The accuracy of normal approximations to the binomial test was evaluated with and without a continuity correction, regarding control of Type I errors and power. Both tests exhibited substantial power loss in comparison to the exact binomial test, although they are easier to apply and are sometimes desirable. (SLD)

Three potential models for the 2 X 2 contingency table are discussed: (1) the hypergeometric Independence Trial; (2) the double-binomial Comparative Trial; and (3) the multinomial Double Dichotomy Trial. Subsequently, the critical regions of seven statistical tests are evaluated within each of these models. (TJH)

A least squares statistical procedure for fitting utility functions is extended to truncated normal and extended beta functions. Implications for educational decision-making are discussed. (JKS)

The purpose of this paper is to calculate the upper limit of the correlation between normal and dichotomous variables. An empirically obtained correlation should be evaluated in view of this limit, instead of the usual limit of Pearson correlation. (Author)

The use of a quasi-equiprobability model in the measurement of observer agreement involving dichotomous coding categories is described. A measure of agreement is presented which gives the probability of agreement under the assumption that observation pairs reflecting disagreement will be equally probable. (Author/JKS)

Data evolving from processes which are developmental or hierarchical in nature are often analyzed by using latent class or latent structure models. A procedure for estimating such models when the model is not "identifiable" is presented. (JKS)

The relationship between quantitative predictor variables and the probability of occurrence of one or more levels of a qualitative criterion variable can be analyzed by quantal response techniques. This paper presents and discusses two quantal response models, comparing them to multiple linear regression and discriminant analysis. (Author/JKS)

A probablistic approach to classification and diagnosis of erroneous rules of operations that result from misconceptions ("bugs") in a procedural domain of arithmetic is introcuced. Variability of response errors is explicitly treated through item response theory. As a concrete example, a signed-number subtraction dataset is analyzed.…

An Edgeworth approximation for accurate significance probabilities for the Wilcoxon two-sample test is substantially simplified. A method is developed that allows quick calculations of very accurate probabilities. Exact formulas are given for most of the remaining cases, and tables are presented comparing the new simplification to likely…

Two stage multiple-comparison procedures give an exact solution to problems of power and Type I errors, but require equal sample sizes in the first stage. This paper suggests a method of evaluating the experimentwise Type I error probability when the first stage has unequal sample sizes. (Author/BW)

The problem of determining which of several populations has the largest mean is considered. The procedure described by Dudewicz and Dalal is extended to the case of unequal sample sizes. (JKS)

A Bayesian solution is suggested to the problem of jointly estimating "k is greater than 1" binomial parameters in conjunction with the problem of testing, in a Bayesian sense, the hypothesis "H" of parametric homogeneity. Applications of the estimates are illustrated with several types of data, including ophthalmological…

Statistical methods are presented for studying "correlates of diversity," defined as characteristics of educational organizations that predict dispersion on the dependent variable. Strategies based on exact distribution theory and asymptotic normal approximation are considered. (TJH)

The well-known problem of cumulating error probabilities is reconsidered from a general epistemological perspective, namely, the concepts of severity and of fairness of tests. It is shown that not only Type 1 but also Type 2 errors can cumulate. A new adjustment strategy is proposed and applied. (Author/JAZ)

The observed probability "p" is the social scientist's primary tool for evaluating the outcome of statistical hypothesis tests. The small-sample accuracy of nonnull asymptotic distributions of several functions of "p" was studied. Implications for use of the approximations are discussed. (SLD)