In This Story
Currently, 12.5% of U.S. adults smoke cigarettes. At the same time, more than one-third of U.S. adults seek health information online, making social media a potentially powerful platform for anti-tobacco campaigns. However, limited research has been done on effective social media strategies for anti-smoking campaigns.
An interprofessional Mason research team led by Associate Professors in the Department of Health Administration and Policy Hong Xue and Gilbert Gimm found that the most popular anti-tobacco campaigns on Facebook were informational and discussed the negative effects of smoking. New information about harmful chemicals and the risks of second-hand smoke on pets were the most engaged with posts.
“Our results show that people respond to messaging on how smoking negatively affects the lives of those they care about, including pets. Messages that are personally meaningful to smokers can help to generate positive behavioral changes among smokers,” said Xue, the principal investigator. “Anti-tobacco campaigns can use these findings to improve their campaigns, better engage the public, and more effectively promote reasons to stop smoking.”
This is the first large-scale social media data mining study that examined key anti-tobacco campaigns in the United States. Researchers also found that large campaigns from government and nonprofit organizations had more user engagement compared to smaller and local campaigns. Facebook users were much more likely to engage in campaign messages with videos.
“Social Media Data Mining of Antitobacco Campaign Messages: Machine Learning Analysis of Facebook Posts” was published in The Journal of Medical Internet Research in February 2023. The project was funded by a research grant from the Virginia Foundation for Healthy Youth.
From Mason, Professor Xiaoquan Zhao in the Department of Communication, Professor Hemant Purohit in the Department of Information Sciences and Technology, and doctoral student Shuo-yu Lin were also researchers on the study. Researchers from Virginia Commonwealth University, Slippery Rock University of Pennsylvania, and Changshu Institute of Technology (China) also contributed to the study.