Mapping the Landscape of COVID-19 Crisis Visualizations

Title of the Paper

Mapping the Landscape of COVID-19 Crisis Visualizations

Bibliographic Information

• Subject Areas: Information Visualization, Crisis Informatics, Public Health Communication
• Keywords: Visualization, COVID-19, Crisis Informatics, Data Quality, Risk Assessment, Information Dissemination, Data Visualization Techniques, Social Impact, Public Health Communication, Information Sources

Research Background and Issues

• Issues and Challenges:

  • During the COVID-19 pandemic, the rapid growth and diversity of visualized public information lacked systematic review and categorization.
  • Information quality, data uncertainty, and design choices significantly impacted the public, potentially misleading critical decisions.
  • Existing studies often focus on professional users, with limited understanding of visualization designs for the general public.

• Significance:

  • Visualization tools can quickly help the public understand risk models and pandemic dynamics during public health crises, serving as a vital component of information dissemination.
  • Effective information dissemination can alter public attitudes and behaviors, thereby influencing the trajectory of the pandemic.

• Research Motivation and Related Work:

  • Literature reviews indicate that historical pandemic visualizations primarily focused on professional applications.
  • The unprecedented scale and impact of the COVID-19 pandemic have driven widespread public engagement, which lacks systematic research and categorization.

Solution

• Method/Solution:

  • Collected and analyzed 668 crisis visualizations created during the COVID-19 pandemic.
  • Proposed a conceptual framework based on Lasswell's communication model and Munzner's nested model, exploring COVID-19 crisis visualizations across multiple dimensions, including "who creates the content, what data is used, what information is conveyed, what presentation methods are adopted, and the specific temporal context."

• Innovation:

  • The conceptual framework integrates communication and visualization research models while emphasizing the dynamic temporal context in visualization design.
  • This framework not only facilitates systematic analysis of existing crisis visualizations but also serves as a theoretical tool for future research.

• Implementation Steps and Techniques:

  • Data Collection: Opportunistic sampling of visualization examples through blogs, search engines, and social media contributions.
  • Data Cleaning: Removal of duplicates and invalid entries, followed by coding of visualization content.
  • Analysis Strategy: Combined inductive and deductive coding methods, designing and applying a codebook with 61 codes for systematic evaluation.

Research Findings

• Specific Findings:

  • Identified six categories of information dissemination: pandemic severity, trend forecasting, crisis nature explanation, risk mitigation guidance, risk and equity communication, and multifaceted impact assessment.
  • Summarized current trends, technical coding methods, and common issues in public health crisis visualization design (e.g., insufficient data normalization).
  • Beyond traditional charts and maps, innovative multivariate visualizations and narrative data presentations showed significant potential during the crisis.

• Comparison with Existing Solutions and Advantages:

  • Comprehensive analysis revealed common issues in existing crisis visualization designs (e.g., "misleading" design examples) and introduced an innovative framework for more systematic design.
  • The conceptual framework aids in developing more timely and audience-oriented dissemination strategies.

• Experimental or Evaluation Results:

  • Found that COVID-19 visualization design content often evolved over time, such as maps transitioning from simple bubble charts to complex choropleth maps.
  • Public engagement, trend changes, and data-driven approaches emerged as critical feedback sources for crisis visualization design.

• Limitations and Future Directions:

  • Limitations:
    • The dataset primarily sourced from Western (especially U.S.) COVID-19 visualizations, potentially lacking global coverage.
    • Did not retrospectively review all historical crisis visualizations, limiting a long-term perspective.
  • Future Directions:
    • Extend the framework to accommodate different types of crisis contexts.
    • Investigate the impact of crisis visualizations on public behavior, emotions, and trust.
    • Explore the differences in social influence between various information sources and design choices.
    • Develop evaluation methods for rapidly changing and long-term impactful graphics.