Study finds fake news spread by social media users not bots
Researchers at the Massachusetts Institute of Technology have had a report published with Science Magazine called “The spread of true and false news online”. The study used a data set of rumor cascades taken from Twitter between 2006 and 2017. Around 126,000 news stories that had been tweeted by about 3 million people over 4.5 million tweets were analyzed by the team who used six independent fact-checking organizations (snopes.com, politifact.com, factcheck.org, truthorfiction.com, hoax-slayer.com, and urbanlegends.about.com) to verify the shared information. Their investigative work led them to this statement:
When we estimated a model of the likelihood of retweeting, we found that falsehoods were 70% more likely to be retweeted than the truth [...]
It appears the novelty of fake news makes it more attractive to share: Often fake news is considered to be shocking, dramatic, sensational, or controversial. Fake news can be extremely damaging, as demonstrated by an AP tweet in 2015 which stated Barack Obama had been injured in an explosion. The tweet is believed to have caused US$130 billion to have been wiped from the stock market before an AP representative reported that their account had been hacked.
Fake news is classified as yellow journalism, which can be loosely defined as items using headlines to attract readers but offering little factual research in the text, akin to the abundant clickbait articles found on social media sites. There have been suggestions that the proliferation of fake news on social media is caused by bots; however, the study also came to the conclusion that bots share true and false news at the same rate, and that the real reason behind the spread and popularity of fake news is human behavior.