Construct validity patterns of test-criteria and item-criteria correlations are shown to be inconsistent across samples. The results of an investigation of construct validity patterns on two published personality scales is presented. (JKS)

The effects of nonnormally distributed perceived distance between pairs of objects on the accuracy of recovered configurations of traditional multidimensional scaling were studied via simulation methods. Multidimensional scaling is shown to be robust with respect to violations of the assumption of normality. (Author/JKS)

An exact probability test for use with certain Likert-type scales is presented. The procedure assumes equally-spaced points, independence of subjects' responses, and that each point has an equal likelihood of response for each subject. Tables for critical values are presented. (Author/JKS)

Two models for scaling of paired comparison data are compared to the Thurstone case III model. Two goodness of fit indices are presented for each model for five data sets. The results illustrate the inability of the Thurstone model to adequately account for data when the scale includes extreme stimuli. (Author)

A procedure is specified for testing the significance of predicted trends in k independent correlations. An example is also provided for illustrative purposes. (Author)

A technique for selecting alpha based on the particular circumstances of a hypothesis testing problem is illustrated. In selecting an alpha, the technique takes into account the prior probability that the null hypothesis is true and the penalties associated with Type I and Type II errors. (Author)

The Ross Educational Philosophical Inventory was psychometrically examined through the use of expert judges and factor analysis of a sample of responses to the inventory. Results are negative in general, but moderately positive results were found in relation to two of the four categories of the inventory. (JKS)

Profile stability of the Kuder Occupational Interest Survey was assessed for profiles obtained twelve and nineteen years apart for persons between thirteen and twenty years of age at the time of first administration. Reported reliabilities ranged from .40 to .80 for various sub-samples. (JKS)

Student points of view in their ratings of courses and instructors were analyzed within three specific college courses and then generalized by analysis of a sample of responses from over 400 classes in five colleges. A factor analysis yielding three different student points of view is presented. (Author/JKS)

A simple and economical method for estimating initial parameter values for the normal ogive or logistic latent trait mental test model is outlined. The accuracy of the method in comparison with maximum likelihood estimation is investigated through the use of Monte-Carlo data. (Author)

The theoretical background underlying the mathematical development of different measures of the degree of attitudinal polarization applicable to bipolar questions and a Fortran IV computer program that provides a rapid means for calculation of four of those measures are presented. (Author/JKS)

A Fortran IV computer program is presented which will unambiguously partition the explained variance of a dependent variable into those parts due uniquely to each independent variable and to all possible combinations of independent variables through commonality analysis. Tests of significance and documentation are provided. (Author/JKS)

Description of a computer program performing rotated canonical variate analysis was presented. The program is compatible with the IBM Scientific Subroutines Package. Examples of rotated and unrotated solutions of a sample problem were given, together with their interpretations illustrating the advantages of the rotated solution. (Author)

A Fortran IV program which computes either of the coefficients of congruence (psi or phi) used in comparison of factors in factor analysis is presented. Output consists of a non-symmetric matrix of factor coefficients. Listings of the program, results and test data are available. (Author/JKS)

A Fortran program is presented for discriminant analysis of discrete variables. The program assumes discrete, nominal data with no distributional, variance-covariance assumptions. The program handles a maximum of fifty predictor variables and twelve outcome groups. (Author/JKS)