Abstract
Ambiguity aversion-the tendency to avoid options whose outcome probabilities are unknown-is a ubiquitous phenomenon. While in some cases ambiguity aversion is an adaptive strategy, in many situations it leads to suboptimal decisions, as illustrated by the famous Ellsberg Paradox. Behavioral interventions for reducing ambiguity aversion should therefore be of substantial practical value. Here we test a simple intervention, aimed at reducing ambiguity aversion in an experimental design, where aversion to ambiguity leads to reduced earnings. Participants made a series of choices between a reference lottery with a 50% chance of winning $5, and another lottery, which offered more money, but whose outcome probability was either lower than 50% (risky lottery) or not fully known (ambiguous lottery). Similar to previous studies, participants exhibited both risk and ambiguity aversion in their choices. They then went through one of three interventions. Two groups of participants learned about the Ellsberg Paradox and their own suboptimal choices, either by actively calculating the objective winning probability of the ambiguous lotteries, or by observing these calculations. A control group learned about base-rate neglect, which was irrelevant to the task. Following the intervention, participants again made a series of choices under risk and ambiguity. Participants who learned about the Ellsberg Paradox were more tolerant of ambiguity, yet ambiguity aversion was not completely abolished. At the same time, these participants also exhibited reduced aversion to risk, suggesting inappropriate generalization of learning to an irrelevant decision domain. Our results highlight the challenge for behavioral interventions: generating a strong, yet specific, behavioral change.
Highlights
Whether we choose a dish from a menu, a medical treatment or a retirement plan, the outcome of our choices is seldom certain
To control for additional factors, including understanding of the Ellsberg Paradox, age, gender, and individual random effects, we modeled choices of the ambiguous lotteries using logistic regression including phase and intervention (AC, non-active calculation (NC), and Control; between-subject) as categorical factors (see all factors included in the model in section Data analysis: Mixed-effect generalized linear model (GLM) approach)
To further control for factors including understanding of the Ellsberg Paradox, age, gender, and individual random effects, we modeled choices of the varying risky lotteries using Logistic regression, including phase and intervention (AC, NC, and Control; between-subject) as categorical factors (see all factors included in the model in section Data analysis: Mixed-effect generalized linear model (GLM) approach)
Summary
Whether we choose a dish from a menu, a medical treatment or a retirement plan, the outcome of our choices is seldom certain. While in some cases we can estimate the likelihoods for different potential outcomes (e.g. when tossing a coin), such accurate estimates are rare. Outcome likelihoods are at least partially unknown, or “ambiguous” [1]. Ambiguity aversion and learning about the Ellsberg Paradox. ) grant R21AG049293 and NIH/NIA (National Institute on Aging, https://www.nia.nih.gov/) grant R56AG058769 to Ifat Levy. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
