An Expected Utility-User's Guide to Nonexpected Utility Experiments

Recent experimental evidence suggests that standard expected utility is violated in a wide variety of ways: losses are treated differently from gains, people are generally risk averse over gains and risk loving over losses, fanning and curvature effects exist, problem representation matters, and preference reversals are pervasive. An effort is made to suggest how these effects will change results from models based on expected utility theory, and how researchers who use expected utility in their analysis should react to these findings.