The SAT discriminates (less)

Bombarded with accusations of racial and socioeconomic discrimination and dropped as a requirement from almost all selective colleges, the Scholastic Aptitude Test now seems to be a dead letter on its way out in college admissions. Yet today some universities are re-implementing standardized testing as a requirement, notably the Massachusetts Institute of Technology.

Standardized tests are out of style, deemphasized in favor of other parts of the college application process — like extracurriculars, Advanced Placement classes, grade point averages, teacher recommendations, and essays. None are free from reflecting inequality. But the SAT tends to be better than other facets of the college application.  

Take extracurriculars: The robotics team, for example, is well organized and competitive at Palo Alto High School. Maybe a school in a less privileged district might have one, but it likely has much less funding and oversight. 

Or teacher recommendation letters — at a school where students often ask for recommendation letters (wealthy schools), teachers inherently get more practice writing them. Oftentimes, teachers are also more experienced and qualified at affluent schools. Consequently, the recommendation letters they write are likely better.

As for GPAs, often touted as a better long-term measurement than SAT scores, grade inflation is more prevalent at schools attended by wealthier students than at schools attended by low-income students, according to a study by the Fordham Institute. For students performing at the same level, students at wealthier schools may tend to have higher GPAs. And for availability of AP classes: They’re not offered in every school, and the specific APs offered vary across school districts. 

That’s why colleges often use school profiles to compare students at the same high school. But with this disparity between GPAs at different schools, using a school profile essentially pits students against their classmates and not high schoolers across the nation. The SAT is the only way to measure performance on a standard scale.

And strangely, students’ choice of college essay topic, like whether they discussed sports injuries or community involvement, are more closely associated with socioeconomic status than SAT scores are, a Stanford study finds.

But for the SAT, students who have a free Saturday morning and access to a test center can try it twice, since fee waivers are available (for the SAT, they waive fees for the Student Answer/Question Answer Service, which gives information about incorrect and correct answers given on the test itself). While location and time are still significant barriers to access to standardized testing, they pose a smaller disadvantage as compared to access to AP classes or extracurriculars, which require a much larger initial time commitment and are far harder to access.

So what can disproportionately help affluent students do better at the SAT? Test prep courses are everywhere — anything from group proctored SAT practices to individualized math prep at $160/hr. Can free online Khan Academy really compare? When controlling for the fact that students who sign up for test prep are not representative of students as a whole, those who pay for test prep tend to see some improvement in their scores — somewhere around 10 to 35 points. And the nationwide score improvement, on average, for taking the SAT a second time is about 40 points. Test prep doesn’t seem to help all that much compared to simply retaking the test, which is free for those eligible.

Perhaps more subtle is the culture around taking standardized tests. At Paly, test-taking culture can be intense, with years-ahead plans on optimal test dates. Access to and knowledgeability about standardized tests in different schools is hard to measure. But it’s plausible that test-taking culture is weaker at lower-income schools, in which students take standardized testing at lower rates. However, this isn’t cause for going test-blind but for raising awareness of testing at lower-income schools.

Lastly, does going test-optional really improve diversity? A recent study finds that in a sample of 100 schools that went test-optional, the relative frequency of Latino, Black and Native American students only rose by about 1%, according to the Hechinger Report. It’s something, but not much at all.

In the end, the SAT is far from perfect. But it should be reinstated as a requirement for college applications until a standardized measure even less-reflective of money comes along.