Validation of Risk/Needs Assessments in the Pennsylvania Department of Corrections Final Report, December 2006

• NCJ Number: 221255
• Author(s): Dr. David Simourd
• Date Published: December 2006
• Length: 35 pages
• Grant Number(s): 15288
• Sponsoring Agency: Pennsylvania Department of Corrections
• Sale Source: Pennsylvania Department of Corrections
• Publication Format: Document
• Publication Type: Measurement/Evaluation Device
• Language: English
• Country: United States of America

Annotation

The study examined the psychometric properties, reliability, and validity of the Level of Service Inventory-Revised (LSI-R), the Criminal Sentiments Scale-Modified (CSS-M), and the Hostile Interpretations Questionnaire (HIQ) among Pennsylvania Department of Corrections (PADOC) offenders.

Abstract

The results of the study indicate the psychometric properties of the three main PACOC assessment instruments have structural integrity and are comparable to other existing databases. The report describes the outcome of a large scale validation of three assessment instruments used by the PADOC for classification and treatment decisionmaking for offenders. The overall pattern of results of the validation of the LSI-R, the CSS-M, and the HIQ among PADOC offenders are positive; the instruments’ technical integrity among PADOC offenders is consistent with existing research, and is being properly administered. However, the relationship to outcome criteria showed weak relationships between the instruments and the predictive validity criteria, and were below comparable data. The PADOC should not be concerned by the predictive criteria results of the project since the history of appropriate correctional risk assessment instruments suggests the present results are an anomaly rather than a definitive conclusion. The application of these instruments in the case management practices of PADOC offenders is appropriate and relevant instruments for this purpose. Although it is not clear as to why the correlations with outcome criteria were so weak, a combination of methodology (archival data collection) and recent timeframe of adopting the instruments into routine practice might have played a role. Three chapters provide details for each individual instrument including data collection procedures, descriptive statistical and psychometric properties, validity, and a summary of the results. Various explanations and suggestions for future investigation are offered. Tables, references, appendices