Testing the predective validity of the time trade-off and the Stardard Gamble

This paper tests the consistency of health utility measurements with individual preferences. We compare three methods, the time trade-off, the standard gamble and a version of the standard gamble that corrects for the deviations from expected utility modelled by prospect theory. Individual preferences are measured both through a ranking task and through a choice task. In decisions involving no risk the time trade-off is most consistent with people’s preferences with the standard gamble a close second. In decisions involving risk the corrected standard gamble is most consistent with people’s preferences. Our data do not support the common assumption in health economics that utility is transferable across decision contexts.