Optimal Sampling Strategies for Cluster-Randomized Trials Detecting Moderation and Main Effects (Poster 2)

• In Event: Innovations in Statistical Power, Intervention Planning and Design, and Causal Effects

Abstract

Cluster-randomized trials investigating moderation effects can answer key policy-relevant research questions like under what conditions, for whom, and where treatment effects are most salient. When designing these experiments, researchers must consider (a) both average/main and moderation effects and (b) the sample size allocation and its design efficiency for both effects. However, the literature has not developed a statistical framework to guide the optimal sampling strategies for both effects. The present study develops an optimal design framework for cluster-randomized trials investigating both moderation and main effects, which can identify the jointly optimal sample allocation that uses the minimum budget to detect both effects. We illustrate the methods through examples and implement the methods in the R package XXX.
 

Authors

• Zuchao Shen, University of Georgia

• Benjamin Kelcey, University of Cincinnati

• Zhenqiu Laura Lu, University of Georgia

• Eunji Lee, University of Georgia