Eliciting preferences and the construction of indifference maps: A comparative empirical evaluation of two measurement methodologies

The paper reports results of experiments conducted to evaluate the following methods of preference measurement: (a) the method of eliciting certainty equivalents to gambles, for obtaining a von Neumann Morgenstern utility function; and (b) the dominance method for obtaining indifference maps. The following attributes were used for method comparison: (1) test-retest consistency; (2) linearity of trade-offs; (3) confidence in the method; (4) ease of judgment required by method; and (5) goodness of method predictions. The study also investigated the associations between method reliability and several behavioural and experimental factors such as subjects' acceptance of rationality axioms, propensities of subjects for judicial modes of evaluation, perceived realism of scenarios and subjects' discrimination bands for probabilities.