>So they've found that encouraging students to take CS courses based on their skin color or genitals is less effective than encouraging students who have an interest or aptitude for the subject? Gee, I never would have guessed that result.
Yes, this is well known.
What traditionally happens is that teachers are very concerned with their pass rate, so they filter kids out of their class that they think won't pass the AP test.
I worked for a College Board program for four years designed to address this problem, as a lot of the people getting filtered out might very well pass anyway, and therefore be denied an opportunity for an advanced class and college credits for no other reason than the teacher's ego.
So they stopped talking about pass rates entirely, and heavily discourage teachers from using the term, instead quantifying teacher success based on *numbers of students who pass* instead. So even if little Timmy only has a 50% chance to pass, it would still encourage the teacher to let him try, since the expected value of letting Timmy stay in the class is better than if the teacher filtered him out.
Unfortunately, the fucking article perpetuates the old model of thinking, which is to emphasize the pass rate over the actual number of kids passing the AP test, and demonstrating that they have a freshman in college-ish level of understanding of the subject.