Benchmark Effect Sizes and Design Parameters for Planning Cluster Randomized Trials and Detecting Moderator Effects
Planning a rigorous evaluation requires researchers to make informed assumptions about expected effects, sample sizes, and other key study design parameters. Although education research provides a growing body of evidence to guide these decisions, relevant findings are often dispersed across studies and can be difficult to locate and apply. This project brings that evidence together, helping researchers make study design decisions grounded in prior research.
The project examined more than 600 moderator effects across seven large-scale education studies. Findings revealed substantial variation in moderator effects across studies, including some effects large enough to be detected in future evaluations without major increases in sample size. These benchmarks can help researchers make more informed decisions about statistical power, sample size requirements, and the magnitude of effects their studies are designed to detect.
The project also supports PowerSpec, a free, open-access Shiny App that makes study design parameters easier to find and use. PowerSpec provides researchers with convenient access to evidence from prior studies that can inform evaluation planning, improve power analyses, and strengthen study design. Researchers can use the tool to plan evaluations that examine not only whether an intervention works, but also for whom it works and under what conditions.