The non-profit organ­i­sa­tion Teach For Amer­ica has, for two decades, been track­ing huge amounts of data on its thou­sands of teach­ers and the results they get from their stu­dents. By min­ing the data, test­ing hypothe­ses and refin­ing hir­ing and train­ing prac­tises con­stantly, the organ­i­sa­tion says it is now start­ing to cre­ate a reli­able pro­file of a suc­cess­ful teacher.

For years, Teach for Amer­ica selected for some­thing called “con­stant learn­ing.” As [Steven Farr, head of train­ing and support,] and oth­ers had noticed, great teach­ers tended to reflect on their per­for­mance and adapt accord­ingly. So peo­ple who tend to be self-aware might be a good bet. […]

But in 2003, the admis­sions staff looked at the data and dis­cov­ered that reflec­tive­ness did not seem to mat­ter either. Or more accu­rately, try­ing to pre­dict reflec­tive­ness in the hir­ing process did not work. […]

The results are spe­cific and sur­pris­ing. Things that you might think would help a new teacher achieve suc­cess in a poor school—like prior expe­ri­ence work­ing in a low-income neighborhood—don’t seem to mat­ter. Other things that may sound trifling—like a teacher’s extracur­ric­u­lar accom­plish­ments in college—tend to pre­dict greatness.

Other fac­tors that indi­cate whether a prospect would likely become an excel­lent teacher:

  • A mod­icum of knowl­edge on a sub­ject (Bachelor’s-level study pre­dicts bet­ter results in the class­room, whereas a Master’s in Edu­ca­tion has no impact).
  • Con­stantly re-evaluation.
  • Avid recruit­ment of stu­dents and their fam­i­lies into the process.
  • Ensur­ing that every­thing con­tributes to stu­dent learn­ing (main­tain­ing focus).
  • Exhaus­tive, pur­pose­ful planning—for the next day or the year ahead—by work­ing back­ward from the desired outcome.
  • Relent­less work ethic (“refus­ing to sur­ren­der to the com­bined men­aces of poverty, bureau­cracy, and bud­getary shortfalls”).
  • A track record, rather than just an atti­tude, of perseverance.
  • The best indi­ca­tor: a mea­sur­able past per­for­mance of achieve­ment (GPA and “lead­er­ship achieve­ment” specifically).

Update: Cedar Riener points to a short video (3m 44s) cre­ated by his col­league, Dan Will­ing­ham, on why merit pay based on test scores is a bad idea: “there is not a way to eval­u­ate teach­ers fairly by using test scores”.