Today Mathematica Policy Research is releasing a new study on the effectiveness of Teach For America (TFA). TFA places exceptional recent college graduates in teaching positions in underserved communities. This is the third such major study done.*
Here is the punchline: The study — examining TFA teachers in six regions (Chicago, Los Angeles, Houston, New Orleans, and the Mississippi Delta region) used random assignment to gather data from 17 schools, 100 classrooms, and about 2000 students — found that TFA teachers were as effective as the general population of teachers in these districts in teaching reading and more effective in teaching math. Make no mistake, TFA teachers are not outpacing all other teachers by leaps and bounds, yet this study confirms that there is not a downside (and probably an upside) to hiring TFA teachers in the present context of these communities.
Paul Decker, the lead researcher, put it this way, “TFA teachers not only had more success than other novice teachers but they had more success than teachers with an average of six years of experience in the classroom”
You’ll hear the following objections:
It was a weak control group because it included some uncertified teachers. Yes, but it was not an abstract control group, it was the actual pool of teachers teaching in these communities right now. Grumbling about the overall quality of the pool does not change the reality for kids there right now. And, TFA is offering as good, or better.
This just shows that we need to get serious about getting certified teachers into these communities. It shows we need to get serious about dealing with teacher equity because poor and minority students are getting shortchanged. But, because TFA teachers are as effective, or more so, than the average teacher in these communities right now, they are making a vital contribution to this effort right now, not down the road in some idealized environment. And remember, certified does not equal qualified. It’s worth noting that the study did not find that certified teachers outperformed the TFA teachers. Actually, though the subgroup samples are too small, it indicates the opposite for math and no difference for reading.
Teach for America teachers just teach for two years and then leave. Yes, many do, but many stay on teaching as well and 60 percent stay involved in education professionally. Besides, turnover for all teachers in these communities is high, it’s demanding work. It’s worth noting also that organizations like New Leaders for New Schools, New Schools Venture Fund, the Broad Foundation, and KIPP (which was founded by two TFA alums) are populated with former TFA’ers, that’s impact. In any event, no one castigates the Peace Corps because volunteers don’t live abroad permanently or other service opportunities that involve a limited time commitment.
What this study should do is shift the burden off of Teach for America to prove why TFA’ers should teach and onto critics of TFA to show, because they’re as good or better, why they shouldn’t. It should also spark a renewed debate about how we train and license teachers because it’s frankly not a ringing endorsement of the status quo that kids just out of college with a five week crash course turn in results like this. Just think about the results from a system that gave schools more flexibility about hiring, encouraged mentoring and support for new teachers, and included rigorous preparation…
Straw Man Alert: It’s also worth pointing out that TFA has never claimed to be a replacement for larger efforts to improve teacher preparation, that’s a mantle foisted on it by critics. Instead, it’s an effort to get disadvantaged kids good teachers now.
Turn the Knife Alert: Hopefully, some reporter will point out that while education schools fight accountability and reporting tooth and nail, TFA invites evaluation and is refreshingly transparent about their results. That, along with what’s proving to be a pretty good model for evaluating candidate potential and their emphasis on cohort support for new teachers are a few things that could be extrapolated to the larger policy debate…
*The other two studies…this one, (PDF) by CREDO, has the same limitations as this one in terms of the control group. But it found pretty similar results. The other, by researchers at Arizona State castigated TFA but be sure to read this quick, concise, and independent (PDF) review of it too, it’s got some pretty serious methodological flaws (read it’s a hack attack).
In the too often missing spirit of balance and debate, there was a critical review of the CREDO study on the National Commission for Teaching and America’s Future website that Eduwonk would post, but in a Kremlin-like airbrushing it seems to be gone. If any readers or comrades have a link, please send it along. Без перевода
Event: The 21st Century Schools Project had planned to host a forum on this study Friday featuring its lead author, Mathematica’s Paul Decker, Gary Galluzzo the former Dean of the Ed School at George Mason University, and Wendy Kopp, the founder of TFA. Because of President Reagan’s funeral on Friday the forum has been moved to Monday at 2PM at PPI’s offices.
Monday’s event will feature a presentation by Paul Decker of Mathematica about the study itself and reaction and discussion about policy implications from (a) David Imig, President and Chief Executive Officer of the American Association of Colleges for Teacher Education (b) Abigail Smith of Teach For America and (c) Kevin Carey of the Education Trust. It’s filling up fast so RSVP now to education AT dlcppi.org.
Update: They like ’em in LA…