Fig. 8: Post-error perseveration.

Proportion of incorrect trials (PT > 1000 ms) in which the participant pressed the same key as on the previous trial, grouped by previous trial accuracy and experiment. Points are means, and error bars are 95% confidence intervals (±SEM * 1.96). Experiment 1: n = 46 participants; Experiment 2: n = 37 participants; Experiment 3: n = 47 participants; Experiment 4: n = 46 participants.