Echo chambers and how you can avoid them

In a poll of registered U.S. voters, 14% of Donald Trump supporters believed reports that a Washington D. C. pizzeria was the HQ of a child sex ring run by Hillary Clinton. 32% of respondents were “not sure”. The remaining 54% thought the story was spurious, and were right to think so; the claim originated from an unverified online conspiracy. ¹ It was “fake news”. The consequences of this “fake news”, however, were atrociously real. An armed 25-year-old man, moved by the story’s fabrications, went to the pizzeria to investigate its claims. He opened fire inside the building.

While this is an extreme example, it signifies a disturbing “post-truth” trend in society. The U.S. Presidential election exemplified the preponderance and believability of fake news. Social media is the commonly cited culprit, charged with encouraging the spread of untruth and intensifying groupthink. On sites like Facebook, bogus news is given the same weighting and is often superficially indistinguishable from real news.

Content is not filtered into fact and fiction; rather, the organising force is the individual user’s preconceived belief. It is easy to disregard articles that don’t align with your beliefs and subscribe to news sources or befriend individuals who share your biases. What ensues is the establishment of echo chambers, collections of self-same individuals between whom fake news, unchallenged by anyone or anything external to the chamber, is misguidedly corroborated as truth. This is dangerous, because misinformation inhibits our ability to know what’s true, make choices based on the most accurate information, and resist those who might not have our interests in mind.

A group of psychologists researching attitudes to global environmental change over Twitter tracked this process. ² On the platform, they found that users quickly segregate into “polarized ‘sceptic’ and ‘activist’ groups”, most often interacting only with like-minded others. The researchers also found mixed-attitude communities in which sceptics and activists frequently interact. Occasionally, cross-community discussion appeared to be constructive; however, in general, messages between sceptics and activists carried negative sentiment. This sort of dynamic is likely to increase antipathy rather than bridge divide, leading to irresolvable standoff between factions. The research also indicates that, rather than being exclusively “tools, traps, or foibles of the right or the left”, fake news and echo chambers are, as psychologist Joe Pierre writes, “vulnerabilities for all of us, on either side of the political fence”. ³

Perhaps the chief irony of social media is revealed: while it enables connection of otherwise disconnected global communities, it is also plays up to our tendency to organize ourselves into silos or echo chambers. The problem with this is that individuals gain zero exposure to oppositional thinking and merely entrench their prejudices. On social media one can, in the words of Simon and Garfunkel, hear what they want to hear and disregard the rest.

The proclivity to hear what we want to hear is known in psychological terms as confirmation bias, our built-in neural bias towards preferring information that confirms our pre-existing beliefs. Facebook enhances this human susceptibility, offering increased opportunity for the confirmation of biases and investment in delusional, erroneous perspectives of reality. This inclination, as professional fake news writer Paul Horner observes, means we don’t fact-check what we consume. Instead, we “just keep passing (untruthful) stuff around”. ⁴ Yet mindlessly copying and pasting and blindly believing what gets shared can have dire consequences.

Alex Pentland, a professor at MIT, discovered this when studying social learning and behaviour on the financial trading site eToro. ⁵ Because sharing and copying others’ activity often proves beneficial in trading, eToro incorporates a social network platform, known as OpenBook, through which users can easily look up other users’ trades, portfolios, and past portfolios. By examining over 10 million transactions on the site, Pentland found that the best financial outcomes were produced by patterns of exploration that display high “idea flow” – that is, the rate at which new trading strategies propagate from user to user along the social network as individual users decide to copy another user, and are in turn copied by others. Crucially, however, sustained success relies on the diversity of the ideas and social network; the more diverse the information available to an individual is, the more informed their decision.

If a social network is too similar in its behaviours or beliefs, too much of an echo-chamber, overconfidence often ensues. Information circles in a “feedback loop” and because ideas are only minimally modified as they pass from person to person they go unrecognized as repetitions of the same idea. It is easy to believe that everyone has arrived at the same idea independently, which leads to unwarranted confidence in the idea. This behaviour is perilous; as Pentland notes, it “is the source of financial bubbles”. Pentland gives the example of a Latvian trader in eToro, whose strategy yielded a long winning streak and was subsequently copied by many, who in turn were copied by others etc. Individual traders might have thought they were following the patterns of several different “gurus” in the network, but they were merely copying copiers, meaning they unwittingly put their faith in the success of one individual. This ended in financial disaster for those who gambled their whole portfolio on the Latvian’s strategy; in Pentland’s words: “pop goes the bubble”.

Pentland shows that the disastrous consequences of echo-chambers and the proliferation of erroneous information within them can be combated by concerted attempts at social diversification and exposure to oppositional viewpoints. The cognitive dissonance and creative abrasion that results in such an environment is highly productive.

There are lessons here for organisations to learn. Pentland’s research shows that companies should really go out of their way to improve diversity, since it breeds success. The latest findings from big data, behavioural science, and social psychology researchers are revealing the importance of diverse networks to performance, but really we’ve known about the link since 1985, when researchers from Bell labs at Carnegie Mellon found that star performers invested in establishing a diverse set of relationships with experts in related fields. Conversely, average performers limited their relationships to those in similar roles.

Luckily nowadays the ability to build a diverse expert network is far easier than in 1985. We have the technology to do so, but it is necessary that we use it smartly. Instead of limiting the perspectives we are exposed to, the best thing to do is undoubtedly to make a concerted effort to break down echo chambers. Generating idea flow across groups is vitally important in both politics and business. However, in order to be open to new perspectives we need to enhance our cognitive flexibility. That means being able to perceive oppositional viewpoints not with derision but with intrigue and framing confrontations with oppositional individuals not as threatening encounters but as opportunities for learning to occur.