Measuring Adversity

Many colleges say they want to admit more applicants who are socioeconomically disadvantaged. But how do colleges define disadvantage -- while making sure that only those who grew up with real hardships get the extra edge in admissions decisions?

The College Board has just finished the first two pilots of a system -- the Environmental Context Dashboard -- that is designed to help colleges be more precise when deciding who deserves that edge. One of the colleges that employed the system in a test review of applications after decisions had already been made reported that up to 20 percent of admissions decisions might have been made differently using the dashboard.

Here's how the system works: when students take the PSAT or the SAT, they enter information about the high school they attend and where they live. Then the College Board uses that information to create an index on adversity based on three calculations: the high school environment, the neighborhood environment and the family environment.

For high school, the formula is based on such factors as the rates at which students "undermatch" (fail to apply to colleges that they could get into), percentage of students receiving free lunches and the availability of a rigorous curriculum. For the neighborhood, calculations would be based on crime rates, housing values and the percentages of homes that are vacant. For family, the index is based on education level of parents, whether the student is being raised by a single parent and other factors. The family figures are based on the average for the neighborhood, so the information is not self-reported.

The idea behind the index is that colleges need help in determining which of their applicants have faced serious adversity and which may not, said Connie Betterton, vice president for higher education access and strategy at the College Board.

Admissions officers "want to evaluate students within the context of where applicants went to school and grew up," Betterton said. Currently they can do so in a "qualitative way," but many want more data to assure fairness and accuracy. Particularly for colleges with a national recruiting base, it may be hard to know much about the high school or neighborhood of an applicant in a way that allows for comparisons to other applicants, she said.

The new index does not consider race. Betterton said the College Board wanted a tool that could be used nationally, including in states where public colleges are barred from considering race and ethnicity in admissions.

The College Board completed two tests of the new system in which colleges used the index for a second review, after having already made admissions decisions. In the experiment at a moderately selective public institution, where admissions decisions are made by formula, admissions officers said the index wouldn't have changed anything (unless the college changed its approach to admissions), but that the information was helpful. At another institution, described by Betterton as a competitive private college that uses holistic admissions (where all applicants' various application materials are examined individually), officials said that using the index might have changed 20 percent of decisions. The College Board declined to identify either of the institutions.

The next step will be a larger pilot with about 10 institutions (again using the system after admissions decisions have been made) before testing when admissions decisions are being made.

No decisions have been made on the business model to be used, Betterton said. That will likely come after the next round of pilots. The College Board currently licenses the names of students with certain characteristics (including first generation and low socioeconomic status) to colleges, and this new data could eventually be offered as well, but no decisions have been made.

Philip Ballinger, associate vice provost for enrollment and undergraduate admissions at the University of Washington and a trustee of the College Board, said he viewed the adversity index as a significant advance. He said his university -- in a state where consideration of race in admissions is banned -- has been trying to gather some of the same information the College Board will use. But he said he was concerned about issues of whether data were comparable, and whether self-reported data on students' families was accurate.

Ultimately, he said, his admissions team "wants a better sense of place" on where applicants come from and the adversities they have faced, and the index seems like it could provide that context.

Richard D. Kahlenberg, a senior fellow at the Century Foundation, said the program is "a huge step" for the College Board. Kahlenberg has long been an advocate of class-based, rather than race-based, affirmative action, and in the past has written that those behind standardized testing were not doing enough to identify smart, talented, low-income students.

"Universities say they want to provide a leg up to disadvantaged students who have overcome obstacles, yet today, because of firewalls between financial aid and admissions departments, admissions officers often have to guess about who is economically disadvantaged," Kahlenberg said. "Colleges that adamantly reject the idea of being race blind in admissions have to make critical decisions in a way that is largely class blind. The environmental dashboard will help them see better."

But Shirley J. Wilcher, executive director of the American Association for Access, Equity and Diversity, said she was concerned about anything that might discourage colleges from appropriate ways of considering race and ethnicity in the admissions process.

Institutions that engage in holistic review "cannot justifiably ignore the question of race," she said.

While the new College Board index would not preclude colleges from also considering issues of race, Wilcher said she feared such an approach might not happen. "Metrics are important. What gets measured gets action," she said via email. "But to not consider race and ethnicity is a rollback, not movement forward. It is precisely in higher ed where we have contributed significant and impactful scholarship about the intersection of [socioeconomic status] with both race/ethnicity and gender. To not also collect the race/ethnicity data means that in fact, they are not fully examining the 'environmental context.'"