Extended Data Fig. 5: Behavioural effects of the presentation format of monetary magnitudes.

(a) Population posterior distributions of the (a, b) intercept, ẟ, as well as (c, d) the indifference point, θ, for both (a,c) perceptual magnitude and (b, d) risky choice tasks. The intercept during perceptual magnitude is no different from zero (indicated here by the vertical dashed line) while it is significantly larger than zero during risky choice in both visual displays. Similarly, the indifference point in perceptual magnitude is no different from one while in risky choice, it is significantly larger than the threshold, \(\frac{1}{0.55}\) (the vertical dashed line). Distributions in pink represent data from the perceptual magnitude task, in blue represent data from risky symbolic payoffs, and in yellow-orange for risky nonsymbolic payoffs. One-sided Bayesian ‘p-values’ were calculated: The light pink-shaded mass of the highest density interval (HDI) covers 95% of the posterior distribution while the dark-shaded tail-ends represent the most extreme 5% probability mass of the posterior distribution. Bayesian comparison between posteriors reveal that the posterior distribution is significantly different from zero (represented here as a vertical dashed line) if the light-shaded mass does not cross zero. (e) Individual measures of the indifference point for nonsymbolic and symbolic payoffs are positively correlated (n = 62). The shaded area around the regression line represents 95% confidence intervals. The black dashed line represents the identity line. p-values were estimated from one-sided Pearson correlations.