When I first started here at ABCTE in 2002, I had little knowledge of anything in the education policy world. I hope it’s apparent that I have come a long way since then. Regardless, one of the first things I learned about was the idea of value-added measurement to determine how well a teacher is teaching. Simply put, measure how well that teacher’s students are performing from year to year and determine the average gain that takes place in his/her classroom. Step two is to adjust compensation accordingly for those teachers and, maybe, get rid of XX percent at the bottom of the curve.
At the time, this seemed brilliant to me and I was ready to adopt it all across the U.S. Since then, my feelings have become more nuanced on the subject. On the whole, I still strongly believe that some type of meritocracy must be attempted for teachers. There has been an experiment here and an experiment there, but no true merit pay system has been implemented for any substantial period of time in public schools. I still support value added measures, but also concede that these should only be one part of any meritocratic compensation.
This short video from Daniel Willingham points out six issues with value added measures. I agree with 1 and 3. 2 is a bit above my head statistically, but I believe that the statimagicians could work around it. I think that 4 and 6 are a bit of a copout and that 5 is highly dependent on the design of the assessment instrument (as is 3). All that being said, there are enough potential issues and areas of uncertainty to restate that I no longer believe that a teacher should be measured by the test scores of their students along.