Cognitive-Behavioral Reciprocity: Testing the Bidirectional Relationship Between Antisocial Cognition and Delinquency

Objective: Although reciprocity between variables is a topic of interest in the field of criminology, we cannot simply assume that all or even most criminological relationships are bidirectional without testing them empirically. The objective of the current investigation was to test whether delinquency and antisocial cognition are reciprocally or bidirectionally related. Hypotheses: The hypotheses evaluated as part of the present study proposed that antisocial cognition would predict delinquency, delinquency would predict antisocial cognition, and bidirectional models would display significantly better fit than the unidirectional models on which they are based. Method: Using data from the Pathways to Desistance study (1,354 serious justice-involved youths), I explored whether antisocial cognition predicts delinquency and a change in delinquency and whether delinquency predicts antisocial cognition and a change in antisocial cognition. I paired two forms of antisocial cognition-moral neutralization and cognitive impulsivity-with delinquency to predict a single future outcome with a zero-order correlation and a lagged outcome or change with a partial correlation. Results: Findings showed that 40 out of 40 prospective zero-order correlations and 36 out of 40 prospective partial correlations achieved significance, with moderate and small effect sizes, respectively. Structural equation modeling revealed that the bidirectional models linking moral neutralization to delinquency and cognitive impulsivity to delinquency using lagged outcome measures both achieved significantly better fit than the unidimensional models on which they were based. Conclusion: The results of this study are congruent with the conclusion that the relationship between antisocial cognition and delinquency is reciprocal and that antisocial cognition is as much a predictor of delinquency as delinquency is a predictor of antisocial cognition. Thus, both patterns need to be taken into account for the purposes of theory integration in criminology, clinical practice in forensic psychology, and policy implementation in criminal justice.