Is Teach For America Effective? What Our Research Found

Interpreting research on teacher quality is becoming more and more difficult. A paper presented by Linda Darling-Hammond and her colleagues at the April 2005 meeting of the American Educational Research Association (AERA) has renewed interest in what research says about the effectiveness of Teach For America (TFA) teachers. Although the AERA paper focuses primarily on teacher certification, the paper's title and some reporting on it have raised questions about the impacts of TFA teachers on student achievement, impacts that were the focus of a study by Mathematica first released in June 2004.

Here we clarify the Mathematica study. Our evaluation asked whether TFA teachers generate higher or lower student achievement than teachers who would have been hired in the absence of TFA. We compared TFA teachers to those who would otherwise teach their students, including novices and veterans, and certified and uncertified teachers, all of whom were willing to work in America's hardest-to-staff schools.

The Mathematica study found that TFA teachers had a positive impact on students' math achievement-average math scores were significantly higher in classes taught by TFA teachers than in classes taught by non-TFA teachers. The size of this effect was 0.15 standard deviation, which is equivalent to one additional month of instruction. In contrast, students in TFA and non-TFA classrooms had similar reading scores.

As part of the Mathematica study, we compared TFA teachers to the certified non-TFA teachers and found the same result: a positive impact on math scores and no impact on reading scores.

Our study provides the most rigorous evidence to date on the effectiveness of TFA teachers. Randomized trials such as this are widely regarded as the gold standard in scientific research. We randomly assigned students to classrooms within the same grades and schools to ensure that teachers had equivalent groups of students and faced the same working conditions.

In addition, the Mathematica study was national in scope. It included seven school districts, representing urban and rural areas across the country in which TFA was operating, and was based on high quality data collected specifically for the evaluation by Mathematica's own staff, who administered achievement tests uniformly in all study classrooms.

Back to Top