Identifying "Race Unknown" Postsecondary Students in Survey Data (original) (raw)

Higher education researchers struggle with how to handle missing student ethnoracial data. Without race data, they have trouble making comparisons across groups. However, dropping "race unknown" students from analyses potentially introduces bias into conclusions. There is little methodological discussion about how to proceed when faced with large numbers of missing student ethnoracial data. An important but so far unstudied question is: what are the characteristics of those college students that end up in the "race unknown" category? This research note examines how these students compare to others who do provide a racial identification to get a better sense of how counting or not counting these students may affect results. Additionally, using students' responses to two different questions about racial identification, we are able to examine differences within the proportion of students who might be counted as "race unknown." Our results suggest that Multiracial and White students make up significant percentages of "race unknown" college students in surveys. We also find that there are differences within the "race unknown" population. Every study of higher education will encounter student non-response on race and ethnicity, so this study is a useful one for researchers and policymakers alike.