A proposed social network analysis of Twitter as it pertains to the 2016 presidential debates.

Twitter: A Network Divided

Kadushin talks about homophily being prevalent feature within social networks - that people tend to associate with others who are similar to themselves. Anecdotally, this can create an echo chamber of sorts in which online social network users are only exposed to thoughts and opinions consistent with their prior beliefs which probably, in turn, contributes to confirmation bias. I find this to be the case when I check Twitter during the presidential debates. Most of what I see in my feed during the debates are quotes and interpretations of the statements made by Mr. Trump as seen through the lens of someone(s) who view(s) him as a an authoritarian, bigoted, dangerous buffoon and as a disgrace to his country and to his political party. I also see a lot of awesome hashtags like #MuslimsReportStuff or (from a simpler time) #BindersFullOfWomen that spring up in response to candidates’ statements.

As such, I was momentarily taken aback when I clicked on the #MAGA hashtag in one of Governor Mike Pence’s post-debate tweets congratulating Mr. Trump for a job well done. Suddenly, I found myself in a bizarre twilight-zone in which Mr. Trump’s inability to answer questions and his bastardization of facts and history (and, indeed, to speak about anything at all with any modicum of cogency) was glorified and touted as a great victory for the United States and for the his party.

Interestingly, the respective meta-debate narratives and conversations coming from liberal and conservative twitter had almost no overlap - it appeared that there was little awareness of how the opposition was interpreting the debate. This gave me an idea for my final project. I’m interested in visualizing several aspects of the meta-debate on Twitter. First, I want to find out how much overlap there is between the denizens of liberal and conservative twitter, respectively, perhaps by examining who follows whom. From there, I want to track hashtags originating at either end of the Clinton-Trump spectrum and see if and how they spread across the network. I wonder whether the likelyhood of / degree to which a given user uses a particular hashtag or tweets about a particular aspect of the debate, combined with that person’s location on the network diagram, is indicative of that person’s preferred candidate and their location on the left-right political spectrum (I’m not quite sure yet how I’d validate any such claim). Other than wanting to do this because I find the meta-debate on Twitter to be really interesting, I’m interested in what the findings of such an investigation say about how political information flows through online social networks, and how/whether this influences elections.

Pretty much everything involved in this project is new territory for me. To begin with, I think I’m going to try to collect tweets during the next presidential debates. I haven’t chosen a tool for this, yet. From there, I’ll dive into the data and explore possible visualizations.