Assessing the Predictive Utility of the Positive Achievement Change Tool at a Texas Juvenile Justice Agency
Abstract
In the United States, millions of youth are processed in the juvenile justice system each year. Juvenile probation officers, case managers, and clinicians have the daunting task of making critical case-processing decisions for this vulnerable population including program placement, supervision, and treatment. Moreover, juvenile justice professionals must be mindful of numerous factors such as risk to public safety and the juvenile’s risk to reoffend when making these decisions. In recent years, juvenile justice agencies have employed the Risk-Need-Responsivity model, the pre-eminent evidence-based model for offender assessment and management, to aide in the decision-making process. Although implementation of risk assessment tools has improved the accuracy with which juvenile offenders are assessed and treated, lack of tool validation and poor implementation may have a negative impact on their predictive utility. The present research examines the predictive utility of the Positive Achievement Change Tool currently in use at a local Texas juvenile justice agency. Results demonstrated the tool’s failure to predict recidivism at the state and technical violations level. The tool did demonstrate predictive utility at the agency-level, however the magnitude of effect was small. Regression analyses demonstrated the PACT Overall Risk Scores predictive power across recidivism types, however, the variance explained did not exceed 2%. Additionally, Criminal History Scores were not a significant predictor of recidivism. The practical and theoretical implications of the findings are discussed.
Subject Area
Developmental psychology|Criminology|Behavioral psychology
Recommended Citation
Hutchins, Elizabeth Perez, "Assessing the Predictive Utility of the Positive Achievement Change Tool at a Texas Juvenile Justice Agency" (2019). ETD Collection for University of Texas, El Paso. AAI13883824.
https://scholarworks.utep.edu/dissertations/AAI13883824