A thought-provoking study from Virginia Tech reveals how social media rhetoric during COVID-19 influenced public health, showing the predictive power of language. Eugenia Rho‘s research team examined how opposition to public health measures spread through social media, impacting real-life behaviour & health decisions.
The study [not behind a paywall- yay!] is the first to empirically link social media linguistic patterns to public health trends using AI, suggesting that large language models can effectively identify critical discussion patterns & enhance public health communication strategies.
The research team focused on Reddit due to the availability of data from discussions, particularly from banned subreddit groups that opposed COVID-19 prevention measures [other social media platforms were generally not accessible for this type of research due to restrictions on data sharing]. The team trained a GPT-4 chatbot to analyse posts from these banned subreddits.
The study applied Fuzzy Trace Theory, which emphasizes the idea that people process information through “gists” or simple cause-&-effect relationships. Valerie Reyna, a collaborator from Cornell University, has shown that this method of processing information is effective, even if the information itself is inaccurate.
Key Findings:
- The study found that social media posts that created a cause-&-effect gist (e.g., “I got the COVID vaccine” linked with “I’ve felt like death ever since“) could significantly predict real-world health outcomes.
- The volume of such gists in social media discussions was strongly correlated with the total & new daily COVID-19 cases, highlighting a direct link between online rhetoric & offline behaviour.
- Social media opposition to health measures led to higher infection rates & over 692,000 preventable hospitalizations in a month, costing $13.8 billion.
Possible takeaways for teachers?
- Critical Media Literacy: discuss the need to critically analyse media & social media rhetoric.
- Threats to democracy: discuss the role of media bias & echo chambers & their potential impact on democracy.
- Online vs offline: discuss the link between online speech & offline behaviour.
- Free speech: discuss the tension between free speech & the regulation of online platforms.
Did something similar happen in another language you know?
Do you think social media punches above its weight compared with ‘traditional’ media in terms of influencing individuals?
Leave a Reply