Heterogeneity in the development of differential reinforcement: examining predictive power of definitions and differential association from a social learning perspective

Abstract

Social learning theory is a prominent framework for understanding the relevance of socialization processes for predicting behavior. While recent research has focused on placing these processes within developmental contexts, there remain gaps in understanding how differential reinforcement for involvement in antisocial behavior may develop across adolescence and emerging adulthood and how definitions and differential association predict development. All 11 waves of the Pathways to Desistance dataset were analyzed, comprising the responses of 1354 justice-involved youth across seven years post-adjudication. Group-based trajectory modeling was used to identify developmental patterns of differential reinforcement and regression modeling was used to examine predictors of development. Results indicated that a two-group model of development best fit the reinforcement data (High, Low). Greater endorsement of antisocial definitions and greater differential association with deviant peers and a significant and positive interaction between these constructs predicted increased odds of assignment to the High group.