2025 Theses Doctoral

Towards a Richer Model of Social Perception: Evidence from Crowd Perception, Social Categorization, and Impression Dynamics

Chwe, John Andrew Hanju

The social world is wildly complex and multiply determined. In the present dissertation, I present a series of experiments and future directions that attempt to incorporate more of this complexity into how we think about and study social behavior.

Chapter 1 begins by studying the perception of groups, rather than individuals, demonstrating our perceptual ability to extract group-level estimates of high-level social information from quick glances of a group. These estimates occur in just 500ms of presentation and impact decision-making in a trust game.

In Chapter 2, I investigate biases in the categorization of Multiracial individuals, showing a reliance on more traditional monoracial categories. Using MouseTracking, I find that categorizations that ultimately end in a Multiracial judgement still show a systematic deviation towards more traditional monoracial categories.

In Chapter 3, we establish the neural basis of face-based impressions, illustrating two complementary systems working in tandem to represent social face-based information.

Finally in Chapter 4, we present preliminary results from a novel paradigm for studying impression formation, one that leverages natural language elicited by dynamic videos to explicitly study the timeseries-dependent nature of impression formation. I conclude with a synopsis of all work, theoretical contributions, and future directions.