Social media and misinformation

Social media platforms (Facebook, Twitter, TikTok, YouTube, Reddit, etc.) have become dominant sources of news and information for billions of people. The structural features of these platforms—algorithmic ranking, engagement-driven feeds, ease of content creation, minimal editorial oversight, and ideological homophily—make them especially conducive to the production, distribution, and consumption of misinformation and fake news.

Key observations

Social media as the primary distribution channel for fake news:
Fake news sites derive approximately 80% of their traffic from social media, compared to only 10% for top news sites. This asymmetry reflects both the platform affordances (low barriers to entry, viral potential) and user behavior (preferential sharing of partisan content) Allcott & Gentzkov (2017)

Ideological segregation amplifies partisan effects:
Facebook's friend-network structure and algorithmic ranking create ideologically segregated information diets. Users are more likely to encounter and share fake news aligned with their political ideology, and ideological homophily moderates the effect Allcott & Gentzkov (2017)

False news spreads faster and farther than true news:
On Twitter, false claims reach more people, spread to greater depth, and achieve higher velocity than true claims. Novelty (not bots) is the primary driver Vosoughi, Roy, & Aral (2018)

Trust in traditional media declines as social media use increases:
There is a strong inverse relationship between the share of Americans who trust mainstream media and increases in social media usage over the past two decades Allcott & Gentzkov (2017)

Platform-specific dynamics

• Facebook: Largest source of fake news traffic; friend networks are highly politically segregated; algorithmic ranking rewards engagement, which favors emotionally charged false content
• Twitter: Rapid real-time diffusion; retweet cascades propagate false claims quickly; trending topics create information asymmetries
• TikTok: Algorithmic recommendation is opaque; short-form video is harder to fact-check; user base skews young
• YouTube: Long-form recommendations can create rabbit holes; algorithm may amplify extreme content; monetization incentivizes sensationalism

Related concepts

• Fake news — content definition and production
• Misinformation spread and diffusion — viral propagation mechanics
• Political bias — ideological homophily and filtering effects
• User profiles — user susceptibility and engagement patterns
• Credibility assessment — assessing platform and source trustworthiness

Key papers in this wiki

• Dou et al. (2021) — User Preference-aware Fake News Detection (UPFD): Models user endogenous preferences from Twitter historical posts alongside propagation graphs via GNNs. Shows confirmation bias drives news sharing on social media; jointly encoding user behavior and graph structure improves fake news detection by ~1% on Twitter-derived Politifact/Gossipcop benchmarks.
• Shu et al. (2017) — Fake News Detection on Social Media: A Data Mining Perspective — comprehensive survey of fake news on social media; discusses malicious accounts (bots, cyborgs, trolls), echo chamber effects, and both detection methods and characterization through social context features (user-based, post-based, network-based).
• Sharma et al. (2018) — Combating Fake News: A Survey on Identification and Mitigation Techniques — survey of detection methods and mitigation strategies on social media; discusses feedback-based detection exploiting social dynamics (propagation patterns, user engagement, temporal dynamics) and intervention strategies for different platform environments.
• Lazer et al. (2018) — The Science of Fake News — foundational multidisciplinary review documenting platform-level mechanisms of misinformation spread and algorithmic amplification via bots and selective exposure
• Nelson & Taneja (2018) — The small, disloyal fake news audience — applies audience measurement to show fake news reaches small, disloyal audiences; demonstrates that social media platforms (especially Facebook) drive most fake news traffic; shows audience availability is a key driver of misinformation exposure.
• Bail et al. (2018) — Exposure to opposing views on social media can increase political polarization — Field experiment on Twitter showing that algorithmic or deliberate exposure to opposing political views can exacerbate polarization; demonstrates partisan asymmetry (Republicans show backfire, Democrats do not).
• Treen, Williams & O'Neill (2020) — Online misinformation about climate change — analyzes social media platform characteristics amplifying misinformation: homophily, echo chambers, algorithmic bias; identifies influencers' echo chamber role in climate denial; describes feedback loops where users share misinformation back to producers; examines how platform affordances enable coordinated inauthentic behavior.
• Lukito (2019) — Coordinating a Multi-Platform Disinformation Campaign: Internet Research Agency Activity on Three U.S. Social Media Platforms, 2015 to 2017 — reveals how state actors exploit platform differences; Reddit as testing ground, Twitter as primary channel; demonstrates platform-specific messaging strategies in coordinated campaigns
• Linvill & Warren (2020) — Troll Factories: Manufacturing Specialized Disinformation on Twitter — production-side analysis of state-sponsored Twitter disinformation; shows Russia's IRA operated as specialized factory with five distinct account types, each exploiting platform affordances for different purposes
• Grinberg et al. (2019) — Fake news on Twitter during the 2016 U.S. presidential election — individual-level Twitter analysis; shows fake news as niche interest consumed by concentrated, older, conservative subpopulation; mainstream media dominated exposures across all political groups; fake news sources formed distinct network cluster.
• Guess, Nagler & Tucker (2019) — Less than you think: Prevalence and predictors of fake news dissemination on Facebook — linked survey to Facebook profile data; finds fake news sharing was rare (8.5% of users), with strongest predictor being age (65+ users shared 7× more than 18–29); robustly identifies digital literacy and cognitive aging as mechanisms.
• Allcott & Gentzkov (2017) — Social Media and Fake News in the 2016 Election — empirical evidence on fake news traffic, partisan asymmetry, and platform role
• Vosoughi, Roy, & Aral (2018) — The Spread of True and False News Online — Twitter-based study of diffusion mechanics
• Helmus et al. (2018) — How to Counter Russian Social Media Influence in Eastern Europe — analysis of state-sponsored social media campaigns; documents how platforms enable coordinated inauthentic behavior at scale
• Zhou & Zafarani (2020) — A Survey of Fake News — reviews detection methods applied to social media
• Fighting an Infodemic: COVID-19 Fake News Dataset — COVID-19 fake news dataset collected from social media (Twitter, Facebook) and fact-checking websites; 10,700 posts with binary labels; benchmarks ML detection methods
• Lu & Li (2020) — GCAN: Graph-aware Co-Attention Networks for Explainable Fake News Detection: Detects fake tweets by analyzing retweet propagation patterns; uses only source text and retweet user sequences (no comments, no explicit network structure); achieves 87.7–90.8% accuracy on Twitter15/16; provides interpretable explanations of suspicious users and linguistic markers, useful for platform intervention and moderation.
• Wilson & Wiysonge (2020) — Social media and vaccine hesitancy — cross-national study demonstrating social media's role in public health outcomes; shows both organic misinformation (user organization activity) and state-sponsored disinformation influence vaccination attitudes and behavior

Open challenges

• How do algorithmic changes (e.g., Facebook's 2018 pivot to "meaningful interactions") affect misinformation spread?
• What is the causal effect of exposure to misinformation on offline behavior (voting, protest, vaccine hesitancy)?
• How do platform-specific features (character limits, retweets, recommendation algorithms) shape misinformation ecology?
• What role do influencers, celebrities, and trusted accounts play in legitimizing false claims?
• How do bots, trolls, and coordinated inauthentic behavior interact with organic user behavior?