Gates/MET releases first set of findings on teacher effectiveness

See all posts

It seems we might all save ourselves a whole lot of time, money and aggravation now being invested in fixing teacher evaluations by just turning the job over to kids. Heck, that sure would save states and districts an awful lot of money.

That may be the most interesting takeaway from the first of (at least) four sets of findings released on December 10 by the ginormous Measures of Effective Teaching (MET) Project, undertaken by the Bill and Melinda Gates Foundation. After all, even the best evaluators catch only a glimpse of a teacher in action, when it's the kids who observe teachers six hours a day, day in and day out.

Student responses to MET's nicely-crafted survey--which avoids questions that would elicit responses about a teacher's popularity--lined up with teachers' value-added scores. Of all the responses, one in particular had the highest correlation: Students in this class treat the teacher with respect.

In addition to releasing the portion of the study dealing with student surveys, Gates also reported on the robustness of value-added measures themselves, looking at questions such as if state tests--from which most value-added scores are determined--correlate with other types of tests, including tests with open-ended responses; if a teacher's high performance one year will predict the next; and if a teacher who posts strong performance in one course section will be strong in another.

The short answer to all of these questions is YES. More specifically:

  • If a teacher produces solid gains one year, she is as likely to produce solid gains the next. All the talk about volatility of teacher's value-added scores, while true, is a bit over played: Previous years scores are in fact strongly predictive of current performance.
  • State tests measuring students' language arts achievement seem to be least able to identify the full range of teacher ability, raising some concerns among the research team about what these tests are measuring. They may just be lousy tests, which is something teachers have said for years.
  • And our favorite finding: Teachers whose students report that they spend a lot of time in the class practicing for the state test are by no means the better teachers.