The "Program Evaluation Standards" supplies a useful framework for generating questions to raise about any evaluation plan or evaluation report to assess its pros and cons. It is a valuable "how-to" for graduate students and professionals. This second edition incorporates changes in the field in the last decade. (SLD)

Two studies focused on aspects of the rating process in performance assessment. The first, which involved 15 raters and about 400 medical students, made the "committee" facet of raters working in groups explicit, and the second, which involved about 200 medical students and four raters, made the "rating-occasion" facet…

Multiple regression analysis shows that both a response-production component and an evaluation component are involved in answers to a free-response synonym task by 299 Belgian college students. Format differences between the multiple choice evaluation task and the synonym task are explained in terms of verbal abilities measured. (SLD)

Two procedures are presented for detecting violations of the unidimensionality assumption made by latent trait models without requiring factor analysis of inter-item correlation matrices. Both procedures require that departures from unidimensionality be hypothesized beforehand. This is usually possible in achievement tests where several content…

Six undergraduate and three graduate classes (300 students) were given multiple-choice tests twice with subsequent evaluation of answer changes. Results indicate that, although instruction leads to a change in attitude in answer changing, the number of changes and overall gain due to changing do not change. (SLD)

College professors and elementary and secondary teachers were compared relative to their perspectives on preservice educational measurement courses. In five of eight content areas on a questionnaire, the relative emphases given by professors differed from teachers in the areas of nontest evaluation, statistical analysis, and formative and…

Six procedures for combining sets of item response theory (IRT) item parameter estimates from different samples were evaluated using real and simulated response data. Results support use of covariance matrix-weighted averaging and a procedure using sample-size-weighted averaging of estimated item characteristic curves at the center of the ability…

Three variations of a least squares regression model are presented that are suitable for determining and correcting for rating error in designs in which examinees are evaluated by a subset of possible raters. Models are applied to ratings from 4 administrations of a medical certification examination in which 40 raters and approximately 115…