NCJ Number
29770
Date Published
1975
Length
16 pages
Annotation
SEVERAL TRADITIONAL METHODS OF MEASURING TREATMENT SUCCESS ARE REEXAMINED; AS THEY ARE EMPLOYED, THESE METHODS ARE FOUND TO BE PROCEDURALLY UNSOUND IN THAT THEY YIELD UNJUSTIFIABLY HIGH SUCCESS RATES.
Abstract
THE AUTHOR STATES THAT THE MEASURES OF RECIDIVISM AND PAROLE VIOLATIONS ARE FAULTY SINCE THEY REFLECT ONLY THOSE OFFENDERS WHO ARE NOTICED, AND SINCE NOT ALL THOSE NOTICED RECEIVE THE OFFICIAL SANCTIONS OF REARREST OR PAROLE REVOCATION. IT IS ALSO NOTED THAT MANY PROGRAMS DO NOT COUNT THOSE WHO DO NOT COMPLETE TREATMENT AS FAILURES - THEY ARE SIMPLY NOT COUNTED AT ALL. SINCE THE RATE OF NONCOMPLETION CAN BE AS HIGH AS 30 PERCENT IN SOME PROGRAMS, FAILURE TO COUNT THESE CASES CAN LEAD TO CONSIDERABLE STATISTICAL ERROR. ALSO DISCUSSED ARE PROBLEMS IN ASSURING THE COMPARABILITY OF CONTROL AND EXPERIMENTAL GROUPS AND THE REPORTING BIASES THAT MAY BE INTRODUCED BY THOSE INVOLVED IN THE PROJECT. IT IS CONCLUDED THAT REGARDLESS OF THE TYPE OF PROGRAM INVESTIGATED, RESIDENTIAL INSTITUTIONS FOR DELINQUENTS ARE CHARACTERIZED BY HIGH RATES OF POTENTIAL FAILURE. IN SPITE OF THIS LACK OF IMPACT ON FAILURE RATES, THE AUTHOR CONTENDS THAT THESE PROGRAMS ARE OF VALUE SINCE THEY PROVIDE HUMANITARIAN TREATMENT WITHOUT INCREASING THE RISK OF FAILURE.