Assumptions

This from John Thompson at Schools Matter:

Assumptions that 'Just Seem Obvious' to Steve Brill

There is a long line of education experts exposing the factual errors in Steve Brill's Class Warfare so, instead, I set out to illuminate the assumptions underneath his brand of "reform." I was overwhelmed, however, by the number of times that Brill, and the non-educators he fawned over, discovered findings that "just seemed obvious." So, I only want to recount Brill's tale of the naive assumptions, opinions, and assertions, based on numbers from one district, that supposedly propelled Bill Gates into the experiment that Gates said was, "the riskiest thing we have ever done."

Brill begins with "Identifying Effective Teaching Using Performance on the Job," by Tom Kane, Robert Gordon, and Douglas Staiger. Kane et. al wrote, "the current credential-centered regime is built upon two questionable premises ..." They then presented evidence that was solid enough to contribute to the minor league debate over teacher certification. But then these economists made another leap, "the second premise is that school districts learn nothing more about teacher effectiveness." Without making any effort to link their data (and that assertion) to reality, Kane and company called for dangerous and revolutionary changes to the entire nation's educational systems. When I first read their opinions, I chalked them up to being academic theory.

If Brill's account is trustworthy, though, I was naive. Either Brill was loose with his words, or he was celebrating a classic "bait and switch." According to Brill, the authors asserted that the need for "paper qualifications" is a "bedrock of public education." Worse, Kane et. al supposedly claimed that the second core assumption of our system is that "districts can learn nothing more of teaching effectiveness after the initial hire," and "this was why almost all teachers initially receive satisfactory evaluations." (emphasis mine.) Even worse, they successfully convinced Bill Gates that a "surprising" lack of volatility in teachers' performances, measured by standardized tests, was grounds for overturning decades of social science research on teaching and learning, and the way that organizations change.

Rereading the paper, I fear that Brill accurately reported the authors' true intentions. Now I see the report as a grab bag of political soundbites. Kane et. al predicted an impending shortage of teachers, which would be most damaging for high-poverty schools. Their solution - based on the presumption that the lowest scoring teachers should be fired using a methodology that is systematically biased against teachers in high-poverty schools - would supposedly attract talent to those schools!?!?! To their credit, Kane and company recognized the need to adjust for circumstances beyond the teachers' control. For example, if construction was going on outside the window on testing day... But, they argued that test score growth models that would be used for evaluations should not be controlled for poverty.

Fortunately, the Gates team has subsequently distanced themselves from their initial overreach, but it still raises a huge question. What did they not know about peer effects and when did they not know it? ...