Task set-up and behavioral performance.
(A) Task. Participants were shown a sequence of five arrows that were briefly shown and subsequently masked. Masks were constructed such that the arrows were either highly visible (HV) or near the threshold of awareness (low-visibility, LV). Participants had to decide whether the predominant direction of the arrow sequence was left or right. (B) Evidence accumulation diagram. Participants start with no evidence for either direction. Each incoming arrow moves the sum up or down in the diagram. Solid lines show the transitions for the HV example in panel A. States at which no decision can yet be made are highlighted in light-grey; states at which enough evidence is available for the decision are highlighted in dark-grey. (C) Individual performance during the six-choice discrimination task (Experiment 2) for HV and LV stimuli. (D) Decision-making performance as a function of accumulated evidence for HV and LV stimuli. Negative and positive numbers denote evidence for a left and right response, respectively (number of right-arrows minus number of left-arrows). (E) Reaction times as a function of accumulated evidence for HV and LV stimuli. (F) Impact of each arrow on the final decision (i.e., the extent to which the direction of the arrow determined the decision) over time for HV and LV trials. (G) Influence of the last arrow on the final decision as a function of the amount of previously accumulated evidence for HV and LV stimuli. (H) Confidence ratings of visibility for HV and LV arrows (Experiment 3). (I) Confidence ratings for correct and error trials for HV and LV arrows (Experiment 3). (J) Relationship between confidence level and performance for HV and LV stimuli (Experiment 3). (K) Priming strength of single HV or LV arrow as measured in a distinct masked priming task, in terms of error rate (left panel) and reaction times (right panel, Experiment 4).