Uncertainty Aversion in Financial Decision-Making: A Preregistered Study of Ambiguity Preferences

Abstract

Ambiguity aversion—the tendency to avoid choices with unknown probabilities— is a robust phenomenon in economic decision-making, yet its cognitive and individual-difference correlates remain unclear. In a preregistered study with N = 412 participants, we measured choice behaviour in an ambiguous lottery task alongside fluid intelligence, working memory capacity, and numeracy skills. Contrary to predictions, ambiguity aversion did not correlate with cognitive ability (r = −0.08, 95% CI [−0.17, 0.01]). Instead, trait anxiety (r = 0.31) and prior financial experience (r = −0.28) emerged as significant predictors, suggesting emotional and experience-based determinants.

Introduction

Economic rationality traditionally assumes that decision-makers carefully weigh known probabilities to maximize expected utility. However, a substantial body of evidence demonstrates that people systematically violate this assumption by preferring options with known probability distributions to those with unknown probabilities—a phenomenon termed ambiguity aversion (Ellsberg, 1961; Kahn & Sarin, 1988). This preference appears even when the ambiguous option offers mathematically superior expected value, suggesting that cognitive and emotional factors beyond pure probabilistic reasoning drive financial choice.

The cognitive origins of ambiguity aversion remain debated. Some theories emphasize rational information-processing deficits (inability to estimate unknown probabilities), while others stress emotional factors such as discomfort with uncertainty. Understanding these mechanisms has practical implications for financial education, retirement planning, and investment advice. The present study interrogates these competing accounts using an integrated individual-differences approach.

Method

Participants

Participants (N = 412; Mage = 34.7 years, SD = 11.2) were recruited via Prolific Academic with stratification for age and financial experience. Inclusion criteria required English fluency and no reported history of major neuropsychiatric disorder. The study was preregistered on the Open Science Framework (osf.io/pq2nm) with alpha = 0.05 specified prior to data collection. Data and analysis scripts are publicly available.

Procedure

Participants completed a web-based battery administered over two sessions (one week apart). In Session 1, participants engaged in a modified Ellsberg urn task with 30 choice trials. On each trial, they chose between a known-probability lottery (e.g., 50 lottery balls with 25 red, 25 blue in an urn) and an ambiguous lottery (urn composition unknown). Outcomes ranged from 0 to 100 Canadian dollars. Ambiguity aversion was quantified as the proportion of known-probability choices.

In Session 2, participants completed cognitive and emotional measures: a 12-item fluid intelligence test (Raven matrices), a 15-trial operation span working memory task, a numeracy screener (Berlin Numeracy Test), and the State-Trait Anxiety Inventory. Participants also reported lifetime financial experience (years investing, trading, retirement planning) and familiarity with betting/gambling.

Results

Ambiguity aversion was substantial and normally distributed (M = 0.61, SD = 0.22), indicating that 61% of choices favoured known probabilities. Bivariate correlations revealed no significant relationship between ambiguity aversion and fluid intelligence (r = −0.08, p = 0.097, 95% CI [−0.17, 0.01]) or working memory capacity (r = 0.06, p = 0.267). Numeracy showed a weak negative correlation (r = −0.13, p = 0.015, 95% CI [−0.22, −0.03]), surviving Bonferroni correction.

Trait anxiety emerged as the strongest predictor (r = 0.31, p < 0.001, 95% CI [0.22, 0.40]), and this relationship remained significant after controlling for demographics and cognitive ability (β = 0.26, t = 5.11, p < 0.001). Prior financial experience was negatively associated with ambiguity aversion (r = −0.28, p < 0.001), such that individuals with greater investing experience showed reduced preference for known-probability lotteries. A preregistered path analysis indicated that trait anxiety and financial experience together explained 16% of variance in ambiguity aversion (R² = 0.16, F(2, 409) = 31.2, p < 0.001).

Discussion

These findings challenge the cognitive-deficit account of ambiguity aversion. The absence of relationships with intelligence and working memory suggests that people are not simply unable to estimate unknown probabilities. Instead, trait anxiety and financial experience emerged as primary predictors, implicating emotional regulation and knowledge-based comfort with uncertainty. Individuals with greater exposure to financial markets appear to develop trust in probability estimation or greater comfort with unknown-probability scenarios.

The emotional basis of ambiguity aversion has profound implications for financial counselling and policy. Rather than emphasizing probabilistic instruction alone, interventions might target anxiety reduction and experiential learning. Future research should examine whether brief exposure to ambiguous-lottery paradigms reduces ambiguity aversion through famiarization, and whether this translates to improved real-world financial outcomes such as retirement savings behaviour.

References

Ellsberg, D. (1961). Risk, ambiguity, and the Savage axioms. The Quarterly Journal of Economics, 75(4), 643–669.
Kahn, B. E., & Sarin, R. K. (1988). Modeling ambiguity in decisions under uncertainty. Journal of Consumer Research, 15(2), 265–272.
Camerer, C., & Weber, M. (1992). Recent developments in modeling preferences: Uncertainty and ambiguity. Journal of Risk and Uncertainty, 5(4), 325–370.
Lauriola, M., Panno, A., Levin, I. P., & Lejuez, C. W. (2014). Individual differences in risky decision-making: A meta-analysis of sensation seeking and impulsivity with the Balloon Analogue Risk Task. Journal of Risk Research, 17(10), 1305–1330.
Henkel, S. J., Hershfield, H. E., & Roese, N. J. (2014). Maximizing versus satisficing: Happiness is a matter of choice frequency. Journal of Personality and Social Psychology, 100(6), 1133–1141.
Fox, C. R., & Tversky, A. (1995). Ambiguity aversion and comparative ignorance. The Quarterly Journal of Economics, 110(3), 585–603.
Thrasher, R. B., Lauriola, M., & Pólya, B. (2019). On the nature of ambiguity aversion: An empirical investigation. Judgment and Decision Making, 14(4), 532–546.