Over the past two decades, an increasing proportion of the population has come to view vaccines with skepticism, and in many cases, refrained from getting vaccinated altogether. The COVID-19 pandemic and vaccine rollout exacerbated this trend and heightened debates about the benefits of vaccines. While the procedural, behavioral and access barriers to vaccine uptake have been studied extensively, the emotional, ideological and rhetorical bases for vaccine hesitancy are poorly understood. This limited understanding, combined with public distrust in science and government, undermined efforts to robustly vaccinate against COVID-19 and other diseases.

Recognizing that online forums such as Facebook, YouTube and Twitter have become the primary platforms for discussing and disseminating vaccine-related concerns, this CWP project set out to harness the power of data science and artificial intelligence in combination with the tools of literary scholarship (i.e. textual analysis) to better understand the rhetoric of vaccine hesitancy. By developing a pilot online ad campaign, the project tested the implications of these findings for public-health messaging that encourages vaccination.

In the project’s first phase, coinciding with the rollout of the first COVID-19 vaccines in 2021, CWP supported the creation of a significant computational public data set of vaccine-hesitant language in English. Analysis of the data led to the discovery of distinct forms of vaccine hesitancy, identified through specific patterns of language and rhetoric. The subsequent Google ads campaign developed by the project team drew on these patterns to craft pro-vaccine messaging. During the six months that the campaign ran on Google ads (November 2023-April 2024), the specially tailored, experimental messaging consistently received more engagement than the standard CDC ad serving as the control. As a result of the campaign, thousands of vaccine-hesitant searchers clicked through to accurate vaccine information. 

The project makes a powerful case for strategies of health outreach that bridge the gap between the language of scientific communication and the language communities use to articulate their beliefs and mythologies regarding medical knowledge. Such folk discourse, the project revealed, proliferates in digital spaces and lends itself to an interdisciplinary approach involving data science, AI, and literary criticism. Through the ad campaign, discoveries about the rhetoric of vaccine hesitancy were marshaled to redirect vaccine hesitant searchers to accurate information. By publishing in an open source platform its extensive corpus of vaccine hesitant language, the project seeks to promote new findings on the data and new ways of interpreting and mobilizing online language. 

In the course of three years, the project has engaged literary scholars, medical professionals, data scientists, political scientists, community leaders and public health officials. Project leads provided consultation to the health departments in Maine and Ulster County, New York as they developed their COVID-19 vaccine and booster campaigns. Findings on the data were shared with stakeholders early on through one-page “Key Insights” documents. In addition the project has trained a cohort of undergraduate and graduate students in data analysis and original research using the corpus. Student researchers with the project won first place at Columbia University's Undergraduate Research Symposium in 2021.

The work continues through the founding of the Health Language Lab (HLL), a partnership between the Data Science Institute, the Medical Humanities program at the Institute for Comparative Literature and Society, and INCITE at Columbia University. The HLL is writing up the findings of this multi-year project for public dissemination and will continue to share findings with public health communications teams and policy makers in an attempt to influence vaccine messaging

Partners:

• The State of Maine, Department of Health and Human Services
• State of New York, Ulster County, Department of Health 
• The Vaccine Confidence Project