Methodological issues of using a class of neural networks called Mixture Density Networks (MDN) for discriminant analysis are discussed. MDN models have the advantage of having a rigorous probabilistic interpretation, and they have proven to be a viable alternative as a classification procedure in discrete domains. Both classification and…

This paper presents a detailed description of maximum parameter estimation for item response models using the general EM algorithm. In this paper the models are specified using a univariate discrete latent ability variable. When the latent ability variable is discrete the distribution of the observed item responses is a finite mixture, and the EM…

The use of the 2 x 2 matrix in qualitative data analysis and theory generation is discussed, embracing the perspective that the objective of qualitative research in general and the analysis of qualitative data in particular is the development of theory. A 2 x 2 matrix is considered to be a tabular representation of the relationship of two…

This theme issue features five articles profiling Indiana University faculty whose work on various campuses continues to broaden and advance knowledge about cognitive science. The articles in the journal are: "A Matter of Time" (Karen Grooms) which discusses the work of Robert F. Port; "Perceiving as a Complex System" (Tom…

Recent empirical studies of actual research practice demonstrate that general linear models, such as regression, have become increasingly popular. The present paper explores issues related to the interpretation of regression results, with illustrations drawn from studies in the counseling psychology literature. The primary focus of this paper is…

This paper provides an introduction to basic issues concerning structural equation modeling (SEM), a research methodology increasingly being used in social science research. First, seven key issues that must be considered in any SEM analysis are explained. These include matrix of associations to analyze, model identification, parameter estimation…

Different ways that dependent variables can cause multivariate significance are illustrated. It is argued that univariate tests should not be used as follow-up procedures to identify the variables that can contribute to multivariate significance. Methods that can be used to determine the contributions individual variables make to multivariate…

This paper discusses the issue of multiple testing and overall Type I error rates in contexts other than multiple comparisons of means. It demonstrates, using a 5 x 5 correlation matrix, the application of 5 recently developed modified Bonferroni procedures developed by the following authors: (1) Y. Hochberg (1988); (2) B. S. Holland and M. D.…

In many large-scale educational assessments it is of interest to compare the distribution of latent abilities of different subpopulations, and track these distributions over time to monitor educational progress. B. Junker, together with two colleagues, has developed a simple scheme, based on the proportion correct score, for smoothly approximating…

Multivariate methods are being used with increasing frequency in educational research because these methods control "experimentwise" error rate inflation, and because the methods best honor the nature of the reality to which the researcher wishes to generalize. This paper: explains the basic logic of canonical analysis; illustrates that…

Model justified the expectation that the transfer from learning one task to another is linearly dependent on the coefficient of intercorrelation between the two studies when the coefficient was corrected for attenuation. Supportive studies are cited. Regression curves yielded negative intercepts. Testable explanations were offered. (Author/BJG)

The fundamental results of integration theory are shown to follow from an axiomatic definition of the integral as a real valued function whose domain is a subset of the set of real valued functions of a real variable. (SD)

Geometric analysis of a cube can motivate formulae for factoring algebraic expressions involving sums and differences of cubes. (SD)

A network model is developed to determine the optimum number of shelf heights to use which will minimize the shelving cost for a given collection of books. (Author)