Predicting Recidivism Using the Offender Screening Tool

Abstract

An estimated 60% of sex offenders serve a term of probation or parole, and the level of risk posed by a particular offender has important implications for the amount of supervision required by probation. Thus, it is especially crucial that probation offices can accurately assess this risk. Probation officers in a growing number of jurisdictions now conduct risk assessments, help create case plans addressing criminogenic needs, match offenders to treatment services, and use motivational techniques to encourage behavior change. The purpose of the current project was to evaluate the Offender Screening Tool (OST), a risk assessment measure used by the probation department in a southwest county in Arizona to determine risk of reoffense and assign a level of supervision to sex offenders on probation. OST scores and recidivism data was collected from 174 adult male sex offenders supervised by the probation department. Participants were also scored on the Static-99R to compare the OST's performance with a well-established risk measure. The OST's accuracy in predicting sexual, violent nonsexual, and nonsexual nonviolent recidivism was measured. Results supported use of the OST for predicting nonsexual, nonviolent recidivism, while the Static-99 performed superiorly in predicting sexual reoffending. Implications for the use of the OST are discussed.