Constructing Long COVID Risk Communication in the US: A Topic Modeling-Based Critical Discourse Analysis of News Coverage
This study conducts a topic modeling-based critical discourse analysis of 111 articles from The New York Times and 89 articles from The Washington Post between 2020 and 2022. Incorporating a Latent Dirichlet Allocation (LDA)-based machine learning technique into the critical discourse analysis, we identified five major themes in news coverage: disparities in healthcare accessibility, vulnerable populations in vaccine response, government response in vaccine efficacy and policy, medical complexities and post-viral symptoms, and future looking clinical trials and research initiatives. Findings show that while both newspapers highlight social inequities, they differ in framing financial, racial, and systemic barriers. Results contribute to risk communication and health equity research by demonstrating practical implications for technical communication researchers, journalists, policymakers, and healthcare providers.
