https://www.google.com/search?q=SAT+score+STEM+attrition+cor...
Also, such a paper should analyze dropout rates at elite colleges, as it's possible that the correlation degrades after a certain threshold.
And yes, it is limited to an elite college (Duke University).
> However, even though we have shown substantial evidence that Duke does indeed possess private information about the student’s future performance, we cannot conclude that there is mismatch.
It (claims to) disprove the hypothesis that there is no mismatch because Duke possess no private information, as private information is a necessary (but not sufficient) condition for mismatch.
> In sum, Duke’s private information on student performance is important beyond the student’s grade point average, affecting both the student’s subjective well-being as well as the likelihood of persisting in a difficult major. It is important to note, however, that we have still only established that some of the necessary conditions for mismatch are met. For mismatch to exist, information must not only be important but the revelation of the information must change the institution the individual would have chosen to attend. Hence, even if individuals find out they are not as academically strong as they thought, the information still may not have changed their decisions.
Let's not lose track of what I initially claimed: that lower levels of academic preparedness lead to students changing out of science/engineering majors, and that this was consistent across race. The study did find that this was true. And this was limited to students at an elite university. Mismatch is about what students would have done when given this information, which is not directly related to my claim
Is it possible that it's actually a predictor of who will predict later extenuating circumstances inhibiting their studies? And one that works with remarkable consistency across all demographics? Maybe but I feel confident enough in the former that I'm going to expect to see data on those circumstances to which you allude before shifting conclusions.