The manner in which researchers classify biracial subjects can seriously skew their results, according to a study presented last month at the American Educational Research Association's annual conference and discussed in depth in a Chronicle of Higher Education article.
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, which 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.With one approach, they classified 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, they gave biracial research subjects dual classifications reflecting their backgrounds, such as “white-black” or “white-Hispanic.” For the sake of keeping the number of categories manageable, they disregarded data from any biracial subset that accounts for less than 1 percent of the total sample studied, with the result being that all dual classifications that did not have "white" on one side of the hyphen were excluded from their analysis.
The researchers found that their choice of classification scheme had a profound impact on their results, with some schemes painting a much bleaker picture than others for certain racial or ethnic groups. In using the second approach, for example, and classifying students who had identified themselves as white and Native American as being Native American, they drastically overestimated the percentage of Native American students who were receiving merit-based aid.
Because each classification scheme had strengths and weaknesses, the researchers concluded “there is no single solution to this empirical dilemma."