Figure 2.

Example participants from behavioral experiment. We used data from the behavioral session to determine subsequent inclusion in an fMRI study based on criteria of consistency and diversity of preferences for different giftcards (assessed using the predictability of BDM ratings), relationships between BDM and paired-choice tasks, and correlations between quantity and subjective value. A, First, we examined quantity–bid relationships for the 13 different giftcards. The slope of the quantity–bid relationship for each giftcard is a measure of that the quality of the giftcard, with higher slopes corresponding to more valuable brands. Here, the participant in i has diverse but noisy preferences, the participant in ii is consistent but has similar preferences across giftcards, while the participant in iii displays an acceptable level of consistency while maintaining diverse preferences. B, To assess preference stability, we compared the slope of lines estimated from the BDM task with the number of times each giftcard was chosen in the paired-choice task. The participant in i shows a weak relationship between choices in each session; the participant in ii is consistent but shows little variability; and the participant in iii is both consistent and displays diverse preferences. C, For the fMRI experiment, we selected three giftcards for each participant that differed maximally in quality. Here we show BDM plots for selected cards. As before, i is noisy but shows diverse quality preferences, ii has similar preferences over giftcards, and iii has consistent and diverse preferences over giftcards.