Journal of Offender Rehabilitation Volume: 25 Issue: 1/2 Dated: 1997 Pages: 129-141
This study analyzed the accuracy of two psychologists' predictions of criminality as a measure of future violent behavior by juveniles.
Other studies have examined how accurately predictions of violence could be made for large populations; this study examined whether a subgroup within a general population could accurately be labeled as very likely to reoffend violently. The juveniles in this study were residents of Ethan Allen School in Wales, Wis., which is a secure correctional facility for male juveniles. Two psychologists were asked to select juveniles with whom they were familiar in their work at the facility and whom they believed were very likely to become re-involved in a crime, particularly violent crime, within a year after release. They were asked to use their clinical judgment alone in making these selections. The group predicted to reoffend consisted of 29 juveniles. The control group consisted of 36 youths who could be followed closely for 1 year after release. Follow-up analysis for the subjects predicted to reoffend was done through the central records office, which provided information regarding convictions. Conviction of a felony was the criterion used to determine whether the prediction of a future criminal offense was accurate. The findings show that the psychologists accurately identified a subpopulation significantly more likely to engage in future acts of violence within a 1 year follow-up period, and they were able to select a group significantly more likely to reoffend than the general population. 3 tables and 26 references
United States of America