Ok, my bad on the transfer students (I must have just looked at the applicant numbers, not enrollees). I still consider their “20yr study” of test submitters vs non-submitters to have a lot of holes in it, especially relative to what they want to actually prove (the kooky multiple intelligence theory).