Migration Background in PISA’s Measure of Social Belonging: Using a Diffractive Lens to Interpret Multi-Method DIF Studies

Abstract

This paper investigates measurement invariance as it relates to migration background using the Program for International Student Assessment measure of social belonging. We explore how the use of two measurement invariance techniques provide insights into differential item functioning using the alignment method in conjunction with logistic regression in the case of multiple group comparisons. Social belonging is a central human need, and we argue that immigration background is important factor when considering how an individual interacts with a survey/items about belonging. Overall results from both the alignment method and ordinal logistic regression, interpreted through a diffractive lens, suggest that it is inappropriate to treat peoples of four different immigration backgrounds within the countries analyzed as exchangeable groups.