Within-subject Preference Reversals in Description- and Experience-based Choice

Abstract

Numerous studies using between-subject designs have found that different decisions can be made about identical binary choice problems depending on whether the options are described or experienced. Using a within-subjects design we examined this Description-Experience ‘gap’ at the level of the individual. We found that: (1) the gap could be observed both at the group and the individual levels, (2) the gap was eliminated, at least at the group level, when controlling for sampling variability, and (3) riskier decisions were made by those with more positive risk attitudes, regardless of format. We conclude that the gap is likely a statistical phenomenon due to biased samples.


Back to Thursday Posters