In this article, Frederick Mosteller, Richard Light, and Jason Sachs explore the nature of the empirical evidence that can inform school leaders' key decisions about how to organize students within schools: Should students be placed in heterogeneous classes or tracked classes? What is the impact of cclass size on students learning? How does it vary? Since tracking (or skill grouping, as the authors prefer to call it) is widely used in U.S. Schools, the authors expected to find a wealth of evidence to support the efficacy of the practice. Surprisingly, they found only a handful of well-designed studies exploring the academic benefits of tracking, and of these, the results were equivocal. With regard to class size, the authors describe the Tennessee class size study, using it to illustrate that large, long-term, randomized controlled field trials can be carried out successfully in education. The Tennessee study demonstrates convincingly that student achievement is better supported in smaller classes in grades K-3, and that this enhanced achievement continues when the srudents move to regular-size classes in the fourth grade and beyond. The authors suggest in conclusion that education would benefit from a commitment to sustained inquiry through well-designed, randomized controlled field trials of education innovations. Such sustained inquiry could provide a source of solid evidence of which educators could base their decisions about how to organize and support student learning in classes and schools.
The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education