Neural mechanisms of predicting individual preferences based on group membership

Contains fulltext : 224793.pdf (preprint version ) (Open Access)
Successful social interaction requires humans to predict others' behavior. To do so, internal models of others are generated based on previous observations. When predicting others' preferences for objects, for example, observations are made at an individual level (5-year old Rosie often chooses a pencil), or at a group level (kids often choose pencils). But previous research has focused either on already established group knowledge, i.e., stereotypes, or on the neural correlates of predicting traits and preferences of individuals. We identified the neural mechanisms underlying predicting individual behavior based on learned group knowledge using fMRI. We show that applying learned group knowledge hinges on both a network of regions commonly referred to as the mentalizing network, and a network of regions implicated in representing social-knowledge. Additionally, we provide evidence for the presence of a gradient in the posterior temporal cortex and the medial frontal cortex, catering to different functions while applying learned group knowledge. This process is characterized by an increased connectivity between medial prefrontal cortex and other mentalizing network regions, and increased connectivity between anterior temporal lobe and other social-knowledge regions. Our study provides insights into the neural mechanisms underlying the application of learned group knowledge.