The Most Effective Online Fact-Checkers May Be Your Peers, Study Suggests

Social media platform X’s community fact-checking system has proven remarkably effective at combating misinformation, according to a new study published in Information Systems Research. Researchers found that when community-generated correction notes appear beneath misleading posts, authors are 32 percent more likely to delete their content.

The study, conducted by researchers from the University of Rochester, University of Illinois Urbana–Champaign, and University of Virginia, challenges initial skepticism about X’s “crowdchecking” approach, formally known as Community Notes.

“Trying to define objectively what is misinformation and then removing that content is controversial and may even backfire,” explained co-author Huaxia Rui, the Xerox Professor of Information Systems and Technology at the University of Rochester’s Simon Business School. “In the long run, I think a better way for misleading posts to disappear is for the authors themselves to remove those posts.”

The research team analyzed 264,600 X posts that received community notes during two time periods – before a U.S. presidential election (June-August 2024) and two months after (January-February 2025). These periods were strategically chosen to capture both high and normal levels of misinformation circulation.

Community Notes operates through a threshold mechanism, where corrections must achieve a “helpfulness” score of at least 0.4 to appear publicly. Notes below this threshold remain visible only to contributors. This design created a natural experiment, allowing researchers to compare deletion rates between posts with public versus private notes.

The study’s methodology employed causal inference through regression discontinuity, comparing posts just above and below the visibility threshold to measure the direct effect of public exposure on author behavior.

“You worry that it’s going to hurt your online reputation if others find your information misleading,” said Rui, explaining why users delete their posts when they receive public correction notes.

This reputation effect proved especially powerful for verified users with blue check marks, who demonstrated greater concern for maintaining credibility. When faced with public correction, these influential users deleted their posts more quickly than average users.

The research also revealed that timing matters significantly. The faster a correction note appeared publicly, the sooner authors tended to retract their misleading posts. This finding highlights the importance of rapid response in combating misinformation, which typically spreads faster than corrections.

“For people to be willing to retract, it’s like admitting their mistakes or wrongdoing, which is difficult for anyone, especially in today’s super polarized environment with all its echo chambers,” Rui noted. Initially, the research team wondered if correction mechanisms might backfire by causing users to become defensive rather than receptive to feedback.

What makes Community Notes particularly effective is its emphasis on diversity of viewpoint. The system uses a “bridging algorithm” that prioritizes ratings from users who have disagreed in past evaluations. This approach prevents partisan manipulation by ensuring notes must be deemed helpful across the political spectrum to reach public visibility.

The study positions crowdchecking as a potentially powerful middle ground in the ongoing debate about content moderation. Rather than relying on centralized censorship or algorithmic filtering, it leverages collective judgment and peer accountability.

“Ultimately, the voluntary removal of misleading or false information is a more civic and possibly more sustainable way to resolve problems,” Rui concluded.

The findings arrive amid growing concerns about misinformation on social media platforms and debates about how to address it without infringing on free speech. X’s Community Notes model suggests that empowering users themselves may be more effective than top-down enforcement approaches.

By demonstrating that social pressure and reputation concerns can effectively combat misinformation, the study provides valuable insights for other platforms developing their own content moderation strategies in an increasingly polarized information landscape.