The "risk elicitation puzzle" summarizes the lack of understanding of why people's stated risk preferences often do not match their actual risk-taking behavior. Among the several potential explanations, one possibility is that stated risk preferences may not accurately reflect their true preferences because of cognitive biases such as oveconfidence, lack of self-awareness, or social desirability. Another possibility is that people's risk-taking behavior is influenced by situational or contextual factors that are not sufficiently internalized when assessing risk preferences in the lab.
The second explanation would also account for the commonly observed instability of the measurement of risk preferences.
In this paper we focus on one factor behind the low correlation between risk elicitation methods and actual risk taking behavior, namely measurement error. In particular, we investigate whether this correlation can be substantially increased by reducing measurement error through the aggregation of behaviors and stated preferences over time (and the latter also across methods).
To do so, we track people’s daily risk taking behavior for 14 days using a daily reconstruction method, that includes also questions about the perceived level of riskiness of the reported activities The same subjects play four different risk elicitation methods (Holt-Laury, BRET, Investment Game and a loss aversion multiple price list) and reply to two widespread risk-taking questionnaires (DOSPERT and SOEP) on seven different occasions each. We expect to observe that the risk elicitation methods become much better predictors of risk-taking behavior as more decisions under risk and game rounds are aggregated. Moreover, we expect that this predictive power is further increased by using statistical methods designed to reduce measurement error.