Figure 1.

Motion Discrimination Task

Each block of 15 to 42 trials was assigned one of six possible base rates: the prior probability that the motion is rightward. In all blocks, the subject discriminated the direction of random dot motion. No feedback about individual decisions or the base rate was provided until the end of the block.

(A) Sequence of events within a trial. After fixation and a random delay, random dot motion was displayed for 0.1–0.9 s (truncated exponential). Subjects then positioned a cursor on the left or right arcs to indicate both choice (left versus right) and confidence (100% certainty [top] to guessing [bottom]). After the motion decision, the subjects reported whether they believed that the block had a rightward or leftward bias (placing the cursor in the left or right half of the line), together with the confidence in this belief (from unsure at the center to certainty at the extremes).

(B) Example of the feedback display provided at the end of a block.

(C) Example of a sequence of trials within a block. Lower two graphs show variables controlled by the computer: motion strength, direction, and duration. Upper two graphs show the subject’s reports: direction choice, confidence that the choice was correct, and belief that the base rate favored rightward. An example of the task is shown in Video S1.