Feng Miao, Carter Chandler C, Page Simon, Emery Sherry L, Tran Hy, Kostygina Ganna
Social Data Collaboratory, NORC at the University of Chicago, Chicago, USA.
J Health Commun. 2025 Mar 28;30(sup1):39-49. doi: 10.1080/10810730.2025.2462682. Epub 2025 Feb 6.
This study identifies and analyzes X (formerly Twitter) posts related to 14 e-cigarette use prevention campaigns from 2014 to 2020, assessing message volume, content, sources, potential reach and engagement. Using supervised machine learning, we classified 618,965 tweets, finding 43% contained opposition messaging. Two regional campaigns received the highest levels of opposition, with over 99% of related tweets classified as opposition. However, prevention/neutral messages exhibited 92% higher potential reach than opposition messages. Geolocation analysis suggested that regional campaigns may have struggled to focus their impact within targeted jurisdictions. These findings illustrate the dual role of social media as both an amplifier of prevention messages and a platform for oppositional narratives, underscoring the need for public health practitioners to develop adaptive strategies to enhance the impact of digital campaigns.
本研究识别并分析了2014年至2020年期间与14项电子烟使用预防活动相关的X（前身为推特）帖子，评估了信息数量、内容、来源、潜在覆盖面和参与度。通过监督式机器学习，我们对618,965条推文进行了分类，发现43%的推文包含反对信息。两项地区性活动收到的反对意见最多，超过99%的相关推文被归类为反对意见。然而，预防/中立信息的潜在覆盖面比反对信息高出92%。地理位置分析表明，地区性活动可能难以将其影响集中在目标辖区内。这些发现说明了社交媒体作为预防信息放大器和对立叙事平台的双重作用，强调了公共卫生从业者制定适应性策略以增强数字活动影响力的必要性。
