Social Evaluation Learning Task

The social evaluation learning task is a reinforcement learning task within a social context. Participants are told that they have to learn how much the computer liked themselves and a fictional other. Separate blocks are completed for each referential condition. On each trial, participants are presented with positive-negative social evaluation word pairs and are asked to select the word that represents the computer’s opinion using the computer keyboard. No time limits are imposed. Participants are then given feedback on their selection (correct/incorrect). For each referential condition, participants learn a positive ‘like’ rule (80% of positive words correct), a neutral rule (50% of positive words correct), and negative ‘dislike’ rules (20% of positive words correct). Rule order is randomised. 24 trials are completed per block. At the end of each rule-block participants are asked to provide a global rating of how much the computer likes them, ranging from "complete dislike" to "complete like".

Learning can be assessed by calculating the number of errors made before reaching the criterion of eight consecutive rule-congruent responses. Relative bias scores can then be calculated by subtracting errors to criterion made when learning the negative rule from the positive rule. A positive value indicates a negative bias, as fewer errors are made learning the negative vs. positive rule. .

Creative Commons Attribution (CC BY)

Button, K. S., Kounali, D., Stapinski, L., Rapee, R. M., Lewis, G., & Munafò, M. R. (2015). Fear of negative evaluation biases social evaluation inference: Evidence from a probabilistic learning task. PloS one, 10(4), e0119456.
10.1371/journal.pone.0119456

Hobbs, C., Sui, J., Kessler, D., Munafò, M. R., & Button, K. S. (2021). Self-processing in relation to emotion and reward processing in depression. Psychological Medicine, 1-13.