Sandra González-Bailón, Carolyn Marvin Professor of Communication and Director, Center for Information Networks and Democracy, at the University of Pennsylvania
Abstract
On social media, human choice and platform affordances – including algorithmic recommendations – intertwine in a complex fashion. Algorithms matter not only for the content users see in their newsfeed, but also for who they friend and which accounts they follow. In this talk, I will discuss research using social media data investigating the complex interaction between social and algorithmic forms of curation and how it shapes exposure to information. This research analyzes political news consumption during the US 2020 election using aggregated data for 208 million US Facebook users. The analyses compare the inventory of all political news that users could have seen in their feeds with the information that they saw (after algorithmic curation) and the information with which they engaged. The results show that the net result is highly– and asymmetrically– segregated news consumption on Facebook.
