A generalization of the Rasch model is employed to quantify statements on a scale in the Thurstone tradition, as well as to measure a person's attitudes in the Likert tradition. An illustration of the technique is provided. (JKS)

The use of the standard deviation of item ratings as a criterion of scale anchor selection is criticized. An alternative criterion based on item intercorrelation procedures is suggested and tested. Results indicate that the alternative criterion does not lead to the negative outcomes of the standard deviation criterion. (Author/JKS)

An analysis of variance model is recommended for data which display a peaked trend. These data are particularly common in psychological studies of stimulus generalization. A variety of hypotheses are suggested, and the testing of these hypotheses is illustrated. (Author/JKS)

Two types of quantitative Rorschach scales are discussed: first, those based on the response categories of content, location, and the determinants, and second, global scales based on the subject's responses to all ten stimulus cards. (Author/JKS)

Hoyt's analysis of variance procedure for estimating reliability assumes that the residual mean square estimates error variability. If, however, an individual's true score varies across items, it is argued that residual mean square estimates two components--error and interaction--and hence Winer's modification of Hoyt's formula, understimates the…

Results of discriminant analysis are of restricted value in the counseling setting because discriminant functions often lack psychological interpretability. A technique is proposed for obtaining rotated discriminant function scores which should have the same clarity of meaning as rotated factors from factor analyses. (Author/JKS)

Computerized answer sheets in mail surveys are examined for their effects on rate of return and response bias. Results of an empirical study of job satisfaction suggested that computerized answer sheets may be used in mail surveys without significantly affecting rate of return or producing response bias. (Author/JKS)

Attitude changes in undergraduates enrolled in a foundations of education course assigned to two different types of reading material--controversial and persuasive, and traditional--were examined. Lasting attitudinal changes were found in students under the first condition. (Author/JKS)

Past practice of operationalizing the concept of fakeability of psychological tests is reviewed. The strengths and weaknesses of these indices are discussed in the light of a proposed new definition of fakeability based upon Naylor's model of measurement accuracy. (Author/JKS)

A FORTRAN program for clustering variables using the alpha coefficient of reliability is described. For batch operation, a rule for stopping the agglomerative precedure is available. The conversational version of the program allows the user to intervene in the process in order to test the final solution for sensitivity to changes. (Author/JKS)

A FORTRAN computer program was derived for an IBM series 360/370 computer system that provides a factor analytic solution for large three-dimensional data matrices. The computational procedures employed are based upon those presented in Method III by Tucker. (Author/JKS)

TMFA, a FORTRAN program for three-mode factor analysis and individual-differences multidimensional scaling, is described. Program features include a variety of input options, extensive preprocessing of input data, and several alternative methods of analysis. (Author)

Two computer programs (one batch, one interactive) are designed to provide statistics for a weighted linear combination of several component variables. Both programs provide mean, variance, standard deviation, and a validity coefficient. (Author/JKS)

Three algorithms for selecting a subset of originally available items, to maximize coefficient alpha, were compared on the size of the resulting alpha and computation time required with nine sets of data. The characteristics of a computer program to perform these item analyses are described. (Author/JKS)

This program computes specific category agreement levels for both nominally and ordinally scaled data. For ordinally scaled data, an option is available for collapsing the original scale to a smaller number of categories, with the goal of improving the level of interrater reliability for the rating scale. (Author)