Fig. 3: Change in belief updating profiles for individuals with high psychosis proneness.
From: Individual differences in belief updating and phasic arousal are related to psychosis proneness
a As Fig. 2c, but now for subgroups of participants occupying the lowest (gray) and highest (black) quintiles of P-scores. b Kernel half-difference summary measure (subtraction of mean weighting of first 5 samples from mean weighting of last 5 samples) capturing degree of recency in evidence weighting, plotted for both lowest and highest P-score quintiles. c As Fig. 2d, e, but now for subgroups occupying lowest and highest P-score quintiles. d Summary measure capturing summed strength of modulations of evidence weighting by CPP (mean modulation weights over sample positions 4–10, significant cluster in Fig. 2d) and -|ψ | (mean modulation weights over sample positions 3–8, significant cluster in Fig. 2e), plotted for lowest and highest P-score quintiles. Red shadings, mean +/− s.e.m. of normative model fits. Horizontal greyscale lines in b and d, mean of data from each subgroup; circles, individual participants. P-values, two-sample permutation tests (two-tailed). CIL-H, 95% confidence intervals (bootstrapped) around difference of means between lowest and highest P-score quintile subgroups. n = 18 participants per subgroup for calculation of means, s.e.m. and statistical tests in all panels.
