Skip navigation
Please use this identifier to cite or link to this item: http://arks.princeton.edu/ark:/88435/dsp014b29b907g
Title: Networks, Collective Behavior, and Information in Social Systems—From Ant Colonies to Social Media
Authors: Tokita, Christopher Kai
Advisors: Tarnita, Corina E
Contributors: Ecology and Evolutionary Biology Department
Keywords: collective behavior
division of labor
political polarization
self-organization
social dynamics
social networks
Subjects: Ecology
Applied mathematics
Sociology
Issue Date: 2021
Publisher: Princeton, NJ : Princeton University
Abstract: From the ant hill and fish school to the modern city and social media platform, social systems are complex collectives that often display self-organization, the emergence of group-level organization from individual-level behavior and interactions. Thus, in social systems, order can emerge in a group without a leader and instead as a result of individuals using simple behavioral rules to react to their social neighbors and local environment. In this dissertation, I explore social systems through the rich interface between biology and social science: I study self-organization in both human and animal (i.e., non-human) social systems in an attempt to find mechanisms that could potentially organize societies broadly, regardless of species or specific system. To do this, I primarily use computational modeling—e.g., simulations of simple mathematical or agent-based models—to create predictions for how a particular individual-level behavioral rule would affect group-level organization. Either through collaboration or my own efforts, I also attempt to test some of these predictions with experiments or observational data collection. In the first chapter of my dissertation, I examine how a simple behavioral rule—response thresholds—could create rudimentary division of labor in simple ant colonies and provide the fitness benefits necessary to allow for the evolution of greater social complexity. In the second chapter, I extend the response threshold model to account for social interactions, allowing individuals to influence each other's behavior. When introducing social interactions, I borrow general social dynamics that are thought to cause opinion polarization in human socio-political systems, and thus I suggest a common social dynamic—the combination of social influence and interaction bias—that could organize behavior and social networks across social systems, including those that display division of labor, political polarization, or emergent personalities. In the third chapter, I turn to social dynamics related to political polarization and explore how individuals may adjust their social networks to avoid conflicting signals between their social connections and their preferred news source. Through both a computational model and analysis of social media data, I show that this simple behavioral rule can cause individuals to inadvertently create social network "echo chambers" when in a polarized information ecosystem—i.e., when news sources present very different coverage of the same topic. Finally, in the fourth chapter, I further explore the modern polarized information ecosystem and attempt to quantify how misinformation is spreading on social media. By analyzing Twitter data and conducting data-driven simulations, I estimate both the extent to which social media users may be seeing and/or believing misinformation and the extent to which platform-level interventions could reduce the impact of misinformation. Overall, this dissertation spans social systems and disciplines in the pursuit of interesting and perhaps universal dynamics that may organize complex societies.
URI: http://arks.princeton.edu/ark:/88435/dsp014b29b907g
Alternate format: The Mudd Manuscript Library retains one bound copy of each dissertation. Search for these copies in the library's main catalog: catalog.princeton.edu
Type of Material: Academic dissertations (Ph.D.)
Language: en
Appears in Collections:Ecology and Evolutionary Biology

Files in This Item:
File Description SizeFormat 
Tokita_princeton_0181D_13667.pdf26.85 MBAdobe PDFView/Download


Items in Dataspace are protected by copyright, with all rights reserved, unless otherwise indicated.