Twitter users retweet fake news almost twice as much as real news

President Obama was not injured in a White House explosion in April 2013. The explosion never happened. But according to users on Twitter, it did. 

A fake news tweet had been sent out from a hijacked Associated Press account. The news spread rapidly, so much so, that the Dow Jones Industrial Average plummeted by 100 points.

The retweet is powerful. It facilitates the rapid spread of news — whether a story is completely bogus or real. And according to MIT research published Thursday in the journal Science, false news is 70 percent more likely to be retweeted than real news stories.

“I think these results are troubling,” said Sinan Aral, a professor at MIT’s Institute for Data, Systems, and Society and co-author of the study, in an interview.

“There are real world and potentially negative consequences if decisions are going to be made based off falsity,” said Aral. 

The MIT researchers call this news “false” as opposed to “fake” because “fake news” is a term used more leniently, Aral explained. Fake news, he said, isn’t always based on the independently-checked veracity of the news itself. 

“Politicians use the term fake news to describe stories they just don’t like,” he said.

Although bots certainly play a role in disseminating completely phony news — including the fictitious story that Hillary Clinton ran a sex-trafficking ring in the basement of a pizza shop — bots weren’t responsible for the dramatic disparity between retweets of false versus real news. The researchers found that bots spread fake and real news in roughly equal proportion. “So the massive differences cannot be because of bots,” Aral said. 

The MIT team spent two years studying the role Twitter plays in spreading false news worldwide. They examined around 126,000 stories that had been tweeted out by some 3 million people globally. And to determine if a story was real or fake, they verified the stories’ veracity by using six independent fact-checking groups, including politifact.com, snopes.org, and factcheck.org.

Image: The Washington Post/Getty Images

Social media sites — not exclusively Twitter — allow fake or false news to be “weaponized” in a new way, according to Tim Weninger, a computer science engineer who researches network science and social media at the University of Notre Dame, in an interview. Prior to the internet-age, when most news was printed, professional news editors made decisions about what news should be published.  

Now, we do — often rashly.

Retweeting or liking a tweet is similar to voting for it, making it more likely to be seen (and retweeted again) by others.

“The consumer now decides what’s important,” said Weninger, who was not part of the MIT study. “Editors don’t make that decision.”

Exacerbating matters, people often just read headlines before sharing news stories, rather than opening up the story to see whether or not it’s completely fabricated nonsense. 

In a study published last year, Weninger found that 70 percent of Reddit users “upvoted” news posts without bothering to read them. 

“If you can design headlines that gathers eyeballs rather than purveys truth — that is the weaponization of news,” said Weninger.

In the MIT study, the fake news tended to be novel, meaning news that Twitter users hadn’t encountered before. This novelty is problematic in the Twitter realm because humans everywhere are already poor at discerning between truth and falsehood.

“They’re horrible at it,” said Andrew Butler, an associate professor of psychological and brain sciences at Washington University in St. Louis who played no role in the study, in an interview.

“Even when people have knowledge that directly contradicts false information, they fail to detect that it is false information,” said Butler. “It’s very powerful.”

In a way, we’re also naturally gullible.

“People have a bias to assume truth,” said Butler, noting that much of what we experience in our daily lives is actually happening. The train arrives (around) when it should. Government weather forecasters accurately predict the arrival of powerful nor’easters.

“A lot of what you’re experiencing is true,” said Butler. 

Getting most people to be better judges of the veracity of their news before hastily sharing it doesn’t have a simple solution. Twitter may make efforts to stomp out fake-news disseminating bots — but it can’t alter our rash behavior. 

“News consumers ought to be more careful with the news that they rate, like, or retweet,” said Weninger. “They should actually give [it] proper thought and consideration.”