A Longitudinal Examination of the Association Between Intelligence and Rearrest Using a Latent Trait-State-Occasion Modeling Approach in a Sample of Previously Adjudicated Youth

摘要

Recidivism remains a serious issue in the modern criminal justice system, with over 80% of those previously incarcerated being rearrested within 9 years of release (Alper, Durose, & Markman, 2018). Although previous studies have identified risk factors that increase the probability of rearrest, much remains unknown regarding the full constellation of risk factors. One potential risk factor that has received limited attention is intelligence, as individuals with lower IQ scores have been found to be more likely to come into initial contact with the criminal justice system. Collectively, previous studies have provided preliminary evidence of intelligence as a risk factor for rearrest but have not fully explored this association. More specifically, it remains unclear whether the association between IQ and recidivism persists after controlling for time-invariant, individual-specific sources of variance in criminal behavior. The current study aimed to address this limitation and more closely examine the longitudinal association between IQ and rearrest with data from the Pathways to Desistance Study (N = 1,331 individuals). To distinguish variance in intelligence from time-stable, individual-specific variance in criminality, we estimated a latent trait-state-occasion model. A subsequent series of survival models, which included the previously estimated measure of criminality as a covariate, revealed a small and negative association between IQ and rearrest (hazard ratio = .95; 95% confidence interval [.92;.98]), suggesting that IQ may play only a minor role in recidivism.