Stan is a new Bayesian statistical software program that implements the powerful and efficient Hamiltonian Monte Carlo (HMC) algorithm. To date there is not a source that systematically provides Stan code for various item response theory (IRT) models. This article provides Stan code for three representative IRT models, including the…

The present article considers a fundamental question in evaluation research: "By how much do program effects vary across sites?" The article first presents a theoretical model of cross-site impact variation and a related estimation model with a random treatment coefficient and fixed site-specific intercepts. This approach eliminates…

In educational technology and learning sciences, there are multiple uses for a predictive model of whether a student will perform a task correctly or not. For example, an intelligent tutoring system may use such a model to estimate whether or not a student has mastered a skill. We analyze the significance of data recency in making such…

General diagnostic models (GDMs) and Bayesian networks are mathematical frameworks that cover a wide variety of psychometric models. Both extend latent class models, and while GDMs also extend item response theory (IRT) models, Bayesian networks can be parameterized using discretized IRT. The purpose of this study is to examine similarities and…

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.…

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 computerized mastery test is described that builds on the Lewis and Sheehan procedure (sequential testlets) (1990), but uses fuzzy set decision theory to determine stopping rules and the Rasch model to calibrate items and estimate abilities. Differences between fuzzy set and Bayesian methods are illustrated through an example. (SLD)

Context for the application and specialization of a Bayesian linear model is m-group regression and the application to the prediction of grade point average. Specialization involves the assumption of homogeneity of regression coefficients (but not intercepts) across groups. Model's predictive efficiency is compared with that of the full m-group…

A short review of the different estimation procedures that have been used in association with the Rasch model is provided. These procedures include joint, conditional, and marginal maximum likelihood methods; Bayesian methods; minimum chi-square methods; and paired comparison estimation. A comparison of the marginal maximum likelihood estimation…

In a situation where the population distribution of latent trait scores can be estimated, the ordinary maximum likelihood estimator of latent trait scores may be improved upon by taking the estimated population distribution into account. In this paper empirical Bayes estimators are compared with the liklihood estimator for three samples of 300…

An alternative to the Rasch model is introduced. It characterizes strength of response according to the ratio of ability and difficulty parameters rather than their difference. Joint estimation and marginal estimation models are applied to two test data sets. (SLD)

The purpose of this paper is to link empirical Bayes methods with two specific topics in item response theory--item/subtest regression, and testing the goodness of fit of the Rasch model--under the assumptions of local independence and sufficiency. It is shown that item/subtest regression results in empirical Bayes estimates only if the Rasch…

In this paper a Bayesian procedure is developed for the simultaneous estimation of the reading ability and difficulty parameters which are assumed to be factors in reading errors by the multiplicative Poisson Model. According to several criteria, the Bayesian estimates are better than comparable maximum likelihood estimates. (Author/JAZ)

The economic utility of the following 5 weighting methods for evaluating consumer loan applications was determined using a sample of 443 loans: (1) unit; (2) weighted application blank; (3) chi square; (4) Bayes; and (5) regression. The unit and weighted application blank procedures were the best approaches. (SLD)

Despite advances in common factor analysis, a review of 89 studies published in four selected journals between 1963 and 1976 indicated that behavioral scientists preferred principal components analysis, followed by varimax or orthogonal rotation. Resultant row sums of squares of factor matrices from principal component analyses of real data sets…