NCJ Number
13242
Date Published
1973
Length
42 pages
Annotation
REVIEW OF THE PROBLEMS ASSOCIATED WITH CONSTRUCTING PREDICTION TABLES FOR OPERATIONAL AND RESEARCH USE AND SOME ALTERNATIVE METHODS OF PREDICTION.
Abstract
FOLLOWING A DISCUSSION OF THE NEED FOR ACCURATE PREDICTION TABLES FOR PAROLE DECISION-MAKING, THE AUTHOR FOCUSES ON POSSIBLE IMPROVEMENTS IN PREDICTION TABLE BUILDING. SUGGESTIONS ARE MADE FOR CODING DATA ON OFFENDERS ACCORDING TO SEQUENCE OF EVENTS, USING THE INDIVIDUAL AS HIS OWN CONTROL, AND EXAMINING DEVIATIONS FROM HIS NORM. SUGGESTIONS ARE ALSO MADE FOR PROCESSING DATA SO THAT THEY NO LONGER RESEMBLE THE CASE FILES AND THE WAY THE INFORMATION WAS CONSIDERED BY THE CLINICAL DECISION-MAKER. THE DISCUSSION THEN CENTERS ON NEW METHODS OF STATISTICAL ANALYSIS, SUCH AS THE USE OF PROBABILITY STATEMENTS AND OTHER KINDS OF PREDICTION. THE AUTHOR CONSIDERS THE DESIGN AND USE OF INSTRUMENTS VERSUS RESEARCH INVESTIGATIONS, MODELS AND SIMULATION OF THE DECISION PROCESS FOR USE BY PAROLE BOARDS, AND PURE AND APPLIED RESEARCH.