An Examination of the Power of Experiments in Education to Detect Intervention Effects Across Contexts

• In Event: Roundtable Session 60
  In Roundtable Session: A Mixture of Papers on Heterogeneity

Abstract

Since 2002 the Institute of Education Sciences (IES) has funded more than 100 experiments to evaluate educational interventions in an effort to generate scientific evidence of program effectiveness on which to base education policy and practice. These studies are typically designed with the goal of having adequate statistical power to detect the average treatment effect. However, the average treatment effect is of limited utility to a practitioner in a particular site if the treatment effects vary substantially from site to site. This paper examines the power of studies to detect differences in intervention effects across contexts or sub-populations. The findings will help inform the planning of future experiments.

Author

• Jessaca K. Spybrook, Western Michigan University