The fact that previous studies have shown that an individual's value function is stable across sources of uncertainty (Abdellaoui et al., 2016 Armantier & Treich, 2016) points to the importance of the probability weighting function for explaining expertise-dependent variation in risk attitudes.Īllowing for probability weighting is also important to better understand risk attitudes and risky choices in the insurance domain (Hansen et al., 2013 Kairies-Schwarz et al., 2017). These findings are remarkable given that the amount of actual information concerning the underlying choices was held constant across levels of perceived expertise. Individuals with higher perceived expertise are more willing to pursue risky investments (Hadar et al., 2013), hold larger fractions of risky assets in their portfolios (Frijns et al., 2008), and prefer securities they feel more knowledgeable about (Ackert et al., 2005 Fellner et al., 2004). Recent experimental evidence, however, suggests that the perception of one's own expertise crucially alters economic behavior in decisions under risk. These models therefore claim that an individual's risk attitude and hence probability weighting function does not vary across decisions within the domain of risk. Even though current decision-making models do account for nonlinear probability distortions, they typically assume that risky events, that is, events where objective probabilities are known to the decision maker, constitute a unique source of uncertainty 1 1 Throughout this paper, the term uncertainty captures both risk (known outcome probabilities) and ambiguity (unknown outcome probabilities). This behavior results in an inverse-S-shaped probability weighting function. Instead, probabilities are transformed nonlinearly, that is, people tend to overweight low (tail) probabilities and underweight high probabilities (e.g., Kahneman & Tversky, 1979 Quiggin, 1982 Tversky & Kahneman, 1992). It is widely accepted that individuals do not process probabilities as normatively prescribed by Subjective Expected Utility Theory (Savage, 1954). We document that ignorance 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. This result suggests that individuals are subject to ignorance illusion in decisions under risk, constituting expertise-dependent risk attitudes. Even though all probabilities are explicitly provided, we find that individuals exhibit more pronounced inverse-S-shaped probability weighting if they perceive their level of expertise regarding a gamble to be lower. We conduct two experiments involving different gambles, that is, risky games where objective probabilities are known, no further information-based advantages exist, and outcomes are independent of knowledge. This paper provides evidence that challenges this assumption. Hence, a decision maker's processing of probabilities and the resulting degree of probability weighting should not vary within the domain of risk. Current decision-making models assume that an individual's attitude towards risk is unique.
0 kommentar(er)
