A comparison of pseudo-Bayesian and joint maximum likelihood procedures for estimating item parameters in the three-parameter IRT model

This study compared pseudo-Bayesian and jointmaximum likelihood procedures for estimating itemparameters for the three-parameter logistic model initem response theory. Two programs, ASCAL andLOGIST, which employ the two methods were comparedusing data simulated from a three-parametermodel. Item responses were generated for sample sizesof 2,000 and 500, test lengths of 35 and 15, and examineesof high, medium, and low ability. The resultsshowed that the item characteristic curves estimated bythe two methods were more similar to each other thanto the generated item characteristic curves. Pseudo-Bayesian estimation consistently produced more accurateitem parameter estimates for the smaller samplesize, whereas joint maximum likelihood was more accurateas test length was reduced. Index terms:ASCAL, item response theory, joint maximum likelihoodestimation, LOGIST, parameter estimation, pseudo-Bayesian estimation, three-parameter model.