This paper examined the relationship between adolescent religious belief and delinquency using multivariate analysis. Religious belief was theorized to be one element of social controls that can constrain the innate tendency toward delinquent behavior. Although a number of researchers have suggested that adolescent religious belief might be an inhibitor of delinquency rather than a contributor, the nature of a relationship between religiosity and delinquency is still controversial and unclear. This study intended to overcome the limitations of previous studies by adopting advanced statistical analysis with longitudinal data.

The main purpose of this study was to examine: (1) whether adolescents who have religious beliefs are less likely to commit delinquent behavior; (2) whether the effect of religious beliefs vary across different types of delinquent behavior such as minor offenses and serious offenses; (3) whether religious beliefs strengthen social bonding to conventional institutions such as family, school, and neighborhood; and (4) whether religious beliefs decrease a harmful influence of delinquent peers.

The data for this study was drawn from a subset of the National Longitudinal Study of Adolescent Health, a longitudinal study of adolescents in grades 7 through 12. This study utilized two waves of in-home interviews conducted with adolescents in 1995 and 1996. A total of 4,412 adolescents were included. The structural equation model was the primary analytic method utilized to examine a series of relationships among variables simultaneously without being influenced by measurement errors. In order to strengthen confidence in the proposed causal directions among variables, exogenous variables are chosen from the wave 1 data and endogenous variables were taken from the wave 2 data. A covariance matrix is used as an input data form. The Maximum Likelihood method was utilized for a method of model estimation. Overall model fit was assessed with several goodness-of-fit indexes because a single index reflects only a particular aspect of fit. Five fit indexes were utilized: Chi-square statistics, Comparative Fit Index, Goodness of Fit, Adjusted Goodness of Fit and Root Mean Square Error of Approximation.

The results of the study indicated adolescent religious beliefs played significant roles in deterring serious delinquent behaviors. Yet, the results were somewhat different for minor offenses. There was no association between religious beliefs and minor delinquent behaviors. The findings provided support that the effect of religious beliefs might be more strongly linked to certain types of offenses. Also, the study indicated that adolescent religiosity might affect many aspects of their social lives. Religious adolescents were more likely to bond to conventional institutions such as family and school and were less likely to associate with delinquent friends.

The findings of this study have some significant implications for practice. As adolescents cannot be juxtaposed with their environment, it is essential for social workers to have comprehensive understanding of adolescents' social environments, which includes the spiritual aspects of adolescents' lives. Advances in knowledge on religious belief-delinquency will enable social workers to develop a deeper understanding of delinquent behaviors among adolescents, which will enable the application of a more holistic practice approach.