New
(2025) AI Processed Human Approved
A Multi-Method Analysis of the Evolution of Social Media Discourse in Prolonged Crises:through the Lens of Crisis and Emergency Risk Communication (CERC)
This study investigates how public discourse on social media evolves across different phases of a prolonged health crisis using the COVID-19 pandemic as an empirical context. Employing a multi-method approach combining Natural Language Processing, Guided LDA, and text network analytics, the authors analyzed over 24 million Facebook posts from 2020 to 2022. The research maps evolving public concerns onto the Crisis and Emergency Risk Communication (CERC) framework to inform public health messaging strategies.
Problem
Unlike sudden disasters, prolonged crises unfold gradually over time, causing public stress to accumulate and social media discourse to continuously shift. Little theory-driven research has examined how public topics and text network connections change across specific phases of an ongoing crisis. As a result, public health agencies lack phase-tailored communication strategies to address shifting public concerns, manage anxiety, and combat misinformation.
Outcome
- Public discussions shifted over time from initial preventative measures and viral impacts to broader concerns surrounding emerging variants.
- Upward variant phases featured more focused and interconnected social media discussions, whereas downward phases exhibited more varied conversations.
- Results highlight the critical need for phase-tailored messaging strategies by government agencies to effectively mitigate anxiety, foster trust, and counter misinformation during prolonged crises.
- Upward variant phases featured more focused and interconnected social media discussions, whereas downward phases exhibited more varied conversations.
- Results highlight the critical need for phase-tailored messaging strategies by government agencies to effectively mitigate anxiety, foster trust, and counter misinformation during prolonged crises.
What it means for you
- CIO / IT Executive: On Monday morning, direct your data science team to explore leveraging sentiment analysis and topic modeling tools to categorize incoming social media data related to ongoing public health issues, enabling proactive identification of shifting public concerns.
- IT Manager: On Monday morning, schedule a brief meeting with your social media monitoring team to review the existing dashboards and identify if they can be reconfigured to visually highlight shifts in discussion topics and the interconnectedness of conversations, aligning with the research's emphasis on phase-tailored messaging.
- Business Strategist: On Monday morning, initiate a brainstorming session with your communications and marketing teams to consider how your organization's messaging can adapt to different phases of a prolonged public health crisis, drawing parallels from the COVID-19 discourse shifts identified in the research.
- Researcher: On Monday morning, begin outlining a research proposal that analyzes social media discourse in another prolonged crisis (e.g., climate change, economic instability) using a similar multi-method approach to the one described, focusing on mapping public concerns onto relevant communication frameworks.
- Policymaker: On Monday morning, task your policy analysis unit with reviewing existing crisis communication guidelines and recommend updates to incorporate phase-specific strategies informed by the research on evolving social media discourse, particularly concerning anxiety mitigation and misinformation countering.
Transcript
Host: Welcome to A.I.S. Insights — powered by Living Knowledge. I'm Anna Ivy Summers. Today, we're diving into an insightful study titled "A Multi-Method Analysis of the Evolution of Social Media Discourse in Prolonged Crises: through the Lens of Crisis and Emergency Risk Communication". Joining me to unpack this research is our expert analyst, Alex Ian Sutherland. Welcome, Alex!
Expert: Thanks, Anna. It's great to be here.
Host: Alex, when a sudden crisis hits, like a hurricane or a power outage, crisis communication feels fairly straightforward—deliver immediate, urgent facts. But what happens during a prolonged crisis, like a multi-year pandemic? What is the main problem this study addresses?
Expert: That distinction is key. Sudden disasters demand immediate, short-term responses. But prolonged crises unfold gradually over months or years. Public stress accumulates over time, and social media discourse constantly shifts. The problem is that most organizations and health agencies rely on static, static-phase communication models. They treat the crisis as one long, uniform event. Without understanding how public discourse evolves across different phases, leaders lack phase-tailored strategies to manage public anxiety, maintain trust, and counter misinformation effectively.
Host: That makes a lot of sense. So how did the researchers analyze such a massive, evolving public conversation over time?
Expert: The study took a multi-method computational approach. They analyzed nearly twenty-five million public Facebook posts from January 2020 to April 2022. They divided the timeline into seven distinct phases based on seven-day moving averages of new COVID-19 cases from Johns Hopkins University. Then, they combined Natural Language Processing, a topic modeling technique called Guided LDA, and text network analytics. Finally, they mapped these discourse patterns onto the Crisis and Emergency Risk Communication, or CERC, framework.
Host: Analyzing twenty-five million posts across two years must have revealed some fascinating trends. What were the key findings?
Expert: First, public discussions underwent a major thematic shift. Early in the crisis, conversations centered heavily on immediate preventative measures and direct health impacts. Over time, attention shifted toward broader concerns about emerging variants and long-term implications. Second, the study revealed clear network differences between rising and falling case phases. During upward variant phases, public discussions became highly focused and tightly interconnected. People rallied around very specific concerns. In contrast, during downward phases, conversations became far more dispersed and varied.
Host: That is really intriguing—that public focus actually tightens during a surge and expands when things calm down. Why does this matter for business leaders, policy makers, and communication professionals?
Expert: It fundamentally changes how we should approach crisis communication. A one-size-fits-all message simply does not work over a long timeline. During an upward phase, when public attention is hyper-focused and anxiety is high, organizations need concise, highly targeted messaging that directly addresses core public fears. During downward phases, when discourse scatters, it's the ideal time to communicate broader forward-looking strategies, build long-term trust, and address emerging topics before the next surge.
Host: And what about combating misinformation? How does this research help with that?
Expert: By using text network analytics to monitor public discourse in real time, organizations can identify rising misinformation topics early in specific phases. Instead of playing catch-up after a rumor goes viral, leaders can deploy fact-based, transparent messaging tailored to the specific phase of the crisis. It turns crisis communication from a reactive scramble into a proactive strategy.
Host: It really highlights that effective communication isn't just about what you say, but when and how you adapt your message to the audience's mindset. Alex, thank you so much for breaking down this study for us today.
Expert: Glad to be here, Anna.
Host: And thank you for tuning in to A.I.S. Insights — powered by Living Knowledge. Be sure to subscribe so you never miss an episode. Until next time, stay informed and stay curious.
Expert: Thanks, Anna. It's great to be here.
Host: Alex, when a sudden crisis hits, like a hurricane or a power outage, crisis communication feels fairly straightforward—deliver immediate, urgent facts. But what happens during a prolonged crisis, like a multi-year pandemic? What is the main problem this study addresses?
Expert: That distinction is key. Sudden disasters demand immediate, short-term responses. But prolonged crises unfold gradually over months or years. Public stress accumulates over time, and social media discourse constantly shifts. The problem is that most organizations and health agencies rely on static, static-phase communication models. They treat the crisis as one long, uniform event. Without understanding how public discourse evolves across different phases, leaders lack phase-tailored strategies to manage public anxiety, maintain trust, and counter misinformation effectively.
Host: That makes a lot of sense. So how did the researchers analyze such a massive, evolving public conversation over time?
Expert: The study took a multi-method computational approach. They analyzed nearly twenty-five million public Facebook posts from January 2020 to April 2022. They divided the timeline into seven distinct phases based on seven-day moving averages of new COVID-19 cases from Johns Hopkins University. Then, they combined Natural Language Processing, a topic modeling technique called Guided LDA, and text network analytics. Finally, they mapped these discourse patterns onto the Crisis and Emergency Risk Communication, or CERC, framework.
Host: Analyzing twenty-five million posts across two years must have revealed some fascinating trends. What were the key findings?
Expert: First, public discussions underwent a major thematic shift. Early in the crisis, conversations centered heavily on immediate preventative measures and direct health impacts. Over time, attention shifted toward broader concerns about emerging variants and long-term implications. Second, the study revealed clear network differences between rising and falling case phases. During upward variant phases, public discussions became highly focused and tightly interconnected. People rallied around very specific concerns. In contrast, during downward phases, conversations became far more dispersed and varied.
Host: That is really intriguing—that public focus actually tightens during a surge and expands when things calm down. Why does this matter for business leaders, policy makers, and communication professionals?
Expert: It fundamentally changes how we should approach crisis communication. A one-size-fits-all message simply does not work over a long timeline. During an upward phase, when public attention is hyper-focused and anxiety is high, organizations need concise, highly targeted messaging that directly addresses core public fears. During downward phases, when discourse scatters, it's the ideal time to communicate broader forward-looking strategies, build long-term trust, and address emerging topics before the next surge.
Host: And what about combating misinformation? How does this research help with that?
Expert: By using text network analytics to monitor public discourse in real time, organizations can identify rising misinformation topics early in specific phases. Instead of playing catch-up after a rumor goes viral, leaders can deploy fact-based, transparent messaging tailored to the specific phase of the crisis. It turns crisis communication from a reactive scramble into a proactive strategy.
Host: It really highlights that effective communication isn't just about what you say, but when and how you adapt your message to the audience's mindset. Alex, thank you so much for breaking down this study for us today.
Expert: Glad to be here, Anna.
Host: And thank you for tuning in to A.I.S. Insights — powered by Living Knowledge. Be sure to subscribe so you never miss an episode. Until next time, stay informed and stay curious.