Knowledge Illusion in Decisions under Risk : The Impact of Perceived Expertise on Probability Weighting

Current decision-making models assume that an individual's attitude towards risk does not vary across different sources. A decision maker's processing of known probabilities and the resulting degree of probability weighting should therefore be unique. This paper provides evidence that challenges this assumption. We conduct two experiments involving different gambles, i.e., risky games where objective probabilities are known, no further information-based advantages exist, and outcomes are independent of knowledge. Even though all probabilities are explicitly provided, we find that individuals engage in less severe probability weighting if they perceive their level of expertise regarding a gamble to be higher. This result suggests that individuals are subject to knowledge illusion in decisions under risk, constituting source-dependent risk attitudes. We document that knowledge illusion stems from the wrongly assigned importance of perceived expertise in the decision-making process and that it occurs in both the gain and the loss domain