Fake News Exploits Our Obliviousness to Proper Sourcing

In the age of social media, we fall short in evaluating the sources of our news.

Posted Jul 29, 2018

Jonas Pereira/Wikimedia Commons
Source: Jonas Pereira/Wikimedia Commons

Some news stories spread better than others, and what differentiates well-spread news from less-spread news is not accuracy. False stories, such as that Hillary Clinton ran a child sex ring out of a Washington DC pizzeria or that climate scientists were caught tampering with data to exaggerate the rise of sea levels, spread farther and faster than true stories, partially because false stories elicits more surprise, more disgust, and more fear. Social media, paired with poor scrutiny of emotionally-charged information, has brought about an epidemic of fake news.

Fake news is, by definition, false, but it has other characteristics as well. It relies on low-quality sources; it relies on few sources; and it elevates sources espousing implausible claims to the same level as sources espousing plausible ones.

Consider the recent spate of blogs by climate change deniers rebutting the discovery that polar bear populations are shrinking because their prime hunting grounds—Artic sea-ice, from which they hunt seals—are also shrinking. A recent analysis of such blogs revealed that 80% cite one particular source: Susan Crockford. Despite never having conducted any research on polar bear populations, Crockford describes herself as “one of the world’s foremost experts on polar bears,” and she runs a website devoted to refuting true experts, describing their research as “bogus,” “lame,” and “dangerous.” Blogs that cite Crockford not only treat her opinions as fact but also ignore contradictory findings from the peer-reviewed literature on polar bear science.

The polar bear blogs are typical of fake news in general. Its creators play fast and loose with sources, and they get away it because people turn a blind eye to the quantity and quality of source information.

Several recent studies have begun to reveal exactly how bad we are at evaluating sources. In one such study, conducted by Yale psychologist Sami Yousif and his colleagues, adult participants were given five articles to read about the future of the Japanese economy. One of the five took a negative stance, and the other four took a positive stance. The negative article was the same for all participants, but the positive articles varied. Some participants read positive articles that each cited a different source, whereas others read articles that all cited the same source, akin to how dozens of polar bear blogs all cite Susan Crockford. Yousif and his colleagues found that participants were just as swayed by the single-source collection as they were by the multi-source collection. Everyone tended to side with the positive argument, but the number of sources contributing to that argument played no role in participants’ evaluation of its quality.

To make sure that participants were not generally inclined toward positive views of Japan’s economy, Yousif and his colleagues gave another group of participants just two articles—one positive and one negative—and found that participants did not hold strong views one way or the other in the absence of additional articles on the positive side.

Another concern they addressed was whether participants’ lack of preference for multi-source arguments over single-source arguments reflected a mere failure of memory, such that participants forgot the identities of the sources as they moved from one article to the next. The researchers gave participants a list of sources at the end of the study and asked them to identify the ones they had seen previously. Accuracy in this task was quite high, meaning that those who had seen only one source knew that they had seen only one source. Hearing that source cited four times was apparently as persuasive as hearing from four different sources.

Not only can people be oblivious to source quantity, they can also be oblivious to source quality. In a recent study by University of Melbourne psychologist Amy Perfors and her colleagues, adults were asked to play the role of journalist covering either a political election or climate change. They were given a list of potential sources and asked who they would interview. The sources were labeled with their stance on the issue—whether they supported a particular candidate or whether they believed in climate change—as well as their confidence, expressed as a percentage. The researchers varied whether participants saw five, ten, or fifteen potential sources, and all sources expressed the same view with the exception of one dissenter. The dissenter varied from the others not just in stance but also in confidence. Members of the consensus group were 90 to 96% confident, whereas the dissenters were only 4% confident.

Perfors and her colleagues found that two-thirds of participants chose to include the dissenter in their story, regardless of whether the dissenter was one of five sources, one of ten, or one of fifteen. The dissenter was included at equal rates for stories about political elections and climate change, and participants who identified as progressive were just as likely to include a dissenter as those who identified as conservative.

In a second study, Perfors and her colleagues provided participants with stories on the same topics and asked them to indicate whether the author of the story was impartial, biased, or biased but trying to appear impartial. The researchers varied whether the story’s sources included a dissenter. Participants who read stories without a dissenter—that is, stories in which all sources agreed—were about 30% more likely to say it was biased than participants who read stories with a dissenter. The dissenter was once again described as being only 4% confident, but low confidence did not matter. The dissenter’s mere presence dropped accusations of bias for both types of stories (election-related and climate-related) and by both types of participants (progressives and conservatives).

People thus value the wrong information when it comes to reasoning about sources. They value a diversity of viewpoints over the accuracy of those viewpoints, and they fail to value whether the viewpoints are independent. These findings may help explain why fake news has gained so much traction in recent years, with the advent of social media. Fake news provides the counterpoint that people assume is needed to make a conversation “unbiased,” and its repetition from one media outlet to another is viewed as inherent evidence of its credibility, even if that story originated from a single source.

Social media platforms are working to curtail the spread of fake news through structural changes to the platforms themselves, but another means of stopping fake news is education. Those who consume fake news and spread fake news might be less inclined to do so if they developed a habit of scrutinizing sources. Habits like these need to be taught early and often, introduced in elementary school and reinforced through high school and even college. Questions of what to believe and who to trust have become thornier and thornier in the age of social media, and they will likely require an overhaul of traditional methods for teaching information literacy.

Now share this story with all your friends.