Oxford’s Research-Based Affirmative Action

A large-scale British study, released last week, gives new empirical support for the drive to provide affirmative action to “strivers,” less advantaged students who, despite obstacles, perform fairly well academically.  The research finds that students who attended regular “comprehensive” (public) secondary schools did better in college than those who scored at the same level on standardized admissions exams and attended “independent” (private) or “grammar” (selective public) schools.

Pointing to the study last week, Oxford University’s dean of undergraduate admissions, Mike Nicholson, created waves when he declared that students who do well at poor performing secondary schools “may have more potential” than those from more-advantaged schools, and that universities should consider the context in which students compile an academic record.  In the United States, universities have claimed for years that admissions officers consider socioeconomic obstacles a student has overcome, though evidence suggests that on average, at the most selective 146 institutions, they do not.

The new study, published by the British National Foundation for Educational Research and the Sutton Trust, a private foundation, was five years in the making and examined 8,000 students.  It found that students from independent or grammar schools performed the same in college as comprehensive-school students who scored one or two grades lower on their entrance exams (known as A-levels).  For example, a comprehensive-school student who scored three B’s on these exams would, on average, perform the same in college as an independent- or grammar-school student who scored an A and two B’s, or two A’s and a B.

Part of what makes the study striking is that it contrasts with research in the United States on racial affirmative action.  In the early years of affirmative-action programs, some hypothesized that because under-represented minority students were likely to have overcome obstacles, they would perform better in college than white students with the same raw academic admissions credentials.  In fact, research has consistently found that the SAT’s and other characteristics of minority students over-predict their college grades.  That is, these students perform worse, not better, in college than one would expect given their high school records.  (Recent research on low-income students in the U.S. finds no evidence of over- or under-prediction.)

Of course, the difference in findings between the American and British research might have to do with the different types of exams used for admissions in the two countries.  In the UK, admissions officers rely in part on achievement-based A-levels, whereas U.S. colleges and universities rely on standardized tests like the SAT, which fall closer to aptitude tests on the continuum between aptitude and achievement exams.  Perhaps the reason that standardized tests under-predict the college grades of students from less-advantaged types of high schools in Britain, but not in the U.S., is that the SAT does a better job of finding “diamonds in the rough” than the British A-levels.  Interestingly, however, a key finding of the British National Foundation study is that the SAT was not a better predictor of college success than A-level exams nor did the SAT do a better job of identifying academic talent among disadvantaged students.

The bigger question, of course, is not whether strivers outperform others in college, given that it takes time to overcome years of disadvantage.  Rather, in the long haul of their careers, do strivers outperform others with similar entering credentials, controlling for relevant factors?  In the meantime, however, the British research is part of a growing body of evidence to suggest, at a minimum, that strivers can perform very well at selective institutions.