Criminal Justice and Behavior, Ahead of Print.
This study had two objectives: (1) to evaluate the predictive accuracy of the AZYAS-Reentry tool and (2) to test whether predictive accuracy levels were consistent across sex and race/ethnicity. Data from 249 Arizona adolescents released from secure care on parole supervision between June 2020 and July 2023 were examined. The mean AZYAS-Reentry score classifies much of the sample as moderate risk. Bivariate analyses show a medium effect size for AZYAS-Reentry on parole revocation for the full sample (r = .344). Predictive accuracy is the highest for African American youth (r = .534), followed by Hispanic (r = .398) and white youth (r = .191). Higher predictive accuracy for African Americans is driven by specific domains (e.g., values, beliefs, and attitudes). The one sex-based difference is the stronger effect of juvenile justice history for females. Findings support using AZYAS-Reentry and highlight the utility of targeting domains to plan interventions across sex and race/ethnicity.