In this edition of the literature tips, Filipa M. Ribeiro, PhD researcher and science writer from University of Porto writes about a new paper on academic social networks.
By now, it is not a surprise that we live in social bubbles, especially when it comes to social media. It is not uncommon that we discuss with our colleagues the fact that, in practice, we live in social bubbles, in the sense that we relate in an homophilous way unless organizations and institutions give us no choice to do otherwise.
If no other, this is an excellent reason to read the paper published in mid February, by Dimitar Nikolov and colleagues: Measuring Online Social Bubbles, which is a nice follow-up of a doctoral thesis that was reviewed in an earlier post on the Hedda blog.
The paper by Nikolov and collegaues points to the fact that the current web-based systems (eg.: recommendation systems and search engines) based on previously declared social relations amplify the effect of a social bubble, which has previously been argued by Figueiredo (2014). The authors studied data from an American University online traffic, looking at applications of social media, email and search engines. The conclusion is that the level of diversity of information, both at individual and collective levels, is low in the first two and higher in the third (search engine for words or texts). The diversity of information is measured by the number of necessary clicks to connect two content.
“The traffic dataset gives us the unique opportunity to carry out this analysis. We are not aware of any other methods of contrasting different information access patterns produced by the same set of users, in the same time period. We have presented evidence of a collective social bubble: the diversity of information reached through interpersonal communication and social media is significantly lower than through searching. Our results suggest that social bubbles exist at the individual level as well: by using social media we are exposed to narrower sets of information sources.”
Thus, the main claim of this paper is that “by using social media we are exposed to narrower sets of information sources”. As seen in the previous work by Figueiredo (2014), we already know why that happens and how we can fix that issue.