Examining the Effectiveness of Test Accommodation Using DIF and a Mixture IRT Model

Abstract

This study examined the validity of test accommodation in third–eighth graders using differential item functioning (DIF) and mixture IRT models. Two data sets were used for these analyses. With the first data set (N = 51,591) we examined whether item type (i.e., story, explanation, straightforward) or item features were associated with item difficulty, item discrimination, or DIF. The second data set (N = 3,452) was used to investigate whether observed DIF was related to students' accommodation status, gender, race, disability, or latent academic ability. Item types and features were significantly related to item difficulty but not to DIF. Story items were not always more difficult than explanation or straightforward items. Observed DIF was partially explained by an interaction of accommodation use and academic ability. Limitations and implications of this study are discussed, and suggestions for future research are offered.