Randomized Controlled Trials in Education

Absract

Randomized controlled trials (RCTs) are viewed by many researchers as the best design for causal inference. In medicine, RCTs have been regarded as the strongest design for causal inference for decades. Since its creation in 2002, the Institute of Education Sciences (IES) has stressed the importance of conducting rigorous research in education and has identified RCTs as the appropriate design to establish causal links between interventions and outcomes. Largely due to funding opportunities offered by the IES’s Investing in Innovation program and its Education Innovation and Research program, numerous RCTs have been carried out in education in the last two decades. IES has funded more than 350 RCTs to date, and thus RCTs are increasingly considered the strongest design for causal inference in education. This chapter provides a discussion of RCTs with a focus on education research and policy. The discussion covers the history of RCTs, the definition and advantages of RCTs, threats to the validity of RCTs, the replication problem, RCT registries, the basic types of experimental design, blocking, clustering, and statistical power. The chapter concludes with remarks about the present and future of RCTs in education. 

Keywords: experimental design; randomized controlled trials; causal inference; internal and external validity of randomized controlled trials; statistical power