NCJ Number
72758
Journal
Journal of Criminal Justice Volume: 8 Issue: 3 Dated: (1980) Pages: 185-188
Date Published
1980
Length
4 pages
Annotation
The ravalidation of thesalient factor score used by the United States Parole Commission is described in this article.
Abstract
The score serves as a predictive instrument in the Federal parole selection process. Revalidation is required in order to assure that the prredictive items are still effective. The score was originally constructed by using information obtained for a random sample of 2,483 Federal prisoners who wre released in 1970.Initial validation was achieved by analyzing data for random samples of prisoners released in 1971 (1,138 prisoners) and 1972 (1,011). The revalidation was based on information obtained for 1,260 prisoners who were released in 1976. The 1976 sample differed in the following ways: while the previous samples each consisted of cases selected from prison releasees, the 1976 sample included a subsample of 697 randomly selected prisoners released through community treatment centers and a comparative but nonrandom subsample of 563 prisoners who were eligible for community treatment center processing but were released directly to the community; and while the previous samples were not geographically restricted, the 1976 sample was limited to releasees in selected metropolitan areas. For each case in the 1976 sample, a I-year followup period from the month of release was available. The following criterion measure was used: favorable outcome -- no new arrest and no parole violation warrant issued. The results showed that the percentage of cases with favorable outcome withing each salient factor score group remained stable over the samples. Calculation of the statistic used to measure predictive power also demonstrated stability over the time periods. A data table, footnotes, and 12 references are included. An appendix contains a list of the scoring items.