We certainly have had to read more than our fair share of dry, academic policy pieces on the issue du jour, teacher evaluation. Not a one of them has the power of a good old-fashioned feature story to drive at the core of what makes evaluation so hard to do right—which is why if you're at all involved in this issue, you need to read a story out of The Washington Post earlier this month. Reporter Stephanie McCrummen gives us a fly-on-the-wall view of a real debriefing session between a District of Columbia Public Schools "master educator" and a 4th grade teacher and, we, at least, are the smarter for it.
Perhaps because both the evaluator and teacher are male—and there isn't an emotional display that might evoke a more cynical response—you can cut with a knife the terrible, palpable awkwardness of their conversation. The evaluator has no choice but to confront the teacher with his weaknessesthe same teacher who had for years thought that he had been doing a pretty great job. You can't help but cringe over what has to be said, recognizing that this conversation times thousands can't be avoided if we're all going to buy into this new world view of teaching. Professional exposure, assessment, and critical reflection must replace teaching as a personal and private affair.
Now that we've trashed policy research, we do want to call your attention to a useful piece from The Aspen Institute looking at the early implementation of the District's IMPACT evaluation system. Teachers report that the framework provides a pretty clear articulation of what is expected of them. And despite widespread rumors in D.C. about wild scoring inconsistencies across evaluators, an analysis of the first year IMPACT data shows that the evaluation ratings of only a few dozen teachers (<1.0 percent) had scoring fluctuations that might call the results into question.
IMPACT also appears to be providing a more meaningful picture of teacher performance and teacher needs, if you're looking to see much better alignment between teacher and student performance. Before IMPACT, only 0.2 percent of all D.C. teachers were rated as "ineffective," with an additional 5 percent on average found to be failing to meet expectations. Under IMPACT, 19 percent of teachers were rated "ineffective" or "minimally effective," which seems far more realistic given D.C. student performance. There was little difference in the distribution of those ratings based on tenure.
But there also are some real concerns. Some teachers report pressure to manipulate the evaluation process by sending challenging students out of their classroom or changing their planned lessons to boost their scores—not unexpected given that D.C. has consequences for consistently poor performance. Most importantly, while evaluation scores on the teaching standards doled out by master educators correlated more with value-added student achievement results than evaluation scores given by principals, overall the assessments of classroom instruction in year one were only modestly related to value-added gains. D.C. will need to address these challenges carefully, and all eyes will be on its 2010-11 results.