Schedule of Events | Symposia

A Game Theoretic Foundation for the Psychophysical Study of Social Interactions

Poster Session B - Sunday, March 30, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom

Vered Kurtz-David¹ (kurtzv02@nyu.edu), Adam Brandenburger^2,3,4, Paul Glimcher^1,5; ¹NYU Grossman School of Medicine, ²Stern School of Business, NYU, ³NYU, Tandon School of Engineering, ⁴NYU Shanghai, ⁵Center for Neural Science, New York University

Finite cognitive capacity limits social cognition. However, how do specific cognitive resources interact and limit social reasoning? Here, we investigate how humans reason in social encounters when forced to integrate across distinct cognitive domains. Combining cognitive neuroscience with paradigms from game theory we created an experimental protocol that decomposed each interaction into social and non-social cognitive demands. We used this separation to develop a new psychophysical model that rests on economic definitions of production functions. We posit that when one increases the cognitive demands in a social encounter, they incur a cognitive cost that is distributed amongst multiple domains. We recognize that individuals have distinct innate capacities in each of these domains, and allow for these capacities to trade-off against each other within an individual. We tested this approach in two studies. In the first, we presented subjects with an array of social interactions, which varied parametrically in their social and arithmetic cognitive demands. In the second, we manipulated the total cognitive resources available to subjects by varying processing times. Both social and arithmetic demands were revealed as contextual factors that determine the depth of the social analysis. Subjects traded-off these capacities in a constrained manner – with a great heterogeneity across individuals – as task demands varied. Our results extend classical psychophysics and theories of individual decision-making into the domain of social reasoning. We are currently applying this framework to a recently collected neuroimaging dataset, aiming for a network-level mapping of subjects’ utilization of social and non-social cortical resources.

Topic Area: THINKING: Reasoning