skip navigation


Register for Latest Research

Stay Informed
Register with NCJRS to receive NCJRS's biweekly e-newsletter JUSTINFO and additional periodic emails from NCJRS and the NCJRS federal sponsors that highlight the latest research published or sponsored by the Office of Justice Programs.

NCJRS Abstract

The document referenced below is part of the NCJRS Virtual Library collection. To conduct further searches of the collection, visit the Virtual Library. See the Obtain Documents page for direction on how to access resources online, via mail, through interlibrary loans, or in a local library.


NCJ Number: 157410 Add to Shopping cart Find in a Library
Title: Accuracy of Recidivism Prediction Scales: A Technical Description of Data and Research Methods
Author(s): J Cohen; S E Zimmerman
Date Published: 1995
Page Count: 75
Sponsoring Agency: Carnegie Mellon University
Pittsburgh, PA 15213
National Institute of Justice (NIJ)
Washington, DC 20531
National Institute of Justice/
Rockville, MD 20849
NCJRS Photocopy Services
Rockville, MD 20849-6000
US Dept of Justice NIJ Pub
Washington, DC 20531
Grant Number: IJ-CX-86-0039
Sale Source: National Institute of Justice/
NCJRS paper reproduction
Box 6000, Dept F
Rockville, MD 20849
United States of America

NCJRS Photocopy Services
Box 6000
Rockville, MD 20849-6000
United States of America

Carnegie Mellon University
H. John Heinz III School of Public Policy and Management
Pittsburgh, PA 15213
United States of America
Document: PDF
Type: Research (Applied/Empirical)
Language: English
Country: United States of America
Annotation: This study was conducted to examine the transportability of four recidivism prediction scales (RAND, INSLAW, SFS81, and CGR) across populations and applications, based on the assumption that it is important to know whether scales developed to predict offender behavior in one population can be employed in another population without substantial degradation in performance and whether scales developed for use at one criminal justice processing stage can be used in another stage.
Abstract: The four recidivism prediction scales were originally designed to predict different outcomes. Offender attributes in each scale included adult criminal record, juvenile record, drug/alcohol use, age at target arrest, educational attainment, and employment history. All data sets associated with the scales contained longitudinal information on individual offending, as indicated by criminal justice interventions. Although the four recidivism prediction scales represented useful approaches to making comparisons across data sets, no scale performed very well or very consistently. Researchers found that, without considerable improvement in measurement ability, theoretical considerations about what factors affect recidivism are less important than statistical characteristics of the sample and the scale. Appendixes contain scale scoring rules and supporting analytical data. 17 references, 14 footnotes, 9 tables, and 6 figures
Main Term(s): Corrections
Index Term(s): Comparative analysis; Comparative criminology; Crime patterns; Longitudinal studies; Recidivism prediction; Research methods; Statistics
To cite this abstract, use the following link:

*A link to the full-text document is provided whenever possible. For documents not available online, a link to the publisher's website is provided. Tell us how you use the NCJRS Library and Abstracts Database - send us your feedback.