An Intersectional Approach to DIF: Do Initial Findings Hold across Tests?

Abstract

ABSTRACT Differential Item Function (DIF) analysis is commonly employed to examine potential bias produced by a test item. Since its introduction DIF analyses have focused on potential bias related to broad categories of oppression, including gender, racial stratification, economic class, and ableness. More recently, efforts to examine the effects of oppression on valued life-outcomes have employed an intersectional approach to more fully represent a person’s identity and capture the multiple, and often compound, impacts of oppression. The study presented here replicated an intersectional approach to DIF analyses to examine whether findings from a previous study that focused on a single grade-level achievement test generalized to other subject areas and grade levels. Findings indicate that the use of an intersectional approach is more sensitive to detecting potential item bias and that this increased sensitivity holds across the subject areas and grade levels examined.