Effect Size Benchmarks for Social and Behavioral Outcomes
- Designing RCTs with sufficient power requires researchers to make assumptions about design parameters.
- Abt is helping create an encyclopedia of reference values.
- The encyclopedia will help researchers benchmark impacts and conduct accurate power analyses.
In order to design randomized trials with sufficient power to detect intervention effects with precision, researchers need to make reasonable assumptions about the design parameters to estimate the required sample sizes. In addition, it is helpful to use empirical benchmarks for interpreting the effect sizes observed in research studies.
Working with the University of North Carolina, Chapel Hill, and the University of Missouri, Abt is creating an encyclopedia that will provide reference design parameter values to inform power analyses for two- and three-level cluster randomized and multisite randomized trials. The encyclopedia also will include empirical benchmarks for interpreting the magnitude of the intervention effects on social and behavioral outcomes.
The goal is to develop a comprehensive reference for researchers that will provide critical information needed to design trials that can detect effects and help researchers interpret their findings for their intended audience.