Rethinking reference and racial/ethnic categories: implications and insights for education researchers

Racial/ethnic categories used in education research are frequently aggregated across groups and often used in regression models as relative to white individuals, (un)consciously centering whiteness. We employ Quantitative Critical Race Theory (QuantCrit) to consider the implications of collapsing racial/ethnic categories and of reference group category selection. We propose a decision tree framework to support researchers in choosing a reference category and illustrate its application using data from the Early Childhood Longitudinal Study (ECLS-K). We describe students’ interest in math and estimate a series of models that vary the reference group, including comparisons to white students but also 1) to the average of all students, 2) to the average of all students excluding white students, 3) to a universal target, and 4) that analyze within group differences. We demonstrate how choices around racial/ethnic categories and reference groups can produce different interpretations of the same underlying data.