Specification of Priors in Bayesian Estimation Under the Rasch Model

• In Event: Advances in Rasch Measurement

Abstract

A review of various priors used in Bayesian estimation under the Rasch model is presented together with clear mathematical definitions of the prior distributions. An example that compared the Bayesian estimation method with conditional, marginal, and joint maximum likelihood estimation methods is presented using empirical data. Evaluation of the priors and their specifications on both item and ability parameter estimates is performed with simulation data with two small sample sizes and three test lengths. The computer program OpenBUGS that implemented the rejection sampling method is the main program employed in the study. A discussion is presented for related issues in Bayesian estimation and use of priors in item response theory.

Authors

• Seock-Ho Kim, University of Georgia

• Allan S. Cohen, University of Georgia

• Minho Kwak, University of Georgia - Athens

• Juyeon Lee, University of Georgia - Athens