across various life domains is very compatible with the notion of gender differences in adolescent development. Josselson. Greenberger. & McConochie (1977a. 1977b) describe how high school females differ from males in their use of interpersonal relationships to facilitate personal differentiation. increase self-esteem. and relieve anxiety. Hodgson and Fischer (1979. 1981) concluded from their studies of college students that males are more developmentally advanced in occupational choices while females seem to be further along in sex ideology and sex- role conceptualizations. Alishio and Schilling (1984) found college-age males focusing on occupational issues while females focused upon interpersonal and sexual issues. Gender differences in achievement orientation and in interpersonal relationships has been uncovered by others (Stein 5 Bailey. 1973: Rosenberg & Simmons. 1975) and have 31 been highlighted by Gilligan's (1982) criticisms of current developmental theories to adequately account for these differences. In response to the burgeoning evidence for gender differences in adolescent identity development. Grotevant. Thorbecke. and Meyer (1982) have extended Marcia's (1966) identity status interview. which focused on ideological and occupational content areas. into the interpersonal domain covering areas of friendship. dating. and sex roles. Assuming that there are intra-individual and gender differences in the developmental processes of late adolescents. it is predicted that for the group scoring high on the Personal Religious Beliefs variable. the females will more frequently approach or attain the identity achievement status as defined by Marcia (1966) in the interpersonal domain (Hypoth. c1). The males in this high belief group are expected to be more developmentally advanced in the ideological domain (Hypoth. c2). It is expected that the males and females in this high personal belief group will differ from each other and from their same sex counterparts in the low religious belief group according to these patterns. 59552225 91 39118192 The final objective of the proposed study is to gather psychometric data for the Religious Involvement Survey (RIS) developed in the previous study by Antosz (1988). While there is no shortage of measures of 32 religiosity. there is a need for a psychometrically-sound instrument that can serve as an effective research tool. One of the most popular means of measuring religiosity is the use of one or two forced-choice questions usually focusing on church attendance. formal religious affiliation. or self-rated religiousness. Campbell and Coles (1973) argue that religiosity and religious affiliation are relatively independent dimensions of religion. and surveys that rely solely on church membership data can result in misleading inferences about religious factors. In their review of such national surveys. Carroll and Roozen (1973) identify serious shortcomings with such measures including the "fuzzy" definition of religion. a unidimensional approach to measurement of a complex construct like religion. and a bias toward traditional religious institutions. (p. 332) Next to such single item measures of religiosity. the next most prevalent assessment tool has been Allport‘s (1966) Religious Orientation Scale (ROS). Allport conceived of an Intrinsic (I) and Extrinsic (E) orientation to religion. An extrinsic orientation refers to an instrumental motivation for religious behavior that includes using religion to attain goals like personal solace and social standing. An intrinsic orientation. however. describes an ultimate motivation for religious belief and practices. whereby religion is a value in and of itself. Through 1984. the ROS has been used in nearly seventy 33 published studies (Donahue. 1985). One of the early critiques of the ROS by Hunt and King (1971) argued that the I-E orientation was not a unidimensional. bipolar contimuum as conceived by Allport (Allport and Ross. 1967). but rather contained several component variables. Hunt and King felt that while the Extrinsic orientation was clearly defined as a selfish. utilitarian approach to religion. the Intrinsic orientation was not operationally defined. Despite Donahue's (1985) defense of the ROS in his review and meta- analysis of research on the instrument. it is clear that while the Extrinsic subscale has been consistently correlated with variables like prejudice and dogmatism. the Intrinsic subscale has only been related to other measures of religiosity. thus supporting Hunt and King's (1971) original criticism about this scale's limited research utility. Batson and Ventis (1982) built on Allport's (1966) I-E scale by adding a third orientation. that of Quest. According to Batson. Quest is "an approach that involves honestly facing existential questions in all their complexity. while resisting clear-cut. pat answers" (Batson & Ventis. 1982. p. 149). Although Batson & Ventis present studies exploring the conceptual utility and psychometric properties of their revised version of the Religious Life Inventory (RLI-R: Batson & Ventis. 1982). there are still serious misgivings about this scale and its theoretical underpinnings (Finney 5 Maloney. 1985: Hilty. Morgan. & 34 Hartman. 1985: Hood 5 Morris. 1985: Spilka. Kojetin. 5 McIntosh. 1985). King and Hunt (1972. 1975) and others (Cornwall. Albrecht. Cunningham. 5 Pitcher. 1986: DeJong. Faulkner 5 Warland. 1976: Himmelfarb. 1975: Law. 1974: Glock 5 Stark. 1965) have constructed scales that measure religion as a multidimensional construct. However. these scales are either specific to a particular denomination or to mainstream. traditional Christianity. thus limiting their use and the generalizability of their results. Additionally. these scales were constructed with the common objective of more closely studying the structure of religious belief and practice. with litle or no emphasis on developing an effective research tool that could be used with non-religious variables. Other more topic-specific measuring instruments like Caird 5 Law's (1982) scale for non-conventional religious beliefs and Hunt's (1971) LAM (literal. anti-literal. and mythological) scale measuring religious meaning and commitment have been developed with little follow-up work or research application while others like Hood's (1975) Mysticism scale have at least been utilized in later research. The current study attempted to refine the Religious Involvement Survey (RIS: Antosz. 1988) and systematically collect data on the psychometric properties of the scale. Reliability was assessed with measures of internal consistency and item-total correlations. Test- 35 retest reliability were gathered over a one month period with a different sample. Construct validity was determined through confirmatory factor analysis. Discriminant validity was tested through the differential relationship of the R15 subscales to the current study's outcome variables. Concurrent validity was investigated by including another established religious measurement tool in the study. A moderately high correlation between the two religious measures. with the R15 displaying better predictive ability with the outcome measures would be reasonable support for continued development of the R15. In summary. for the current study. several hypotheses have been made regarding religious variables and the adaptational outcome measures of physical and psychological health. These hypotheses are summarized in Table 1. In the process of testing these hypotheses. further psychometric data on the Religious Involvement Inventory was systematically collected. 36 Table 1. A Summary of the Major Hypotheses. Iaeeeeaeeat Qireetiee e: Qereaeeet Eerieble Eiieet Eerieble Ao Qireei 5119215: al. Prayer Positive Psychol. well-being a2. Personal Negative Physical symptoms Religious Beliefs a3. Church Activi- Positive. Psychol. well-being ties. social integration a4. Church Activi- ties. social integration as. Church Activi- ties. social integration B. Indirect Effects: b1. Prayer and Personal Religious Beliefs b2. Prayer and Personal Religious Beliefs b3. Personal Religious Beliefs b4. Personal Religious Beliefs c. Gender ere Idea: c1. Personal Religious Beliefs (Hi Belief) c2. Personal Religious Beliefs (Hi Belief) especially for females Negative Negative. especially for males Positive through Hassles 5 Uplifts Negative through Hassles 5 Uplifts Positive through Purpose-in-Life Negative through Purpose-in-Life 11! 51199153 Positively related for females Positively related for males Psychol. distress Physical symptoms Psychological well-being Psychol. distress Physical symptoms Psychol. well-being Psychol. distress Identity Achievement (Interpersonal) Identity Achievement (Ideological) 89:899. §BEJ§EE§ The late adolescent sample (3:440) utilized in this study was comprised of male (g=139) and female (nz300) college undergraduates recruited from psychology classes at a large midwestern state university. Although a strictly college sample limits the generalizability of results to all similar-aged adolescents. there are some important benefits in utilizing this particular kind of sample. The aim and (the results of the previous study by this author support the use of a similar sample group to avoid introducing external confounds. The late adolescents of a college sample often are experiencing their first extended separation from their families and facing the complexity and anonymity of a large university setting. The milieu for this group may be quite different than that for adolescents who choose alternate routes upon graduation from high school. Additionally. the presence of a student health center can serve as a somewhat uniform criterion for measuring one type of illness-related behaviors. The target size of the sample was six hundred subjects to provide an adequate pool of responses with which to test the measurement model especially since gender differences on some of the variables were hypothesized. Characteristically. in undergraduate samples of this nature. female volunteers are often overrepresented. Thus to compensate for the smaller percentage of male subjects. a 37 38 larger. overall sample size was planned to insure an adequate sampling of males. Subjects were recruited from various sections of introductory psychology courses as well as from two smaller upper level psychology classes. The student volunteers were unpaid but received research credits that applied to their course grade. All the students participated during the Spring Term of 1988. Because of the limited size of the available subject pool at this time. only four hundred and forty subjects were recruited. The composition of the resulting sample was 300 females and 139 males. with the data for one subject uncoded for gender. The age range of the sample was from 17 to 23. with a mean age of 19.2 years old. 83% of the students were either in their first or second year of college. The subjects were predominantly white (89%) and almost all (96%) lived away from home during the school year. 15.7% of the students reported that they were g9; currently affiliated with any formal religious organization. 45.52 of the sample identified themselves as Catholic. 30% as a variety of major Protestant denominations. 5.92 as Jewish. and the remaining as inter-demoninational or nondenominational Christians. Given the method of voluntary recruitement through a predetermined human subject pool in a college setting. there is always some question as to the representativeness of the sample with respect to critical variables. For this particular sample. there are some differences in religious 39 affiliation in comparison to national surveys for this age group (Gallup. 1987). While the figures for affiliation with the Catholic church or various Protestant denominations are different in a national poll (Protestantz47z and Catholic=332). the percentage of no stated religious affiliation among this age group is similar (14%). The religious affiliation of the sample recruited for the previous study by this author (1988) through a recruiting process similar to that used for the current study. the religious affiliation broke down into 362 Catholics and 32% Protestants. with 182 reporting no religious affiliation. Thus it appears that the current sample is more heavily represented by Catholics. but generally the same percentage of students in this sample in comparison to a national sample are reporting some religious affiliation. Meeeeree Eelisieee Inxelxeeen: §erxex. A revised version of the Religious Involvement Survey (RIS) developed for the last study to measure the independent variable of religiosity served the same function in the current study (Appendix A). Although only seventeen of the original forty-seven items of the R15 were ultimately employed for the five religious scales in the previous study. other items from the initial pool were also used in this study. The five scales from the previous study were derived from four interpretable factors of an exploratory factor analysis (principal components analysis with a 40 varimax rotation) of the R15. Thirty-five of the initial 47 R15 items loaded on the initial four factors that were then further refined using a confirmatory factor analysis routine. For the current study. the 35 items of the four initial factors were retained. The twelve items that were deleted included Batson's Quest items. which did not correlate well with other RIS items or with each other. and most of the "consequential" items (behavioral manifestations of religious belief other than public worship activities). Consequential items more appropriate to this specific—age sample were written and included with the remaining 35 R15 items (for example. "my personal religious beliefs are reflected in the way I treat the people around me. both friends and strangers"). In addition. some of the existing items were slightly modified. All of the items referring to beliefs and practices of organized religions were given a "No Church" response alternative to differentiate those individuals who state a specific religious affiliation but score low on these formal religion items from those subjects who do not claim membership in any organized religion. Given the importance of the religiosity variable in this study and the basically untested psychometric properties of the RIS. another. more established measure of religiosity was also included. For the purposes of this study. the DeJong et al. (1976) Religiosity Scale seemed most appropriate. considering its multidimensional approach to religiosity and its empirically derived factor structure. 41 which are characteristics of the RIS. This 38-item scale was developed by DeJong. Faulkner. and Warland at Pennsylvania State University. The original samples used for the scale construction consisted of 542 American and 400 German college undergraduate students. Based on factor analyses. six similar subscales were identified for both samples and included the religious dimensions of belief. knowledge. experience. practice. individual moral consequences and social consequences. Second-order factor analyses combined the belief. experience. practice. and individual moral consequences into a generic religiosity factor. The remaining two subscales of religious knowledge and social consequences of religion remained relatively independent higher order factors. While DeJong et al. (1976) do not report any reliability coefficients. a recent unpublished study using 155 college undergraduates computed Cronbach alphas on the six subscales that ranged from .54 to .89. The total scale alpha was .89 and the summed reliability coefficient for the belief. experience. and practice subscales was .93 (Eckert. 1984). The data from the original samples gathered by DeJong and his collegues have been subjected to cluster analyses to determine an empirical taxonomy of religious indviduals (Filsinger. 1981: Filsinger. Faulkner. 5 Warland. 1979). §é§§l§§ £29 QEliIE§° While the Hassles and Uplifts scales were not the major dependent measures in the current study. they were included 42 to test the hypotheses that the perception of daily events is a mediating process in the relationship between religion and physical and psychological outcomes. While some studies (Kanner et al.. 1981: Weinberger et al.. 1987) have not shown a clear association between scores on the Uplifts scale and stress-related outcomes. the particular relationships that emerged betweem the Uplifts scale and the personal religious belief variable in the previously described study (Antosz. 1988) justified this subscale's use in the current study. In fact. the unexpected results from the last study required some form of replication with this scale. The Hassles subscale was also included since a smaller but still significant relationship was observed between the the frequency of hassles endorsed and the personal religious belief variable. Eexeheleaieel exerteee -99 -- The major dependent variables of psychological and physical symptoms were measured by the Mental Health Index (MHI. Veit 5 Ware. 1983) and a brief questionnaire/checklist of somatic problems and perceived physical health status constructed for this study. The MHI is a 46-item questionnaire measuring psychological distress and well-being during the past month. This instrument has 8 imbedded items controlling for socially desireable response sets. Scoring results in 2 higher order factors of positive and negative subjective well-being and 5 lower order factors including anxiety. depression. emotional ties. general 43 positive affect. and loss of behavioral control. This instrument was developed as part of a larger battery of tests to assess general health status in the Health Insurance Study by the Rand Corporation under a grant by the 0.5. Department of Health. Education. and Welfare. Beginning in 1974. over 8000 individuals in 2750 families were enrolled in the study in six selected national sites for periods of three to five years. Like the other instruments in the battery. the MHI was validated on over 4000 subjects. Because the MHI provides scores for negative as well as positive subjective well-being. it is especially compatible with the concept and format of the Hassles and Uplifts scales. Although the MHI has not been extensively used in other academically based research. Veit and Ware (1983) provide considerable psychometric support for this instrument. The internal consistency coefficients based on 5.089 subjects range from .83 to .91 for the five lower order factors and from .92 to .96 for the two higher order factors and the total MHI index. Stability coefficients over a one year interval range from .56 to .64. Intercorrelations among the five subscales range from .34 to .75. Cassileth et a1. (1984). using a sample of over 800 medical patients. report a correlation of .82 between the MHI depression scale and the Beck Depression Inventory and a .84 correlation between the MHI anxiety subscale and the Spielberger State Anxiety scale. In a study of over 1500 adults. Manning. Newhouse. 44 and Ware (1982) found that the MHI was a better predictor of the utilization of general medical services than other measures of health status. A similar study by Ware et al. (1984) found the MHI to be an effective predictor of the use of outpatient mental health services. Ehlfiiéél §¥EEIQE§£ QfllE§~ Physical symptoms were measured by a modified version of the Cohen-Hoberman Inventory of Physical Symptoms (CHIPS). The CHIPS is a 39 item list of common physical symptoms that are rated according to how much that problem bothered the individual during the past two weeks. Items are rated from 1 (not at all) to 5 (extremely). The authors of the scale state that although many of the symptoms included on the scale have been traditionally viewed as psychosomatic problems. they excluded any symptoms of "an obviously psychological nature (e.g.. felt nervous or depressed)." (Cohen 5 Hoberman. 1983. p. 106) In the same article. the authors report that for two separate college student samples (3:331 and 114). the CHIPS correlated significantly (;=.22 and .29) with visits to the student health center within a 5-week followbup period. In another sample of 70 undergraduate students. Cohen 5 Hoberman (1983) report an internal reliability (Cronbach's alpha) of .88 for the CHIPS. Additionally for this last sample. the CHIPS correlated .44 (p<.001) with a life events checklist for college students. In addition to the 39 symptoms on the CHIPS. 6 other 45 physical problems common to college-age samples were included (Ebbin 5 Blankenship. 1986: Hoffman 5 Madsen. 1977). Other health related items added to the checklist included questions about the frequency of use of various medical services (e.g.. student health center. family physcian) as well as health habits and the use of alcohol. tobacco. and recreational drugs. Two items concerning general health status as suggested by Davies 5 Wars (1981) were also included. To control for the effect of ongoing or chronic health problems on the reports of current symptoms (Grant. Patterson. Olshen. and Yager. 1987: Billings 5 Moos. 1982). the CHIPS was modified to include data about symptoms from the past year as well as from the past month. The other questions about health habits. use of medical services. and concern about overall health also gathered data about these two points in time. In order to more directly determine the relationship of religious identity to other aspects of the general adolescent developmental task of identity formation. the Extended Version of the Objective Measure of Ego Identity Status (EOMEIS-Z. Grotevant 5 Adams. 1984) was used. The EOMEIS-z is a 64-item self-report instrument based on Marcia's (1966) 4-category continuum of identity achievement status. including diffusion. moratorium. foreclosure. and identity achievement. The EOMEIS-z measures identity as defined by the presence or absence of crisis and commitment 46 in several domains. Ideological identity is assessed in the domains of occupation. politics. religion. and philosophical life-style. Interpersonal identity is measured in the areas of sex roles. friendship. recreation. and dating. Protocols are scored so that subjects can be classified into one of Marcia's identity status categories for either or both the ideological and interpersonal content areas. The EOMEIS-z is the second and latest revision of this instrument. The three versions of this test have been the subject of eight psychometric studies and have been used as a research tool in thirty other published studies (Adams. Bennion. 5 Huh. 1987). This instrument has been used with subjects ranging in age from 14 to 56 years of age. Norms for the different versions of the instrument are based on samples of nearly 700 college students and over 2000 high school students. The internal consistency of the interpersonal and ideological subscales ranges from .30 to .89 across thirteen studies with a median alpha of .66. Test-retest reliability of time intervals of up to one month range from .59 to .93 with the median stability coefficient of .76 (Adams et al.. 1987). Validity data from thirty- eight studies have shown that the EOMEIS-z (and the two previous versions of the test) scales relate predictably to variables such as cognitive development. rigidity. authoritarianism. intimacy. locus of control. self-esteem. academic achievement. conformity behaviors. and involvement 47 in social activities. This instrument also correlates moderately well with other ego identity measures. and more importantly. shows moderate to high agreement in identity classifications determined with the Marcia Ego Identity Interview (Adams et al.. 1987). Eereeee in life: BIL. The Purpose-in-Life (PIL) test by Crumbaugh (1968) was used as a measure of personal meaning. Crumbaugh developed the PIL in an attempt to operationalize Frankl's (1962) existential concepts. The PIL is a self-report scale made up of twenty items rated from 1 (low purpose) to 7 (high purpose). Thus total scale scores range from a low of 20 to a high of 140. with scores usually skewed toward the high end of the scale (Robinson 5 Shaver. 1972). Crumbaugh (1968) initially validated the scale on four normal (3:805) and five psychiatric (Nz335) groups. Based on the scores of these validation groups. Crumbaugh determined that scores below 92 indicate a lack of clear meaning and purpose: scores between 92 and 112 represent an indecisive range: while scores above 112 indicate the presence of definite purpose and meaning in life (Crumbaugh 5 Maholic. 1969). Crumbaugh (1968) reported a split-half reliability coefficient of .85 based on a subsample (3:120) of one of the original normal groups. No test-retest figures were reported. The PIL correlated significantly (r:-.65) with the Depression scale of the MMPI and also correlated -.40 with the Srole anomia (social alienation) scale. Since its 48 development. the PIL has been used in over 100 theses and dissertations (Crumbaugh. 1982) as well as being utilized as a pre-post measure of treatment manipulations (e.g.. Gruner. 1984: Jacobson et al.. 1974). §QEiél §HEEQEE~ In order to investigate the social support function of religion. three items measuring social integration were used in addition to the church activities subscale of the RIS. These three items are taken from Roberts 5 Davidson (1984) to measure social interaction within a religous group. The items survey how well the individual fits in socially with other church members: how helpful church membership has been in meeting the right kind of people: and how many of the individual's closest friends are also members of the same church. The last item was included in the original RIS item pool. The other two items were also imbedded in the RIS for this study. In order to acommodate the response format of the other RIS items. the Roberts 5 Davidson items were modified from three to four choice responses. In their sample of over 500 adults in two church congregations. Roberts 5 Davidson found that their three item scale on social relationships in the church had an internal reliability (Cronbach's alpha) of .74 and was the strongest predictor of church involvement among variables like religious belief. religious meaning. age. education. and denomination. 49 922952222122: Finally. a brief demographic background questionnaire (Appendix B) was included to gather information about age. gender. and ethnicity. Additional questions concerned current and previous religious affiliation. parttime employment. academic load and major. and involvement in campus and off-campus organizations and activities. Five items that were used in the previous study (Antosz. 1988) to measure the perceived organization of the subjects' church or synagogoue along the social control dimension defined by Pargament et al. (1979) were also included. According to Pargament and his collegues. "social control can be defined on a continuum ranging from nonparticipative/individually restrictive social control to participative/individual1y enhancing social control" (pp. 650-651). For descriptive purposes. the latter type of organizational structure and process can be called "horizontal" and the the former. "hieararchical" (Pargament et al.. 1979. p. 651). All of this information was collected to provide a descriptive profile of the sample as well as to gather data on possible control variables. 229929922 To avoid any further sampling error than that usually incurred with volunteerism. subjects were recruited to participate in a study entitled "Coping with Stress." Meeting in small groups no larger than 25 students. the subjects were briefly informed that the current study was an 50 attempt to determine how this particular age group dealt with daily stressors and that religion was one of several areas of coping resources being investigated. The subjects then filled out a questionnaire packet containing all the self-report measures previously described. The sequence of measures in the packets were set-balanced as a safeguard against systematic error variance from order effects. Other aspects of the test procedure such as the completion of the questionnaire packet in one sitting. and emphasizing the importance of complete and accurate data were all planned attempts to improve the quality of the data. Results In general. the data analyses for this study followed the general principles and strategies outlined by Hunter and Gerbing (1982). who argue for the "construction of two interrelated models. causal and measurement models" (p. 267). According to this approach. it is critical to initially determine the adequacy of the measurement model. Toward this end. most of the measurement instruments in this study were subjected to at least an initial factor analysis and a test of internal consistency using Cronbach's coefficient alpha. Bezeheeetrie Ereeertiee e: the heeeeree 891181999 19291299991 992292 1819)- Considerable attention was given to the RIS since it is a new. relatively untested instrument. An initial factor analysis (principal components analysis with a varimax rotation using communalities) resulted in four factors that accounted for all of the 42 items. The first factor consisted solely of items pertaining to institutional religion and included all six items of the previous Church Beliefs (REL2) and Church Worship (REL5) subscales. The standard score alpha of this first factor was .94. The second factor contained predominantly items referring to personal beliefs including four of the five items from the previous Personal Religious Beliefs (REL4) subscale. The alpha for this factor was .93. Together these two factors accounted for 362 of the total score variance. The third 51 52 and smaller factor included all three of the items from the previous Church Nonworship Activities (REL3) subscale and had an alpha of .62. Finally. all three of the items from the previous Personal Prayer (RELI) subscale loaded on the final factor which had an alpha of .73. In effect. the original subscales of the RIS received preliminary support in this blind factor analysis procedure. The intercorrelations between these four factors ranged from -.02 to .88. for a mean factor intercorrelation of .53 using Fisher's g-Transformation). A confirmatory factor analysis using Hunter's PACKAGE computer program (Hunter. Gerbing. Cohen. 5 Nichol. 1980) tested the five subscales of the RIS which were derived during the initial development of the RIS (Antosz. 1988). Standard score alphas for the five subscales of RELI to REL5 were .67. .83. .73. .88. and .74 respectively. Factor intercorrelations ranged from .07 to .81 with a mean of .57. Further confirmatory factor analyses were performed in an attempt to refine these scales using the conceptual and statistical criteria outlined by Hunter and Gerbing (1982) which include homogeneity of content. internal consistency. and external consistency or parallelism. Only minor modifications were made as the result of these analyses. One item with a weak reliability. as judged by its estimated communality value. was replaced in the Personal Prayer (RELI) subscale. One item from the Personal Religious Beliefs (REL4) subscale was deleted because it lacked 53 sufficient discriminating value between scales. An extra item was added to each of the Church Worship (REL5) and Church Nonworship Activities (REL3) subscales to improve their internal consistency. The standard score alphas for the final five subscales of RELI to REL5 are .73. .83. .77. .87. and .80 respectively. Factor intercorrelations ranged from -.06 to .69 with a mean of .45. Data for test-retest reliability of the RIS subscales were collected from a sepearate volunteer undergraduate sample. This sample (3:42) was predominantly female (3:29) and somewhat older (M=22 years old) than the original study sample. The interval between administrations was approximately four weeks. For the whole RIS. the correlation between testing periods was very high (;=.96). The correlations for the individual subscales were: Personal Prayer (2:.88). Church Beliefs (52.87). Church Nonworship Activities (;=.92). Personal Religious Beliefs (5:.92). and Church Wbrship (2:.91). eeheer 91119299999 19 the 319- Because of gender differences that were observed during the initial development of the RIS subscales (Antosz. 1988). the internal consistency and interscale correlations for the revised subscales were also computed separately for males and females. The results indicate that the revised subscales have adequate internal consistency for each gender group with alphas ranging from .70 to .87. While some of the interscale correlations are larger for the female 54 subsample. only one of the ten intercorrelations exceeds the .70 criterion of independence suggested by Stark and Glock (1968). These results then allow for meaningful comparisons between genders on the religious scales themselves as well as gender comparisons on the relationships between this set of religious scales and the outcome variables. A list of the items in each RIS subscale as well as the standard score alphas for each subscale for the whole sample and for each gender are presented in Table 2. The RIS subscale correlations and as well as means and standard deviations for each subscale for the whole sample and for each gender are detailed in Table 3. A series of t tests were performed on the five RIS subscales to determine gender differences. As it can be seen in Table 4. males scored significantly lower as a group than females on four of the subscales (Church Beliefs. Church Non-worhsip Activities. Personal Religious Beliefs. and Church Worship). However. the males scored higher on the Personal Prayer scale. with the t-value approaching significance. Additionally. Personal Prayer (RELl) is the only subscale where there is a significant difference in the within group variances (5:1.35. p<.05) with males having considerably more variation in scores on this particular subscale. Examining the pattern of interscale correlations for each gender. the only major differences between genders is on the relationship of Personal Prayer to the other subscales. While the magnitude of the relationship between 55 Table 2 RIS Subscales and Reliabilities. Eereehel Erexer £89911 --I experience peace and joy during my private prayers or meditations. --About how many times in a week do you pray at home privately or with your family (other than grace before meals)? --My private prayer is one of the most important and satisfying aspects of my religious experience. Alphas for: whole sample=.73. males=.77. females=.70 999299 9911919 12299) --I believe in what my church teaches about the nature of God. --Based on what I know about the major doctrines of my church. I would say I strongly believe...(response choices from "all of them" to "none"). --I believe my church's teachings about what is required to gain salvation. Alphas for: whole sample=.83. males=.80. females=.84 Qhereh heheerehie 1911211199 18913) --How much help has your church or religious group membership been in meeting the right kind of people? --I attend Bible instruction classes. prayer groups. or other such groups sponsored by my church that help me grow in my religious faith. --List the church offices. committees. or jobs of any kind in which you served during the past twelve months. --Church activities (meetings. committee work. etc.) are a major source of satisfaction in my life. Alphas for: whole sample=.77. males=.73. females=.78 56 Table 2 (cont'd). 22299991 921151999 9211219 129911 --I would say that my personal religious beliefs affect the way I look at everyday events. --I do not think about my personal religious beliefs very often. --My personal religious beliefs do not effect my decisions in the various areas of my life (e.g.. work. family. friends). --I try to carry my personal religious beliefs over into all my dealings in life.. Alphas for: whole sample=.87. males=.85. females:.87 Qhereh Berehie 139151 --The church is not a very important part of my life. --My church's worship services help me to feel close to God. --I do not enjoy attending the worship services at my church. --If not prevented by unavoidable circumstances. I attend church services:...(responses from "more than once a week" to "twice a year of less"). Alphas for: whole sample=.80. males=.76. females:.81 Table 3 RELl REL2 REL3 REL4 REL5 Mean SD Maximum Score RELl RELZ REL3 REL4 REL5 Mean SD RELI RELZ REL3 REL4 REL5 Mean SD REL1=Personal Prayer RELl 9.64 3.15 12.00 RELI 10.08 30‘s RELl 9.45 57 RIS Subscale Intercorrelations. 92912 929212 REL2 .03 9.31 2.09 13.00 RELZ -.27 REL2=Church Beliefs REL3=Church Nonworship Activities REL4=Personal Religious Beliefs REL5=Church Worship REL3 .16 .57 7.20 2.44 12.00 REL4 -.06 .66 .63 9. 82 2.94 16.00 REL4 -.18 .57 REL4 .02 .70 10.02 2.94 REL5 .68 10.75 2.64 Table 4 I Tests by Gender heeh 99 Eereehel Erexer £89911 Females 9.45 2.98 Males 10.08 3.46 999999 9911919 12992) Females 9.55 1.99 Males 8.71 2.23 58 Groups for RIS Subscales. 9hereh Heheerheie hetieitiee £89991 Females 7.38 2.48 Males 6.71 2.27 Females 10.02 Males 9. 999999 2999919 IRELél Females 10.75 2.64 Males 9.53 2.56 T 22192 2299291111! -1.84 p<.07 3.22 p<.01 2.30 p<.01 £9991) 2.12 p<.05 3.93 p<.001 59 Personal Prayer and Church Beliefs is about the same for males (:=-.27) and females (2:.20). the relationship is negative for males and positive for females. Using Fisher's g-Transformation and comparing the correlations between Personal Prayer and Church Nonworship Activities subscales (males: ;=-.01 and females: 3:.26). there is a significant difference between the genders (g=2.10. p<.05). indicating that Church activities are more related to prayer for females than males. It appears that prayer for males is either not related at all or even negatively related to the other religious variables. Thus while prayer seems to be more important to males. it seems to function differently for each gender with respect to the other measured aspects of religion. 999998 Eelaieeitx EeeIe Looking next at the DeJong et al. Religiosity Scale. an initial factor analysis of the instrument resulted in four factors. The first factor accounted for 21 percent of the variance and contained all the items from the Belief. Experience. and Practice subscales that resulted from DeJong et al.'s (1976) original analyses. The alpha for this first factor was .93 which replicates Eckert's (1984) findings and supports the argument that these three subscales are closely related and are also the most reliable of DeJong‘s subscales. The remaining three factors were each composed of one of the remaining three subscales of Individual Morality. Social Morality. and Religious Knowledge. These 60 factors had coefficient alphas of .82. .59. and .71 respectively. The interfactor correlations ranged from .09 to .57. Using confirmatory factor analysis to breakdown the first factor into the three original component scales resulted in alphas of .91. .87. and .74 for the Belief. Experience. and Practice subscales respectively. The three intercorrelations for these subscales were .86. .79. and .73. suggesting that these three subscales together may be measuring a single construct. As a result. these subscales were combined into a single scale for subsequent analyses with the outcome variables. Given the content of these subscales. the combined scale was labeled "Orthodoxy" to reflect the fact that a high score indicates a greater degree of adherance to orthodox Christian beliefs and practices. The coefficient alpha for each gender was .93. The internal consistency estimates of the six original subscales were very similar for each gender with alphas ranging from .91 to .58. The reliabilities of the social and individual morality subscales were consistently low for the whole sample as well as for the gender subsamples. calling into question their utility especially in light of very similar findings by Eckert (1984). As a test of concurrent validity for the new RIS scales. comparisons were made between the RIS subscales and similar scales from the DeJong measure and are presented in Table 5. The Church Beliefs (RELZ) subscale of the RIS correlated the highest with DeJong's Belief scale (3:.70). DeJong's 61 Table 5 Correlations Between RIS and DeJong Subscales. 8911:1999 19291299991 eurvez 18191 99199: 2911819911! Belief ”:34.70 §Q=8.83 Experience 5:11.88 §p=4.41 Practice M:9.12 §Q:3.65 Individual Morality ”:20.16 $924.17 Social Morality ”218.48 §D:2.65 Religious Knowledge M=5.53 SQ=2.95 a: p<.05 RELl -.29c -.24c -.11a -.07 b: p<.01 REL1=Personal Prayer REL2=Church Beliefs REL3=Church Nonworship Activities REL4=Personal Religious Beliefs REL5=Church Wership RELZ .70c .63c .56c .51c .15b .19c c: p<.001 REL3 .48c .54c .75c .48c .21c .26c REL4 .60c .71c .66c .48c .17c .33c REL5 .62c .61c .65c .47c .19c .20c 62 Religious Experience scale correlated .71 with the Personal Religious Beliefs (REL4) subscale. and the DeJong Practice scale correlated .75 with the Church Activities (REL3) subscale. It is noteworthy that the Prayer (RELI) subscale of the RIS correlated negatively with all six of the DeJong scales. However. only the correlations between Personal Prayer and the Belief. Experience. and Practice subscales were statistically significant. If these three DeJong scales are representative of orthodox Christian beliefs and practices. then there is further support that prayer. as measured by the RIS. is a unique aspect of religion for this sample. It should be noted that while concurrent validity for the RIS can be established by comparing it with the DeJong measure. the overall "religiousness" of the current sample based on the DeJong measure cannot be determined since no normative data (means and standard deviations for the six scale scores) have been published. 91992 291151999 99999299- Turning to the other measurement instruments. the three item measure of social interaction within a religious group taken from Roberts 5 Davidson (1984) resulted in very low inter-item corrleations and an alpha of .58. making it unacceptable for inclusion in this study. However. during the initial refinement of the Religious Involvement Survey. one of the items was added to the Church Nonworship Activities (REL3) subscale because of its conceptual and 63 statistical match with the other items in this subscale. The addition of this item to the Church Nonworship Activities subscale more clearly defines a social interaction dimension that apparently is being tapped by this subscale. The five item scale to measure the perceived organization of the subjects' church or synagogue had an alpha of .73 for the whole sample. with alphas of .70 and .78 for the female and male subsamples respectively. Higher scores on this scale (Involve) describe a more horizontal church structure that exercises social control through increased member participation and involvement. With a maximum score of 20. the mean and standard deviation for the whole sample was 15.62 and 2.40. When this scale was constructed for this author's previous study (Antosz. 1988). similar scores were attained with a comparable college undergraduate sample (M=15.43. 52:2.07). 292999921929119 £219). The Purpose-In-Life (PIL) questionnaire by Crumbaugh (1968) was factor analyzed. resulting in two major factors which were conceptually indistinguishable from each other. These two factors were unlike the future and present purpose factors that Cote and Levine (1983) identified in their factor analysis of this instrument. The current data support the unidimensionality of the construct being measured by the PIL. In addition. the alpha of the full scale PIL for the whole sample is .89. with alphas of .88 64 and .89 for the male and female subsamples respectively. The mean score for the current sample was 103.16 with a standard deviation of 14.06. This mean score for the whole sample falls within the "indecisive" range of 92 through 112 described by Crumbaugh 5 Maholic (1969) based on normative data gathered on 1.151 subjects. According to Crumbaugh 5 Maholic. scores above 112 indicate clear purpose and meaning in life while scores below 92 indicate a lack of purpose in life. Factor analysis of the measure of ego identity status (EOMEIS-Z) turned up one major and several minor factors. Among the nineteen items on the first factor were fourteen of the sixteen Foreclosure status items. The remaining factors were uninterpretable combinations of the other subscale items. The reliabilities of the four ideological and four interpersonal subscales described by Adams et al. (1987) ranged from .56 to .85. with six of the eight alphas falling below .70. Because of these unacceptably low estimates of internal consistency. it was decided not to utilize these particular subscales which empahsized the distinction between interpersonal and ideological dimensions of identity status. When the eight items from each of these two dimensions were combined for each status grouping according to the alternate scoring rules provided by Adams et al. (1987). the alphas improved considerably. The alphas for the sixteem item subscales were .70 for the Identity 65 Diffusion cluster. .73 for Moratorium. .75 for Identity Achievement. and .89 for the Foreclosure grouping. The reliabilities of these larger scales by gender were very similar to each other and to the alphas for the whole sample. The interfactor correlations for these four scales as presented in Table 6 ranged from -.35 to .42. with most of these relationships predictable by the psychosocial theories of Erikson (1968) and Marcia (1966) which are the conceptual bases for these scales. For instance. the Achievement scale has significant negative correlations with both the Diffusion (r=-.38. p<.001) and Moratorium (::-.34. Table 6 Intercorrelations Among the Ego Identity Scales. ACH FOR MOR DIF Achievemnt (ACH) -- .06 —.34c -.38c Foreclosure (FOR) -- .11a .27c Moratorium (MOR) -- .46c Diffusion (DIF) -- a: p<.05 b: p<.01 c: p<.001 2<.001) scales. The Moratorium and the Diffusion scales are positively and significantly correlated (3:.46. p<.001). As a result of these findings. only the sixteen item subscales of the EOMEIS-2 were utilized in subsequent analyses. The means and standard deviations for the whole sample for these four scales along with normative data provided by Adams et al. (1987) from two college samples are presented 66 in Table 7. Looking at this data. it appears that the means for the Achievement and Foreclosure scales are relatively lower for the current sample while the Moratorium and Table 7 Current Sample and Normative Data on the EOMEIS-II. 9999199 Current Texas Utah 99919 M 99 2 99 9 99 Achievement 62.6 8.6 65.4 8.2 65.5 8.3 Moratorium 54.1 8.9 54.1 9.7 52.6 9.9 Foreclosure 37.2 11.3 39.9 11.1 43.5 10.8 Diffusion 44.1 8.7 44.4 9.2 43.2 9.3 Diffusion scales are very similar across samples. Based on scoring rules which will be explained later for determining identity status groups. these comparisons suggest that there may be fewer subjects in the current sample who fall into the "pure" identity status categories of Achievement and Foreclosure. As a result. there may be more subjects in the current sample who might be classified as belonging to transitional identity status groups. Since Adams et al. (1987) do not provide data on the percentage of their samples classified in the various identity status groups. this point cannot be conclusively tested. 999191 999119 12192 1991)- The Mental Health Index (MHI) was factor analyzed and resulted in one large and four smaller interpretable factors. The first factor consisted primarily of the items 67 from the Positive Affect subscale as well as items from the Depression and Loss of Behavioral/Emotional Control scales which had strong negative loadings on this factor. The second factor was a replication of the complete Anxiety scale. The remaining three factors were also unidimensional and represented the Positive Affect. the Loss of Control. and the Emotional Ties subscales. The alphas for the five first order subscales ranged from .80 to .89. The second order factors of Psychological well-Being and Distress had alphas of .91 and .94. respectively. The eight-item Social Desireability Scale. with very low inter-item correlations and a resulting alpha of .26. was dropped from the current study. The reliability of all the scales for the gender subsamples were very similar to each other and to the full sample figures. It is interesting to note that there is a considerable discrepancy between the mean scores on the MHI for this current late adolescent sample in comparison to the normative data for this measure collected on over five thousand subjects in primarily adult samples. 0n the Psychological well-being scale. the mean score for the current sample (3:51.03. §Q=10.65) was lower than that for the normative samples (5:59.16. 59:12.16). Conversely. the current sample (5:63.81. 59:15.91) scored much higher on the Psychological Distress scale than the normative samples (M=47.54. §Q=15.39). The mean age of the normative samples was 32.2 with a range of 13-69. while the mean age of the 68 current sample was 19.2 with a range of 17-23. From a life span developmental perspective which would assume that adults generally have a broader range of experiences for developing personal satisfaction as well as more coping resources. the pattern of differences between the scores for this sample and the normative samples would be expected. H§§§l§§ £99 Eeliii§~ A factor analysis was performed on the Hassles and Uplifts subscales. For the Hassles subscale. seventeen factors were extracted. but only seven were interpretable. The items on the first factor pertained to general social and interpersonal concerns and were labeled "Shy". The second factor encompassed complaints about too many responsibilities and too little time to fulfill them and was labeled "Pressure". Together the items on the third factor referred to difficulties making decisions and was labeled "Decisions". The following two factors. which referred to financial worries. were combined and labeled "Money". The last two interpretable factors related to obstacles in visiting family members and concerns about the future. Only the first three factors had adequate internal consistency. Combining the two factors referring to financial issues increased the internal consistency estimate to an acceptable level (alpha=.80). The coefficient alphas for the Shy. Pressure. Decide. and Money factors were .74. .80. .75. and .80 respectively. Factor analysis of the Uplifts subscale was unsuccessful 69 as varimax and oblique factor rotations failed to converge. As a result. only one large factor was identified through the initial principal components analysis. Sixty-eight items loaded on this factor which was conceptually uninterpretable. Similar results with the Uplifts subscale were found in a previous study with a college undergraduate sample (Antosz. 1988). For the purposes of this study. only the frequency score (sum total of items endorsed) and mean intensity score (the sum of severity ratings divided by the frequency score) described by Kanner et al. (1981) were used in the primary analyses of the current study. The four hassles factors described above were used only for ancillary descriptive comparisons since no preliminary hypotheses were made regarding specific clusters of hassles and uplifts. The rationale for utilizing the frequency and mean intensity scores as opposed to the severity score (sum of severity ratings for the two subscales) is the possibility that these scores represent different response styles. The high correlation within these two categories of scores across the subscales (: of hassles and uplifts frequency scores=.72. : of hassles and uplifts mean intensity scores=.4l) and the much lower inter-category correlations across subscales (g of Hassles frequency and Uplifts intensity=-.15 and I of Uplifts frequency and Hassles intensity =-.11) suggest distinct styles of reporting stressful and satisfying minor daily events. To the extent that different response styles 70 might be involved in the coping process. using a total severity rating score for each subscale would only mask important differences. Gender differences in the frequency and mean intensity scores will be discussed later. In order to get a cleared picture of the current sample's performance on the Hassles and Uplifts scale relative to other samples. the means and standard deviations of the four frequency and intensity scores for the current and four other samples are presented in Table 8. The original normative group for the Hassles and Uplifts scale is the middle-aged sample (ages ranging from 45 to 65) used by Kanner et al. (1981). The data for this sample represent Table 8 Current Sample and Normative Data on the Hassles and Uplifts. flssslss sad ueliiis Ssszss Hassles Hassles Uplifts Uplifts Frequency Intensity Frequency Intensity §§EElQ§ Current g 34.6 1.72 52.3 1.88 SD (21.7) (.35) (26.7) (.35) Kanner et al. (1981) g 20.5 1.47 49.5 1.77 SD (17.7) (.39) (27.8) (.40) Antosz (1988) M 39.1 1.66 53.4 1.80 SD (26.8) (.37) (29.0) (.33) Bernardo (1988) * M 27.1 1.86 ---- ---- 59 (14.3) (.98) ---- ---- XX 5 23.8 1.54 ---- ---- SD (13.8) (.37) ---- ---- x Subject pool participants XX Religious groups participants 71 the average scores for cumulative responses to the scale given each month for nine consecutive months. The remaining three samples have an age range very similar to the current sample. The Antosz (1988) data is taken from this author's previous research with a sample recruited from the subject pool of undergraduate students in introductory psychology courses. The first sample for Bernardo (1988) represents a very similar sample to this author's current and previous samples in terms of recruitment procedures from the same midwestern university. The second sample from Bernardo (1988) consists of subjects recruited from on and off-campus religious groups. The Uplifts subscale was not used in the Bernardo (1988) study. In general. it can be said that the current and three other late adolescent samples endorse more frequent and intense Hassles and Uplifts than the middle- aged sample originally studied by Kanner et al. (1981). These discrepancies in scores may be due to general response style differences between age cohorts or may reflect the developmental differences previously suggested for similar discrepancies on the MHI. However. it is less clear why the mean of Hassles frequencies is noticeably greater for the current sample compared to Bernardo's sample when both samples utilized similar recruitment procedures and both were initially presented to students at studies of how individuals cope with "stress" (Antosz sample) or "daily hassles" (Bernardo sample). Given the larger standard deviation for Hassles frequencies in the current sample. the 72 the variability of responses was also much greater in this sample. Hséissl iaéisss- The Cohen-Hoberman Inventory of Physical Symptoms (CHIPS) was factor analyzed with a varimax rotation resulting in fifteen small factors. most of which were not conceptually unidimensional. The coefficient alpha for the whole sample was .88. with alphas of .85 and .91 for the female and male subsamples. The score for the CHIPS was computed by summing the ratings for each item. which results in a severity score. similar to that discussed for the Hassles and Uplifts. This type of score seems especially appropriate considering the above discussion about response styles with the Hassles and Uplifts Scale. Initial inspection of the data reveals that the severity score for the CHIPS correlates similarly to both the frequency and mean intensity scores of the Hassles and Uplifts Scale. Meanwhile. computing frequency and mean intensity scores for the CHIPS and correlating them with the Hassles and Uplifts scores results in the same pattern of higher intra-category and lower inter-category correlations. further supporting the notion of response styles. Since the CHIPS was modified for this current study. there are no appropriate normative data against which the responses of this sample can be compared. The mean scores for this sample were as follows: frequency of symptoms reported (g=l6.5. §Q=7.2); intensity of symptoms (3:1.68. 73 §Q=.46): and severity of symptoms (M:28.2. 59:15.9). In addition to the CHIPS summary score. two other health related scores were utilized in the subsequent analyses. A score representing the frequency of medical services use in the past month was computed by summing responses about the number of visits within the past month to seven medical settings. including the university student health center For the whole sample. the mean scores was .90 with a standard deviation of 1.78 with a range of scores from 0 to 12. Finally. two questions with forced-choice responses about the subjects' general state of health and their concern over their health in the past month were used as a measure of overall health status. For the whole sample the mean score was 5.77 with a standard deviation of 1.29. For the sake of interpretability. a maximum score of 8.0 indicates "excellent" health with no health concerns. 99299! Qi££§E§BE§§ Since it appears that overall the measurement model used in this study adequately applies to each of the gender subsamples. the sharp differences in group means between males and females on the RIS subscales suggests possible gender differences in the overall structure and function of religion. Gender differences were also observed in a similar sample in the pilot study (Antosz. 1988) and a comparison of that late adolescent sample with an adult sample indicated that these gender differences may be unique to this particular developmental group (Antosz. 1989). To 74 further explore gender differences in the other dependent and independent variables. a series of 1 tests were performed to compare the male and female group means on all the relevant scale scores. As Table 9 indicates. in addition to the significant differences on four of the five RIS subscales. males and females differ significantly on twenty other variables. including two of the mental health indicators. two of the identity scales. and most of the indicators of physical symptoms. Given these considerable gender differences across most of the variables and the increasing support for unique developmental pathways for males and females (Alishio G Schilling. 1984: Gilligan. 1982: Thorbecke & Grotevant. 1982: Josselson. Greenberger. & McConochie. 1977). it appears that the data in this study can be more meaningfully utilized by investigating specific gender differences rather than merely statistically controlling for them in the group data. Thus most of the remaining analyses will be carried out along gender lines. Ezesihssss Isssiss sad Rslsisé Aaslzsss To test the major hypotheses of this study. zero-order correlation matrices for each gender were constructed. These matrices. which include the religious variables as well as the intervening and outcome variables are presented in Table 10. In addition to the RIS variables. the Orthodoxy scale containing the first three subscales of the DeJong Religiosity measure is included. 75 Table 9 I Tests by Gender Groups for Other Major Variables. Essa §;Q; I xslgs Ershséilitz EQ!§I§:£ Foreclosure Females 36.30 11.26 -2.53 p<.01 Males 39.26 11.21 Diffusion Females 42.98 8.21 -3.73 g<.001 Males 46.31 9.29 ELL Females 104.2 14.0 2.16 p<.05 Males 101.1 14.0 fl§§§l§§ é Heliiié Average Intensity of Hassles Females 1.74 .35 2.83 p<.01 Males 1.65 .32 Average Intensity of Uplifts Females 1.92 .35 3.60 p<.001 Males 1.80 .33 Qsisnz leixissitx 5951s Belief Females 36.10 8.23 4.93 p<.001 Males 31.72 9.37 Experience Females 12.42 4.41 3.78 p<.001 Males 10.73 4.23 Practice Females 9.37 3.75 2.12 p<.05 Males 8.57 3.41 Individual Morality Females 20.82 4.13 4.81 p<.001 Males 18.81 3.85 Social Morality Females 18.89 2.49 4.89 p<.001 Males 17.58 2.80 76 Table 9 (cont'd). EIQPEEiliEX p<.05 p<.05 p<.001 p<.001 p<.001 p<.05 p<.01 p<.01 p<.05 p<.01 EEQB §;Q; I 25122 881 Loss of Behavioral/Emotional Control Females 21.27 6.25 2.10 Males 19.91 6.42 Emotional Ties Females 7.83 2.58 2.21 Males 7.25 2.50 QUIE§ Number of Symptoms Females 17.44 6.86 4.24 Males 14.35 7.62 Average Intensity of Symptoms Females 1.73 .47 3.47 Males 1.57 .42 Total Severity of Symptoms (Number x Average Intensity) Females 30.43 15.28 4.35 Males 23.47 16.33 Number of Visits to Medical Providers (past month) Females 1.00 2.02 2.11 Males .69 1.09 Frequency of Over-the-counter Medications (past month) Females 3.62 1.48 2.72 Males 3.21 1.45 Frequency of Prescription Medication Use (past month) Females .47 .50 2.88 Males .32 .47 Recreational Drug Use (past month) Females 1.43 .96 -.2.30 Males 1.73 1.36 Alcohol Use (past month) Females 3.48 1.35 -3.06 Males 3.91 1.31 Exercise Females 3.47 .94 -2.80 Males 3.73 .88 B<’01 77 oo.v~ ou.vo~ .............Ouuancunoeocssmuoz—z «convex .0 seneeZumh~m~>xu .su on.n as.m ........s:asam sense: assocoouahqmm .nn an. mo.~ ..suuu—aa uo aassceusu asexuwmoz«.ac00~uh_a .on n~.o -.vn ............E:.50ussox >«.uc0e_nuox .o o~.n~ om.on ...........os:oo_ueuou >«.osoo_nuou .0 on.» v«.no ...........«s0§0>o.nu< >v.«c0e_u=u< .s mo.vn no.5m ....onsum aa.eo.u«_eu usOHOOuozonuo .o '0." ”Foo." accuse-oooo-oooond‘.1—g ‘ULaabUHNsfluu on ve.« ~o.on ....suom_om ssosu.—eu .scossomuvquu .v ev.« on.s ..sosas>.au< assssoacoz couscounauu .n “a.” ”moo oseeeloeeoeeeOOO.u.‘—.n ‘0“:‘UH~JN~ ON ”Mu-IN n'lflfi ......oooooooooh.§.ht —.co.nbotufi\du“ oH an x noo.va "u "o.va "A mo.va "a nu oz~zx¢ .sn onu. un«.n ueu. Av~.n un :hqcmz .un oo«.n unv. 0on.u now. usn.n nu xm>um .ou uou. mo.u can. Ann. sun. no. un mm=zz osouaao use eeouuu~om use ucoe< snowed—OLLOULOuc~ on smash 78 Uhn.n Una. mn.n awn. Don.l can. so. Io~.n Ao~.n ~«.u Onn.n 00. Dan. and. Inn. no. on. «a. «0. ON ous.u so. a-.n use. na—.u «a. on". va. ~o.u ass. as". -.u as“. so.u na~. .0. new. a-.u an no.l AON. Unn.n Ohm. no. Avn.n Auu.n co. ANN.I nc.n Con. 09. v". «0.! NH. no.n no. on su Mn.l mo. no. me. nu.u an. AO~.I no.n AvN.l so.n DO. we. no. co. an.l nu. Umn.l Nu. o~.u n°.u Inn.u Con.l oo.u mo. u«.u Ou.n no.n hc.u oo.u Nu. an no.v~ oo.no~ no.o~ on.uo o~.ou nu.~n oo.n no. em." on.m nn.ou nv.n~ mm. on.~ on.o« n~.~n an. no.— o~.o« va.mn oN.e ~n.uv vn.e so.nn -.- o~.mn o~.o mm.~o on.m~ ~o.—n on.” no.9 no.~ on.o s~.~ us.o mm." “5.9 owhn coho“ am 3 doo.va "u ao.n n one. co. un no. oo. o". on. «o.n so". Nu. oo.n an. no". «a. no. on. no. no. ou. v«.n ~o. on. no.n Na. saw. no. mg. «n. no.u ss~.n «n.n ~o.u sn—.u nu - Ha mmmmm .............ossqucsuouoauaauez_zecnnha ...-a.e«> .su.vox .0 assasznmh_m—>xm ........e:asum :a_so: ~ssocoon=ha¢mz .eEOunsam .eumsagm no anuse>omuxm>um ...euu¢_a= .0 he‘sceas— csexumaaztux ........suu«—a: uo auceececmumma>z«.eceuc~ cIOXumczzz¢w“=.vuuh~o sauces-sooooaeaflbov.Lox >“«“:.‘~H“ox ............0L:eo_uOLOL >o.~coe_nuou ............»:oso>os:u< su.«coe_u=o< .....0—IUm >a«lO«I«-O¢ UCOHODHUZOHND cos-cocoooocs-oos““.nh°’ ‘ULI-‘Uunl—Nm .....euos_om snow-“.0: .ssoesemnvqmm ...-o.a.>_au< a.o:toacoz gusaonnaum ...-osossu-ooouoo.h°m~on go‘s‘UHNJU“ successes-o-eoooh".ah& _.:°.I~0mufld~“ Ho.va "a can. un so". me. An~.u Am~.u uvn.n so.u u-.n n~.u unn.u No. o~.u o~.n -.u oo.u co. n-.u on a "Q oz~zx8 Shamm mm=zzv 151-158. Carroll. J.w.. & Roozen. D.A. (1973). National sample questions on religion: An inventory of material available from the Roper Public Opinion Research Center. Journal for the Scientific Study 9; Religion. 1g. 325-338. Cassileth. B.. Lusk. E.J.. Strouse. T.B.. Miller. D.S.. Brown. L.L.. Cross. P.. & Tenaglin. A.N. (1984). Psychosocial status in chronic illness: A comparative analysis of six diagnostic groups. 599 9951999 goggngl 91 99919199: 91;. 506-511- Chamberlain. R.. & Zika. S. (1988). Religiosity. life meaning and Wellbeing: Some relationships in a sample of women- 1999991 199 199 9919911119 91912 91 89113199: 12(3). 411-420. Clayton. R.. & Gladden. J. (1974). The five dimensions of religiosity: toward demythologizing a sacred artifact. 1999991 199 199 9919911119 91999 91 1911:199v 19v 135- Coe. G.A. (1916). IEE psychology 99 99liS199° Chicago: University of Chicago Press. Cohen. S.. & Hoberman. H. (1983). Positive events and social supports as buffers of life change stress. Journal of 999119d 999191 2929991999, ;§. 99-125- Cohen. S.. a McKay. (1984). Social support. stress and the buffering hypothesis: A theoretical analysis. In A. Baum. J.E. Singer. & S.E. Taylor (Eds.). 39399993 9: Esxghglggx and health (Vol. 4. pp. 253-267). Hillsdale. NJ: Erlbaum. Cohen. S.. & Wills. T.A. (1985). Stress. social supportr. and the buffering hypothesis. Psychological ggllgtig. 98(2). 310-357. Coleman. J.C. (1978). Current contradictions in adolescent theory. 1999991 91 19919 999 9991999999. 2. 1-11- °°1°manv J°C~ (1974)~ 3919119999199 19 9991999999- 30510“: Routledge and Kegan Paul. Comstock. L.K.. 9 Slome. C. (1973). A health survey of students: 1. Prevalence of perceived problems. Joggggl 99 99991999 9911989 999199 99999199199v 22. 150’155- 142 ). Adeleseaee and xguibi Esxebglgaisal g 5 ghhhgihg wghlg. New York: Harper & Row. Cornwall. M. (1987). The social bases of religion: A study of factors influencing religious belief and commitment. 89219! of Religions Research. 22(1). 44-56. Cornwall. M.. Albrecht. S.L.. Cunningham. P.H.. & Pitcher. B. (1986). The dimensions of religiosity: A conceptual model with an empirical test. 8921:! of Religiggs Research. 21(3). 226-243. Crawford. M.. Handal. P.. & Wiener. R. (1989). The relationship between religion and mental health/distress. 89219! of 82111199: Research. 31(1). 16-22. Crumbaugh. J.C. (1982). Bibliography 9; 21h ligghgtggg. Gulfport. Mississippi: Originality Press. Crumbaugh. J.C. (1981). Logotherapy: New help for problem gzinkersv international {923$ {Qt LQIQIDEIEEXv 1(1). 29- Crumbaugh. J.C. (1968). Cross-validation of purpose-in- life-test based on Frankl's concepts. JQBEBQL of lndixidnal Eaxghglgzx. 21. 74-81. Crumbaugh. J.C.. & Maholic. L.T. (1969). £99251 9; instructions 19! ihg EEEEQéQ in L119 1§§1~ Munster. 1”: Psychometric Affiliates. Davies. A.R.. & Ware. J.E. (1981). hgggggihg hgglgh BEIEQBEEQBE 19 EDQ 595135 iBEBEEBEQ 9529219992 (Report No. R-27ll-HHS). Santa Monica. CA.: Rand Corp. D'Antonio. W.. & wright. S. (1982). Religion and family life: how social scientists view the relationship. 4922251 for the §sieaiifis §12§2 9: Religion. 2l(3). 218- 225. Davidson. J.D. (1972). Religious belief as an independent variable- iguana; 19: the §gieatiiig §12éz 9i Eeliaignv ll. 67-75. DeJong. G.. Faulkner. J.. & Warland. R. (1976). Dimensions of religiosity reconsidered: Evidence from a cross- cultural study. §Qgihl Eggggg. §1(4). 867-888. DeLongis. A.. Coyne. J.C.. Dakof. G.. Folkman. 5.. and Lazarus. R.L. (1982). Relationship of daily hassles. uplifts. and major life events to health status. flfiéllh Esxghglgax. l, 119-136- 143 Dohrenwend. B.P.. & Shrout. P.F. (1985). "Hassles" in the conceptualization and measurement of life stress variabIBS- Aeezieee Eexebeleaiei. 19(7). 780-785. Donahue. M.J. (1985). Intrinsic and extrinsic religiousness: Review and meta-analysis. Journal of Eereeeeliix e99 §eeiel Eexeeeleaxv i§(2). 400-419. Dudley. R.. Mitch. P.. & Cruise. R. (1987). Drug usage among Seventh-Day Adventist Youth in North America. Jeeznel fer the §eieeiiiie §ieéx 9i Reliaieav 29(2). 213-233. Duke. J.T.. & Johnson. B.L. (1984). Spiritual well-being and the consequential dimension of religiosity. Regiew e: Reliaieee Eeeeereb' 29(1). 59-72. Ebbin. A.J.. & Blankenship. E.S. (1986). A longitudinal health care study: International versus domestic studentS- 4922991 9: Amerieee Qelleae Health: 32. 177- 181. Eckert. P.A. (1984)- ghzieiiee ertheéesx eeé BremeIiLel SQEQaI 9993219! EEOB‘ h91§29§°§2 l 991 Kg § HQQBI§' Unpublished master’s thesis. Micfiigan égate niversity. East Lansing. Michigan. Elkind. D. (1970). The origins of religion in the child. Bexiey 9i Eeliaieee Eeeeezeh. 13. 35-42. Erikson. E-H- (1968)- lieeiiixe xeeih and ezisie- London: Faber a Faber. Erikson. E.H. (1963). thléhQQQ hhg §Q§i§§x. New York: Norton. Filsinger. E. (1981). Differences in religious structures between empirically derived types. gggigw Q; Religighg EQEQQEEhv 22(3): 255-267. Filsinger. E.. Faulkner. J.. & Warland. R. (1979). An empirical taxonomy of religious individuals. §eeieleaieel Aeelxeiev 19' 420-429- Finney. J.R.. & Malony. H.N.. Jr. (1985). Means. end. and quest: A research note. Keeley 91 Reliaieue Reseezehv 29(4). 304-307. Folkman. 5.. & Lazarus. R.S. (1985). If it changes. it must be a process: A study of emotion and coping during three stages of a college examination. geezeel g; E§£§Qflélill £99 §9§i§l EEXEDQLQSEv i§v 150‘170~ 144 Folkman. S.. & Lazarus. R.S. (1980). Coping in an adequately functioning middle-aged population. JQQIQQL 91 health the §eeiel hehexier' 21. 219-239- FOWIBT: J. (1931)- §1§8§§ QI ifiiihi ID: 252999198! 91 human QQXQlQEEQBI and 1h: SHEEI £92 mflfinini- New York: Harper & Row. Franklv v.2. (1962). Hehie eeereh fer eeehiha; Ah ihtreéeetteh te teaethereexo Boston: Beacon Press. Gallup (1987)- Eeliaieh ih aheriee (Report NO- 259)- Princeton. NJ: Gallup Organization. Gallup (1981). ggli‘igg 12 hmggigh (Report No. 184). Princeton. NJ: Gallup Organization. Ganellen. R.J.. & Blaney. P.H. (1984). Hardiness and social support as moderators of the effects of life stress. Jeerhel et Eereehelitx éhé §eeiel Eexeheleax. 11(1). 155- 153. Gilligan. C- (1982). 1h e hittereht xeieet Eexeheleaieel theerx and weeehie eezeleehehto Cambridze. MA: Harvard University Press. Glik. D.C. (1986). Psychosocial wellness among spiritual healing participants. §eetel §etehee the heétethev 22(5). 579-586. Glock. C.. & Stark. R- (1965)- Reliaieh éflé.§9§i§i¥ in 1995192- Chicago: Rand McNally. Godwin. D. (1988). Causal modeling in family research. Jeerhel et herrieae the the Eehilx. £9. 917-927. Gore. 5.. (1978). The effect of social support in moderating the health consequences of unemployment. JQQEDQL 91 flgéllh é §Q§i§l RQDQXIQI' 12. 157‘155- Gore. 5.. (1981). Stress-buffering functions of social supports: An appraisal and clarification of research models. In 8.5. Dohrenwend & B.P. Dohrenwend (Eds.). Stressful life events and their contexts. New York: Prodist. Gorsuch. R.. & Smith. C. (1983). Attribution of responsibility to God: An interaction of religious b°1i°f3 and °“t°°m°" Jeerhet ter the eetehtitie e229: 9: EQliliQBv 22(4). 340-352- Grant. 1.. Patterson. T.. Olshen. R.. & Yager. J. (1987). Life events do not predict symptoms: Symptoms predict 145 symptoms. Jeerhei ei hehexierei heiieihev 19(3). 231- 240. Greeley. A. (1972). Th; Qéflgminéiigflél society. Glenview. IL: Scott Foresman. Grossman. N. (1975). On Peter Berger's definition of rolixi°n~ Jeerhei ier the §eiehtiiie §teix ei Eeiiaiehv 11(3). 289-292. Grotevant. H.D.. & Adams. G.R. (1984). Development of a measure to assess ego identity in adolesence: Validation and PePlicati°no Jeerhei ei teeth the hieieeeheev iiv 419-438. Gruner. L. (1984). Heroin. hashish. and hallelujah: The iggrigsfor meanina- sexiew ei Reliaieee Reeeereh. 29(2). Hadaway. C.K.. & Roof. W.C. (1978). Religious commitment and the quality of life in American Society. Eggigy g; heiisieee Eeeeerehv 12. 295-307- Hay. D.. & Morisey. A. (1985). Secular society. religious meanings: A contemporary paradox. gggigg g; ggligighg Eeeeerehv 2§(3). 213-21?- Henderson. 5. (1981). Social relationships. adversity and neurosis: An analysis of prospective observations. hritieh Jeerhei ei Eexehietrxv lit, 391-398- Henderson. 5.. Byrne. D.G.. Duncan-Jones. P.. Scott. R.. & Adcock. S. (1980). Social relationships. adversity and neurosis: A study of associations in a general population sample. British Jeerhei ei Bezehietrx' lie. 354-583~ Hilty. D.M.. Morgan. R.. Hartman. W. (1985). A structural equation modeling analysis of the means. end and quest dimensiODS- leerhel fer the Seientifie etude ei Eelixieh. 21(4). 424- 435. Himmelfarb. H.S. (1975). Measuring religious involvement. §Q§1§l Egrces. §§(4). 606-618. Hodgson. J.W.. & Fischer. J.L. (1981). Pathways of identity development in college women. 5;; 89195. 1. 681-690. Hodgson. J.w.. & Fischer. J.L. (1979). Sex differences in identity and intimacy development in college youth. Jeerhei ei teeth the hieieeeehee. t. 37-50. Hoffman. W.K.. & Madsen. M.K. (1977). College student health care needs at the University of Wisconsin- 146 “11"‘“*°°- {attest 9i tastiest 9911252 hesithv 22' 296’ 301. Holahan. C.J.. & Moos. R.H. (1987). Personal and contextual determinants of coping strategies. {thhhl 91 2225295113! 599 §QEiél EEXEDQLQSXv §Z(5)v 946-955. Holahan. C.K.. Holahan. C.J.. & Belk. 5.5. (1984). Adjustment in aging: The roles of life stress. hassles. and self-efficacy. agglhh Eggghglggx. g. 315-328. Hood. R.W. Jr. (1975). The construction and preliminary validation of reported mystical experience. gghhhh; :9; the §eiehtiiie §teex ei Eeiiaieh. it. 29-41- Hood. R.W. Jr.. a Morris. R.J. (1985). Conceptualization of quest: A critical rejoinder to Batson. Eggigw Q; Eeliaieee Eeeeereh. 2§(4). 391-397. Hunt. R.A. (1972). Mythological-symbolic religious commitment: The LAM scales. lQBEDél {9; 3h; fiEIQBiiiiE §teix 91 89118192: 11. 42‘ 52. Hunt. R.A.. & King.M. (1971). The Intrinsic-Extrinsic concept: A review and evaluation. Journal for the §eiehtiiie §teex ei Eeiiaieh. 19(4). 339-356. Hunter. J.E. (1980). Factor analysis. In P. Monge & J. Cappella (Eds.). MQlLiXEIiEEQ l§§hn1929§ in harsh ggmmuhigatioh :ggghhgh. New York: Academic Press. Hunter. J.E.. Cohen. S.. & Nicol. T. (1982). 2393;89- Unpublished manual. Michigan State University. Hunter. J.E.. & Gerbing. D. (1982). Unidimensional measurement. second order factor analysis. and causal models. In Eeeeereh ih Qraehieetiehei hehexier (Vol. 4. pp. 267-320). Greenwich. CT: JAI Press Inc. Hunter. J.E.. & Hamilton. M. (1986). Eéihi h ggggghh 1h §§§iE§~ Unpublished manual. Michigan State University. Hurlock. 5.8. (1973). hgglggggh; gggglgggghi. New York: McGraw- Hill. Husaini. S.A.. Neff. J.A.. Newbrough. J.R.. & Moore. M.C. (1982). The stress-buffering role of social support and personal confidence among the rural married. Jeerhel Q; thhehitx Eexeheieax' 19' 409-425- Jacobson. G.R.. Ritter. D.P.. & Mueller. L. (1977). Purpose in life and personal values among adult alcoholics. Jeerhei ei Qiihieei Eexeheleszv §§<1)v 314-316. 147 Jamesv W- (195131902. The zerietiee ei reiiaieee expgrighgg. New York: Collier Books. Jemmot. J.B.. III. & Locke. S.E. (1984). Psychosocial factors. immunologic mediation. and human susceptibility to infectious diseases: How much do we know? Bezeheieaieei heiietihv 2t: 78-108. Joselson. R.. Greenberger. E.. & McConochie. D. (1977a). Phenomenological aspects of psychosocial maturity in adolesence. Part 1. Boys. 1992221 9; I221! hhg heeieeeheev t. 25-55- Josselson. R.. Greenberger. E.. & McConochie. D. (1977b). Phenomenological aspects of psychosocial maturity in adolescence. Part II. Girls. gggzhhl 9; 19329 534 ééQlQEQQBEEv 6. 145-157. Kahoe. R.D. (1989). A concluding unscientific postscript on relizious development- Jeerhel ier the Seiehtiiie Steer 9i Eeiiaiehv 2§(2). 233-236. Kanner. A.. Coyne. J.. Schaefer. C.. & Lazarus. R. (1981). Comparison of two modes of stress measurement: Daily hassles and uplifts versus major life events. lggghh; 92 heheyierel MEéiEiDQv 1(1), 1‘39- King. M.B.. & Hunt. R. (1975). Measuring the religious variable: National replication. 2922951 {9; ghg §eiehtitie §teéx Qt Eeiiaieho 11' 13’22- Klinaerv E- (1977)- heehiha ehi zeie; ihher ereerieeee she the iheehtizee ih eeeeieie iixee- Minneapolis: University of Minnesota Press. Kobasa. S.C. (1979). Stressful life events. personality. and health: An inquiry into hardiness. Journal of EQEEQDQliE! £99 §QEi£l BENEDQLQllv §Zv 1’11- Kobasa. S.C.. Maddi. S.R.. a Courington. S. (1981). Personality and constitution as mediators in stress- illnese relations. Jeerhel ef Health and ieeiel Behexier. 22. 368-380- Koenig. H.. Kvale. J.. & Ferrel. C. (1988). Religion and well-being in later life. figzgnhglggist. 2§(1). 18-28. Kopplin. D- (1976). Religieee er ................ EIQQQDEE £99 reietei eereeheiitx ehereeterietiee~ Paper presented at American Psychological Association. Washington. D.C. 148 Krantz. D.S.. Grunberg. N.E.. & Baum. A. (1985). Health Psycholoayo hhheei Beeiee 91 BeethQLQKXv §§. 349'383- Kuhn. M.. & McPartland. T. (1954). An empirical investigation of self-attitudes. hhghighh Sggiglggigg; Reeieev 12. 68-76. Law. H.G.. (1974). 51229122: and meeeereheht Qt helieie sheet reiiaieh the eexe te lieeo Unpublished PhoDo dissertation. University of Queensland. Lazarus. R.S.. DeLongis. A.. Folkman. S.. & Gruen. R. (1985). Stress and adaptational outcomes: The problem of confounded measures. hhghiggh Egyghglggist. 10(7). 770- 779. Lazarus. R.S.. Kanner. A.D.. & Folkman. S. (1980). Emotions: A cognitive phenomenological analysis. In R. Plutchick & H. Kellerman (Eds.). Iheorieg of emotion- New York: Academic Press. Leuba. J.H. (1912)- A REXEDQlQIIEél §19d¥ 9i reliaieho New York: Macmillan. Lindenthal. J.. Myers. J.. Pepper. M.. & Stern. M. (197). Mental status and religious behavior. leerhei {gr hhg §eieetiiie §teex 9t EELiSiQQv 2(2). 143-150. Lorch. B.. 6 Hughes. R. (1988). Church. youth. alcohol. and drug education programs and youth substance use. ggurnal ei hieehel end Dre: Edeeetieh. 33(2), 14-25, ----- Manning. W.G.. Newhouse. J.P.. & Ware. J.E. (1982). The status of health in demand estimation: Or. beyond Excellent. Good. Fair. and Poor. In V.R. Fuchs (Ed.). Eeeheeie eeeeete ei heeitho Chicaao= University of Chicago Press. Marcia. J.E. (1966). Development and validation of ego- ‘d9ntity ‘tetuso Jeerhel ef Eereenelite the Seeiel Eeeeheieae. t. 551- 553. McClure. R.F.. & Loden. M. (1982). Religious activity. denomination membership and life satisfaction. Eeeeheieaxt h Qeerterie Jeerhei ei heeeh heheeier' 12(1). 12-17. McCrae. R.R. (1984). Situational determinants of coping responses: Loss. threat. and challenge. gggrggl 9; Bereeheiite the §eeiel Eeeeheleae. it. 919-928- 149 McKay. D.A.. Blake. R.L.. Colwill. J.M.. Brent. E.E.. McCauley. J.. Umlauf. R.. Stearman. G.W.. & Kivlahan. D. (1985). Social supports and stress as predictors of illness. The Jeerhel ei Eehiie Ereetiee. 29(6). 575-581- Mischey. E.J. (1981). Faith. identity and morality in late adolescence. 93525922! EQEEEEiEl' 1. 175-185. Monroe..S.M. (1983). Major and minor life events as predictors of psychological distress: Further issues and findinSS- Jeereei ei heheeierei heeieihev 9. 189-205. Najman. J.. Williams. 6.. Keeping. J.. & Morrison. J. (1988). Religious values. practices and pregnancy outomes: A comparison of the impact of sect and mainstream Christian affiliation. §Qgihl Sgighgg hhg heeieihev Z§(4). 401-407- Newcomb. M.D.. & Harlow. L.L. (1986). Life events and substance use among adolescents: Mediating effects of perceived loss of control and meaninglessness in life. Jeerhei ei Eereeheiite eeé §eeiei Eexeheleae. §1(3). 554- 577. Nordin. K.D. (1972). Youth fleeing churches--but not the valueS- Qhrietieh §eiehee hehitero Nudelman. A. (1976). Measuring religiosity by direct and indirect methods: A multiple replication. Eggigg g; Eeiiaieee Eeeeereh. 11(2). 102-106. Parsament. x.1. (1987). get heir he; Ieeeree e ireheeerh 9! 992158 £92 £59 BEXEDQIQBX 91 re Paper presented at the American Psychological Association. New York City. Pargament. K.I.. Echemendia. R.J.. Johnson. S.M.. & McGath. C.A. (1984). Assessing the religious needs of college students: Action-oriented research in the religious context» Eeeiee et Beiiaieee Beeeerehv 2t(3). 265-285. Pargament. K.I.. & Hahn. J. (1986). God and the just world: Causal and coping attributions in health situations. gggreei ier the eeiehtitie eteie ei Reliaieh. at. 193- Pargament. K.I.. Kennell. J.. Hathaway. W.. Grevengoed. N.. Newman. J.. & Jones. W. (1988). Religion and the problem-solving process: Three styles of coping. Jeereei ier the §eiehtiiie eteee ei EélifliQBv 21(1). 90-104. 150 Pargament. K.I.. Sullivan. M.S.. Tyler. F.B.. & Steele. R.E. (1982). Patterns of attribution of control and individual psychosocial competence. Egyghglggighl Reports. §;. 1243-1252. Parker. M.S. (1985). Identity and the development of religious thinking. In A.S. Waterman (Ed.). 1922111! in heeieeeeheet Ereeeeeee ehe eehtehte- San Francisco: Jossey-Bass. Pearlin. L.I.. Menaghan. E.G.. Lieberman. M.A.. & Mullan. J.T. (1981). The stress process. 1292951 9; HEQLEQ ghg §eeiai Beheeier. 22. 337-356- Pearlin. L.I.. & Schooler. C. (1978). The structure of copins- Jeerhal of health ehe social Beheeier. 12. 2-21- Peterson. L.R.. & Roy. A. (1985). Religiosity. anxiety. and meaning and purpose: Religion's consequences for psychological well-being. Eeeiee ei Reliaieee Eeeehreh, 22(1). 49-52. Poloma. M.. a Pendleton. B. (1989). Exploring types of prayer and quality of life: A researh note. gggigw g; Eeliaieee EQEQQIEDv §l(1)v 45-53. Reich. J.W.. & Zautra. A. (1981). Life events and personal causation: Some relationships with satisfaction and distr°88~ Jeerhei ei Eereehhiite ehe §eeiel Zeeeheiezev 11. 1002-1012. Ritzema. R.J. (1979). Attribution to supernatural causation: An important component of religious gggméggent? JQBEDEI 91 E§2§h9198¥ QDQ IDQQIQKX' Zv Roberts. M.K.. & Davidson. J.D. (1984). The nature and sources of religious involvement. Eggigg 9; ggligighg Research. 25(4): 334-350. Robinson. J.P.. & Shaver. P.R. (1972). Mghghhgg 91 gggih; § ghggo 19a; tti hgeg. University of Michigan. Ann grfior. M1: Inngtu%e for Social Research. Roos. R.E.. & Cohen. L.H. (1987). Sex roles and social support as moderators of life stress adjustment. JQBEBél ei Eereehhlite ehe fieeiai Eeeeheleae. §2(3). 576-585. Rosenberg. F.R.. A Simmons. R.G. (1975). Sex differences in the self-concept in adolescence. $9! 39195. 1. 147-159. Rosenberg, M. (1955). §QE£2£¥ EQQ 192 ééQlEEEEBE §§l£2 ghhgg. Princeton: Princeton University Press. 151 Sanderson. H.W.. & Crawley. M. (1982). Characteristics of successful family-care parents. 95221952 499225; Q; Mental Defigighgy. fig (5). 519-525. Sarason. I.G.. Levine. H.M.. Basham. R.E.. A Sarason. B.R. (1983). Assessing social support: The social support 9°96ti°nnair°° ieerhei ei Eereeheiite the §eeiei Beecheieaev 11' 127-139. Schaefer. C.. Coyne. J.C.. & Lazarus. R.S. (1981). The health-related functions of social support. JQBEBEL g; heheeierhi heeieihe. 1(4). 381-406- Sevensky. R. (1984). Religion. psychology. and mental health. heerieeh Jeerhei ei Eerehetherehe' it: 73-86. Siddique. C.M.. & D'Arcy. C. (1984). Adolescence. stress. and psycholoaical well-beinz- Jeerhel 91 teeth the heeieeeeheev 1§(6). 459-472. Spilka. 8.. Shaver. P.. & Kirkpatrick. L.A. (1985). A general attribution theory for the psychology of religion~ Jeerhel t9: the §eiehtitie §teet et Eelixieh' gi(1)v 1-20. Starbuck' E~ (1899)- the eexeheieex ei reiisieht he ehhirieel eteee hi the areeth ei reliaieee QQD§§19¥§B§§§- New York: Charles Scribner’s & Sons. Stein. A.H.. & Bailey. M.M. (1973). The socialization of achievement orientation in females. Egyghghggigh; helletihv fig: 345-356- Studer. M.. & Thornton. A. (1987). Adolescent religiosity and contraceptive ucaae- Jeerhei et herrieae the the Family. 12. 117-128. Strunk. O. (1976). The world view factor in psychotherapy. Jeerhel et Reliaieh 5nd health. 19. 192-196- Thorbecke. W.. & Grotevant. H.D. (1982). Gender differences in adolescent interpersonal identity formation. Jeerhei ei teeth the heeieeeeheev 11(6). 479-492. Veit. C.T.. & Ware. J.E. (1983). The structure of psychological distress and well-being in general Populations. eeerhel et theeltiha ehe Qiihieei Eexeheieae. 51(5). 730-742. Ware. J.E.. Manning. W.G.. Duan. W.. Wells. K.B.. & Newhouse. J. (1984). Health status and the use of outpatient health services. hmgrighh E§y§hglggi§t. 32. 1090-1100. 152 weinberger. M.. Hiner. S.L.. a Tierney. W.M. (1987). In support of hassles as a measure of stress in predicting health outcomec- eeerhei ht heheeierei heeieihe' 19<1>v 19-31. Williams. A.W.. Ware. J.E.. & Donald. C.A. (1981). A model of mental health. life events. and social supports applicable to general populations. JQHEDQL Qt thlih ehe Seeiel Behheier. 23. 324-336- Witter. R.A.. Stock. W.A.. Okun. M.A.. & Haring. M.J. (1985). Religion and subjective well-being in adulthood: A quantitative cynthecic- Eeeiee et Eeliaieee Research: 2e(4). 332-342. Zarski. J.J. (1984). Hassles and health: A replication. health reeeheieae. e. 243-251- Zika. S.. & Chamberlain. K. (1987). Relation of hassles and personality to subjective well-being. teerhel Q; Eereehelite hhe eeeiei Eeeeheieae. e§(1). 155-162. Zimmerman. S.L. (1984). The mental retardation family subsidy program: It‘s effects on families with a mentally handicapped child. Eghily Relggighg. §§. 105-118. Zuk. G.H. (1959). The religious factor and the role of guilt in parental acceptance of the retarded child. American Journal Of hehtei Detieiehee. 65. 139-147. u!