Predictions from Assessments of Violent Offenders Under Stress

Abstract

This is a study of violence prediction as carried out in the Stress Assessment Unit of the California Department of Corrections at the Vacaville institution, between 1962 and 1977. The population studied consisted of 573 men, who completed the program with either positive or negative recommendation to the parole board and were paroled or discharged before 1975. Parole and arrest records were used to follow up for a minimum of two years postrelease up to ten years. The percentage of those with negative recommendations who had subsequent arrests for violent acts (38.5) was not statistically significant. Those with positive parole recommendations were more likely, however, to avoid a return to major criminal activity of any sort (30% versus 43%). It is concluded that the method used did not substantially improve the accuracy of predicting future violence for these men with known violent past crimes. It is suggested that efforts such as this one designed on commonsense assumptions should not be repeated without rigorous research design.