Using Computational Methods and Experimentation to Understand the Persuasiveness of Vaccine Messages
Date of Completion
6-20-2018
Embargo Period
12-16-2018
Keywords
Misinformation, vaccine, emotion, topic evolution, popularity
Major Advisor
Ross Buck
Associate Advisor
Leslie Snyder
Associate Advisor
Mark A. Hamilton
Field of Study
Communication Sciences
Degree
Doctor of Philosophy
Open Access
Open Access
Abstract
In the digital age, true and false information can influence health-related decisions differently. It is important to monitor the dissemination of misinformation, understand public interests and develop persuasive messages. Two studies were conducted in order to meet these goals. Study 1 used computational methods to explore topic evolution and popularity in vaccine-related online messages. Topic modeling identified 14 topics in pro-vaccine messages (PVMs) and 12 topics in anti-vaccine messages (AVMs). PVMs that used personal stories received the highest number of shares, reactions, and comments. Pure scientific knowledge received the least attention. The most frequently appearing topic in AVMs was about child death. AVMs that discussed flu shots and government conspiracy were the most popular. Since 2016, even though more PVMs appeared online, AVMs were more popular and the comparative popularity of AVMs increased fast. AVMs that discussed vaccine damage were increasingly popular. Newly-emerged anti-vaccine topics (e.g. false rumors of CDC conspiracy) should be noted.
Based on Study 1’s results, Study 2 designed four PVMs and tested if these messages could increase HPV vaccine acceptability among parents who had not vaccinated their children. A total of 301 parents participated in an online experiment. SEM was used to analyze emotional and cognitive responses after the intervention. Results revealed that only the pro-organization message with low levels of emotions could increase vaccine acceptability. Anti-organization messages had no effect on the outcome. Political views and gender of parents could influence vaccine-related cognitive responses and eventually affected vaccine acceptability. Implications for message design were provided.
Recommended Citation
Xu, Zhan, "Using Computational Methods and Experimentation to Understand the Persuasiveness of Vaccine Messages" (2018). Doctoral Dissertations. 1847.
https://digitalcommons.lib.uconn.edu/dissertations/1847