Criminal thinking and recidivism: Meta-analytic evidence on the predictive and incremental validity of the Psychological Inventory of Criminal Thinking Styles (PICTS)

Two meta-analyses were performed on the Psychological Inventory of Criminal Thinking Styles (PICTS: Walters, 1995) as a predictor of recidivism. Seven samples from six different studies (5 published, 1 unpublished) were included in the first meta-analysis. The results revealed a pooled mean effect size (r) of .20 for the reconstructed General Criminal Thinking score (GCT_rc), a pooled mean effect size of .17 for the reconstructed Proactive Criminal Thinking score (P_rc), and a pooled mean effect size of .17 for the reconstructed Reactive Criminal Thinking score (R_rc). Data from three of these studies (4 samples) were sufficient to perform a meta-analysis of the incremental validity of the GCT_rc. In all four samples, the GCT_rc score predicted recidivism above and beyond the contributions of age and criminal history, with a mean effect size (odds ratio) of 1.27. Whereas the PICTS GCT_rc score satisfied two of the three criteria for dynamic risk (i.e., dynamic risk factors are statistically and clinically significant predictors of outcome and are incrementally valid relative to well-established static risk factors), there is a need for additional data to assess its standing on the third criterion (i.e., changes in dynamic risk factors predict outcome and changes in outcome risk).