The neural correlates of social and non-social decision-making in resolving complex, dynamic uncertainty

Understanding how people form appraisals under uncertainty is a foundational challenge for social cognition research. Naturalistic stimuli are a powerful tool for modeling these processes, but the cognitive and neural dynamics that govern real-time appraisal during complex narratives remain underexplored. Across three studies, this dissertation investigated how individuals dynamically appraise social and non-social ambiguity using a 45-minute episode of a murder mystery drama. In the first fMRI study, participants continuously rated their certainty regarding a primary character’s guilt. Neural synchrony during this appraisal task was examined via inter-subject correlation and sliding window approaches, revealing temporally dynamic, content-sensitive neural alignment in social-cognitive networks. The second study extended this design behaviorally, examining the structure and variability of hypothesis generation across individuals. The third study manipulated appraisals directly, instructing participants to assess the informativeness of scenes in relation to a guilt hypothesis. Integrating these findings, the dissertation demonstrates that stimulus complexity and interpretive framing jointly modulate appraisal synchrony, and that these processes are supported by distributed regions in the default mode, salience, and attentional networks. These findings advance our understanding of how people resolve social ambiguity in real time and highlight the value of continuous self-report and dynamic brain-behavior alignment in naturalistic settings. Implications for social inference and real-world decision-making are discussed.