Simple yet Sharp Sensitivity Analysis for Unmeasured Confounding
This work addresses the challenge of unmeasured confounding in causal analysis for researchers and practitioners, offering a simpler and potentially more precise tool, though it appears incremental as it builds on existing sensitivity analysis methods.
The authors tackled the problem of assessing sensitivity to unmeasured confounding in causal inference by developing a method that requires only two intuitive parameters and returns arbitrarily sharp bounds on the true causal effect. They showed experimentally that their bounds can be tighter than those from a prior method, which requires an additional parameter, and extended it to handle measured mediators and unmeasured exposure-outcome confounding.
We present a method for assessing the sensitivity of the true causal effect to unmeasured confounding. The method requires the analyst to set two intuitive parameters. Otherwise, the method is assumption-free. The method returns an interval that contains the true causal effect, and whose bounds are arbitrarily sharp, i.e. practically attainable. We show experimentally that our bounds can be tighter than those obtained by the method of Ding and VanderWeele (2016a) which, moreover, requires to set one more parameter than our method. Finally, we extend our method to bound the natural direct and indirect effects when there are measured mediators and unmeasured exposure-outcome confounding.