Looking at Confirmation bias in the lens of information cascade
As we slowly move away from the trend of mass production, where all goods are alike, the market searches for a more nuanced, individual-specific services. In the context of such a shift and growing accuracy for user specification algorithm that selectively chooses content that we may like, agree, or enjoy, social media has become a dangerous tool and a system that encourages emergence of cascade of opinions.
The article “How to Break Out of Your Social Media Echo Chamber” addresses how social media can be used to simply confirm what we already believe, rather than being a source for new information for a more well-balanced perspective. More specifically, it views such a trend in the lens of confirmation bias. Confirmation bias refers to a phenomenon where people seek to interpret new information in accordance with preexisting beliefs. It points out that as algorithms of social media continuously and exclusively exposes the users to contents that holds similar view as theirs, the information is used to reinforce your prior belief. While mundane the effect may seem, it can pose as a hug obstacle for attempting to hold a balanced view and may encourage extreme opinions to remain unchecked.
The idea of confirmation bias shines a new light onto the emergence and continuation of cascades in real life. When investigating cascade mathematically, we assume an unbiased, rational agent in action. In other words, we presuppose that the agent will treat all information equally and always aim to take optimal measures with the given situation. Therefore, even when the cascade is continuing, the agents do not view it as an endless accumulation of evidence for one side. Rather, they recognize the existence of a cascade and is willing to break the cascade if any new information outside of the cascade is presented.
However, confirmation bias hints that a cascade may be much harder to break in real life. People are not rational; they do not calculate conditional probabilities and consider how cascades may be skewing the response of opinions they see. Moreover, they are biased, selectively choosing to see and acknowledge information. Not all information is the same; some hold different weight. The effect becomes more extreme as companies and algorithms have no incentive to fight against cascades. It results in not only the process of information by the users being biased but also the access and exposure itself.
The article points toward a remedy of this dire situation: intentionality. Fighting to control confirmation bias is necessary on an individual level to prevent wrong cascades from taking actions, especially since all consumers of social media are also the producers of content for social media. While unbiased rational agents are not perfectly protected from cascades, its susceptibility to it reduces dramatically with decreased dependence on single source of information. As social media’s power lies in its ability to connect, an effect on the users’ part to may be enough to prevent extreme cascades.
Reference:
https://www.wired.com/story/facebook-twitter-echo-chamber-confirmation-bias/