NCJ Number
168991
Journal
Justice Quarterly Volume: 14 Issue: 1 Dated: (March 1997) Pages: 167-180
Date Published
1997
Length
14 pages
Annotation
Using 10-year follow-up data and logistic regression techniques, this study estimated the likelihood of recidivism among parolees from a close-security prison in Ohio who had varying levels of education.
Abstract
The study used a quasi-experimental time-series analysis of 318 male inmates who left a medium-security prison in Ohio between 1982 and 1983. Some attended an associate's degree program operated by Wilmington College. Program participants lived in a separate cell block. This study examined the program's impact on recidivism by dividing inmate-subjects into two comparison groups, based on their levels of education. The treatment group consisted of 95 inmates who received associate's degrees in prison. The quasi-control group consisted of 223 randomly selected inmates at the same prison who did not receive associate's degrees but were released between 1982 and 1983. The primary dependent variable was recidivism, which was defined as return to prison for any cause. The independent variables included age, race, county of residence, substance abuse history, juvenile criminal history, violent criminal history, prearrest employment, postrelease employment, and college education. Based on the findings, this study concludes that inmates who completed higher levels of education encountered greater employment success on parole and thereby recidivated less frequently over an extended follow-up period than did their less highly educated counterparts. 2 tables, 24 references, and appended data on the characteristics of comparison groups and zero-order correlations