Statistical procedures based on Bayesian estimation for obtaining estimates of a propensity (which would include estimates of proportions or relative frequencies) are described for the special case where the observer can only note whether the propensity exceeds or does not exceed a constant between 0 and 1. (JKS)

This study investigated statistical methods for identifying errors in Bayesian networks (BN) with latent variables, as found in intelligent cognitive assessments. BN, commonly used in artificial intelligence systems, are promising mechanisms for scoring constructed-response examinations. The success of an intelligent assessment or tutoring system…

This digest introduces ways of responding to the call for criterion-referenced information using Bayes' Theorem, a method that was coupled with criterion-referenced testing in the early 1970s (see R. Hambleton and M. Novick, 1973). To illustrate Bayes' Theorem, an example is given in which the goal is to classify an examinee as being a master or…

It is not unusual for several tests measuring different abilities to be given in one test administration. A common practice is to estimate these abilities independently of each other, but this paper proposes a more efficient method of estimating these abilities that takes into account the correlational structure of the abilities. The method uses a…

The utility of least squares multiple regression in predicting new scores from previously established equations is considered. It is shown that in the absence of useful prior information, and when normality assumptions are not violated, least squares multiple regression weights are superior to alternatives recently presented in the literature.…

Modern statistics are considered as a branch of psychometrics and the question of how the central problems of statistics can be resolved using psychometric methods is investigated. Methods developed in the fields of test theory, scaling, and factor analysis are related to principle problems of modern statistical theory and method. (Author/JKS)

The use of Bayes' estimates for proportions in the Law of Comparative Judgment is suggested to avoid sample proportions of zero and one. (Author)

Using a new coefficient, a rescaling of the Bayes risk is examined and a modification of this coefficient is described which yields an index that always has a value between zero and one. (Author/MH)

A simple example is presented that illustrates advantages of Bayesian and likelihood methods of inference relative to sampling distribution methods of inference. It is argued that Bayesian and likelihood methods of inference should be utilized more generally to analyze real data. (Author/CTM)

Proposed a new nonlinear structural equation model with fixed covariates to deal with some complicated substantive theory and developed a Bayesian path sampling procedure for model comparison. Illustrated the approach with an illustrative example using data from an international study. (SLD)

Presents a Bayesian approach to the evaluation of person fit in item response theory (IRT) models. Works the procedure in detail for the three parameter normal ogive model, but shows that the procedure can be generalized to many other IRT models. (SLD)

Developed and illustrated an analysis for detecting changes in multiple time series for a multilevel setting. Shows how the approach can be used to evaluate intervention effects associated with a treatment to which the lengths of delay in the response may be unique to each treated unit. The approach does not assume a known change point. (SLD)

Implemented a Bayesian factor analysis model to formulate estimators of the hyper-parameters. Simulation results lead to the proposal of new bias-corrected estimates of factor loadings. (SLD)

A Bayesian approach to the testing of competing covariance structures is developed. Approximate posterior probabilities are easily obtained from the chi square values and other known constants. The approach is illustrated using an example that demonstrates how the prior probabilities can alter results concerning the preferred model specification.…

Eleven experiments involving over 200 undergraduate students investigated how categorization of examples influences feature prediction for new examples. Results suggest that category-based prediction generally relies on a single category rather than multiple categories when there is a clear target category. (SLD)