Coefficient Alpha can be used to estimate the reliability of a test when the test is split into several parts. It is known that alpha can severly underestimate test reliability when the parts have an unequal number of items. A generalization of alpha is proposed to correct this defect. (Author/JKS)

In some situations where reliability must be estimated it is impossible to divide the measuring instrument into more than two separately scoreable parts. In such a case, neither Cronbach's coefficient alpha nor Kristof's three-part approach are satisfactory. A technique is developed for estimating reliability in such situations. (Author/BJG)

Finding and interpreting lower bounds for reliability coefficients for tests with non-homogeneous items has been a problem for psychometricians. A computer search procedure is developed for locating such a lower bound in a variety of settings. (Author/JKS)

A model is proposed for the description of ordinal test scores based on the definition of true score as expected rank; its deviations are compared with results from classical test theory. An unbiased estimator of population true score from sample data is calculated. Score variance and population reliability are examined. (Author/BJG)

Finding and interpreting lower bounds for reliability coefficients for tests with nonhomogenous items has been a problem for psychometricians. This paper presents a mathematical formula for finding the greatest lower bound for such a coefficient. (Author/JKS)

A Bayesian Model II approach to the estimation of proportions in m groups is extended to obtain posterior marginal distributions for the proportions. The approach is extended to allow greater use of prior information than previously and the specification of this prior information is discussed. (Author/RC)

An alternative formulation for Guttman scaling is presented. The new formulation is described, and advantages over Guttman's formulation are detailed. The method is assumption-free and capable of multidimensional analysis. (Author/JKS)

Limitations and extensions of Feldt's approach to testing the equality of Cronbach's alpha coefficients in independent and matched samples are discussed. In particular, this approach is used to test equality of intraclass correlation coefficients. (Author)

Methods for estimating the mean and variance of latent ability parameters of a normally distributed population that has been tested with Rasch model-calibrated test items are discussed. Methods for checking the normality of the population are also included. (JKS)

A two-stage testing procedure, a routing test followed by one of several alternative second-stage tests, is studied in the situation where the purpose is measurement, not classification. Models are developed, examined, and compared with conventional tests and up-and-down procedures. (DG)

The problem of determining an optimal passing score for a mastery test is discussed, when the purpose of the test is to predict success on an external criterion. For the case of constant losses for the two possible error types, a method for determining passing scores is derived. (Author/JKS)

An alternative to the uniform probability distribution model for ordinal data is considered. Implications for statistics and for test theory are discussed. (JKS)

Scoring of response vectors to give maximum test-retest correlation is investigated. A general method is given for finding the best scores, deriving them for the normal factor model, and showing that for a standard model for binary response it is easy to approximate the best scores. (SLD)