Twitter

Twitter (now X) is a social media platform where users post short messages ("tweets") and share content. Research on misinformation, polarization, and information behavior frequently uses Twitter data due to its public API and prominence in political and public discourse.

Key papers

• A large-scale COVID-19 Twitter chatter dataset for open scientific research - an international collaboration — Public resource dataset of 800+ million COVID-19 tweets with multilingual coverage, clean and full versions, and processing pipelines; demonstrates large-scale open Twitter research infrastructure supporting misinformation detection and sentiment analysis
• Zannettou et al. (2019) — Characterizing the Use of Images in State-Sponsored Information Warfare Operations by Russian Trolls on Twitter — large-scale analysis of 1.8M images posted by Russian state-sponsored accounts; cross-platform influence analysis of image vs. URL sharing; political imagery 2× more influential
• Zannettou et al. (2018) — Who Let The Trolls Out? Towards Understanding State-Sponsored Trolls — comparative analysis of Russian and Iranian state-sponsored troll operations on Twitter; documents platform-specific tactics, influence measurement via URL amplification, and temporal dynamics
• Garimella et al. (2017) — Quantifying Controversy on Social Media — Develops graph-based methods for measuring controversy on Twitter from conversation networks (retweets, follows, mentions); tests on 20 Twitter topics and validates against external datasets.
• Ferrara et al. (2015) — The Rise of Social Bots — Survey of social bot phenomenon on Twitter; taxonomy of detection approaches (network, crowd-sourced, feature-based); analysis of behavioral features distinguishing bots from humans
• Cinelli et al. (2021) — The echo chamber effect on social media — Demonstrates that Twitter's interaction model (retweets, mentions) combined with algorithmic curation produces homophilic clustering and biased information diffusion; users with similar political leanings preferentially follow each other and receive information from like-minded peers
• Eady et al. (2019) — How Many People Live in Political Bubbles on Social Media? — Surveys 1,496 Americans and links to 642K Twitter accounts and 1.2B tweets; measures extent of political bubbles and filter effects; finds substantial cross-ideological exposure in accounts followed and retweets, challenging filter bubble determinism
• Bail et al. (2018) — Exposure to opposing views on social media can increase political polarization — Large-scale field experiment using Twitter bots to test whether exposure to opposing political content shapes attitudes; finds backfire effects for Republicans
• Castillo et al. (2011) — Information Credibility on Twitter — foundational study assessing news credibility on Twitter; identifies user reputation (registration age, followers, activity) and propagation patterns (retweet tree structure) as key credibility signals; achieves 86% accuracy on binary credibility classification
• Golovchenko et al. (2020) — Cross-Platform State Propaganda: Russian Trolls on Twitter and YouTube during the 2016 U.S. Presidential Election — analysis of IRA hyperlink-sharing behavior on Twitter; shows strategic targeting of conservative news media and use of pre-propaganda strategy
• Linvill & Warren (2020) — Troll Factories: Manufacturing Specialized Disinformation on Twitter — large-scale analysis of Russia's Internet Research Agency operations; identifies five account types with distinct behavioral signatures deployed during 2016 election
• Grinberg et al. (2019) — Fake news on Twitter during the 2016 U.S. presidential election — individual-level analysis linking Twitter behavior to voter registration data
• Shared partisanship dramatically increases social tie formation in a Twitter field experiment — Field experiment showing partisan preference as a driver of social tie formation on Twitter
• Cognitive reflection correlates with behavior on Twitter — Links individual differences in cognitive reflection to real-world Twitter behavior, including accounts followed and news sources shared
• Sahly et al. (2019) — Social Media for Political Campaigns: An Examination of Trump's and Clinton's Frame Building and Its Effect on Audience Engagement — Content analysis of 3,805 Trump tweets and 655 Clinton tweets during 2016 campaign; shows conflict and morality frames drove retweets, while emotional frames had candidate-specific effects