NCJ Number
63465
Date Published
1978
Length
52 pages
Annotation
FOURTEEN PREDICTION SCALES FOCUSING ON VARIOUS ASPECTS OF CRIMINALITY, PAROLE, AND PROBATION, AS DEVELOPED THROUGH DATA COLLECTED IN A 50-STATE SURVEY BY THE VIRGINIA DEPARTMENT OF CORRECTIONS, ARE PRESENTED.
Abstract
THE PROJECT WAS INITIATED TO DETERMINE THE LEVEL OF USE OF PREDICTION IN THE CRIMINAL JUSTICE SYSTEM AND TO GATHER CURRENT DATA RELATIVE TO PREDICTION. THE ACQUIRED DATA INDICATE THAT PREDICTION METHODOLOGY IS NOT USED IN MOST RESEARCH ACTIVITIES OR IN THE CRIMINAL JUSTICE DECISIONMAKING PROCESS. AS A RESULT OF THE INITIAL SURVEY, 14 PREDICTION SCALES WERE DEVELOPED. THEY PURPORT TO QUANTIFY PAROLE SUCCESS AND FAILURE, RISK OF ASSAULT AND PROPERTY CRIMES AMONG PAROLEES, INMATES, PROPENSITY FOR ESCAPE, UNRESPONSIVE INMATE CHARACTERISTICS, FACTORS ASSOCIATED WITH INSTITUTIONAL COMMITMENT, AND THE COMMUNITY RISK ASSOCIATED WITH PROBATION PLACEMENT. THE NUMBER OF CASES FROM WHICH THE SCALES WERE DEVELOPED RANGES FROM LESS THAN 200 TO SEVERAL THOUSAND, AND THE TECHNIQUES EMPLOYED TO PRODUCE THE SCALES VARY IN SOPHISTICATION AND COMPLEXITY. BURGESS, EXPERIENCE, AND BASE EXPECTANCY SCALES ARE INCLUDED. SEVERAL RESEARCHERS USED MULTIPLE REGRESSION ANALYSIS, CONFIGURAL ANALYSIS, AND DISCRIMINATE ANALYSIS. THE ACCURACY OF ALL PREDICTION DEVICES DEPENDS ON THE RELIABILITY AND VALIDITY OF THE VARIABLES IN THE DATA BASE AND THE DEGREE TO WHICH THE CONSTRUCT SAMPLE IS REPRESENTATIVE OF THE TARGET POPULATION. THIS FACT IS TRUE OF THE DEVELOPED PREDICTION SCALES. NO SCALE APPEARING IN THIS REPORT SHOULD BE USED FOR ANY PURPOSE WITHOUT FIRST CHECKING ITS PREDICTIVE VALIDITY ON THE INTENDED POPULATION. TABULAR STATISTICAL DATA, GRAPHS, CHARTS, AND FORMS ARE PROVIDED IN THE SURVEY REPORT. (AUTHOR ABSTRACT MODIFIED--LWM).