The Kaipher Editorial Team 
A University of Washington study published in the Proceedings of the National Academy of Sciences finds that X’s Community Notes feature reduces engagement with false information by nearly 50%. Researchers tracked 40,074 posts that received fact-checking notes between March and June 2023, finding that reposts fell by 46% and likes decreased by 44% once notes were attached (UW News).
Study Methodology and Key Findings
Assistant Professor Martin Saveski and doctoral student Isaac Slaughter employed synthetic control methods to compare engagement before and after Community Notes appeared. Of the 6,757 posts with notes deemed helpful and displayed publicly, they observed:
- 22% drop in replies
- 14% reduction in views
When considering each post’s entire lifespan—including activity prior to note attachment—the declines were 12% for reposts, 13% for likes, 7% for replies, and 6% for views (EurekAlert).
“This shows Community Notes are effective when attached, especially in reducing supportive engagement like reposts and likes,” Saveski said. The study also found that flagged content spread less within social networks, with users further from the original author much less likely to interact with it.
Factors Influencing Effectiveness
The researchers identified key factors that affect Community Notes’ impact:
- Content Type: Notes on altered media (fake photos/videos) yielded greater engagement reductions than those on text-only posts.
- Post Popularity: Highly viral posts saw larger drops in user interaction.
- Timing: Notes published within hours of the original post were significantly more effective; delays nearing 48 hours rendered notes nearly ineffective. “Content spreads rapidly on X, and if a note comes too late, few users will see it,” Slaughter explained (Phys.org).
Broader Implications and Platform Adoption
Following X’s lead, Meta and YouTube have introduced similar community-based fact-checking systems. However, researchers caution about the sustainability of multiple, separate fact-checking programs and whether user participation will be consistent across platforms, Saveski noted.
