Further Investigation of Findings Reported for the Minnesota Sex Offender Screening Tool-Revised

• NCJ Number: 237132
• Journal: Journal of Interpersonal Violence Volume: 23 Issue: 10 Dated: October 2008 Pages: 1363-1379
• Author(s): Calvin M. Langton; Howard E. Barbaree; Leigh Harkins; Edward J. Peacock; Tamara Arenovich
• Date Published: October 2008
• Length: 17 pages
• Sponsoring Agency: Canada Correctional Service; Ontario Mental Health Foundation; Centre for Addiction and Mental Health
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

This study examined whether apparent differences in the prediction accuracy of the MnSOST-R between two studies sharing a subset of sexual offenders were statistically significant.

Abstract

Among a number of widely used risk assessment instruments with adult sexual offenders, the Minnesota Sex Offender Screening Tool-Revised (MnSOST-R) has been subject to relatively few evaluation studies. Only two independent research groups have published replication studies in the peer-reviewed literature with data not provided by the MnSOST-R's developers, and the results regarding the accuracy of predicting sexual recidivism have been mixed. In this article, important differences between the Barbaree et al. and Langton et al. studies are presented. Analyses reported for the various subsets comprising these two samples indicate that coding discrepancies in the Barbaree et al. study account for the different findings, with a moderate level of predictive accuracy using the Receiver Operating Characteristic curve ultimately found for the MnSOST-R in both datasets. (Published Abstract)