NCJ Number
90335
Journal
Free Inquiry in Creative Sociology Volume: 9 Issue: 2 Dated: (November 1981) Pages: 198-203
Date Published
1981
Length
6 pages
Annotation
The use of discriminant analysis after combining several independent variables to classify parole success and failure offers possibilities for improving parole outcome prediction.
Abstract
In using discriminant analysis to distinguish between successful and unsuccessful parolees, the researcher selects a collection of discriminating variables (e.g., education, age at arrest, sentence length) that measure characteristics on which the two groups are expected to differ. The sample for this study was selected from two groups of persons under the Oklahoma Department of Corrections. Fifty selected for the failure category had postrelease convictions and been incarcerated two or more times. Twenty-eight in the success category had just one felony conviction and 18 months or more under successful parole supervision without further convictions. Parole success or failure was the dependent variable, and the independent variables used were age, years of education, age at first arrest, length of sentence, time served, socialization score, responsibility score, and type of offense. Using the t-test, it was found that age had little to do with finding the correct classification in either group. The percent of correct classifications increased from 92 percent to 98 percent when age was excluded. This testing indicates that the kinds of data used in this study could be analyzed by discriminant analysis to be used in demarcation points in the parole decision for each inmate. If total populations of incarcerated and parole offenders were to be used as a statistical base, the discriminant analysis function would adjust constantly to consider every new observation. Tabular data and two references are provided.