In multivariate analysis of variance the canonical variates of one effect may be correlated with the canonical variates of another effect. When the two effects are an interaction and a main effect this correlation interferes with the interpretation of the main effect. When the two effects are both main effects the interpretation of the common…

A definition of a suppressor variable is presented which is based on the relation of the semipartial correlation to the zero order correlation. Advantages of the definition are given. (Author/JKS)

Path analytic techniques for assessing the validity of indicators in multitrait-multimethod matrices represent improvements on Campbell and Fiske's early procedures for testing for validity. The diagnosis of indicator ills by means of confirmatory factor analysis is especially useful. An empirical example of these techniques is presented.…

An artifact encountered in regression models of human judgment is explored. The direction and magnitude of the artifactual effect is shown to depend upon the nature of the experimental task and task conditions. Use of an alternative index is recommended. (Author/JKS)

The strategy for investigating convergent and discriminant test validity, known as the multitrait-multimethod matrix, is investigated. A nonparametric significance testing procedure is suggested and demonstrated. (JKS)

An analytic procedure is presented for testing the homogeneity of unstandardized regression weight vectors when the vectors are correlated. The basic design involves repeated measurements on a dependent variable and a set of independent variables. The method is illustrated with a study of perceived leader behavior. (Author/BW)

A strategy for the pairwise assessment which may be used to evaluate the nature of both "prerequisite" and "transference" relations existing among a set of traits is presented. Both confirmatory and exploratory strategies are presented. An example of how the strategy would be used is presented in the exploratory context.…

Many models have been proposed for examining factors from several batteries of tests. A model for such an analysis is presented which allows for maintaining the distinction among batteries. A discussion of the computational procedures is given, and examples are provided. (Author/JKS)

A mathematical link between test reliability and test validity is derived, taking into account the correlation between error scores on a test and error scores on a criterion measure. When this correlation is positive, the "paradoxical" nonmonotonic relation between test reliability and test validity occurs universally. (Author/BW)

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)

A method is proposed for overcoming the problem of multicollinearity in multiple regression equations where multiplicative independent terms are entered. The method is not a ridge regression solution. (JKS)

Three factors that increase score variability yet can be associated with an increase, a decrease, or no change in Pearson's correlation coefficient (r) are discussed (restriction of range, errors of measurement, and linear transformations of data). The connection between changes in variability and r depends on how changes occur. (SLD)

Large sample chi-square tests of the significance of the difference between two correlated kappas, weighted or unweighted, are derived. Cases are presented with one judge in common between the two kappas and no judge in common. An illustrative calculation is included. (Author/SLD)

An algorithm for obtaining initial values for the minimization process in covariance structure analysis is developed that is more generally applicable for computing parameters connected to latent variables than the currently existing ones. The algorithm is formulated in terms of the RAM model but can be extended. (SLD)

Formal relationships are explored among several methods of linearly constrained correspondence analysis. The seven methods considered fall within one of two classes corresponding to ways of imposing linear constraints: the reparameterization method, and the null space method. A connection between these classes is established through a lemma by C.…