دینداری، خودکنترلی و رفتار ضد اجتماعی: دینداری به عنوان یک عامل ارتقا دهنده و حفاظتی

Abstract Three hypotheses with the potential to provide information on the role of religiosity as a promotive and protective factor in early adolescence were tested. Adolescents (N = 166, M age = 13 years, 49% female, 49% European American, 45% African American) and mothers reported their own personal importance of religion and the frequency of their attendance of religious services. Greater mother importance and attendance was associated with greater adolescent importance and attendance. Mother importance was indirectly linked to adolescent antisocial behavior through adolescent importance. Less adolescent importance and attendance were associated with low self-control and low self-control was associated with more antisocial and rule-breaking behavior. Adolescent importance also moderated the links between low self-control and antisocial and rule-breaking behavior such that low levels of self-control were more strongly associated with more antisocial and rule-breaking behavior among adolescents reporting low religious importance compared to adolescents reporting high religious importance.

Introduction In both adolescence and adulthood, greater religiosity is associated with more positive health-relevant outcomes (e.g., McCullough et al., 2000 and Powell et al., 2003). In adolescence, greater religiosity has been consistently linked with lower levels of involvement in a wide range of undesirable behaviors including alcohol, tobacco, and drug use, delinquency, and risky sexual behavior (e.g., Manlove et al., 2008, Sinha et al., 2007, Wallace and Williams, 1997 and Wills, Gibbons, et al., 2003). However, limited progress has been made toward understanding the processes linking religiosity with less problem behavior or toward understanding whether religiosity also functions as a protective factor. The current study tests three hypotheses to better understand whether and how religiosity functions as a promotive and protective factor in early adolescence.

Results The mean response on the importance of religion items fell between the important, and very important response options with 8% of mothers and 10% of adolescents reporting that religion is not very important compared to 64% of mothers and 40% of adolescents reporting that religion is very important. The mean response on the frequency of attendance item fell between the a few times per month and once a week options with 16% of mothers and 15% of adolescent reporting that they never attend religious services compared with 24% of mothers and 28% of adolescents reporting that they attend services a few times a week or every day. As shown in Table 1, mothers’ and adolescents’ reports of their own personal importance of religion and religious attendance were strongly, and positively, associated with one another. Likewise, mothers’ importance and attendance were strongly associated with adolescents’ importance and attendance, as expected. Furthermore, less adolescent religious importance was associated with lower self-control and with more current adolescent-reported antisocial behavior. In contrast, adolescent attendance, mother importance, and mother attendance were not associated with low self-control, antisocial behavior or rule-breaking behavior. Girls and mothers of girls reported more religious importance than boys and mothers of boys. African American adolescents reported more attendance than European American adolescents, and mothers of African American adolescents reported more importance and attendance than mothers of European American Adolescents. Single mothers also reported more importance and attendance than married mothers. Table 1. Bivariate correlations among religiosity, self-control, behavior problems, and control variables. Variable 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 1. Adol. Religious Importance 2. Adol. Religious Attendance .51⁎⁎⁎ 3. Mother Religious Importance .44⁎⁎⁎ .42⁎⁎⁎ 4. Mother Religious Attendance .33⁎⁎⁎ .58⁎⁎⁎ .60⁎⁎⁎ 5. Low Self-Control ­.20⁎ ­.14 ­.11 ­.05 6. Previous Antisocial Behavior ­.12 .02 ­.01 .10 .39⁎⁎⁎ 7. Current Antisocial Behavior ­.22⁎⁎ ­.04 ­.03 .07 .57⁎⁎⁎ .69⁎⁎⁎ 8. Previous Rule-Breaking -.05 .01 -.05 -.01 .15 .31⁎⁎⁎ .28⁎⁎⁎ 9. Current Rule-Breaking ­.03 ­.07 ­.07 .01 .22⁎⁎ .23⁎⁎ .30⁎⁎⁎ .75⁎⁎⁎ 10. Male Sex ­.24⁎⁎ ­.14 ­.25⁎⁎⁎ ­.12 .23⁎⁎ .14 .20⁎ .17⁎ .16⁎ 11. Single Parent Home .06 .13 .15⁎ .21⁎⁎ .10 .15 .14 .07 .05 ­.06 12. Mother Education ­.01 .08 .01 ­.01 ­.16⁎ ­.13 ­.22⁎⁎ ­.04 ­.06 ­.03 ­.18⁎⁎ 13. African American Ethnicity .14 .28⁎⁎⁎ .26⁎⁎⁎ .31⁎⁎⁎ .07 .19⁎ .20⁎ .07 .04 .02 .37⁎⁎⁎ ­.19⁎⁎ Note. ns = 153–166. ⁎p < .05. ⁎⁎p < .01. ⁎⁎⁎p < .001, two-tailed. Table options Path models To test the three primary hypotheses, two path models were fit in MPlus 5.2. Full information maximum likelihood estimation was used to deal with the small amount of missing data (12 cases were missing values for previous behavior problems; Schafer & Graham, 2002). The two indices of religiosity – importance and attendance – were tested in separate models that were otherwise identical. The path model included five primary variables as shown in Fig. 1. The primary variables were mother religiosity, adolescent religiosity, low self-control, adolescent-reported antisocial behavior and mother-reported rule-breaking behavior. The model included direct paths from mother religiosity to adolescent religiosity, adolescent-reported antisocial behavior, and mother-reported rule-breaking behavior. Likewise, the model included direct paths from adolescent religiosity to low self-control, antisocial behavior, and rule-breaking behavior. Additional paths in the model were from low self-control to antisocial behavior and rule-breaking behavior. The Adolescent Religiosity × Low Self-Control interaction was a multiplicative term computed from centered variables as recommended by Cohen, Cohen, West, and Aiken (2003). The interaction term is represented in the figure by paths from adolescent religiosity to the paths from low self-control to antisocial and rule-breaking behavior. The covariance between the antisocial behavior and rule-breaking residuals also was estimated. Additional covariances (not shown) were estimated among the adolescent religiosity, self-control, and the interaction term residuals to ensure that the model was not misspecified. The model also included six covariates (i.e., previous antisocial behavior, previous rule-breaking behavior, gender, marital status, ethnicity, and mother education level) that are not shown in Fig. 1. The model included paths from all six covariates to each of the five primary variables and the interaction term (see Table 2 for path estimates). Mediation was tested via indirect effects using boot-strapped standard errors and bias-corrected confidence intervals as recommended by McCartney, Burchinal, and Bub (2006) because indirect effects are not normally distributed. Simple slopes calculated according to procedures described by Cohen et al. (2003) aided the interpretation of the interaction term used to test moderation. Path model with standardized path estimates. Path estimates from the importance ... Fig. 1. Path model with standardized path estimates. Path estimates from the importance model appear before the slash; estimates from the attendance model follow the slash. Paths from six covariates to each variable shown are excluded for clarity (see Table 2 for the path estimates). **p < .05. ** p < .01. *** p < .001. Figure options Table 2. Standardized Path Estimates from Covariates to Primary Variables in the Importance and Attendance Path Models. Covariate Primary Variable Importance Model Attendance Model Previous Antisocial Behavior Adolescent Religiosity ­.12 ­.05 Previous Rule-breaking Adolescent Religiosity .02 .02 Male Gender Adolescent Religiosity ­.13⁎ ­.06 Single Parent Home Adolescent Religiosity ­.03 ­.02 Mother Education Adolescent Religiosity ­.02 .11 African American Ethnicity Adolescent Religiosity .07 .15⁎ Previous Antisocial Behavior Mother Religiosity ­.02 .07 Previous Rule-breaking Mother Religiosity ­.02 ­.03 Male Gender Mother Religiosity ­.23⁎⁎⁎ ­.12 Single Parent Home Mother Religiosity .06 .11 Mother Education Mother Religiosity .06 .07 African American Ethnicity Mother Religiosity .25⁎⁎ .27⁎⁎⁎ Previous Antisocial Behavior Low Self-control (LSC) .32⁎⁎⁎ .33⁎⁎⁎ Previous Rule-breaking Low Self-control .03 .03 Male Gender Low Self-control .15⁎ .17⁎ Single Parent Home Low Self-control .06 .06 Mother Education Low Self-control ­.11 ­.09 African American Ethnicity Low Self-control ­.02 ­.13 Previous Antisocial Behavior Religiosity × LSC .09 .20⁎ Previous Rule-breaking Religiosity × LSC ­.04 ­.17⁎ Male Gender Religiosity × LSC .05 .06 Single Parent Home Religiosity × LSC .08 .04 Mother Education Religiosity × LSC .07 .20⁎ African American Ethnicity Religiosity × LSC ­.01 ­.02 Previous Antisocial Behavior Antisocial Behavior .51⁎⁎⁎ .51⁎⁎⁎ Previous Rule-breaking Antisocial Behavior .04 .04 Male Gender Antisocial Behavior .04 .05 Single Parent Home Antisocial Behavior .01 ­.01 Mother Education Antisocial Behavior ­.08 ­.08 African American Ethnicity Antisocial Behavior .07 .06 Previous Antisocial Behavior Rule-breaking ­.03 ­.04 Previous Rule-breaking Rule-breaking .74⁎⁎⁎ .74⁎⁎⁎ Male Gender Rule-breaking .01 .01 Single Parent Home Rule-breaking .02 .00 Mother Education Rule-breaking ­.02 ­.02 African American Ethnicity Rule-breaking ­.03 ­.02 ⁎p < .05. ⁎⁎p < .01. ⁎⁎⁎p < .001, two-tailed. Table options Importance of religion For religious importance, the path model provided a very good fit to the data, χ2(2) = .50, p = .78, CFI = 1.00, root mean square error of approximation (RMSEA) = 0.00. Path estimates for the model including mother and adolescent perspectives on the importance of religion are provided in Fig. 1 before the slash. More mother religious importance predicted more adolescent religious importance but not less antisocial or rule-breaking behavior. More adolescent religious importance predicted less low self-control and less antisocial behavior. Adolescent religious importance did not predict rule-breaking behavior. Low self-control predicted more antisocial and rule-breaking behavior. To determine whether low self-control mediates the association between adolescent religiosity and misbehavior, the indirect effects from adolescent religious importance to antisocial and rule-breaking behavior through self-control were estimated as a test of hypothesis 1. The indirect effect was non-significant for both antisocial behavior, b⁎ = ­.043, 95% CI [­.09, .008], and rule-breaking behavior, b⁎ = ­.016, 95% CI [-.04, .01]. Thus, results are not consistent with hypothesis 1. To determine whether religious importance moderates the association between self-control and misbehavior, the importance × self-control interaction was included as predictor of misbehavior as a test of hypothesis 2. As shown in Fig. 2, and consistent with hypothesis 2, simple slopes indicated that low self-control was less strongly associated with antisocial behavior at high (+ 1 SD) levels of adolescent religious importance, b = .12, SE = .05, p = .026, than at low (­1 SD) levels of religious importance, b = .36, SE = .05, p < .001. Likewise, as shown in Fig. 3, simple slopes indicated that low self-control was less strongly associated with rule-breaking behavior at high levels of adolescent religious importance, b = .02, SE = .08, p = .85, than at low levels of religious importance, b = .24, SE = .08, p = .004. Simple slopes showing the association between low self-control and antisocial ... Fig. 2. Simple slopes showing the association between low self-control and antisocial behavior at high (+ 1 SD) and low (­1 SD) levels of adolescent importance of religion. Figure options Simple slopes showing the association between low self-control and rule-breaking ... Fig. 3. Simple slopes showing the association between low self-control and rule-breaking behavior at high (+ 1 SD) and low (­1 SD) levels of adolescent importance of religion. Figure options To determine whether adolescent religiosity mediates the association between mother religiosity and misbehavior, the indirect effects from mother religious importance to antisocial behavior and rule-breaking behavior through adolescent religious importance were estimated as a test of hypothesis 3. The total indirect effect from mother religious importance through adolescent religious importance to antisocial behavior (directly and through low self-control) was significant, b⁎ = ­.065, 95% CI [­.12, ­.01], as was the specific indirect effect (not passing through self-control), b⁎ = ­.048, 95% CI [­.10 to ­.001]. Both the total indirect effect, b⁎ = ­.009, 95% CI [­.05, .07], and the specific indirect effect, b⁎ = ­.006, 95% CI [­.02, .006], were non-significant predictors of rule-breaking behavior. Thus, results for antisocial behavior, but not for rule-breaking behavior, are consistent with hypothesis 3. Religious attendance For religious attendance, the path model provided a very good fit to the data, χ2(2) = .10, p = .95, CFI = 1.00, RMSEA = 0.00. Path estimates for the model including mother and adolescent attendance of religious services are provided in Fig. 1 following the slash. More mother religious attendance predicted more adolescent religious attendance but not less antisocial or rule-breaking behavior. More adolescent religious attendance predicted less low self-control. Adolescent religious attendance did not predict antisocial or rule-breaking behavior. Low self-control predicted more antisocial and rule-breaking behavior. Once again, the indirect effects from adolescent religious attendance to antisocial and rule-breaking behavior through self-control were estimated as a test of hypothesis 1. The indirect effects from adolescent religious attendance through low self-control to both antisocial behavior, b⁎ = ­.047, 95% CI [­.10, .007], and rule-breaking behavior, b⁎ = ­.014, 95% CI [­.04, .01], were non-significant. Thus, results are not consistent with hypothesis 1. Adolescent religious attendance did not moderate associations between self-control and antisocial behavior or rule-breaking behavior and thus results are not consistent with hypothesis 2. Finally, the total indirect effect from mother religious attendance through adolescent religious attendance to antisocial behavior (directly and through low self-control), b⁎ = ­.035, 95% CI [­.12, .05] was non-significant as was the specific indirect effect (not passing through self-control), b⁎ = ­.01, 95% CI [­.09 to .07]. Likewise, both the total indirect effect, b⁎ = ­.06, 95% CI [­.15, .03], and the specific indirect effect, b⁎ = ­.008, 95% CI [­.02, .008], were non-significant predictors of rule-breaking behavior. Thus, results for religious attendance are not consistent with hypothesis 3.