TWIE: New TfA Study Makes Ed Schools Look Bad

My Monday post at This Week in Ed was unsurprisingly controversial despite reaching the pretty obvious conclusion about Teach for America corps member effectiveness:

So it’s not so much that this new study makes TfA look good, it’s that it makes schools of education and traditional teacher certification look bad.

After all, we as a society dedicate rather a lot of resources to teacher training. It takes a lot of time, money, and effort to get a teacher through a year (or more!) of traditional certification, and many more potential teachers are likely deterred by those commitments.

What are we getting for all of that trouble? Not, apparently, more effective teachers.

First, it’s worth noting that I confusingly conflated this new study with the literature on TfA effectiveness as a whole. So, yes, this particular study focuses only on secondary math teachers, but it’s consistent with numerous other studies of other subjects and grade levels finding that TfA teachers are about as effective as other teachers, and perhaps somewhat more effective in math.

What’s really important to my point about traditional teacher certification is the rough equivalence between TfA and non-TfA teachers in the totality of the research literature. Quibbling with the specifics of the findings – e.g, “The TfA advantage is so small!” – doesn’t really help to avoid the conclusion that ed schools don’t seem to be adding much value. We’ve been told time and again that one of the central problems with the program is that unlike their traditionally-certified counterparts, TfA teachers are woefully under-prepared to enter the classroom. It turns out that, in reality, traditional schools of education are spending a great deal more time “preparing” students but fail to get better results.

Encouragingly, much of the pushback I’ve gotten abandons the “TfA teachers are unprepared” canard and makes the somewhat-more-plausible argument that the real problem is that Teach for America isn’t scalable.

That argument is more plausible just to the extent that it’s true that  TfA currently produces a tiny fraction of our new teachers  and it is definitely not at all clear how big TfA (or similar models) can get. We need a lot of teachers in this country and TfA may not be able to provide all of them.

But how big could TfA get? The fact is that we don’t know, but the “scalability” argument is only really persuasive if you assume the answer is “no bigger than it currently is”, which is doubtful.1

More fundamentally, the scalability argument is not really in direct conflict with my point. My point is that traditional schools of education appear to be relatively ineffective compared to TfA, since TfA seems to get results that are at least as good, but much faster and quite possibly at a lower overall cost. “Ineffective, but scalable!” is both consistent with my thesis and a a very tepid defense of education schools.

  1. Note, for example, that the Mathematica study found that college selectivity doesn’t seem to matter much for teacher effectiveness; the potential TfA recruit pool may very well be larger than critics tend to assume. []
This entry was posted in Education, Education Reform, Teacher Training and tagged , . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.


  1. Posted October 1, 2013 at 2:49 PM | Permalink

    After reading the executive summary for the DOE report, it seems to me that the 0.07 number is something like an effect size. If that is the case, to put this number in context it is helpful to consider the work of John Hattie (author of “Visible Learning: A Synthesis of Over 800 Meta-Analyses Relating to Achievement”). In this book, Hattie finds that across a huge swath of educational research studies, the average effect size of various educational interventions is 0.4. That means that this 0.07 effect size really is very small and that compared to other possible interventions, it is not very effective at all.

    • Posted October 1, 2013 at 5:23 PM | Permalink

      There’s definitely not a lot of evidence that TfA teachers are a great deal better than traditionally-certified teachers, but it’s really pretty damning for ed schools that there’s not a sizable *ed school* intervention effect. For how onerous they are, they really needed TfA to have a large *negative* effect to avoid looking bad here.

Leave a Reply