Religious bonds, low self-control, and deviant behavior: a fixed-effects approach

Abstract

ABSTRACT Prior scholarship has raised concerns over potential selection effects with religion-deviant behavior relationships. Yet, few studies have attempted to account for selection beyond incorporating control measures in statistical models, and fewer still have done so using adult samples. Using data from the National Study of Adolescent to Adult Health (N = 9,799), this study employs fixed-effects techniques to examine if religious bonds are related to deviant behaviors, and if these relationships are mediated by low self-control. Results from the fixed-effects logistic regressions show that increases in religious bonds are related to decreased odds of both substance use outcomes (binge drinking and marijuana use), but do not relate to non-violent deviance and are related to increased odds of violent deviance. Results further show that low self-control, while a significant predictor of each type of deviant behavior, fails to substantially mediate any of the relationships.