Is it possible that the causality goes in the opposite direction - hormonal levels affect job selection?