Response modes and inconsistencies in preference assessments

A general explanation is presented for a broad class of inconsistencies or preference reversal phenomena that occur in assessing preferences or utility functions. Based on detailed, highly controlled experiments, it appears that people systematically attach more importance to the attribute on which they are focusing, relative to other attributes of an object. This relative overweighting of the response dimension emerges as the first general explanation of the different performance reversal phenomena. This result is particularly interesting because it implies ways to obtain truly consistent utility functions