Media data and vaccine hesitancy: a scoping review (Preprint)

Abstract

Media studies are important for vaccine hesitancy research since they analyze how media shapes risk perceptions and uptake of vaccines. Despite a growth in studies in this field due to advances in computing, language processing, and an expanding social media landscape, no study has consolidated the methodological approaches used to study vaccine hesitancy. Synthesizing this information can better structure and set a precedent for this growing subfield of digital epidemiology.

The review aimed to identify and illustrate what media platforms and methods were used to study vaccine hesitancy, and how they build or contribute to the study of media’s influence on vaccine hesitancy and public health.

This study followed PRISMA’s (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) guidelines for scoping reviews. A search was done of PubMed, Web of Science, and SCOPUS for any studies that: used media data (social media and/or traditional media); had an outcome related to vaccine sentiment (opinion, uptake, hesitancy, acceptance, stance); were written in English; and published after 2010. Studies were screened by only one reviewer, and extracted for media platform, analysis method, and theoretical models used.

A total of 123 studies were included, 69 which used traditional research methods and 54 which used computational methods. Of the traditional methods, a majority used content analysis (74.2%) and sentiment analysis (37.1%) to analyze texts, with few using campaign evaluation methods (8.1%) and tracking of search activity and/or information spread (11.3%). The most common platform was newspapers, print media, and online news. Computational methods mostly used sentiment analysis (57.4%), topic modelling (31.5%), and network analysis (27.8%). Fewer studies used projections and feature extraction as methods. The most common platforms were Twitter and Facebook. Theoretically, most studies were weak. Of traditional methods, only 8 had a theory-based approach (11.6%); and for computational methods, only 6 (11.1%). Due to a patchwork of studies resulting from combinations of platforms and methods, it is difficult to draw a coherent conclusion on media’s influence on vaccine hesitancy.

There is heterogeneity in using media data to study vaccine hesitation, evidenced in the medley of combinations of platforms and computer science tools (eg. network analysis, sentiment analysis). Yet, these studies are guided by a preference for using novel methods rather than theory, making their links to public health tenuous. This review suggests and walks through a theory-first approach that can aid in better formulation of knowledge and establish a coherent paradigm in media studies on vaccine hesitancy. It ends with a statement that media data analyses, though groundbreaking in approach, should supplement–not supplant–current practices in public health research.