The race distributions are interesting, sure. But the full report has some better data points for drawing conclusions. Of particular interest to me was the income/parental education differential (obvious increases, but interesting to see by how much) but especially a category for results by when the test was taken - weekend vs in-school. Students who took the test on a weekend were nearly 300pts higher, and the "met requirements" percentage nearly doubled.

It would be particularly interesting to see this parsed out by the same economic and education related parameters in order to see how much an effect just having a smaller quieter "separate" test environment helps a student's score.