a Participants pressed a button timed to the estimated completion of lights moving around a circle. The gray target zone cue displayed error tolerance around the 1 s target interval. An example participant RT distribution is centered at the target interval. Audiovisual feedback is indicated by the tolerance cue turning green for wins and red for losses. A black tick mark displayed RT feedback. On 12% of randomly selected trials, blue neutral feedback was given with no RT marker. b Example recording session for one participant for training (first 35 trials) and experimental blocks. Staircase adjustments of tolerance are plotted as a solid purple line, and the dotted purple line indicates the minimum bound on tolerance at ±15 ms. Accuracy for easy and hard blocks is plotted as white circles on gray backgrounds and black diamonds on white backgrounds, respectively. c Separate staircase procedures resulted in group accuracies of (mean ± SD) 82.7 ± 1.7% for easy and 18.1 ± 2.5% for hard blocks. Error bars indicate standard deviation across participants, with individual participant (n = 32 participants) accuracy overlaid as gray lines. d Tolerance and outcome data for the same example participant. Larger markers show block-level accuracy; smaller markers show binary single-trial outcomes. Model fit using logistic regression provides single-trial estimates of win probability, which can be converted to expected value. Inset shows win probability curves across all 32 participants.
展开▼