The Mean Cost Rating (MCR) measures the differentiation ability of those variables or groups of variables used in construct ing parole prediction or base expectancy tables. The literature of the past two decades suggests that it represents perhaps the most satisfactoiy statistical index of predictive selectivity. Its use has been limited, however, by its complex method of computa tion. The MCR is explained and a simplified procedure for com puting it is offered. In addition, the MCR is compared with Chi square, the J-value, and the Index of Predictive Selectivity, which have also been used to determine the selective ability of factors in the construction of prediction tables.

No abstract

A key decision in the correctional process is determining when to release an offender to parole. Setting the optimum time to be served is difficult. One of the purposes of this study is to see if research based on past experience can provide some future guidelines for determining optimum time for different types of offenders. The Uniform Parole Reports (UPR) program, sponsored by the National Probation and Parole Institutes, provides the research base for this study. The study groups consisted of males convicted of burglary offenses who were paroled and reported to UPR. These groups consisted of 7,200 parolees in 1968 and 7,600 parolees in 1969. The standard UPR definition of parole outcome for a one- year follow-up was used as the criterion variable. An experience table with different offender classifications was developed to achieve the maximum logical differentiation as to parole outcome. The parole outcome rates for the 1968 classifications ranged from 26 percent to 87 percent with a favorable adjustment at the end of one year. The classifications developed for 1968 parolees fairly accurately predicted the 1969 parolees' experience. In the process of developing the offender classifications, sup portive evidence was found for the maturation concept. In 21 of 22 comparisons, the older parolees had more favorable outcome rates than the younger parolees. The number of months served showed no consistent relationship to parole outcome for any classification. This study does not assess imprisonment as a punish ment device or as a means of custody. It does suggest that the vast sums being spent on correctional institutions, as a crime reduction device, need further evaluation if we are to stem the growing crime rate.

The purpose of this study is to examine how different types of students respond t o various drug abuse prevention areas, so that some sensitivity to differences in audiences' need can be known for planning. The students were differentiated through a classification scheme based on three factors-friends' drug use, family cohesion and school interest.This study is based on a school survey completed by 4,000 rural students and 2,700 suburban students in New York State during 1970-71. Students in the classifications most involved with drugs were those least receptive to traditional educational programs.

The present paper reports some results from a winter 1974-1975, New York State survey of junior and senior public high school youths' awareness of and attitudes toward the drug prevention efforts their schools had undertaken. The findings indicate that most youths are not aware of community resources available for assistance with a drug problem. Substantial discrepancies were found between student ratings of the effectiveness of prevention programs and the programs which they were most frequently exposed. Finally, the data highlight the peer context and drug culture aspect of the youths' substance taking, emphasizing that the sociocultural backgrounds of students need to be taken into account in prevention program development.