Sobell Dept, UCL Institute of Neurology, London, UK
When appraising information in order to make a decision, we are strongly influenced by how that information is presented. For example, it has been shown that subjects are more likely to choose a given option if it is presented in terms of gains, rather than losses. Since humans are highly social creatures, and most real-world decisions are made within a social context, one would also expect social cues to influence decision making. To investigate how social and utilitarian cues interact during decision-making, we tested participants on a guessing game in which they were asked to choose one of two faces (one smiling, one angry or sad) on each trial. They were told that one face had a higher probability of a ’win’ than the other, and to try and determine through trial and error which face that was. After making the choice, participants were told whether they had won or lost on that trial, and this information was shown along with the face they had chosen. One face led to a win 40% of the time, the other 60%. After every block the probabilities changed, and the probabilities were balanced across facial expressions.
We compared participants’ responses to an ideal-observer model that predicted the optimal response on any given trial based on the evidence available. Participants’ performance deviated significantly from that of the ideal observer, demonstrating a bias towards selecting the smiling face. By fitting additional model parameters, we found that this bias consists of both a prior bias towards the smiling face, and a tendency to overweight positive outcomes associated with it. Positive outcomes associated with sad or angry faces were systematically underweighted.
These results are in accordance with previous imaging studies. Smiling faces have been shown to act as positive reinforcers, activating OFC, thought to represent the reward value of stimuli. Since lesions of OFC cause reliable deficits in extinction and reversal tasks, OFC has a clear role in guiding decision-making behaviour. Conversely, there is evidence that viewing sad or angry facial expressions elicits activity in ACC, an area associated with error detection. This suggests that sad or angry faces are taken as an indication of disapproval, encouraging a change in behaviour. By modifying our task for use with fMRI we hope to extend these findings. Specifically we will examine instances where a subject should have chosen an angry face (based on the available utilitarian evidence) but chooses a smiling face instead. Thus we hope to elucidate the neural basis of conflict between social and utilitarian cues in decision-making.