In case you hadn't tired of the certification question, there's a new study out comparing performance of Teach For America corps members with their peers in 27 Los Angeles schools.

The TFA teachers appeared to post significantly stronger student achievement gains than their counterparts, even though they were up against teachers with a lot more experience.

Something about this paper didn't add up, so we did some digging. Much like the egregiously misleading results from the Louisiana study of teachers coming out of The The New Teacher Project (see here), there's more to this tale. In this study, it turns out that all of the teachers in the TFA sample had at least one year of teaching under their belts, but not so for the control group. In fact, the control group included plenty of first-year teachers--including, ironically, some first-year TFA teachers.

Apparently, all first-year teachers in the district were included in the control sample regardless of their origins, simply because of a technical problem with the dataset. A comparison was made anyway, even though the research on first-year teachers shows consistently basement-level performance, be they TFA, non TFA or little green Martians.

We've long been critical of deeply flawed research touting the value of certification. Too bad the alternative certification camp can't keep to the high ground.