Investigating Political Polarization on Twitter: A Canadian Perspective

Abstract

This article investigates political polarization in social media by undertaking social network analysis of a sample of 5,918 tweets posted by 1,492 Twitter users during the 2011 Canadian Federal Election. On the one hand, we observed a clustering effect around shared political views among supporters of the same party in the Twitter communication network, suggesting that there are pockets of political polarization on Twitter. At the same time, there was evidence of cross‐ideological connections and exchanges, which may facilitate open, cross‐party, and cross‐ideological discourse, and ignite wider debate and learning as they are observed by nonaffiliated voters and the media at large. However, what appeared to be far less likely was any increased willingness or tendency for committed partisans to shift their allegiances as a result of their Twitter engagements, and we postulate that Twitter usage at present is likely to further embed partisan loyalties during electoral periods rather than loosen them; a dynamic that would seemingly contribute to political polarization.

Measuring international relations in social media conversations

“…It indicates divisions and separation among entities (Watts & Strogatz, 1998). In prior studies, clustering is used to identify political and ideological divides (Kim, Barnett, & Kwon, 2010;Himelboim, McCreery, & Smith, 2013;Gruzd & Roy, 2014). Too much clustering can create blockages in the free flow of information, limiting the scope and variety of influence (Granovetter, 1983).…”

Detecting Offensive Statements towards Foreigners in Social Media

“…In social media networks, they might facilitate homogenous viewpoints by superseding opposing viewpoints [11]. Within such a network, users create mutual connections, for example by friendship or follower relations as well as by forming groups.…”

“…Content published by friends or connected groups is more likely to be displayed than content from other users. More importantly, the resulting content often contains onesided viewpoints as friends typically share similar interests and opinions [11]. In the context of political opinion-formation, echo chambers might be a problem if users are exposed to homogenous opinions favoring an extreme political viewpoint while having restricted access to credible opposing opinions [11,12].…”

“…After the detection of such echo chambers, the beliefs of the users might be challenged by presenting them controversial and well-researched information [11]. The EdgeRank in Facebook, for example, could be adjusted to selectively inject such content.…”

“…Social media platforms intensify this problem by the possibility to anonymously create content rapidly reaching a large number of users [6]. More importantly, content containing one-sided and radical viewpoints might be a problem in political opinion-formation, if users have only restricted access to credible opposing opinions [11,12]. This way, an important concept of democracy is violated: taking informed decisions in the context of competing opinions and ideas [13].…”