Revisiting the Empirical Status of Social Learning Theory: A Meta-analysis of Substance Use

• In Event: Theoretical Predictors of Substance Use

Abstract

Pratt and colleagues (2010) published their meta-analysis on social learning theory which found that variables specified by social learning theory tend to be robust predictors of criminal and deviant behavior. Even so, their results of moderator analyses indicated considerable variation in the magnitude and stability of effect sizes across various model specifications. The current meta-analysis revisits the empirical status of social learning theory by synthesizing research published between 1973 and 2018 that has empirically tested concepts of Akers’ social learning theory in regards to substance use. Specifically, this study assesses the utility of the main variables specified by social learning theory (differential association, definitions, differential reinforcement, and imitation) for predicting both illegal and legal substance use, including alcohol, cigarettes, and opioids. The authors discuss differences in effect sizes found through moderator analyses and conclude with recommendations for possible policy implications.

Authors

• Nathan Kruis, Indiana University of Pennsylvania

• Chunghyeon Seo, Indiana University of Pennsylvania

• Bitna Kim, Indiana University of Pennsylvania