How to read a tweet to determine whether it’s fake
What if you read a news article you thought was fake, only to realize it was just a photo?
If you’re like most people, you probably think of Twitter as a place where you can tweet to find out if a tweet is a hoax.
But a new study published this week suggests that it’s much more than that.
“A lot of people get a lot of misinformation out of it,” says Jessica Muehlhauer, a social psychologist at Michigan State University.
“But this study shows how false information can spread.”
The study, published in the journal PLOS ONE, involved analyzing tweets sent between April 26 and May 5, 2016, between users in the United States and China.
The study found that, on average, the more frequently users posted false tweets, the higher the probability that the tweet was fake.
The researchers also found that the more often users shared the false tweet, the lower the likelihood that the fake tweet was verified.
What’s more, the researchers found that when users shared false tweets to their followers, their followers were more likely to believe the tweets, while their own followers were less likely to do so.
This could be because followers of the fake users tended to trust the fake tweets more.
“What this shows is that there’s this relationship between trust and misinformation,” Muehauer says.
“In other words, when people share misinformation, it’s more likely they’ll believe the misinformation, and it will spread.”
To investigate this, the study researchers took the tweets and divided them into three categories: tweets that were verified, tweets that didn’t have verified verification, and tweets that did have verified authenticity.
The final group was just tweets that had been retweeted.
To determine the proportion of verified tweets, they calculated the number of verified verified tweets in each category.
Then they took a look at how often each group of tweets was shared on Twitter.
They then compared that share of tweets with the share of verified tweet.
The more often they shared fake tweets, Muehrauer says, the greater the likelihood they were verified.
“There was a very high proportion of the tweets that people share that are not verified that are really not verified,” she says.
That’s because the group of verified authentic tweets that went viral was the most common group of fake tweets.
And when people shared fake verified tweets to people who shared verified authentic ones, the group was the least likely to trust it.
Muechauer says it’s important to note that the study didn’t look at tweets from real users.
That was done by using an algorithm, which uses automated detection of tweets to determine if a particular user is tweeting from a particular account.
So the real-world data that was used for this study didn, in fact, reflect the people that were tweeting.
“We could have taken all of the real tweets and looked at them separately,” Mueshauer adds.
“If we had looked at only tweets that came from real people, that would have shown that this is actually an extremely large proportion of fake content.”
Mueachers study also focused on how fake content spreads online.
“Social media can amplify misinformation and misinformation can spread to a much larger audience than the number that is exposed,” Muthhlauer says of her study.
The takeaway is that “the more fake content you see on the Internet, the less likely it is that people will be skeptical of it and the more likely it will be accepted,” she adds.
The real takeaway, Mueschauer hopes, is that when you’re trying to find something to believe in, look at the source of it.
“That way, if you have a tweet that’s just wrong, you’ll be able to debunk it and make sense of it because you’ve looked at the context.”
For more stories like this, subscribe to our weekly newsletter.