One reader writes frequently to boast about the high, high, high test scores at Eva Moskowitz’s charter chain in New York City called Success Academy (previously known as Harlem Success Academy until Eva decided to move beyond the Harlem area).

Another reader offered this response:

Great that the test scores are so high. But these data alone are not enough to reject the null hypothesis that poverty and poor test results are unrelated…as in, like, everywhere. Here are a few things that would make others get more excited about SA’s test scores.1. Make sure the co-located school and the SA school are truly matched pairs in terms of sample. Others on this blog have suggested ways that the student populations might be different.2. You need to isolate the intervention so that the co-located school can serve as your control group. Skeptics say that student population and test prep account for the scores. Proponents say it is about expecting more and believing the kids can do it. Or maybe it is about more total instructional time.If you want people to be more excited, you’ve got to tell them what is in the secret sauce because high test scores themselves need a context. It may be possible to raise the test scores of ELL students 50% in one year by strapping them into a computer adaptive module for 5 hours a day. But, would we want to do that?

Supposedly, the whole point of charters was to be able to scale up innovations. High test scores themselves tell us very little about the specific intervention(s) that are causing them. And, since we are talking about developing human beings here, we really need to get a sense of potential side effects of these interventions and if a less “costly” (financially or otherwise) change would achieve our goal just as well.