Representing ambiguity in the laboratory using a Bingo Blower (which is transparent and not manipulable) and asking the subjects a series of allocation questions (which are more efcient than pairwise choice questions), we obtain data from which we can estimate by maximum likelihood methods (with...