This study reports results from an evaluation of the experimental Rio Hondo driving under the influence (DUI) court of Los Angeles County, California. Interviews and official record checks with 284 research participants who were randomly assigned to a DUI court or a traditional criminal court were assessed at baseline and at 24-month follow-up.…

Evaluation Research Society members (n=160) assessed the attributes of research that produce credibility. The study focused on the acceptability of no-difference findings. Randomization, statistical power, number of measures reported, and equivalent groups were tested to determine their influence on acceptability of both no-difference and…

Estimates of the Comprehensive Employment and Training Act's net impact on participant earnings, using Continuous Longitudinal Manpower Survey data, were compared to a similar sample from the Current Population Survey. The use of multivariate matching and weighting yielded acceptable results. (GDC)

Self-generated codes to insure student anonymity were used in a study of substance use among 2500 adolescents. Matching rates for various cohorts as a proportion of total respondents at different survey periods were less than ideal. The effects of coding and other attrition factors on the tracking are discussed. (JAZ)

Evidence is presented that strengthens the argument against comparison group research strategies. Results from a study of youth attitudes related to the California Conservation Corps involving 528 participants and 668 comparisons indicate that the comparison group may not adequately represent the nontreatment state. (SLD)