When are echo chambers harmful?

The presence of echo chambers may seem synonymous with their potential ill effects.

Existence of echo chambers indicates a polarized state of the society concerning beliefs regarding a given topic. A social network becomes polarized when several communities exist with differing beliefs.

Such communities are termed as echo chambers when people within each community of beliefs communicate mostly only with others having similar beliefs.

However, increase in the “intensity of echo chamber phenomenon” does not necessarily translate to an increase in the “intensity of polarization”.

Credibility of information

A well-known mechanism of the creation of echo chambers online, which has been observed during recent years, is the spread of fake information. Such echo chambers have true and fake versions of a given story propagating in separate communities.

Let us assume that, in a given period, true information propagates with a certain level of credibility. When can fake information create more polarization? Is it possible when such information has no credibility?

Definitely not. Hence credibility of information that propagates in a social network is an important characteristic of that information which determines how strongly it will affect future polarization levels.

Findings from polarized climate change tweets

The above contemplations motivated our recent study published in Palgrave Communications to research the role of information credibility, focusing on the case of climate change discussion on Twitter.

The most surprising finding from the study is that an increase in homophily in communication, where climate change supporters and skeptics communicate mostly with others having similar beliefs as their own, has led to a decrease in polarization about beliefs concerning the reality of climate change during 2007-2017.

Polarization is measured using the degree of bi-modality of the distribution of emotion-adjusted beliefs. Emotion adjustment is made to incorporate the intensity of an expressed belief in addition to whether the belief supports or disregards that climate change is real.

The interaction effect

The study shows, using a simple model of opinion updating, that increase in homophily interacts with information credibility to drive future polarization trends.

In a regime where fake information does not have credibility, an increase in homophily in communication does not lead to an increase in polarization.

In other regimes where fake information carries at least minimal credibility, polarization can never decrease with an increase in homophily.

How increase in homophily in communication in a social network affects future polarization depends on the credibility of propagating information. Photo by Abhishek Samantray & Paolo Pin

Conclusion and perspective

The conclusion that emerges from the empirical investigation of tweets and the nature of the interaction effect, as described above, is that climate change skeptic tweets are not credible .

This is because the empirical investigation shows a negative effect of homophily on polarization, which can happen (according to the opinion updating model) only in the case when climate change-skeptic tweets are not credible.

It appears that the study may have spillovers to the broader thought that connects fake information, echo chambers, and polarization. In particular, we probably need to rethink, among others, the following points.

  1. Echo chambers and polarization may not represent the same phenomenon in all cases.
  2. What offline or online factors are the primary drivers of increase in polarization in social media?
  3. How does the credibility of fake information originate and evolve?