Date of Award

May 2022

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Department

English

First Advisor

David Clark

Committee Members

Stuart Moulthrop, Patricia Mayes, Shevaun Watson, Thomas Malaby

Keywords

Anti-vaccine, Digital Writing, Ethos, Language, Social Media

Abstract

Many scholars attribute social media’s influence with a rise in distrust of expert advice. These scholars have suggested that people are turning to non-experts for advice because those non-experts seem to be more willing to openly discuss medical issues while also providing empathy, as opposed to the experts who have been trained to speak with detached authority. For this dissertation, I have done a study to find evidence supporting these theories. To do this, I looked at the Twitter conversation which has been focusing on anti-vaccination themes. Drawing on tweets from within that conversation, I conducted an inter-rater reliability test to categorize 1,000 tweets as either using a more empathetic and conversational tone versus those with the authoritative tone traditionally favored by experts. I then used those evaluations to conduct machine learning to evaluate over 50,000 additional tweets from the anti-vaccination conversation. I evaluated the relative success of tweets those tweets which used “authoritative” language compared to those that used “dialogic” language. Through this research, I was able to find a correlation between the degree to which the language within a tweet seemed to express empathy and encourage give-and-take forms of conversation and with engagement rates achieved by those tweets. Analysis suggests that the amount of influence this language use has on engagement rates is relatively minor, with tweets using stronger levels of dialogic language earning approximately one additional like for every 5,000 followers an account may have over tweets using primarily authoritative language. This study was done with the intention of considering how an audience’s preference for dialogic language might influence the way we prioritize authoritative voice in academic writing. As the data only marginally confirms this preference, this study shifts focus to ways of teaching students to be more responsible as readers in lieu of relying on experts using a more empathetic voice.

Download

Included in

Rhetoric Commons

Share

COinS