This paper provides a critique of techiniques for estimating distributional weights using the revealed preference (imputational) approach. Stochastic versus deterministic methods are contrasted and a preference for the stochastic methodology is indicated. It argues that Probit is the most...