|
Abstract:
|
In recent years there has been a push towards randomized experiments as the best way to determine "what works" in education. This talk will present examples of randomized experiments in education, including a cluster randomized trial of remedial reading programs, a nationally representative trial of the Upward Bound program, and a charter schools evaluation. When done well, randomized experiments can yield unbiased estimates of program effectiveness. However, in practice there are a number of challenges to doing random assignment studies in education, including recruitment and study retention, noncompliance, and clustering of students within classrooms, schools, and neighborhoods. Many of these issues lead to methodological challenges that have not yet been fully solved.
|
