Indefinite symmetric matrices that are estimates of positive-definite population matrices occur in a variety of contexts such as correlation matrices computed from pairwise present missing data and multinormal based methods for discretized variables. This note describes a methodology for scaling selected off-diagonal rows and columns of such a…

A scaled difference test statistic T[tilde][subscript d] that can be computed from standard software of structural equation models (SEM) by hand calculations was proposed in Satorra and Bentler (Psychometrika 66:507-514, 2001). The statistic T[tilde][subscript d] is asymptotically equivalent to the scaled difference test statistic T[bar][subscript…

Scalability coefficients play an important role in Mokken scale analysis. For a set of items, scalability coefficients have been defined for each pair of items, for each individual item, and for the entire scale. Hypothesis testing with respect to these scalability coefficients has not been fully developed. This study introduces marginal modelling…

Using Lumsden's Thurstonian fluctuation model as a starting point, this paper attempts to develop a unidimensional item response theory model intended for binary personality items. Under some additional assumptions, a new model is obtained in which the item characteristic curves are defined by a cumulative Pearson-Type-VII distribution, and the…

Under conditions commonly met in optimal scaling problems, arbitrary sets of optimal weights can be obtained by choices of generalized universe scores. It is suggested that the invariant parameters of optimal scaling should be interpreted according to latent trait theory, rather than the arbitrary weights. (Author/JKS)

Restricted multidimensional scaling models, allowing constraints on parameters, are extended to the case of asymmetric data. Examples of several models are provided, using journal citation data. Possible extensions of the models are considered. (Author/JKS)

A review of existing techniques for the analysis of three-way data revealed that none were appropriate to the wide variety of data usually encountered in psychological research, and few were capable of both isolating common information and systematically describing individual differences. Such a model is developed and evaluated. (Author/JKS)

A new method to estimate the parameters of Tucker's three mode principal component model is discussed, and the convergence properties of the alternating least squares algorithm to solve the estimation problem are considered. An example is presented. (Author/JKS)

Ideal point discriminant analysis (IPDA) is proposed for the analysis of contingency tables of cross-classified data. Several data sets illustrate IPDA, which combines log-linear and dual scaling models to provide a spatial representation of row and column categories and allow statistical evaluation of various structural hypotheses about…

Current psychometric models of choice behavior are strongly influenced by Thurstone's (1927, 1931) experimental and statistical work on measuring and scaling preferences. Aided by advances in computational techniques, choice models can now accommodate a wide range of different data types and sources of preference variability among respondents…

Several methods for quantifying categorical multivariate data (including multiple correspondence analysis and optimal scaling) are discussed and shown to be equivalent. The method can be applied by variable matrix with categorical variables, by item matrix of multiple-choice data, or by a multi-way contingency table. (NSF)

The multivariate stochastic processes associated with the Marshall-Olkin multivariate exponential distribution are shown to be able to generate several models of similarity or preference data in the literature. (JKS)

A nonmetric coordinate adjustment technique determining scale values for objects whose interobject intervals (differences in subjective value) have been directly compared is developed. In Monte Carlo simulations, the degree of metric determinancy of the scale values appears quite high even when the amount of error is relatively high. (Author/JKS)

The least squares loss function of cluster differences scaling, originally defined only on residuals of pairs allocated to different clusters, is extended with a loss component for pairs allocated to the same cluster. Findings show that this makes the method equivalent to multidimensional scaling with cluster constraints on the coordinates. (SLD)

The ability of three estimation criteria to recover parameters of the Thurstone Case V and Case III models from comparative judgment data was investigated via Monte Carlo techniques. Significant differences in recovery are shown to exist. (Author/JKS)