Fig. 1: Stimuli and behavioral results.

a During both encoding and testing, each trial started with the presentation of a cartoon character together with two objects. During encoding trials, participants randomly selected one object and then received a feedback indicating how much the selected object was liked by the character (see b). c During encoding, three different feedback values were presented indicating whether a selected object was disliked (magenta circle, −1), moderately liked (blue circle, +1), or very much liked (green circle, +5). d Average reward was the average of ten feedback values displayed for a given character. Two characters were associated with high (Hi) average reward (one +1, five +5s), two with medium (Me) average reward (three +1s, three +5s), and two with low (Lo) average reward (five +1s and one +5). e Memory performance (proportion correct) increased monotonically as a function of feedback value. f. Memory performance increased nonlinearly (inverted U-shape) as a function of average reward. ***p < 0.001, **p < 0.01, *p < 0.05, ns not significant (p > 0.05), indicates the p value for paired t-tests. Horizontal lines in f and g indicate mean ± SEM. Source data are provided as a Source Data file. The images used to illustrate the paradigm in a, b were obtained from https://publicdomainvectors.org/ and http://www.freestockphotos.biz/.