Examination of Items in a Teacher-Made Test in the context of Differential Item Functioning

Abstract

The present study aims to examine whether the items in a teacher-made test show a classical test theory and item response theory based differential item functioning (DIF) with different DIF determination methods. For this purpose, final exam scores of 435 students regarding the course of measurement and evaluation were used. Final exam consisted of 50 multiple-choice items. Data were collected online through the learning management system. Data analysis was performed using the ShinyItemAnalysis package on R. As a result of the study, it was found that one item showed DIF according to four different methods by gender. As a result of analysis made by university, on the other hand, none of the items showed DIF according to delta plot method, while five items showed DIF according to Mantel-Haenszel, Logistic Regression and Lord chi(2) techniques. When the statements in the items showing DIF by both gender and university are examined, on the other hand, it was observed that there was no statement that can create DIF.