Conspiracy theorists have burned down mobile phone towers in the United Kingdom, the Netherlands, Canada and other countries in recent weeks. Why? There’s a rumor circulating on social media that 5G technology is causing the global covid-19 pandemic — a rumor without any evidence, which has been debunked by the WHO and prominent fact-checkers.

Misinformation is always a problem. Misinformation about breaking news, like the coronavirus, is particularly problematic. Sharing misinformation about covid-19 is widespread — a white paper we recently published shows that people shared links on Twitter to misinformation sites almost as often as links to credible health websites like the WHO.

Misinformation isn’t going anywhere — but our research, described below, suggests one solution. Experts, social media platforms and social media users can correct others when they see misinformation on social media.

Our recent survey suggests that many people are doing just that and recognizing others are, too. Attitudes about this kind of correction suggest that people generally think we could all be doing even more.

How people tackle misinformation on social media

People often blame social media for propagating misinformation, but these channels also offer an opportunity to mitigate the problem. Our work, and the work of others, demonstrates that observational correction — correction that occurs on social media where people can observe other people being corrected — reduces misperceptions.

Observational correction works because of the nature of social media. Correcting a friend, family member or stranger in person, on email, or in some other relatively private communication may set that person straight about the facts on that topic. But doing the same on social media means that dozens of people — and perhaps thousands — can witness the correction.

In addition to the sheer scale that social media networks offer, other structural elements may improve correction efficacy. First, the correction occurs in proximity (temporally and spatially) to the original misinformation, increasing the likelihood that people hadn’t had a chance to absorb the misinformation at all. Second, witnessing someone else being corrected may be less threatening than being corrected directly, but with all the same benefits.

Expert correction is particularly effective. When highly trusted groups like the CDC directly respond to users sharing misinformation on social media, people are likely to believe the correction.

Correction can also come from social media platforms themselves. For example, Facebook uses its “related articles” function to display debunking information from third-party fact-checkers. Our work shows that exposure to this type of information reduces public misperceptions.

Additionally, social media users can correct one another. Such corrections are most effective when users link to a highly credible expert source, and when multiple users offer a correction.

Do people actually do this?

Yes, our March 2020 survey suggests that Americans are experiencing quite a bit of correction on social media during the covid-19 pandemic.

We surveyed 1,072 participants in the United States from March 27 to March 29, using the Lucid Academic panel (an online non-probability panel that tries to assemble participant pools that resemble the demographics of the United States), then weighted the data using raking weights on gender, race, education, age, partisanship and ideology — essentially trying to make the data look more like the population of the United States. We asked a series of questions about how often people witnessed or engaged in correction, and what they thought about others doing so.

Here’s what we found. In the previous week, 34 percent of people reported seeing someone else being told they shared misinformation regarding covid-19 on social media, and 23 percent reported having told someone they shared misinformation on the topic.

Older adults were less likely to report seeing someone else being corrected or performing a correction themselves, while more educated people were more likely to see and engage in corrections on social media. Research similarly shows that older people struggle to identify misinformation, and are more likely to share it. Likewise, more educated respondents may have the digital literacy skills — or the confidence — to search for, locate and share corrective information.

The frequency of seeing or performing these corrections did not differ by party or ideology. Many people perceive social media as an echo chamber, but research generally finds this is not true. The fact that people across the ideological spectrum are witnessing and engaging in correction regarding covid-19 information is further evidence along those lines.

People value these corrections

So what do people think about the correction behaviors we documented?

A majority of Americans agree that correction is important — 56 percent of respondents agreed at least somewhat that they like it when people correct others on social media. Even more people find this move appropriate: 68 percent agree people should respond when they see someone sharing misinformation, and 67 percent agree that addressing misinformation on social media is everyone’s responsibility.

We also found that this approval is broadly shared. Support for observational correction does not differ by gender, education or partisanship. Older people tend to have more favorable attitudes toward observational correction — even though they notice it less, and tend to do it less frequently.

The public does recognize some risks in responding to misinformation in this way, with 48 percent agreeing that responding to misinformation on social media creates confusion, and 53 percent agreeing that it encourages trolling.

Research-based correction on social media

So when witnessing misinformation on social media, what does research suggest might be an effective response?

First, providing credible information that the misinformation is incorrect — and offering facts in response — is likely to be the most effective solution. A fact check from an independent journalistic organization or information from a credible organization like the American Medical Association is particularly effective.

Making these types of corrections even if someone else has already done so reinforces the true message to those who see it.

Second, it’s also important to note that people generally agree that correction is appropriate, and that it’s a shared responsibility. This may make people on social media more comfortable with correcting others, and more likely to engage in it more often.

Leticia Bode (@leticiabode) is a Provost’s Distinguished Associate Professor in the Communication, Culture, and Technology program at Georgetown University.

Emily Vraga (@ekvraga) is an Associate Professor in the Hubbard School of Journalism and Mass Communication at the University of Minnesota, where she holds the Don and Carole Larson Professorship in Health Communication.