Thread by @James_S_Murphy

The latest WSJ journal piece about the SAT is the usual sloppy mess (what on earth is an adjusted SAT score?) but the data viz of how a high school's average SAT score correlates with socioeconomic advantage powerfully knocks down the myth of meritocracy.

"Of the 1,212 schools that faced the most adversity — schools with adversity scores of 90 and above — only 22 managed to score above the overall median SAT score."

"Of the 10% of high schools with the highest SAT scores, a total of 1,035, just 64 had an adversity score of 50 or higher on the College Board’s scale."

Again, I don't get or buy the analysis of adjusted SAT scores. That's not at all how admissions works. The reporting is terrible. But the data viz is worth spending time with. wsj.com/articles/what-…

The outliers on this dataviz plotting a HS's avg SAT score against how advantaged it is are intriguing. What explains them? Do you have a guess?

They have 2 main things in common--in fact there are only 2 or maybe 1 genuine outlier. (more later)

So maybe you were hoping that the outliers on this SAT graph were schools doing extraordinary work with students so that they outstrip their peers on the exam, that these were the schools that really lift students up rather than just make sure not to drop them. Bad news.

What these outliers have in common is that all but two of them use an exam in their admissions process (2 of them use the SAT! for HS admission!!) and all but 2 are in a city or equally dense population (NJ). h/t to @nsoodik who knew how NYC specialized HSs work.

The moral of the story: If you want good outcomes on tests, enroll students who are good at tests. #outputsareinputs

I'm not saying these schools don't give students a great education, but only 2 schools that don't use exams had similar results. What are they doing?