Understanding Variation in Maladaptive Outcomes Across Adolescents' Religious Affiliations

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - Moral Development

Abstract

Introduction: Prior studies have reported variation in maladaptive adolescent outcomes across religious affiliation (e.g., Smith & Denton, 2005). However, little research has sought to unpack such effects and identify potential mediators. The purpose of the present study was to examine variation in four maladaptive outcomes (sexual risk-taking, alcohol use, depression, and academic cheating) across religious affiliation, and to test whether such effects are mediated by religious worship service attendance and religious salience.
Methods: The data came from the first wave of the National Study of Youth and Religion (NSYR; Smith & Denton, 2005), which is a nationally representative sample of youth (N=3370; ages 13-17, M age = 15). Participant religious affiliation was Evangelical Protestant (31%), Black Protestant (12%), Mainline Protestant (10%), Catholic (24%), Jewish (3%), Latter-day Saint (2%), and None (12%). The remaining 6% did not fit into these categories and thus were not included in the analyses. Two models were run for each of the four maladaptive outcomes of religion: an ANOVA, and then an ANCOVA with religious attendance and religious salience as covariates.
Results: First, for sexual risk taking, ANOVA analysis showed a statistically significant but small effect of religious affiliation on these maladaptive outcomes, F(6, 2750) = 6.44, p < .001, R2 = .014. The F-test of the R2 change was statistically significant, F(2, 2747) = 26.85, p < .001, ΔR2 = .029. Second, alcohol use varied significantly across affiliations, F(6, 3205) = 12.84, p < .001, R2 = .023. The F-test for the R-square change upon adding the covariates was significant as well, F(2, 3202) = 43.3, p < .001, ΔR2 = .039. Third, depression varied significantly across affiliations as well, F(6, 3205) = 2.74, p < .05, R2 = .005. The F-test for the R-square change upon adding the covariates was also significant, F(2, 3202) = 9.36, p < .01, ΔR2 = .009. Lastly, academic cheating did not vary significantly across affiliations, F(6, 3206) = 1.60, p > .05, R2 = .003.
Discussion:
The addition of religious attendance and religious salience as covariates significantly reduced the variance in the three of the maladaptive outcomes (sexual risk-taking, alcohol use, and depression) accounted for by religious affiliation, providing evidence for mediation. These results suggest that part of the reason religious affiliations differ on these adolescent maladaptive outcomes is their differences in religious commitment. Religious affiliations where youth have lower levels of maladaptive outcomes also tend to be religious affiliations with more religiously committed youth. Patterns for all three significant outcomes were complex, with no clear trends observed across affiliations. LDS teens scored statistically lower on alcohol consumption, but due to a small sample size no substantial claims could be made. Further research into these outcomes with larger sample sizes of the smaller sampled affiliations is recommended.

Authors

• McKay Morgan, Brigham Young University
  Presenting Author

• Sam A Hardy, Brigham Young University
  Non-Presenting Author