A Content Analysis of Social Media Campaigns in the 2016 US Presidential Election

Peter Susko

Major Professor: Gary L. Kreps, PhD, Department of Communication

Committee Members: Xiaomei Cai, Samuel R.Lichter

Online Location, Online
November 18, 2020, 01:00 PM to 03:00 PM


The use of social media has become a staple in political campaigns. However, research on social media has been reactionary and has not begun to apply theory to social media messaging. This dissertation analyzed the social media posts of the 2016 US presidential candidates Hillary Clinton and Donald Trump. The overall goal of this research was to explore the different tactics and strategies candidates used on social media. Study 1 in the dissertation used a content analysis approach for studying social media posts (Facebook, Instagram, and Twitter) using the Extended Parallel Process Model (EPPM) to identify if the candidates’ social media posts contained the components to have an effective fear appeal. Study 2 in the dissertation used the same data using the Image Repair Theory to identify tactics and strategics and looked at the candidates’ social media posts during a point where each candidate faced a crisis. Study 2 also looked at Trump’s responses after the “Access Hollywood” tape was released and at Clinton’s responses after FBI Director James Comey’s letter on Clinton’s emails was leaked two weeks before the election. The author found both candidates used various tactics under the fear appeal model. The author did not find conclusive evidence from study 2 to reach strong conclusions. This research is the first application of the EPPM in a political communication context. The dissertation also provides a new context to the Image Repair Theory, bringing the theory into the age of social media. Findings from each study are discussed along with practical applications, theoretical implications, methodological and conceptual limitations, as well as directions and considerations for future research.