Primary Submission Category: Sensitivity Analysis
A Split-Sampling Framework for Powerful Design of Observational Studies under Unmeasured Confounding
Authors: William Bekerman, Abhinandan Dalal, Dylan Small,
Presenting Author: William Bekerman*
Observational studies are valuable tools for inferring causal effects in the absence of controlled experiments. However, these studies may be biased due to the presence of some relevant, unmeasured set of covariates. The design of an observational study has a prominent effect on its sensitivity to hidden biases and the best design may not be apparent without examining the data. One approach to facilitate a data-inspired design is to split the sample into a planning sample for choosing the design and an analysis sample for making inferences. This procedure has been shown to enhance power when it is assumed that at most one among multiple outcomes is affected by the treatment and a single outcome is chosen in the planning sample. We devise a powerful and flexible method for selecting outcomes in the planning sample when more than one outcome may be affected by the treatment. We investigate the theoretical properties of our method and conduct extensive simulations that demonstrate pronounced benefits, especially at higher levels of allowance for unmeasured confounding. Finally, we demonstrate our method in an observational study of the multi-dimensional impacts of a devastating flood in Bangladesh.