In my prior post, I argued that the bill to re-authorize ESEA mostly reinforced the principles of NCLB 1.0 and their implications for practice. Here, I am going to focus on a much narrower issue that I have written a lot about in the past: value-added measures.
One element of the Congressional bill clearly reinforces value-added--the continuation of annual testing. Beyond that, the effect on value-added depends completely on a distinction that receives far too little attention: teacher value-added versus school value-added. Let’s start with teacher value-added since that’s what is on the mind of most educators right now.
I made a distinction in the prior post between NCLB 1.0, where teacher accountability was absent and credentials like certification dominated, and NCLB 2.0, which used federal waivers to encourage states to adopt teacher accountability. In particular, the NCLB 2.0 waivers encouraged states to implement more extensive teacher evaluation systems based on a mix or “basket” of measures including value-added, classroom observations, and others selected by states and districts.
The new law means the states will no longer have to abide by these waivers and there is a good chance that some states will scale back aggressive evaluation and accountability. In several cases, teacher accountability appears so closely tied to the federal waiver that it will quickly disappear because of NCLB 3.0 (the Every Student Succeeds Act).
Other states will have a decision to make. I hope, as they proceed with those decisions, that they recognize that the debate about teacher value-added is only a small part of the debate about teacher accountability. As I wrote in my book, teachers and teacher unions have long complained--rightfully so--that they receive too little useful feedback for their instruction. Perhaps the most important contribution of the rise of teacher evaluation and accountability is that it has helped address that feedback problem.
But what role for teacher value-added? Advocates for value-added correctly argue that value-added measures are more objective than many alternatives, that student test scores are important outcomes, and that teachers with high value-added measures also produce better long-term outcomes on average. On the other hand, critics are equally correct that value-added measures yield extremely limited feedback, are confusing to teachers, and cannot be calculated for most teachers. Also, teachers do not believe value-added measures are accurate or fair and teacher evaluation and accountability will not be successful without at least some buy-in from teachers themselves. It is difficult to weigh these pros and cons, however, because we have so little evidence on the effects of using value-added measures on teaching and learning.
Partly because of limitations of value-added measures, I’ve also argued for many years that perhaps the best use of teacher value-added is as the first stage of a valid evaluation process, a process where personnel decisions are ultimately based on measures other than value-added. This approach would address several problems with value-added, including some of the validity and reliability concerns. There are other ways to use value-added as well. The point is that it would be a shame if we went back to a no-feedback world, and even more of a shame if we did so because of an over-stated connection between evaluation, accountability, and value-added.
What about school value-added? Seven years ago, I wrote a piece with Tom Toch that was critical of the NCLB 1.0 focus on proficiency levels, which punish schools serving disadvantaged students because these students start school far behind their peers. As a result, they tell us relatively little about actual school performance.
In this respect, the Congressional bill is a great opportunity. States now have considerable flexibility in how they design their performance measures, allowing them to add a school value-added component if they wish. My report on the Louisiana educational system and other writings also provide suggestions on how to do that, e.g., how to use value-added and levels-based measures in tandem.
NCLB 3.0 is also an opportunity because it requires schools to move beyond test scores in accountability. This will mean a reduced focus on value-added to student test scores, but that’s OK if the new measures provide a more complete assessment of the value that schools provide.
Whatever you might think of the rest of ESSA/NCLB 3.0, it is a chance to get value-added right and correct the problems with how we measure teacher and school performance.