Presents a method to improve client classification. Reviews clinical prediction models and statistical prediction issues. In classifying clients for remedial services, utility of the prediction scheme depends on gains accrued by its use over program base rates and relative cost. This assumes that differential treatments parallel diagnostic…

This overview of both quantitative and qualitative methods of educational forecasting is introduced by a discussion of a general typology of forecasting methods. In each of the following sections, discussion follows the same general format: a number of basic approaches are identified (e.g. extrapolation, correlation, systems modelling), and each…

Discusses the relationship between changes in delinquency prevalence over time and prediction error. Data from a 21-year British longitudinal study exemplify the formal relationships that are derived. Results indicate statistically, as prevalence increases, a relative increase in false negative errors and a decrease in false positive errors…

Explores the limitations associated with the expectancy formula approach, using a logical and mathematical analysis. Expectancy tables are predicted on the questionable assumption that achievement follows a straight line growth pattern. The approach is biased in the direction of applying a more stringent underachievement criterion for older,…

Describes the practice of federal evaluation of research impact through three approaches: retrospective methods; qualitative methods, including peer review; and quantitative methods. Recommended areas for study in federal research impact assessment are suggested, including predictive reliability, comparative studies, time and cost estimates,…

Argues that G. A. Cziko's conclusions about the inadequacy of mainstream educational research result from misunderstandings of statistical method, philosophy of science, and other topics. Perceptual control theory, which he espouses, may contribute for educational research but its potential does not result from shortcomings of mainstream…

Three versions of a probabilistic model adapted from the theory of information retrieval--a binary version, a version using the full value of the data, and a version using principal components--were tested and applied to data available from application forms to predict graduate school performance of library school students. (EM)

Alternative mathematical models for scoring and decision making with criterion referenced tests are described, especially as they concern appropriate test length and methods of establishing statistically valid cutting scores. Several of these approaches are reviewed and compared on formal-analytic and empirical grounds: (1) Block's approach to…

Discusses the adoption of computer technology at universities and examines reasons why some professors don't use computers. Topics discussed include computer applications, including artificial intelligence, social science research, statistical analysis, and cooperative research; appropriateness of the technology for the task; the Computer Aptitude…