Research OVERVIEW

How do we learn who to trust, how to cooperate, when to reciprocate, and what to do when we have been treated unfairly? The lab seeks to disentangle the cognitive, affective, and neural processes behind the complex choices that form the basis of human social behavior, whether during one-on-one interactions, or amongst the multitude of relationships that comprise our complex social communities. We ask questions that span levels of analysis, from representation to choice to planning. The lab’s research focuses on two main questions, broadly construed:

How do humans learn about their social world to behave adaptively?

The social world is rife with uncertainty. Given that we do not have access to the inner workings of others’ minds, we must constantly assess how to behave with another person given very limited information (e.g., is our new friend trustworthy?). This problem becomes even harder when we consider the fact that all of our social lives operate against the backdrop of large, evolving, and complicated social networks, where our actions can reverberate far beyond a single interaction with a friend. The lab tries to understand how the mind and brain represent information about other people, how that information is leveraged to trust or to cooperate with the right people, and how these processes can go awry, leading to suboptimal behaviors. We tackle these questions about representation, learning, and behavior using a variety of testbeds. From trust and cooperation, to political polarization, we measure both individual interactions and the most complex social environments, our social networks.

Topics include: Altruism, anxiety, cognitive maps, computational modeling, cooperation, generative replay, information transmission, latent structure learning, naturalistic contexts, neuroimaging, political polarization, politics, punishment, structure of social networks, trust

How do emotions bias social learning and behavior?

At the heart of every social thought, choice, or interaction is an emotion. A long tradition of affective science argues that emotion contributes to both the representation and computation of value. The lab extends these ideas by leveraging various tools, from machine learning to neuroimaging, to better understand how emotion and affect drive how we represent and interact with others in our social world. Most recently, we have argued that affective error signals likely play a critical role in shaping social behavior, helping us to learn efficiently and therefore behave adaptively. We are just now tackling the question of how the brain encodes affective prediction errors to efficiently compute social value.

Topics include: Affect, depression, EEG, emotion, empathy, learning, machine learning, physiological measurements, reward, social value