"To question education is really dangerous. It's the absolute taboo..."

[sakky]

<p>

Umm, that didn’t really answer my question. How do you know that the people who get rejected from these schools are so hard-working?

</p>

<p>Look, obviously, we’ll never know the counterfactual unless we devise a quantum engine to visit the other universe where we admitted that person and viewed just how hard-working he would be.</p>

<p>But the fact is, colleges reject plenty of people right now who might have been successful at that college had they not been rejected. Yet why aren’t my detractors concerned about them?</p>

<p>My stance is, given that the colleges have to reject somebody, they should do so based on the best statistically predictive model about future performance that they have. This is precisely the same calculus that insurance companies perform every day. Obviously no insurance company knows exactly when somebody is going to die or have a car accident or have their house struck by a hurricane. But they make statistical predictions based on when those events are likely to occur, and then charge you premiums - or reject your insurance application entirely - accordingly. A guy with numerous DWI’s and other serious moving violations and owns a sportscar isn’t easily going to find auto insurance. </p>

<p>

Indeed, some such students did not become hard working until community college, sometimes after doing something else like military service between high school and community college (i.e. they were not ready to study at the college or university level right out of high school, but became committed students later).

</p>

<p>I have never advocated that we shut down the transfer admissions process. Schools could continue to admit transfer students…but again, based on a predictive statistical model. For example, A’s in community college creampuff classes may not be highly predictive of ability to graduate, but A’s in community college math/sci courses might be. Certain kinds of community college math courses might be more predictive than others. </p>

<p>Again, keep in mind that just as not all high school senior applicants are admitted, not all community college transfer applicants are admitted either. Indeed, many (usually most) are rejected. Since schools will need to reject plenty of applicants anyway, it should behoove them to do the best job of admitting the best ones that they can. </p>

<p>Obviously no system will ever be entirely perfect. Statistical models naturally have room for error. But it will surely be better than the current admissions system, which is clearly fraught with error. Even a 1% improvement in graduation rate would represent plenty of new graduates.</p>