Figure 2.

(A) Graphical description of the Bayesian model fit for time perception ratings and circle-size ratings. $\widehat{y}$ is calculated with either a linear, a quadratic, or a power function, with the respective participant-specific parameters a, b, and c (see Materials and Methods in the text for more details). π is the participant-specific precision parameter and y is the measured ratings of the participant. (B) Graphical description of the Bayesian model fit for temporal discounting. μ_β is the group mean prior of the β parameter and λ_β is the respected precision parameter. μ_k is the group mean prior of the k parameter and λ_k is the respected precision parameter. Participant-specific β and k parameters are drawn from separate group distributions for HC and PG. β and k parameters determine the choice probability p in every trial, which is connected to the actual choice y. Shaded nodes indicate observed variables, while latent nodes are non-shaded. Round nodes are continuous quantities and squared nodes are discrete quantities. Double-bordered nodes are fully determined by their parent nodes.