You should probably follow Nassim's idea of not trying to measure x (leadership) vs output of leadership f(x) and try to measure the exposure and how it impacts

You should probably follow Nassim's idea of not trying to measure x (leadership) vs output of leadership f(x) and try to measure the exposure and how it impacts it. Since probably the most significant effort has been pulled into probability theory and trying to get a measure of x that's probably the place to look. And you would NOT try to measure it as a point estimate as many have reminded us, you would try to set bounds lower&upper.