Computational Social Science Working Group
The CSS Working Group uses computational and mathematical approaches to investigate social behavior. We have two main goals. First, we are interested in the empirical analysis of social phenomenon in core areas of interest. My own lie in social media and politics, and am looking for undergraduates interested in:
Elections research: Visual communication dynamics on Facebook, Instagram, and TikTok.
Game-theoretic modeling of social behavior.
GitHub collaboration dynamics.
Second, we work on the development of methodological tools to aid analysis, such as deep learning for image analysis, social network analysis, ChatGPT & LLM workstreams, simulation, and integrating journalistic and creative approaches.
The group is growing and I'm accepting applications for a post-doc and Ph.D student in the 2023 cycle. If you are interested in social media, modeling/machine learning, or AI & game theory, please get in touch at email@example.com with your C.V.
I'm excited to work in tandem with students to develop their core social science areas through student-led projects, whether it's in politics, public policy, AI, or sociology. We can always find the right data and right computational tools for each project, with the right research questions.
I am also committed to helping students succeed later in their careers, whether it is for further study or transition to industry.
Yu (Sunny) Fang is a data science and political science student at Barnard College, Columbia University. Her research employs computational methods to study democratic values, such as media portrayal of geopolitics and online discourse about Taiwan. Previously, she has written about the legal and global implications of Taiwan's semiconductor dominance in the Columbia Undergraduate Law Review and the Asia-Pacific Affairs Council Journal of Columbia. She hopes to further utilize computational methods to explore the interdependence of technology and society.
Joanna Olagundoye ' 24
Joanna Olagundoye is a Senior at Dartmouth College where she double majors in Quantitative Social Science (QSS) and Astronomy. Within QSS she focuses on building a firm understanding in economics, enjoying the topic of game theory learning how businesses and markets run. But the major has also offered her the opportunity to delve into a more personal side of the economy – how consumer thoughts and actions affect the numbers and trends that economists study. In her honours thesis in QSS she aims to examine economy shifts to the perceptions of consumers.
Sean Noh '28
I'm from Virginia and will be attending Dartmouth next year to study Computer Science, and I'm also interested in education, QSS, and psychology. I hope to develop educational technology and learn application development. I love playing tabletop games, I am an Eagle Scout, and I recently published a habit-tracking app, Cogo.
Sam Winchester ' 24
Sam Winchester is majoring in Quantitative Social Science with a minor in Sustainability. At Dartmouth, he is a member of the Dartmouth Ski Patroller and of the Committee of Standards. He previously served as the Director of Strategy at The Dartmouth and as a First Year Fellow and Researcher at the Rockefeller Center for Public Policy. He is interested in social networks, sustainability, sports analytics, and government-sponsored aid programs, and is currently completing a Senior Thesis on the impact of social media sentiment on the performance of professional athletes.
Datasets and Research Streams
These are existing streams of research, based on existing databases. We have capabilities to scrape TikTok, Instagram, and YouTube. Students' own interests, questions, and data sources are encouraged. For internal tools, please consult our Slack channel.
COVID-19 Twitter Dataset
The largest public Twitter dataset for COVID-19 (n > 4 billion). Can be used to analyze discourse and specific topics of interest. Past examples include politics, K-pop, and public health.
Instagram & the Floyd Protests
Dataset of more than 1.59 million public Instagram posts during the #BLM2020 protests following the murder of Mr. George Floyd. This stream combines social justice and visual machine learning. *Seeking RAs.
News & Media
For students interested in journalism, media framing, and partisan bias, there are rich opportunities to cultivate your own area of research. Combines NLP, LLM and ChatGPT workstreams, and partisan bias. Examples include Gender in the 2020 Democratic Primaries and Debiasing AI from LGBTQ+ Bias.
Multimodal dataset across YouTube, Instagram, TikTok, and Twitch of tobacco and alcohol abuse. Entry points in deep learning, social network analysis, and surveys.