Decisions based on descriptions of the exact probabilities and
magnitudes of the outcomes are called description based decisions.
Decisions based on past experience without the benefit of such descriptions are called
The current dissertation makes an attempt to compare behavior of the same decision maker
in description based and experience based tasks. Since individuals in a group may differ in many respects, comparing individuals’ behavior allows us to make more valid and exact conclusions about human behavior in different kinds of decision making tasks, than examining aggregate behavior. Seventy Five participants performed a battery of 40 description-based tasks and 10 experience-based tasks. The results were analyzed using different cognitive models which were evaluated for their ability to predict each single choice as well as for their generalizability at the individual level between different tasks. The results show that the best fitting models in both the description and experience based tasks contained direct weighting of gains and losses and did not require the additional assumption of diminishing sensitivity to payoff magnitude, that is often made in the literature of description based tasks. Additionally, consistency in some of the model parameters, particularly the weighting of gains and losses was observed at the individual level between description and experience based tasks. These findings shed light on some of the well-known differences in decision making behavior in the two tasks types.