The Outcomes of Hospitalized and Incarcerated Sex Offenders: A Study of Offender Types, Recidivism Rates, and Identifying Characteristics of the Repeat Offender (longitudinal, Incest, Rape, Pedophilia)

Abstract

The purpose of this study was threefold: (1) to provide further knowledge regarding the characteristics of sex offenders and their offenses, (2) to evaluate the effectiveness of treatment, and (3) to identify predictors of recidivism. Subjects were 150 sex offenders who had been either committed to a treatment facility or imprisoned. The treated and imprisoned offenders were matched on type of sex offense, prior sex convictions, and security status at release. , Characteristics of pedophiles, rapists, and incest offenders were compared. Differences found were largely limited to characteristics of their offenses. Pedophiles were less likely than rapists to assault strangers, and more likely to engage in lengthy exploitation of victims. Incest offenders and pedophiles were less physically aggressive than rapists, and encountered less resistance from victims. Incest offenders were older than pedophiles or rapists. Personal histories, family backgrounds, and most aspects of criminal history were similar across offender types. , Recidivism was evaluated 1 to 9 years after release. The proportions of recidivists in the treated and imprisoned groups did not differ. Examination of uncontrolled variables revealed initial differences that may have threatened the adequacy of using imprisoned offenders as a comparison group. , A discriminant function analysis successfully identified predictors of recidivism. One set of predictors distinguished between types of recidivism, while the other distinguished between recidivism and non-recidivism. The primary predictors were age at first sex conviction and relationship to victim. Both discriminant functions were significant, and together accounted for 22.5% of the variation in outcome. Although the predictive equations correctly classified 77% of the subjects, false positive rates were high. , It was concluded that heterogeneity among these confined sex offenders may be largely limited to characteristics of their criminal sexual behaviors. Secondly, it was concluded that matching procedures may not sufficiently override dissimilarities that initially form the basis for differential processing of sex offenders, and thus do not guarantee the adequacy of comparison groups formed in this manner. It was also concluded that statistical prediction shows some promise as a tool for identification of potential recidivists, but high error rates limit its practical utility at this time.