THE ROLE OF UNCERTAINTY IN SOCIAL COGNITION: INTEGRATING FMRI AND BEHAVIORAL APPROACHES
Social decisions and inferences involve high levels of uncertainty. We cannot directly observe others’ feelings, beliefs, and intentions. Additionally, we only have limited information regarding the past behavior and personality traits of a given person or what situational factors bear on their actions at a given moment. Therefore, in order to make social predictions, we need to rely on probabilistic inferences. In this dissertation, I examine the role of uncertainty in social inferences, integrating fMRI and behavioral approaches. In Study 1, I examine whether social tasks commonly used in the social cognition literature involve higher levels of uncertainty than nonsocial comparison tasks and whether this difference in uncertainty can account for some of the neural findings that came out of these studies. This investigation showed that i) social tasks used in the literature are perceived to be more uncertain than nonsocial control tasks, and ii) the DMPFC, a key brain region that is consistently shown to be more activated under social tasks, might show a domain-general response to uncertainty instead of a domain-specific response to social information processing. In Study 2, I empirically tested this idea by designing an fMRI task where participants made mental and nonmental inferences under varying levels of uncertainty. The results confirmed the hypothesis that the DMPFC responds to uncertainty in both social and nonsocial inferences, suggesting a domain-general role for this brain region in processing and/or reducing uncertainty. In Study 3, I focused on how people update their impressions of others when their predictions about them turn out to be incorrect. Specifically, I investigated whether the presence of situational factors that are congruent (versus incongruent) with unexpected observations can lead to lower levels of impression updating. The results showed that individuals can indeed differentiate between situational factors that are congruent and incongruent with their observations, and guide how much they update their impressions of someone accordingly. Together, this work helps us better understand the role of uncertainty in social inferences and can have implications for our understanding of the cognitive architecture of the social brain and of disorders of social function.