U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

Randomized Experiment Testing Inmate Classification Systems

NCJ Number
200102
Journal
Criminology & Public Policy Volume: 2 Issue: 2 Dated: March 2003 Pages: 215-242
Author(s)
Richard A. Berk; Heather Ladd; Heidi Graziano; Jong-Ho Baek
Date Published
March 2003
Length
28 pages
Annotation
This article describes the development and testing of a new inmate classification scoring system.
Abstract
When inmates enter the California prison system, they are classified according to the level of supervision required for them; the higher the security classification, the more restrictive the prison environment. The authors developed a new classification system to better predict inmate misconduct. The development process involved removing outdated items that have been shown to have no association with prison misconduct and adding several new variables that have proven useful in predicting inmate behavior. In order to test the new classification system, 20,000 inmates were randomly assigned to have their security level determined by either (a) using the standard scoring system or (b) using the new scoring system. Researchers focused on examining the key outcomes of amount and type of misconduct in prison and prison overcrowding. Results indicated that the new system performed better at sorting inmates according to their level of risk, especially in the case of predicting serious misconduct in prison. Although there is the potential of crowding problems, the new system is better than the old system in terms of accurate sorting and its ease of use. Tables, references