This will come as no surprise, but if you’re looking for the capital-T truth, do not try to find it on Twitter. At least not during disasters, according to a study at the University at Buffalo.
We need not look further than Hurricane Sandy in 2012 for the wildest, most ridiculously menacing examples of the phenomenon: Massive storm approaches nation’s most populous corridor, people freak out, anonymous Twitter users prey on their fear for kicks.
Twitter user @ComfortablySmug — unmasked as a GOP operative and hedge-fund analyst — was Sandy’s greatest villain, the Atlantic wrote:
While people tried to make sense of a deluge of incoming Sandy news, they might have heard CNN report that the floor of the New York Stock Exchange was flooded three feet deep. Maybe they read a Reuters brief about Con Edison employees being trapped in a power station. Maybe they scrolled past a panic-stricken all caps retweet announcing a preemptive power shut down throughout Manhattan. None of those reports were correct, and they can all be traced back to Comfortably Smug’s anonymous account, cultivated around the persona of a cocky, Romney-supporting 1-percenter.
Those might be strong words, but the chaos Shashank Tripathi — Smug’s real name — generated was very real in a moment when people in New York City had every reason to be frightened.
Nearly six years later, what have we learned? Little to nothing. Tripathi has amassed nearly 80,000 followers, many of whom are some of politics’ and journalism’s biggest influencers. And a study found that we’re still really bad at detecting lies on Twitter.
“To the best of our knowledge, this is the first study to investigate how apt Twitter users are at debunking falsehoods during disasters,” said Jun Zhuang, associate professor of industrial and systems engineering at the University at Buffalo and the lead author of the study. “Unfortunately, the results paint a less-than-flattering picture.”
The University at Buffalo researchers looked at three possible responses users could have when confronted with false information — they could spread it further, seek to confirm it or cast doubt on it. Guess which one we do most often?:
- 86 to 91 percent of the users spread the false news by retweeting or liking.
- 5 to 9 percent sought to confirm the false news, typically by retweeting and asking whether the information was correct.
- 1 to 9 percent expressed doubt, often by saying the original tweet was not accurate.
It gets better (and by better I mean worse). Even after the false information had been proved false:
- Fewer than 10 percent of users who spread the false information deleted the tweet.
- Fewer than 20 percent of those users corrected the false tweet with a new one.
This is not great, people. Especially during disasters, we need to keep our heads. You might think false information — even on something as “trivial” as Twitter — is not going to make a difference, but emergency managers depend on social media to find out who is actually in trouble. If rescuers are going to a flooded neighborhood that isn’t actually flooded, that’s diverting resources away from the people who actually need help.
Lies on social media are dangerous, and it’s not just the liar who’s to blame — according to this study, it’s all of us.