The Brookings Institution looks at college enrollment disparities by gender, race, and socioeconomic status among students with similar academic preparation — measured by student test scores, high school grades, and course-taking — and find — surprise! — that gaps are much smaller and sometimes reversed among students with similar academic preparation:
We use the restricted-use High School Longitudinal Survey (HSLS 2009), a nationally representative sample of 2009-10 ninth graders.
So, smart people are smart and dumb people are dumb? How much research did that study require?