Skip to main content
. Author manuscript; available in PMC: 2024 Jun 13.
Published in final edited form as: Nat Hum Behav. 2023 Jul 6;7(10):1667–1681. doi: 10.1038/s41562-023-01640-7

Figure 4:

Figure 4:

Absolute Spearman rank correlations between task behavior (y-axis) and symptom measures (x-axis) under different regimes of data screening and participant exclusions. (A) No Screening = no exclusions (N = 386). (B) Accuracy Only = exclusions based on chance-level performance in the reversal-learning task (N = 352). (C) Infrequency Only = exclusions based on invalid or improbable responses to infrequency items (N = 301). (D) Both Types = exclusions based on the previous two measures (N = 283). Only statistically significant correlations are shown (p < 0.05, two-sided, not corrected for multiple comparisons; signed correlations are shown in Figure S1 and Tables S6–S9). Black Xs indicate significant correlations abolished under screening. Acronyms: Acc = choice accuracy; Pts = total points earned; WS = win-stay rate; LS = lose-shift rate; Pers = perseveration errors; β = inverse temperature; ηp = positive learning rate; ηn = negative learning rate; κ = learning rate asymmetry.