ISR-Thompson, 6050 ISR Thompson + Virtual (Zoom Link)
Zuchao “William” Shen
Assistant Research Professor at the University of Kansas Life Span Institute
Experimental studies have been widely used to evaluate the causal effects of interventions in the social sciences. In planning experimental studies, investigators must consider the sample size needed to achieve adequate statistical power and the budget required to conduct these studies. In this presentation, we will start with an example to motivate our discussion about the optimal design framework, which simultaneously considers sampling costs, budget, and statistical power. An optimal sample allocation can be identified by maximizing the statistical power under a fixed budget. Subsequent power analysis results will lead to efficient and effective design of experiments. We will use examples to demonstrate the utility of the flexible framework and its connections with the conventional framework. This line of research has the potential to shift future experimental studies to more efficient and effective ones. The presentation is designed to be accessible to scholars with different backgrounds and emphasizes conceptual understanding and practical applications.