Fig. 1: Experimental design and behavioral results.

a Trial sequence of the Becker-DeGroot-Marschak (BDM) auction task. On each trial, the subjects faced a snack food item and had to indicate their willingness to pay for that item. Subjects first pressed the left button on the mouse to signal that they were ready to indicate their willingness-to-pay. A matrix that indicated possible prices, from 0 to 200 New Taiwan Dollars (TWD, 1 USD = 30 TWD), in 10-dollar increments, would then appear on the computer screen. The subjects’ task was to use the mouse cursor to point to and click on the number closest to their maximum willingness-to-pay. b The distribution of willingness-to-pay from all subjects on all food items. c The impact of the bid offered by the subject on the previous trial on willingness-to-pay on the current trial. Note that sequential items were selected randomly and in an uncorrelated manner for presentation. For each subject, we regressed their willingness-to-pay—a measure of subjective value (SV)—in the current trial against the willingness-to-pay in the previous trial. Here we plot the regression coefficient of the willingness-to-pay in the previous trial. In the majority of subjects, willingness-to-pay in the current trial was positively correlated with the willingness-to-pay in the previous trial. d The relationship between the willingness-to-pay and response time (RT). For each subject, we regressed their willingness-to-pay against the response time in the trial. We plot the regression coefficient of the response time. In the majority of subjects, there was no relation between willingness-to-pay and response time. The * symbol indicates p < 0.05 (one-sample t test, two-tailed). Error bars represent ±1 standard error of the mean. Source data are provided as a Source Data file.