A few months ago, we criticized a much ballyhooed study out of Louisiana. That study claimed that teachers coming out of the Teaching Fellows' alternative certification program were stronger teachers than teachers coming out of other nontraditional programs, but the study failed to ensure that teachers had the same level of experience. In effect, it compared second-year teachers from the Teaching Fellows program, a cohort that included significant numbers of Teach for America corps members, with first-year teachers from the other programs--a patently unfair comparison.

Now TFA has come out with new data to correct the earlier misleading study, and in fact the TFA corps members going through the Teaching Fellows program still look pretty darn good.

This time with the researchers controlling for experience, TFA teachers were found to be more effective, raising test scores by an average of 4.5 more points in math over non-TFA teachers. When compared with the general teaching population, the data also shows TFA teacher to be slightly more effective; however, the differences were not statistically significant.

The study also examines rates of retention, and here the news is unsurprisingly not so good. TFA teachers were much less likely to stay in Louisiana and teach than other teachers. In some years, the TFA retention rate was as low as 4 percent after five years, compared with 65 percent for non-TFA teachers.