For higher-education researchers, the choice of how to classify people who identify themselves as biracial is hardly as simple as black and white. In fact, a paper being presented here on Thursday at the annual conference of the American Educational Research Association says that each of three commonly used approaches to classifying biracial or multiracial people has benefits and drawbacks, and the choice of which one to use can significantly influence a study’s results.
The authors of the study are Karen Kurotsuchi Inkelas, an associate professor of college-student personnel at the University of Maryland at College Park; Matthew Soldner, a doctoral student at Maryland; and Katalin Szelényi, an assistant professor of education at the University of Massachusetts at Boston. They conducted their analysis using data on more than 22,000 undergraduate students at 49 colleges gathered as part of the 2007 National Study of Living-Learning Programs, for which Ms. Inkelas serves a principal investigator. The study provided an ideal source of data for them because it uses a survey instrument that lets students identify with as many races and ethnicities as they please.
The researchers crunched their numbers using three commonly used approaches to classifying biracial and multiracial students, to see how each approach would affect their results.
With one approach, researchers classify subjects who belong to two or more racial or ethnic groups as simply being “biracial” or “multiracial.” With a second approach, subjects who identity with two groups are classified as belonging to the least prevalent one, so that a student who reports being both white and black is designated as black.
Under a third approach, used by the federal Office of Management and Budget, most biracial research subjects are given biracial classifications that reflect their backgrounds, such as “white-black” or “white-Hispanic.” But, for the sake of keeping the number of categories manageable, researchers disregard data from any biracial subset that accounts for less than 1 percent of the total sample studied. (With the student data collected by the National Study of Living-Learning Programs, such an approach left the researchers focused only on those biracial or multiracial students who identified as white and Hispanic, white and Asian American, white and Native American, or white and black.)
Ms. Inkelas and her colleagues sought to see how using each of the three different racial-classification schemes would affect a statistical analysis of students’ responses to survey questions about making the transition to college. They found that the schemes produced sharply different results, and each skewed at least some of their findings. In classifying students who had identified themselves as white and Native American as being Native American, for example, they drastically overestimated the percentage of Native American students who were receiving merit-based aid.
How researchers classify a biracial population, says the paper summarizing the authors’ findings, “can have profound implications” for both the descriptions of students that arise from those researchers’ work and the conclusions that result from their analyses. “Unfortunately,” it says, “there is no single solution to this empirical dilemma. Indeed, each approach has its strengths and its limitations.”
By, for example, classifying all students who identify with two or more groups as being simply “biracial” or “multiracial,” researchers avoid erroneously lumping student together with single-race peers but run the risk of glossing over significant differences between biracial populations.
Although they do not recommend a way around the problem, the authors suggest that it may be worthwhile to conduct studies determining how biracial or multiracial students would themselves prefer to be classified. “After all,” they say, “the constituency affected most significantly by the various classification schemes, one could argue, should have a stake in how they are ultimately treated.”