Increasing the predictive validity of the Oregon youth authority recidivism risk assessment with black juveniles

Abstract

Previous researchers have found that the Oregon Youth Authority’s (OYA) current risk assessment tool for juvenile offenders (the OYA Recidivism Risk Assessment [ORRA]) underpredicted recidivism for Black juveniles. The present study was conducted to determine whether adding dynamic factors to the ORRA, which utilizes static risk factors only, would improve the prediction of recidivism for Black juveniles. Using a sample of 4,090 youths who were released from OYA between January 1, 2006, and December 31, 2010, multiple logistic and hierarchical logistic regressions were conducted. The results indicated that, although the dynamic variables were not statistically significant predictors of recidivism when added to the ORRA score, the addition of the dynamic variables reduced underprediction of recidivism for Black juveniles.