Figure 2: Estimated value sensitivities are comparable between non-psychiatric controls and individuals with MDD, and stable across visits.

We used a standard power utility function and softmax choice rule to identify separate ‘risk preference’ and ‘inverse temperature’ parameters to explain nonlinear and linear value sensitivities in decision-making. (ai, bi) Estimated RP and IT were stable across four repeated visits for both MDD and control participants. Across the repeated visits, both RP and IT were comparable between the control and MDD groups (no main effect of group using mixed-design ANOVA with rank transformation). The gray dotted line indicates risk neutrality (RP = 1). Each point represents an individual participant; group medians are indicated in green. Gray and red shades show the distribution of data points along the y-axis. (aii, bii) Spearman’s correlation coefficients were calculated to test whether the rank order of the parameters among individuals was consistent between visits to the lab (([1^st vs 2^nd visit], [1^st vs 3^rd visit], ... [3^rd vs 4^th visit]). See Supplementary Table S2 for statistical results. Each point represents an individual participant, and the color-coded lines are the robust regression line between measures from two visits. The x- and y-axes each represent the rank order of individual participants at each visit (for simplicity, not labeled here); *P < 0.05, **P < 0.01, ***P < 0.001, uncorrected; all correlations were significant after applying multiple comparison correction (FDR q < 0.0001).