Four models of possible predictors are advanced and tested using data collected from 1,164 faculty in 80 university departments and from published sources. Results indicated that there is no single set of factors that can reliably predict improvement or decline in prestige across all disciplines. (Author/LBH)

Explains multiple regression, demonstrates its flexibility for analyzing data from various designs, and discusses interpretation of results from multiple regression analysis. Presents regression equations for single independent variable and for two or more independent variables, followed by a discussion of coefficients related to these. Compares…

The common approach to studies of predictive bias is analyzed within the context of a conceptual model in which predictors and criterion measures are viewed as fallible indicators of idealized qualifications. (Author/PN)

The purpose of this study was to compare five predictor models, including two least-square procedures, two probability weighting (semi-Bayesian) methods, and a Bayesian model developed by Lindley. (See also TM 501 088, TM 501 089, and TM 501 090) (Author/NE)

Four commonly employed procedures were repeatedly applied to computer-simulated samples to provide comparative data pertaining to two questions: (a) which procedure can be expected to produce and equation that yields the most accurate predictions for the population, and (b) which procedure is most likely to identify the optimal set of independent…

Discusses enrollment projection in terms of the general logic of statistical prediction. (Author)

A meta-analysis of outcomes from 32 studies investigating pretest effects was conducted. For all outcomes the average effect size was +.22, indicating an elevating effect of pretest on posttest. Duration of time between pre- and posttesting was also related to effect size. Researchers should continue to include pretest as a design variable.…

Depending on the nature of the predictor variable in the expectancy table, expectancy ranges may be computed by incorporating the standard error of estimate. The process of developing ranges is illustrated, and steps to be used in evaluating the quality and utility of expectancy data are outlined. (Author)

A critical reanalysis of Barrett, Caldwell, and Alexander's (1985) critique of dynamic criteria. Summarizes and questions Barrett, et al.'s three definitions of dynamic criteria and their conclusion that reported temporal changes in criteria could be explained by methodological artifacts. A greater focus on dynamic criteria as constructs is…

Responds to Austin, Humphreys, and Hulin's (1989) critique of Barrett, Caldwell, and Alexander, suggesting that the burden of proof still rests on the advocates of the concept of dynamic criteria, and that empirical support is lacking for the existence of dynamic criteria as a simplex. Contrary evidence from educational, organizations, and…

Seven methods for measuring and comparing value added in higher education are considered; six are index methods (all found unsatisfactory for lack of empirical support) and a seventh is proposed which calculates value added as the difference between the outcomes achieved and those predicted from national (United Kingdom) data. The advantages of a…

Examines difficulties with citation analysis as it is used to study citation frequency, usually for the evaluation of scientists, publications, or institutions. Topics addressed include selection of a control set of papers, comparisons of different types of papers (e.g., methodological or theoretical), effects of independent variables, and use of…

Nonlinear relationships and latent variable assumptions can lead to serious specification errors in structural models. A quadratic relationship, described by a linear structural model with a latent variable, is shown to have less predictive validity than a simple manifest variable regression model. Advocates the use of simpler preliminary…

Application of the statistical techniques of survival analysis and credit scoring to private education loans extended to law students found a pronounced seasoning effect for such loans and the robust predictive power of credit bureau scoring of borrowers. Other predictors of default included school-of-attendance, school's geographic location, and…