.comment-link {margin-left:.6em;}

Ed Knows Policy

EKP -- a local (Washington, DC) and national blog about education policy, politics, and research.

Who is Ed Researcher?

Tuesday, April 18, 2006

Robert Gordon Recycles

Andy Eduwonk Rotherham being a prolific blogger, covered it much sooner than I could, but the Center for American Progress had an event last week that featured Robert Gordon, discussing his paper with Tom Kane and Doug Staiger called "Identifying Effective Teachers Using Performance on the Job."



I'm all for a think tank that tries to be the Heritage of the left, but the CAP is not quite cutting the mustard yet. Kane or Staiger, who are actual social scientists, could not be bothered to attend this event and I can't say I blame them. Gordon is a classic DC type -- a lawyer who gives policy advice to campaigns in hopes that his guy gets elected and appoints him to something. Fair enough. Folks like that are necessary. Here Gordon is touting bold, innovative ideas, which as far as I can tell have all been tried before or have been in circulation for decades, with the exception of a few of the more naive extensions like firing the teachers who have low value added scores after three years. Ok, I guess it's unusual for Democrats to be pushing them.




Don't get me wrong. I'm not saying the ideas are bad. They're mostly very good with a few clunkers that are not as well thought through. All schools and districts should attempt to use student acheivement data to identify the unique contributions of teachers and schools (value added) to student learning, but this paper is fairly naive. Performance based pay is one of those things that does get a bit easier with the prevalence of more testing data, but as we will soon see in Florida, the policies have a half life of about one school year.



One of the strongest pieces to argue against placing high stakes on teacher-level value added indicators is this one here, by, um, Kane and Staiger, on the volatility of these measures. Their proposed "solution" of aggregating across years and applying a Kalman Filter is much less convincing than their documentation of the problem.



There's more to say on the substance of the paper, but Andy does a good job at pithily summing that up.



By the way, the average age in the room must have been 25, with half the audience made up of what must have been CAP interns. Come on CAP, you have to do better than that.



The discussants were two union people who gave predictable (but still good) responses and two other people -- a child policy advocate and a DC teacher-- who didn't seem to have any social science training. Very disappointing. I wish I had the text of the Joan Baratz-Snowden's (bio here) comments, which she read off the page. Some of the points were standard union tripe but some of them were cleverly worded counterarguments or calling Gordon et al. to task for being naive.


0 Comments:

Post a Comment

Links to this post:

Create a Link

<< Home