Citation

Benda BB. J. Crim. Justice 1989; 17(6): 487-500.

Copyright

(Copyright © 1989, Elsevier Publishing)

DOI

10.1016/0047-2352(89)90079-2

PMID

unavailable

Abstract

This prediction study compared three statistical procedures for predictive accuracy using a criterion of return to Wisconsin training schools among first-admissions to these prisons for youthful offenders. Comparing logit analysis, predictive attribute analysis, and a Burgess method, this study found no clear superiority in prediction among these statistics; however, there were dimensions of these classifications in which one statistical procedure was more predictive than another. Some implications of prediction are discussed.