Panning for teacher talent takes more than one tool

See all posts
Inside Higher Ed is featuring an important new study from Carolyn Hoxby and Christopher Avery that reinforces the need to take a multi-faceted approach to evaluating talent for admission to teacher preparation and teacher hiring. College selectivity has been shown repeatedly to correlate with future teacher effectiveness, but using the selectivity of a college as a proxy for academic smarts simply doesn't tell the whole story for top minority talent. The takeaway for us: Just as our teacher prep standards address more than the selectivity of the college, district HR offices need to use selectivity as only one measure to gauge their ability to compete for top talent. Why? Hoxby and Avery provide compelling evidence that top minority talent isn't only found in selective institutions.

Quoting from Scott Jaschik at Inside Higher Ed:

Indeed, 53 percent of low-income, highly talented students do not apply to a single selective college that doesn't have significantly lower average grades and test scores for admitted applicants than for these students -- and these students do apply to at least one college that is not selective at all. Many of these students apply to a single nonselective college, or to a single nonselective college and one moderately selective college. And more than half of these students do not have a single application sent to what might be considered a "reach" college for them. In other words, students who could probably gain admission to highly selective colleges (and receive financial aid from them) are not even trying to get in.

Importantly, Hoxby and Avery also call attention to the fact that more selective colleges tend to only recruit talented minorities from a few well known high schools in a district, eschewing a strategy of seeking more widely for the bright stars. Successful students at schools lacking that crucial teacher or counselor to give an extra push through the application process may get passed over.

Read the IHE write up here.  The actual paper is from NBER, requiring a subscription.