Though the 2004 Mathematica findings on the "do no harm" effects of Teach For America didn't quite quell critics of the popular program, perhaps the latest study from the CALDER Center will. Researchers looked at the effectiveness (in terms of student performance) of Teach For America secondary teachers compared with traditional teachers and found some impressive results, particularly in math and science.

Not only were TFA teachers more effective in producing higher student performance than traditionally trained teachers (novice and experienced alike), but the difference amounted to .12 standard deviations--roughly the difference in having a novice teacher as opposed to a teacher with three years experience. The effects were especially pronounced because not only did TFA teachers have less experience on average than the comparison group, but they were also assigned to more academically challenged classrooms in already low performing districts and schools.

The authors suggest that TFA teachers were able to offset their lack of experience (and we might add traditional pedagogical preparation) with better academic preparation. They also suggest that perhaps an unmeasured, but nonetheless important factor, motivation, plays a role in the findings.