The effect of scale selection on error in calibrating item and ability parameters was investigated, with particular reference to the standardized mean-squared difference (SMSD) statistic. Through simulation, three scaling methods for selecting the common scale were used to demonstrate their effects on SMSD values. (SLD)

Detailed reanalyses of data reported in Studies in Educational Evaluation: Monograph No. 1 by Y. Kashti and Monograph No. 5 by U. Kattmann, 1979, were performed using an explicitly structurally oriented approach via target analysis (PINDIS). Results contradict those reached by Kashti and Kattmann. (RL)

Uncorrectable skew and heteroscedasticity are among the "lemons" of psychological data, yet many important variables naturally exhibit these properties. For scales with a lower and upper bound, a suitable candidate for models is the beta distribution, which is very flexible and models skew quite well. The authors present…

A two-step procedure is developed for the analysis of pretest-posttest data. Nonlinear canonical correlation analysis is followed by covariance analysis of optimally scaled pretest and posttest measures. The procedure is available in the computer program CANALS. Advantages of the two-step procedure are discussed. (SLD)

A latent-class scaling approach is presented for modeling paired comparison and "pick any/t" data obtained in preference studies. The utility of this approach is demonstrated through analysis of data from studies involving consumer preference and preference for political candidates. (SLD)

Three approaches to the graphic representation of similarity and dissimilarity matrices are compared and contrasted. Specifically, Kruskal's multidimensional scaling, Johnson's hierarchical clustering, and Waern's graphing techniques are employed to depict, in two dimensions, data representing the structure of a set of botanical concepts. Each of…

Multidimensional scaling (MDS) and exploratory and confirmatory factor analyses were compared in the assessment of the dimensionality of data sets, using sets generated to be one-dimensional or two-dimensional and differing in degree of interdimensional correlation and number of items defining a dimension. (SLD)

The algorithm developed by B. A. Schneider (1980) for analysis of paired comparisons of psychological intervals is replaced by one proposed by R. M. Johnson. Monte Carlo simulations of pairwise dissimilarities and pairwise conjoint effects show that Johnson's algorithm can provide good metric recovery. (SLD)

Principal component, image component, three types of factor score estimates, and one scale score method were compared over different levels of variables, saturations, sample sizes, variable to component ratios, and pattern rotations. There were virtually no overall differences among methods, with the average correlation between matched scores…

Given "T" joint observations on "K" variables, it is frequently useful to consider the weighted average or scaled score. L-scaling is introduced as a technique for determining the weights. The technique is so named because of its resemblance to the Leontief matrix of mathematical economics. L-scaling is compared to two…

Many multidimensional item response theory (IRT) models have been proposed. A comparison is made between the so-called full information models and the models that use only pairwise information. Three multidimensional models described are: (1) the compensatory model of R. D. Bock and M. Aitken (1981) using the computer program TESTFACT; (2) a model…

The question of whether test factor structure is indicative of the test item hierarchy was examined. Data from 1,000 subjects on two sets of five bivalued Law School Admission Test items, which were analyzed with latent trait methods of Bock and Lieberman and of Christoffersson in Psychometrika, were analyzed with an ordering-theoretic method to…

Two sets of real sorting data from 50 college students are used to compare results of multidimensional scaling of raw co-occurrence frequencies or dissimilarity measures (D) and profile distances (delta) to determine which yields a better representation of the underlying structure of 2 sets of stimuli. Slight differences are discussed. (SLD)

The following seven statistical procedures are compared in terms of the ability to recover a unidimensional latent trait from Likert-type data: (1) factor analysis based on Pearson correlations (FA-PR); (2) factor analysis based on polychoric correlations (FA-PL); (3) the graded response model in item response theory (IRT-GRM); (4) internal…

Equatings were performed on both simulated and real data sets using common-examinee design and two abilities for each examinee. Results indicate that effective equating, as measured by comparability of true scores, is possible with the techniques used in this study. However, the stability of the ability estimates proved unsatisfactory. (TJH)