Keng-Chi Chang

Keng-Chi Chang

Hello. I am a M.S. & Ph.D. Candidate at UC San Diego. I will be joining the Program in Quantitative Social Science at Dartmouth College as an Assistant Professor in July 2025.

My research interests are centered on computational social science, motivated by a keen interest in exploring the impact of emerging technologies and media ecosystems on group dynamics in politics through the lens of social networks and the dissemination of accurate or false information.

My research is driven by the following questions: (1) How can we leverage computing and large-scale empirical data to learn about fundamental political and societal challenges? (2) What is the role of these technologies in altering behaviors related to news consumption, democratic elections, and authoritarian dissent? (3) What new research methods can we develop while addressing these questions?

My dissertation research has been funded by NSF/APSA and the Rapoport Family Foundation. My recent works have been published in the Proceedings of the National Academy of Sciences, Journal of Communication, PLOS One, and Journal of Quantitative Description: Digital Media, as well as peer-reviewed Computer Science conferences.

Interests

  • Computational Social Science
  • Political Methodology
  • Digital Politics
  • Image as Data

Education

  • PhD Candidate in Political Science

    University of California, San Diego

  • MS Candidate in Computer Science

    University of California, San Diego

  • BA in Economics

    National Taiwan University

Research

The Importance of Prompt Tuning for Automated Neuron Explanations

The Importance of Prompt Tuning for Automated Neuron Explanations

  • Study neuron-level explanability by prompting LLMs with activated patterns.
  • Evaluate different prompting methods with both machine & humans.
Do Imageries Lend Credibility to News Articles?

Do Imageries Lend Credibility to News Articles?

  • Varies image treatment while holding fixed other aspects of news articles.
  • Overall presence of image does not always increase credibility perception.
  • Images with some identified latent treatments (such as photos from press conferences, comics, or visuals of male suits) can alter credibility perception.
Characterizing Image Sharing Behaviors in US Politically Engaged, Random, and Demographic Audience Segments

Characterizing Image Sharing Behaviors in US Politically Engaged, Random, and Demographic Audience Segments

  • Collect images shared by political & random Twitter users.
  • Study how sharing different types of images are predictive of demographic attributes.
Compensation and the Consolidation of Authoritarian Power: Evidence from China’s 2016 PLA Reform

Compensation and the Consolidation of Authoritarian Power: Evidence from China’s 2016 PLA Reform

  • Collect large-scale biographical panel data of PLA officers.
  • Estimate patterns for promotion within PLA before/after PLA reform.
  • Finds that Xi only started to promote followers after power consolidation.
Mapping Visual Themes among Authentic and Coordinated Memes

Mapping Visual Themes among Authentic and Coordinated Memes

  • Collect coordinated IRA memes from Twitter and authentic memes from Reddit.
  • Cluster memes to find visual themes using self-supervised transfer learning.
  • Coordinated and authentic memes share visual themes but with different emphasis.
COVID-19 Increased Censorship Circumvention And Access To Sensitive Topics In China

COVID-19 Increased Censorship Circumvention And Access To Sensitive Topics In China

  • Use geolocated Tweets from China during the COVID-19 crisis.
  • Show that crisis motivates citizens to seek out sensitive information.
  • Gateway to both current and historically sensitive content is not found in countries without extensive online censorship.
The Effect of Streaming Chat on Perceptions of Debates

The Effect of Streaming Chat on Perceptions of Debates

  • Asks whether social chats on livestreams affect debate viewing.
  • Assign subjects to watch debates on ABC, Facebook Live, and 538.
  • Democratic subjects assigned to the Facebook chat condition reported lower affect towards Democrats and a worse viewing experience.
  • The tone of candidate-directed comments also matter.
Using Facebook Data to Predict the 2016 U.S. Presidential Election

Using Facebook Data to Predict the 2016 U.S. Presidential Election

  • Use 19 billion likes to measure dynamic ideological positions of users and fan pages.
  • Guess user’s geolocation by likes and measure state level support rates for candidates.
  • Assume that users would support candidates with closer ideology.
  • FB support rates predict election outcome well, and share similar trends with polls.
  • Polls systematically overestimate Clinton’s support in right-leaning states.
Connective Effervescence and Streaming Chat During Political Debates

Connective Effervescence and Streaming Chat During Political Debates

  • Collect streaming chats of US Presidential debates on Facebook Live.
  • Describe their dynamics, degrees of toxicity, and insults.
Social Mobility in Ming China: Evidence from Twelve Thousand Chin-shih Data

Social Mobility in Ming China: Evidence from Twelve Thousand Chin-shih Data

  • Challenge the conventional wisdom that civil service exam was meritocratic.
  • Use detailed data of past 3 generations to estimate the effect of family background on tests.
  • Multi-staged examination structure provides chances to partially control for ability.