Fig. 2

Behavioral results. A Superiority rating score. B Signal detection theoretic (SDT) parameter estimates for the Ponzo illusion task. C SDT parameter estimates differences (between depth cue and control conditions) for the Ponzo illusion task. In A-C, transparent dots represent individual data points (superiority rating, d’, C, and meta-C) or individual estimates obtained from a single-subject Bayesian model fit (M-ratio and log M-ratio). Larger non-transparent dots and corresponding leftmost values represent the mean values across participants (superiority rating, d’, C, and meta-C) or the group-level hierarchical Bayesian maximum a posteriori (MAP) probability estimates (M-ratio and log M-ratio). Error bars represent 95% confidence intervals of the mean (superiority rating, d’, C, and meta-C) or 95% highest-density intervals (HDI) of posterior distributions (M-ratio and log M-ratio). The rightmost values indicate statistical test values. Asterisks represent statistical significance (* p < 0.05, **** p < 0.0001). M-ratio = meta-d’ / d’. log M-ratio = log(meta-d’ / d’). VAS, visual analog scale. a.u., arbitrary unit