Project Activities
The software was programmed in R and is freely available to applied researchers. In addition to publishing in journals and presenting at conferences, the research team used social media, press releases through NORC, briefs for nontechnical publications, and workshops at conferences, such as AERA and SREE, for applied education researchers.
Key outcomes
The main finding of this project is as follows:
- In anticipating sample size requirements for quasi-experiments, the number of observations required for such studies typically exceed those required for expected impacts from randomized trials. For propensity score-based studies, the sample requirements are often 30 percent larger, and for time-series studies, the requirements depend on the level of autocorrelation. If the autocorrelation between time points is about .5, the sample size requirements range from needing 4 to 16 times more timepoints relative to data without autocorrelation, depending on the model used. (Hedberg, 2023)
People and institutions involved
IES program contact(s)
Project contributors
Products and publications
Publications:
Hedberg, E. C. (2023). How Many Cases per Cluster? Operationalizing the Number of Units per Cluster Relative to Minimum Detectable Effects in Two-Level Cluster Randomized Evaluations with Linear Outcomes. American Journal of Evaluation, 44(1), 153-168.
Hedberg, E. C., & Hedges, L. V. (2026). Computing Statistical Power for the Difference in Differences Design. Evaluation Review, 50(1), 149-180.
Additional project information
The code for the software is available at: https://github.com/hedbergec/plannerapp.
The app is available at: https://steppcenter.shinyapps.io/Planner/
Previous award details:
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
