PCRA revisited: Testing the validity of the Federal Post Conviction Risk Assessment (PCRA)

Area of Science:

• Criminology
• Forensic Psychology
• Sociology

Background:

• The Post Conviction Risk Assessment (PCRA) was established in 2009 for the U.S. Probation System.
• Prior studies confirmed PCRA's predictive validity using development and validation samples.

Purpose of the Study:

• To evaluate the PCRA's predictive accuracy for new arrests (any crime and violent offenses) over 24 months.
• To assess PCRA's validity across different racial, gender, and ethnic groups.

Main Methods:

• Analysis of 113,281 offender assessments conducted by U.S. probation officers.
• Bivariate and multivariate statistical models were employed.
• Receiver Operating Characteristic (ROC) curve analysis was used to determine predictive accuracy (AUC-ROC).

Main Results:

• Area Under the Curve (AUC-ROC) values ranged from .70 to .77.
• The PCRA demonstrated consistent predictive validity across various subgroups and follow-up intervals.
• The tool effectively predicted arrests for both general criminal conduct and violent offenses.

Conclusions:

• The PCRA is a valid and reliable tool for predicting recidivism in supervised offenders.
• The PCRA's predictive performance is equitable across diverse demographic groups.
• This study validates the PCRA's utility in informing probation and parole decisions.