Abstract

Classical Test Theory (CTT) and Item Response Theory (IRT) are two competing theory in measurement field. Superiority of invariance property of item difficulty parameter in item response theory to classical test theory was proved theoretically. This qualitative empirical study is designed to answer how invariant the item difficulty parameter derived from each measurement theory across different samples? Material and Method The present study empirically examined, using norm-referenced large scale data of 39^th comprehensive Pre-training examination of medical students of IRAN with 2075 examinee to show invariance property of the item difficulty parameter under the two competing measurement theory. Winestep version 3.71 and SPSS version 17 softwares were used for item analysis in two models. Joint maximum likelihood procedure with Expectation Maximization was applied in this study.  Results The findings indicated that the degree of invariance of the item difficulty parameter across different samples, usually considered as the theoretical superiority IRT models, also appeared to be similar for the two measurement frameworks. Discussion The findings suggest that across samples the degree of invariance of the CTT item difficulty index was very similar with IRT item difficulty estimates.