Issues Affecting Item Response Theory Fit in Language Assessment: A Study of Differential Item Functioning in the Iranian National University Entrance Exam

Abstract

The assumptions of unidimensionality and local independence are related in that; "items found to be locally dependent will appear as a separate dimension in a factor analysis" (Reeve 2003, p. 12). Besides factor analysis, model fit can provide evidence that this assumption is satisfied. Exclusion of the discrimination power or pseudo-guessing as the possible sources of DIF will result in undetected DIF items and may hence lead to the removal of the most useful items in a measure (Angoff, 1993). [...]applications of the Rasch models limit our understanding of the group differences in responding to items. [...]a likelihood-ratio -statistic for each item is computed by ... where Hj is the number of intervals for item j and rhj is the observed frequency for interval h in item j. The degree of freedom is the number of response categories minus 1; since all items on the INUEE are dichotomously scored, this is always 1. [...]misfit could be due to the fact that the items were too difficult for the target population, leading to the skewed raw score distributions in Figures 1 and 2.