A hierarchical IRT model for criterionreferenced measurement

sampler, posterior predictive checks A hierarchical IRT model is proposed for mastery classification in criterion-referenced measurement. In this model, items measuring the same criterion are grouped, and a difficulty and discrimination parameter of the criterion is estimated on the same scale as the person and item parameters. The level of proficiency of a student with respect to the criterion is determined by the probability of success on the criterion. Cutoff points on the probability scale can be used to classify respondents into masters and nonmasters. The hierar-chical IRT model is estimated using the Gibbs sampler and tested using posterior predictive checks. The model is illustrated with a test measuring the attainment targets of reading comprehension (in Dutch) at the end of primary education. An assumption often made in mastery learning is that mastering a trait or educational objective is a gradual process consisting of a progression along a continuum. A cutoff point on the continuum indicates the point of sufficient proficiency for a student to be classified as a master. This continuum view of This paper was prepared under various grants to the respective authors. The first author worked as a postdoctoral ssistant at the Leuven Institute for Educational Research and under a postdoctoral research grant from the University of Leuven (PDM/97/042). Part of the research was prepared while she was a visiting scholar tETS in Princeton, N J, whic