Valuing A Risky Prospect Less Than Its Worst Outcome: Uncertainty Effect Or Task Ambiguity?

GLW show that when lotteries involve non-monetary outcomes and people are not explicitly prompted to value them based on valuing the outcomes, lotteries with equal or similar probabilities are often valued less than their worst outcomes. The authors' results for Replication suggest that this IA violation is vulnerable to subject pool differences. Their results for Rewording suggest that the IA violation can be completely undone by rewording GLW's lottery instructions in a way that arguably makes the application of the IA more transparent.