A popular methodology for sensitivity analysis offers two concrete proposal to draw conclusions about the severity of omitted variable bias: (a) comparing delta* (the value of the sensitivity parameter, delta, consistent with zero treatment effect) with 1, and (b) constructing identified sets...