Recidivism of criminal psychopaths after therapeutic community treatment

Abstract

Official reconviction rates were examined for one hundred six male offenders participating in a Therapeutic Community treatment program at the Regional Psychiatric Centre in Saskatoon, Saskatchewan. Offenders were assessed by Hare's Revised Psychopathy Checklist (PCL-R, Hare, 1985a, 1991), and were divided into low, medium, and high psychopathy groups. The proportion of offenders reoffending within each PCL-R group, and survival analyses, indicated the high PCL-R group reoffended at a higher rate than the low PCL-R group on most measures of recidivism. These results are consistent with the view that psychopaths should be particularly resistant to treatment. However, because the high PCL-R group had a more extensive criminal history prior to treatment compared with the low PCL-R group, and there was no untreated control group, we cannot be sure treatment did not have an impact on subsequent rates of offending. Nonviolent recidivism was consistently related to the behavioural deviance component of the PCL-R, and less related to the interpersonal-affective component of the PCL-R, but the pattern of results for violent recidivism was less clear. A second major purpose of the study was to compare the predictive accuracy of the PCL-R against two standardized actuarial scales, the statistical

Information On Recidivism (SIR; Nuffield, 1982), and the Salient Factor Score (SFS; Hoffman, 1983). Predictions of nonviolent recidivism by the PCL-R, SIR, and SFS were superior to predictions of violent recidivism, and depending on the measure of recidivism, different predictor scales were more accurate. For practical decision-making, the PCL-R was not able to predict recidivism beyond the contribution of the actuarial scales, as suggested by correlational and logistic regression analyses, Relative Improvement Over Chance statistics, and kappa coefficients. Limitations of the current study and areas for future research are considered in more detail in the thesis.