This Brookings Report highlights the continuing gaps in performance on the SAT and similar IQ tests among racial groups. Former Economics Professor Mike McPherson also gets a mention. Key chart:


Several Ephs tweeted out a link to the related New York Times story:

“Race gaps on the SATs are especially pronounced at the tails of the distribution,” the two authors note. In math, for example,

among top scorers — those scoring between a 750 and 800 — 60 percent are Asian and 33 percent are white, compared to 5 percent Latino and 2 percent black. Meanwhile, among those scoring between 300 and 350, 37 percent are Latino, 35 percent are black, 21 percent are white, and 6 percent are Asian.

Translating those percentages into concrete numbers, Reeves and Halikias estimate that

in the entire country last year at most 2,200 black and 4,900 Latino test-takers scored above a 700. In comparison, roughly 48,000 whites and 52,800 Asians scored that high. The same absolute disparity persists among the highest scorers: 16,000 whites and 29,570 Asians scored above a 750, compared to only at most 1,000 blacks and 2,400 Latinos.

There should be a way to combine this data with what we know about college admissions and applicant preferences to get a more up-to-date estimate of racial distribution of SAT scores at Williams. Start with the latest available Common Data Set (pdf):


Full analysis left as an exercise for the reader! Comments:

1) About 2/3s of Williams students score above a 1400 combined. Speaking very roughly (and using hand-waving as my statistical estimation method of choice), whites and Asian Americans have about the same raw numbers in this pool. (There are, of course, many more white than Asian 17 year-olds in the US, but the whites do much worse on the SATs (and most other IQ tests)). So, why is the ratio of whites to Asians among Williams students almost 4:1? This suggests that Williams might discriminate against Asian-Americans in admissions. Now, there are many other plausible explanations other than discrimination which might explain this, mainly involving student/family preferences. But there is an interesting Record article (or senior thesis!) to write about this topic.

2) The ratio of Asian-Americans (74) to African-Americans (43) in the class of 2020 is not quite 2:1. But the ratio of students with Williams caliber SAT scores between these two groups is at least 20:1. The only thing that could possibly explain this discrepancy is massive preferences for African-Americans (relative to Asian-Americans) in Williams admissions. Taking another hand-waving guess, I would estimate that at least 70 of the Asian-Americans scored higher on the SAT/ACT than at least 40 of the African-Americans. In other words, the two distributions probably have almost no overlap, looking something like:


That couldn’t cause any problems on campus, could it? Below is an example of the sorts of “conversations” that students with radically different SAT scores have at Williams.


Print  •  Email