Glass's estimator of effect size, the sample mean difference divided by the sample standard deviation, is studied in the context of an explicit statistical model. The exact distribution of Glass's estimator is obtained and the estimator is shown to have a small sample bias. Alternatives are proposed and discussed. (Author/JKS)

Assuming conditional normality and independence, standard estimation and hypothesis testing procedures for regression coefficients are the same whether using fixed x values, random x values or mixed fixed and random values. Contrary to Sockloff's suggestion, polynomial regression is useful with either fixed or random models. (See EJ 142 050). (CP)

Tests the "holistic construal" method of validating constructs and testing organizational hypotheses, using examples from organizational theory and data on wholesale distribution companies. Holistic construal is meant to explicitly represent theoretical and empirical concepts, nonobservational hypotheses, and correspondence rules and…