orgtheory.net

teacher performance research

Teaching research is notoriously hard because there are selection effects at every turn. Students select into schools, teachers select into the profession. Students aren’t randomly matched to teachers inside schools and students aren’t randomly assigned teacher attention. Policies aren’t randomly assigned to schools or teachers. It’s nice to see this review of teacher performance research by Michael Podgursky and Matthew Springer in the Journal of Policy Analysis and Management. Seems like there’s been some progress on the issue:

  • There’s a lot of evidence that teachers matter, more so than schools. In most cases, descriptive studies show consistent teacher effects. There’s just tons of classroom/teacher variation in most large data sets from large school districts. In contrast, school effects are all over the place. Sometimes you find them, sometimes you don’t.
  • There’s a lot of evidence that teacher credentials don’t matter. The same data sets show almost negligible effects of credentials. Thus paying people merely for degrees earned is likely not a good idea.
  • Principals seem to have a good idea of who is a good teacher. Multiple studies of qualitative evaluation in different types of schools usually find that people who have recieved good evaluations (and often higher pay) often have better classroom outcomes.
  • It’s very hard to evaluate the exact role that pay has in performance. Aside from the selection issues, there’s also a mixture problem. Performance pay is always mixed in with other pay schemes, like seniority.
  • Despite these problems, most random assignment studies of performance related pay find a positive effect.
  • The downside is that the effects are short lived, suggesting that education needs to be sustained over multiple grade levels.

Bottom line: Getting an ultimate perfectly identified answer is nearly impossible, but the evidence does suggest that teachers matter and they respond to incentives, but that, like most educational interventions, incentives must be sustained over time.

Written by fabiorojas

February 18, 2010 at 4:17 am

Posted in economics, education, fabio

5 Responses

Subscribe to comments with RSS.

  1. “people who have recieved good evaluations (and often higher pay) often have better classroom outcomes.”

    Alternate (and to me more plausible) explanation: Principals protect their favorite teachers from having the most disruptive kids in their classes.

    In my experience, teachers are teaching the best that they can. I’ve never seen, “oh, I can get a bonus? Hey, for that I’ll pull out my *good* teaching!”

    Like

    Em

    February 18, 2010 at 5:06 am

  2. Em: That’s why randomized trials are important because, as the review points out, there is a lot of gaming by teachers and principals. And randomized trial do support performance based pay.

    “In my experience, teachers are teaching the best that they can. I’ve never seen, “oh, I can get a bonus? Hey, for that I’ll pull out my *good* teaching!” ”

    If you are in a school where all the teachers are angels, then that is a heaven that I’d want to live in.

    Like

    fabiorojas

    February 18, 2010 at 5:15 am

  3. Atlantic Monthly covered the efforts from Teach for America to find better teachers. The organization chooses its teachers based on a model of predicted effectiveness in terms of student outcomes (output minus input). The article references a third-party random-assignment study as well.

    My takeaway was less in the reported results but that there is probably some really interesting data TFA has for the person who is interested enough in this topic to try to gain access and make sense of it.

    The article previews this book on the 6 principles of effective teachers, based more on interviews of teachers who were able to improve outcomes vs. were not.

    Like

    ThoughtMarkS

    February 21, 2010 at 6:14 pm

  4. tariely

    March 15, 2010 at 7:39 am

  5. i teach but i dont understand all these jargon. Translate in Kiswahili

    Like

    Mr Masibo M. T

    April 29, 2012 at 11:42 am


Comments are closed.

%d bloggers like this: