Abstract
One hundred and twenty male inmates who had been released from a maximum-security inpatient psychiatric unit of a federal penitentiary were followed for an average of 92 months. Seventy-eight percent of the sample were arrested for any offense, and 50% were arrested for a violent offense. An actuarial instrument developed using Nuffield's (1982) method correlated .43 with violent recidivism, resulting in 32% relative improvement over chance. The predictors of violent recidivism in this instrument were the following: juvenile delinquency, younger age at release, drugs involved in offenses, violent convictions, separation from parents before age 16, alcohol involved in offenses, criminal versatility, short periods of employment, and no psychotic illness. These results support those of earlier follow-up studies, in particular, the negative association between psychosis and violent recidivism in high-risk samples and the use of actuarial instruments for appraising the long-term risk of violent recidivism.
Get full access to this article
View all access options for this article.
