Through a Live Elections Dashboard, Darkly: Managing Expectations and Trust in Progressive Vote Counting During the 2024 U.S. Election

Paper Title

Through a Live Elections Dashboard, Darkly: Managing Expectations and Trust in Progressive Vote Counting During the 2024 U.S. Election

Publication Info

• Topic area: Public-facing progressive visualization for live election data.
• Keywords: Progressive visualization, live dashboards, election trust, vote counting, data journalism, uncertainty communication, public trust, live forecasts, U.S. elections, contested data.

Background and Problem

• Problem / challenge: Existing live election dashboards struggle to contextualize partial, evolving vote counts, leading to misinterpretations and reduced trust in elections. There is limited empirical research on how audiences engage with such dashboards and how they impact trust.
• Significance: Misinterpretations of vote trends can erode public trust in elections and news media, especially in politically charged contexts. Improved dashboard designs could help manage expectations and reduce susceptibility to misinformation.
• Motivation and related work: Prior work has explored progressive visualization and election dashboards but has not adequately addressed how to design for public understanding of partial data or how to empirically evaluate audience engagement and trust. This paper builds on these gaps by designing and testing a live dashboard during the 2024 U.S. presidential election.

Solution

• Proposed approach: A live elections dashboard designed to foster understanding of progressive vote counts, manage expectations about vote counting timelines, and reduce misinterpretations of trends.
• Novelty:
  1. A framework for how audiences form expectations around live data.
  2. A dashboard design emphasizing partiality, progressive trends, and explanatory text.
  3. Empirical evaluation of the dashboard's impact on trust and perceptions of fraud during a real-world election.
• Procedure and key techniques:
  • Conducted interviews with data journalists, election analysts, and lay participants to identify design objectives.
  • Prototyped and iteratively refined dashboard components, including live forecasts, progress indicators, and explanatory text.
  • Deployed the dashboard during the 2024 U.S. presidential election, collecting survey and interview data from 308 participants.
  • Analyzed quantitative survey responses and qualitative interview themes to assess the dashboard's impact.

Results

• Concrete findings:
  • Live forecasts possibly increased confidence in vote counting (0.43, 95% CI = [-0.03, 0.9]) and slightly reduced belief in evidence of fraud (-0.6, 95% CI = [-1.09, -0.1]).
  • Text explanations had no meaningful effect on confidence or belief in fraud.
  • Participants valued the dashboard as a factual reference point and appreciated its emphasis on agency and partiality.
• Advantage over baselines:
  • The dashboard's live forecasts and progress indicators helped participants understand uncertainty and the progressive nature of vote counts, distinguishing it from traditional dashboards.
• Experiments / evaluation:
  • Participants were assigned to one of four dashboard conditions (with/without live forecasts and text explanations).
  • Surveys captured trust, confidence, and fraud perceptions before, during, and after the election.
  • Post-election interviews explored participants' experiences and perceptions in depth.
• Limitations and future work:
  • Participant sample skewed liberal and educated, limiting generalizability.
  • Findings may not apply to other elections or contexts.
  • Dashboard outages during early election results may have impacted retention and engagement.
  • Future work should explore forecast formats, longer elections, and the role of external information and emotions in shaping expectations.

Summary

This paper presents a live elections dashboard designed to improve public understanding of progressive vote counts and manage expectations during the 2024 U.S. presidential election. The dashboard incorporated live forecasts, progress indicators, and explanatory text to address challenges in communicating uncertainty and partiality. Empirical evaluation suggested small positive effects on trust and confidence, with participants valuing the dashboard as a factual, agency-supporting reference point. However, live forecasts were perceived as less useful in this particular election, raising questions about designing forecasts that balance uncertainty, trustworthiness, and utility. The study contributes a framework for designing public-facing progressive visualizations and highlights the importance of considering external information and emotions in contested political contexts.