Bayesian Finite Mixtures for Nonlinear Modeling of Educational Data

In this paper we discuss a Bayesian approach for nding latent classes in the data. In our approach we use nite mixture models to describe the underlying structure in the data, and demonstrate that the possibility to use full joint probability models raises interesting new prospects for exploratory data analysis. The concepts and methods discussed are illustrated with a case study using a data set from a recent educational study. The Bayesian classiication approach described has been implemented , and presents an appealing addition to the standard toolbox for exploratory data analysis of educational data.