Explored are two theoretical approaches that attempt to cope with omitted responses, that is, when an examinee omits (fails to respond to) an item and therefore the item response formula cannot be used. Preliminary considerations are discussed, and it is shown that a conveniently simple application of equivalent items leads to internal…

A mathematical setting based on a statistical sampling probability mechanism is described. In this setting, a mathematical meaning is given for the information function; and it becomes possible to study the relative merits of various ability-estimating procedures. The maximum likelihood estimation procedure under the one-, two-, three-parameter…

There are currently three main approaches to parameter estimation in item response theory (IRT): (1) joint maximum likelihood, exemplified by LOGIST, yielding maximum likelihood estimates; (2) marginal maximum likelihood, exemplified by BILOG, yielding maximum likelihood estimates of item parameters (ability parameters can be estimated…

A new method of multiple discriminant analysis allows a mixture of continuous and discrete predictors. It handles conditional, joint, or separate sampling. Subjects and criterion groups are represented as points in a multidimensional Euclidean space. Advantages of the method, deriving from Akaike Information Criterion model evaluation, are…

Maximum likelihood procedures are presented for a general model to unify the various models and techniques that have been proposed for item analysis. Unconditional maximum likelihood estimation, proposed by Wright and Haberman, and conditional maximum likelihood estimation, proposed by Rasch and Andersen, are shown as important special cases. (JAZ)

This article describes a Bayesian framework for estimation in item response models, with two-stage distributions on both item and examinee populations. Strategies for point and interval estimation are discussed, and a general procedure based on the EM algorithm is presented. (Author/LMO)

A model for longitudinal latent structure analysis was proposed that combined the values of a latent variable at two time points in a two-dimensional latent density. The correlation coefficient between the two values of the latent variable can then be estimated. (NSF)

The Dutch Identity is presented as a useful tool for expressing the basic equations of item response models that relate the manifest probabilities to the item response functions and the latent trait distribution. Ways in which the identity may be exploited are suggested and illustrated. (SLD)

Recently, C. E. McCulloch (1987) suggested a modification of the Morgan-Pitman test for comparing the variances of two dependent groups. This paper demonstrates that there are situations where the procedure is not robust. A subsample approach, similar to the Box-Scheffe test, and the Sandvik-Olsson procedure are also assessed. (TJH)

The robustness of the likelihood ratio difference statistic to the violation of a regularity condition when used to assess differences in fit provided by pairs of latent class models was investigated. Recommendations are made regarding the use of the statistic under violation of the regularity condition. (SLD)

Testing the fit of the Rasch model is examined. Tests proposed are based on the comparison of expected and observed frequencies. Conditional maximum likelihood estimates (MLEs) and marginal MLEs are compared. A statistical testing procedure is proposed that is a diagnostic tool for identifying violations of the Rasch model. (SLD)

Describes algorithms used in the computer program LOGIMO for obtaining maximum likelihood estimates of the parameters in loglinear models. These algorithms are also useful for the analysis of loglinear item-response theory models. Presents modified versions of the iterative proportional fitting and Newton-Raphson algorithms. Simulated data…

An empirical Bayes-maximum likelihood estimation procedure is presented for the application of fuzzy partition models in describing high dimensional discrete response data. The model describes individuals in terms of partial membership in multiple latent categories that represent bounded discrete spaces. (SLD)

Two joint maximum likelihood estimation methods (LOGIST 2B and LOGIST 5) and two marginal maximum likelihood estimation methods (BILOG and ForScore) were contrasted by measuring the difference between a simulation model and a model obtained by applying an estimation method to simulation data. Marginal estimation was generally superior. (SLD)

Based on the usual factor analysis model, this paper investigates the relationship between improper solutions and the number of factors. The properties of the noniterative estimation method of M. Ihara and Y. Kano in exploratory factor analysis are also discussed. The estimators were compared in a Monte Carlo experiment. (TJH)