Leggi in italiano

Social networks are often accused of encouraging polarisation and widening ideological divisions, but this effect has been rarely studied in detail. Now a study in Proceedings of the National Academy of Science shows how different platforms foster different levels of segregation in online communities, depending of their feed algorithms and their social networking features1. The researchers, led by Walter Quattrociocchi from Università La Sapienza in Rome, analysed more than a hundred million pieces of content — posts, as well as interactions such as likes, comments and re-posts — on controversial topics such as abortion, vaccinations, gun rights, and US presidential elections. The content came from four social media platforms: Facebook, Twitter, Reddit (one of the world’s most popular websites), and Gab, similar to Twitter and popular among right-wing extremists.

The researchers looked for evidence of ‘echo chambers’, defined as online environments in which users’ opinions get reinforced by interacting mostly with like-minded sources. Since earlier descriptions of echo chambers were qualitative, the group had to introduce a quantitative definition, relying on two dimensions: each user’s leaning on a given topic, and direct interactions between users. They used slightly different methods on each of the four social networks to measure the political leaning of users (likes for Facebook, links to specific news outlets for Twitter, Gab and Reddit), and to reconstruct interaction network (for example, two Facebook users were considered connected if there was at least one post commented on by both). “We wanted to know whether applying the same definition of echo chamber to each platform we would observe different social dynamics” says Quattrociocchi, a professor of Informatics. The answer was yes.

The group found higher segregation on Facebook than on other platforms, and a clear distinction between social media with a feed algorithm tweakable by the users (such as Reddit) and less tweakable ones. Facebook, for example, does not offer a simple chronological option to see what your contacts share. Twitter lets you opt out of the algorithm by choosing to see ‘most recent’ tweets. Reddit and Gab work differently, and the former, in particular, relies less on algorithms and gives users more freedom as to what posts they see.

The researchers also simulated the spread of information across the four networks using epidemic models similar to the ones that are used to model the spread of COVID-19. The results confirm that, on Facebook and Twitter, users with a particular leaning are much more likely to be reached by information propagated by users with similar leaning - and then disseminate it accordingly. On Reddit and Gab, this effect is not visible. This could be because these social networks are less prone to echo chambers, or because Reddit communities (‘subreddits’) and Gab as a whole are echo chambers themselves, with users who are more politically homogenous. “The insight is that platforms with feed algorithms have a strong impact on polarisation and segregation dynamics” says Quattrociocchi. While even in real life we tend to congregate around people that share our values and our interests, feed algorithms tend to amplify this tendency in ways that were previously not quantified.