The New SAT ‘Adversity Score’ Misses the Point
Apparently the college-admissions industry hasn’t had its fill of controversy for the year.
The College Board, the nonprofit that administers the SAT, has unveiled a new metric — called an “adversity score” by the media but officially named the “Overall Disadvantage Level” — in an attempt to account for the socioeconomic climate of each student’s hometown and school. This score, which forms the core of the College Board’s “Environmental Context Dashboard,” will be provided to college-admissions officers along with students’ SAT scores, but will never be made available to the students themselves. It will fall on a scale of 1–100, with anything above 50 indicating adversity, and anything below 50 indicating privilege.
The score is meant to balance out perceived inequities between students from affluent, stable backgrounds and those from impoverished areas and underperforming schools. Anthony Carnevale, director of the Center on Education and the Workforce at Georgetown University and one-time supervisor of a similar, short-lived College Board program called “Strivers,” has suggested that its “purpose is to get to race without using race.” In other words, the criteria considered for the Overall Disadvantage Level will approximate a racial distinction without, hopefully, risking the inevitable legal challenges to an explicitly race-based policy.
It would be impossible to deny that the results of the SAT and similar standardized tests vary significantly across class and race. But what matters is whether the new score accurately identifies the nature of the problems that lead to those disparities, and whether it presents a real path for overcoming them.
In a column for Business Insider, Leigh Patel argues that correlation between high SAT scores and high family incomes “undermines any claim that the SAT is objective.” Her complaint is patently absurd. The fact that standardized tests do not produce identical results for different people is evidence not that the tests are flawed but that different people are . . . different. It makes a fair amount of sense that a student from a stable family who has spent eleven years in a well-functioning school district might outperform a child of absentee parents who has been stuck for those same years in schools that barely manage to keep their doors open. This is not a failure of the test, and it can’t be fixed by a simple numerical adjustment for circumstances — it is a function of the circumstances themselves. A comprehensive 2011 study by Sean Reardon, an endowed professor of poverty and inequality at the Stanford Graduate School of Education, indicates wide achievement gaps correlate with family income across the academic board.
Even if accommodation for economic inequality did make sense, though, the College Board’s new metric doesn’t actually offer any, because it takes into account only broad factors such as the median income of the student’s geographic area. Thus it penalizes low-income students from high-income areas simply for living in decent neighborhoods. And it has the potential to encourage gentrification, as it incentivizes wealthy families to move into low-income neighborhoods to give their kids an advantage in the admission process. (The absurd bribery scandal that broke this spring is proof that this should-be-ridiculous hypothetical is more than plausible in 2019 America.) On income inequality, the program will almost certainly do more harm than good.
Complaints similar to Patel’s have arisen on the question of race, too, as early as a 2003 article by Roy Freedle in the Harvard Educational Review alleging that the SAT is “both culturally and statistically biased against African Americans, Hispanic Americans, and Asian Americans.” It is a common observation that the average score of white Americans is markedly higher than that of black Americans. This is true, but it is far from the whole truth. Data compiled by the Brookings Institution show a near-perfectly normal distribution of white students across score ranges, leaving the average score of white students relatively close to the overall national average. In the top score range, however, even white students are proportionally underrepresented: they make up 51 percent of test-takers overall, but only 33 percent of those in the highest range. Asian-American students, on the other hand, make up 14 percent of test-takers overall but 60 percent of those in the top range. Let’s break that down: it means that an Asian-American is disproportionately likely by more than a factor of two just to take the test in the first place, and then, within that select group of college hopefuls, is disproportionately likely by more than a factor of four to score highly relative to all test takers. That sure puts a damper on Freedle’s accusation of anti-Asian-American bias in the test.
Nevertheless, black and Hispanic students really are overrepresented in the lowest score ranges and heavily underrepresented in the highest ranges. The facts quickly discredit the bias-implicit-in-the-test narrative, so what is actually going on here? One factor suggested by such scholars as John McWhorter and Thomas Sowell is culture. While many Hispanic and African Americans maintain negative attitudes toward academic achievement, they say, Asian Americans — especially recent immigrants and children of immigrants — often demonstrate remarkable commitment to academic success. Thus, even in the same struggling school districts, students with different cultural attitudes produce staggeringly different results.
Students fail tests because schools, parents, and a number of other social factors are stacked against them. To ignore that reality is to throw young people unprepared into competitive schools and a competitive world, at the expense of everyone involved — the students, the institutions whose standards inevitably decline, and their deserving peers who were passed over — in the name of a forced and false equality. To address it will take a lot more than adding another number to the College Board’s scoresheets.
