An analysis of the factors influencing engagement metrics within the dissemination of health science misinformation

Abstract

Objective: The proliferation of health misinformation on social media platforms presents a significant challenge. Methods: Data were collected from WeChat, video websites, and Weibo in November 2024. A total of 109 health misinformation samples were selected using our team’s “Health Misinformation Screening Criteria.” This study analyzes the activity and influencing factors of this misinformation. Activity indicators, including reads (views), comments, reposts, and likes, were weighted based on communication theory principles. A combined weighting method, using the entropy weight method and the analytic hierarchy process (AHP), was employed, followed by sensitivity analysis to determine activity levels. Non-parametric tests and negative binomial regression models were used to analyze the key influencing factors of misinformation activity. Results: WeChat exhibited the highest proportion of health misinformation (44.95%), with the majority originating from individual authors (55.96%), and primarily positive sentiment (37.61%). Misinformation related to disease prevention and control was most prevalent (54.13%), with declarative sentences being the most common tone (55.04%). Significant differences in propagation activity were observed, with WeChat exhibiting the highest activity, followed by video websites, and then Weibo. Misinformation with negative sentiment had significantly higher interactivity than neutral and positive content. Misinformation published by institutional authors was more likely to spread due to their authoritative advantage. Negative binomial regression analysis indicated that the disease prevention and control theme, three types of tone (interrogative, declarative, and exclamatory sentences), positive sentiment, and institutional authors significantly influenced misinformation activity (p < 0.05). Conclusion: By illustrating misinformation cases and stratified prevention strategies, this study reveals the key roles of themes, expression forms, sentiment, and publishing entities in the spread of health misinformation. It provides foundational data and theoretical support for targeted prevention, follow-up research, and the formulation of relevant management strategies, promoting a comprehensive governance model of “platform technology interception - science education prevention - misinformation source management.”