NCJ Number
221437
Journal
Federal Probation Volume: 71 Issue: 3 Dated: December 2007 Pages: 25-29
Date Published
December 2007
Length
5 pages
Annotation
Using a sample of probationers and parolees in Iowa, this study tested the predictive validity of the Level of Service Inventory-Revised (LSI-R), a risk/needs assessment tool used in correctional settings.
Abstract
The findings show that the total LSI-R score is significantly related to the prediction of future criminal behavior. The higher the total risk score, the more likely that the client would reoffend. Both the bivariate and receiver operating characteristic (ROC) analyses showed that the LSI-R was a valid predictor of reoffending for probationers and parolees. The final analysis will examine the predictive ability of the LSI-R while considering the effect of gender, race, age, marital status, supervision status, and time at risk. Initial LSI-R assessments were completed between May 12, 2003, and November 21, 2003, using a sample of 1,145 cases (902 probationers and 243 parolees). Only one dependent variable, reoffending, was considered in the study. This measure pertained to whether or not an offender was rearrested based on a felony charge or an indictable misdemeanor. Time at risk was measured as the number of days from the start of supervision until the end of the followup period. The paper also suggests how programs and facilities can validate the LSI-R on their own clients. 3 tables and 24 references