Fake News: Why Does it Persist and Who's Sharing it?

There are many features of fake news, in terms of content, its appeal, and its presentation, that make it more likely to influence people’s beliefs. One of the key features that contributes to its virality is often the novelty of the information it pushes to audiences. Vosoughi, Roy, and Aral (2018) studied over 126,000 rumor threads on Twitter (consisting of an original tweet and its retweets), to see how true and false news spreads differently. Information shared in the original tweets were then classified as true, false, or mixed. ^[8]

They saw that false news spread quicker, faster, and to a wider audience than true news. While the top 5% of false stories had well over 200 tweets about them, the top 5% of true stories had around 10. False news was 70% more likely to be retweeted than true news, and it took a sixth of the time it did true news to reach 1500 people. It also took 10 times longer for true news to be retweeted 10 times, than it did for falsehoods to be retweeted 19 times. ^[8]

To understand why this false news spread so quickly, Vosoughi’s team looked into novelty. They selected Twitter users and ran an algorithm comparing the content and topics of false tweets a user was exposed to with all the tweets the user was exposed to before. The algorithm determined that the content of false tweets was significantly different and more unique from previous tweets. After applying another algorithm to assess tweets for words associated with emotions, they found that false tweets were characterized by more feelings of surprise and disgust, which set the tone and appeal of these tweets apart from average news. ^[8]

This shows that fake news appears novel to users because its content is very different from that of tweets users are normally exposed to, and the language that conveys the fake news mirrors the sentiment of novelty, surprise, and shock. People may therefore feel compelled to share fake news because it contains information that they think others do not yet know. This was confirmed in a study by Pennycook et al., (2017) wherein people expressed more willingness to share novel news over familiar news.

Vosoughi et al. (2018) adds that novelty “facilitates decision making since it updates our understanding of the world”. New information provides useful insights that help us learn about a subject or form opinions on an issue, and this attracts users to share this information with others, again aiding fake news in becoming more viral.

Social media companies can incorporate novelty as a metric for flagging unreliable news by implementing algorithms that analyze content for words associated with feelings of surprise and disgust that characterize potential pieces of fake news. Social media sites could ask users “Are you sure?” when sharing news flagged with possibly unreliable content, thereby nudging users to verify content themselves before disseminating it.

As consumers, we should also be alert when coming across news that is unusually sensational or striking. Although that (unfortunately) comprises most of our news today, novelty signals that news could be unreliable — and should be read and shared with caution.

Novelty isn’t the only feature of fake news that may contribute to its abundance. The more people read or come across a piece of information, the more likely they are to believe that it is true. This idea is called the Illusory Truth Effect, a phenomenon explored by Pennycook et al. in 2017. They conducted an experiment with over 500 people, where they asked participants to score headlines based on their accuracy, incorporating headlines coming from actual articles posted to Facebook during the 2016 United States presidential elections. ^[6]

They showed participants different headlines throughout multiple sittings and invited participants back weeks after to perform the experiment again. Participants gave higher accuracy ratings to fake news headlines they already saw compared to those that they had just seen. Participants gave even higher accuracy ratings to news headlines they already saw when showed the statements again one week later. ^[6]

The Illusory Truth Effect operates on the principle that increased exposure to a statement increases one’s fluency and prior knowledge. Fluency is a person’s ability to mentally process statements, and the more they hear a statement, the faster they get at processing it. Prior knowledge of a statement also leads people to confirm the statement the next time they see it, a problem also known as confirmation bias. ^[6]

As shown by the Illusory Truth Effect, repeated exposure to an article, whether real or fake, increases people’s perceptions of its accuracy. Thus, social media companies can track the frequency of appearances of an article on a user’s timeline, and use this as a metric in identifying potential fake news.

Pennycook’s team discovered that the Illusory Truth Effect also applies to warnings. Adding disclaimers under headlines about potentially unreliable content was effective in lowering people’s perceptions of the accuracy of headlines, after seeing these warnings attached to headlines multiple times. This is a viable solution social media companies may adopt to preventing the spread of fake news.

As consumers, it is important to recognize our own biases and consume news consciously. It is easy to fall into the trap of confirmation bias, especially when numerous articles spring up at the height of a certain issue. By reading articles first and verifying the facts however, we can prevent ourselves from falling prey to fake news.

While others fear fake news has permeated social media and gripped the online community, Nelson and Taneja wanted to understand the scope and reach of fake news. Using data from comScore, they analyzed the online usage data of over 1 million US users to track their visits to both real and fake news websites. ^[5]

What they found was that there were far fewer people visiting fake news websites than real ones (e.g. CNN, The New York Times, Fox News). On average, a real news website receives over 40 times the number of monthly visits a fake news website does, and audiences spend twice the amount of time monthly on a real news website than a fake one. ^[5]

However, they saw that people who visited fake news websites generally spent more time on the Internet. This ties into the concept of audience availability — the time people have to use media such as the Internet. Despite all the choices available to us online, we are limited by the time we have to surf the web. When users have limited time, they tend to gravitate towards the most popular sites, leading to a “winner-take-all” effect. This is known as the Law of Double Jeopardy, which states that popular media channels will attract even more loyal audiences, while less known channels attract fewer audiences. ^[5]

Interestingly, the group also noticed that fake news websites were often accessed from a social media site, whereas real news websites were accessed more from web browsers, pointing to the key role of social media in the spread of fake news. ^[5] This shows that people actively seeking news visited real news sites, whereas those opening fake news often merely stumbled upon them through social media.

Social media platforms can look into ways to ensure that even high-use accounts receive verified content, to avoid the further spread of fake news. Seeing as social media are the main sites that refer viewers to fake news outlets, social media companies themselves must be more rigorous in fact-checking and ensuring that content their platforms share does not promote the proliferation of fake news. They can identify and block links to known fake news sites, and implement warnings as well as “Are you sure?” buttons when users share content as described above.

Although fake news may not be as pervasive as it seems, that does not make it any less of an issue. Fake news is a powerful tool for pushing false political agenda and mudslinging, and it hinders society from engaging in an insightful discussion when even the most basic of facts are being debated over. We must hold online platforms accountable for fact-checking and implementing measures to warn users about potentially malicious content, and we as consumers must also be aware of the content we are reading and sharing.