Differential Item Functioning Detection Procedures for Data Fitted to the Multidimensional Graded Response Model

• In Event: Differential Item Functioning and Measurement Invariance

Abstract

Construct validity is necessary to confirm psychometric instruments measure their intended construct and limit measurement error. Differential item functioning (DIF) helps to support construct validity. DIF detection procedures have been thoroughly scrutinized under several item response theory models, but no study has explored DIF detection procedures for data fitted to polytomous multidimensional item response theory models. This study aims to identify the optimal DIF procedures for data that fits the multidimensional graded response model. Comparisons were made on three DIF detection procedures through a Monte Carlo simulation to determine the effect on type I error and rejection rates. General method selection recommendations are provided for psychometric and assessment professionals.

Authors

• John Phillip Walker, Illuminate Education

• Louis Rocconi, University of Tennessee