Bayesian analysis of an m-group model is considered. A convenient stage III prior is proposed, and cases when the posterior distributions take on a simple form are exhibited. The behavior of various point estimators of the linear parameters of the model are explored in a Monte Carlo study. In the simple model considered, the O'Hagan estimator…

The inferences allowable with a significant F in regression analysis are discussed. Included in this discussion are the effects of specificity of the research hypothesis, incorporation of covariates, directional hypotheses, and the manipulation of variables on the interpretation of significance for such purposes as causal and directional…

Following Conger's revised definition of suppressor variables, the universe relationships among two predictors and a criterion is analyzed. A simple mapping of relationships, based on the correlation between two predictors and the ratio of their validities, is provided. The relation between suppressor and part correlation is also discussed.…

The use of multiple regression analysis to detect sex-related salary discrimination is investigated via a simulation study. (JKS)

A unified approach, based on ranks, to the statistical analysis of data arising from complex experimental designs is presented. The rank methods closely parallel the familiar methods of least squares, so that the estimates and tests have natural interpretations. (Author/JKS)

Interpretation of multivariate models requires knowing how much the fit of the model is impaired by changes in the parameters. The relation of parameter change to loss of goodness of fit can be called parameter sensitivity. Formulas are presented for assessing the sensitivity of multiple regression and principal component weights. (Author/JKS)

Three methods that can be used subsequent to a regression analysis to determine the relative effectiveness of schools are Dyer's performance indicators, Scheffe's hyperbolic confidence bands, and Gafarian's linear confidence bands. The purpose of this paper is to investigate the relative usefulness of the three methods under various conditions.…

Multiple cutoff scores are recommended as an alternative to multiple regression for predicting future performance of individuals from educational tests. Tables of the type proposed by Taylor and Russell are provided for evaluating the effects of using this procedure. (Author/JKS)

Multiple regression techniques are a potent tool in conducting student affairs research. Several considerations in employing these methods are reviewed and evaluated. (Author)

An analysis of variance procedure for testing differences in r-squared, the coefficient of determination, across independent samples is proposed and briefly discussed. The principal advantage of the procedure is to minimize Type I error for follow-up tests of pairwise differences. (Author/JKS)

In order to make the parallel analysis criterion for determining the number of factors in factor analysis easy to use, regression equations for predicting the logarithms of the latent roots of random correlation matrices, with squared multiple correlations on the diagonal, are presented. (Author/JKS)

A numerical procedure is outlined for obtaining an interval estimate of the regression of true score or observed score, utilizing only the frequency distribution of observed scores. The procedure assumes that the conditional distribution of observed scores for fixed true scores is binomial. Several illustrations are given. (Author/HG)

The important and sometimes difficult-to-grasp concept of regression suppressor variable effects is explored. An inquiry into the phenomenon of suppressor effects is accomplished via a synthesis of the existing literature and the use of a small heuristic data set to improve the accessibility of the concept. Implications for researchers are also…

Multiple linear regression is used to model the effects of violating statistical assumptions on the likelihood of making a Type I error. This procedure is illustrated for the student's t-test (for independent groups) using data from previous Monte Carlo studies in which the actual alpha levels associated with violations of the normality…