A controlled study of the impact of a juvenile education program on the recidivism rates of juveniles was performed. The program involved introducing the juveniles to prison life. Findings showed that youths categorized as more delinquent were affected differently by program attendance compared to youths categorized as less delinquent. (Author/LMO)

An expectancy-value procedure is developed for delineating the nature of interagency relationships during the operation of a youth-serving agency. Advantages of this approach are presented. (Author/GK)

This article argues that evaluators must go further in mapping how clients' indigenous social structures complement or conflict with the centralized rational structure of program reforms. Evaluation of youth programs in Israel and the United States exemplifies how the respective social rules of program sponsor versus clients may be mutually…

Ways in which multilevel modeling techniques can broaden the kinds of questions asked in multisite evaluations are illustrated through the example of the Transition Mathematics Program for secondary schools. These techniques provide a means of moving beyond estimating effects to investigating how differences across sites relate to differences in…

The Teaching Family Model (TFM) of delinquency treatment has evolved into a national network of community-based behaviorally oriented group homes. A longitudinal summative evaluation of TFM homes and comparison groups from the same or neighboring communities was mounted, and the results of a five year cost-effectiveness study are presented.…

An iterative ad hoc analysis of a variety of existing data bases was undertaken to estimate probable teacher shortages in math and science. This article describes the full set of compromises needed in order to use the data and the conclusions drawn. (Author/BW)

An outcomes-based model was developed to evaluate vocational education programs in Illinois. The model, which evaluates programs based on six vital signs of program quality, uses a series of computer technology expert systems to give local personnel assistance in exploring program problems and improvement. (SLD)

High intercommunity variation in adolescent smoking precluded detection of effects of mass media campaigns on initiation of cigarette smoking in a study with a final sample of 1,637 adolescent subjects. Implications for studies using the individual as unit of analysis and the community as unit of randomization are discussed. (SLD)