Simple yet Sharp Sensitivity Analysis for Unmeasured Confounding
We present a method for assessing the sensitivity of the true causal effect to unmeasured confounding. The method requires the analyst to specify two intuitive parameters. Otherwise, the method is assumption-free. The method returns an interval that contains the true causal effect. Moreover, the bounds of the interval are sharp, i.e. attainable. We show experimentally that our bounds can be sharper than those obtained by the method of Ding and VanderWeele (2016). Finally, we extend our method to bound the natural direct and indirect effects when there are measured mediators and unmeasured exposure-outcome confounding.
READ FULL TEXT