Preliminary – DO NOT CIRCULATE Underweighting rare events in experience-based decisions: Beyond sample error

Recent research has focused on the difference between description and experience based decisions and on their treatment of rare events. While rare events are overweighted in description based decisions, people tend to behave as if they underweight rare events in decisions based on experience. Barron and Erev (2003) and Hertwig et al. (2004) argue that such findings are substantive and call for a theory of decision making under risk other than Prospect Theory when decision are based upon experience. Fox and Hadar (2006) suggest that the discrepancy is due to sampling error: people are likely to sample rare events less often than objective probability implies, especially if their samples are small. Ungemach, Chater, Stewart (2008) and Hau, Pleskac, Kiefer and Hertwig (2008) show that forcing subjects to sample more reduces the description-experience gap but doesn’t eliminate it. The current paper brings stronger evidence in favor of the description-experience gap with two new experiments. The first shows that the gap persists even when people know the actual distribution of outcomes and make a decision under risk rather than under uncertainty. The second shows that the gap exists also when people make repeated decisions controlling for sample biases and the “hot stove” effect. Moreover, while underweighting is observed in actual choices, overweighting is observed in judged probabilities. The results of the two experiments strengthen the finding that descriptive theories assuming the overweighting of small probabilities are not useful in describing decisions from experience. This is true even when there is no sample error for decisions under risk and for repeated choices and their judgment would imply underweighting. Preliminary – DO NOT CIRCULATE