2026
How Data Narratives Go Wrong examines how issues arise, propagate, and compound in data communication. We introduce TIC (Taxonomy of Issues in Data Communication), developed by synthesizing 34 prior works and qualitatively annotating 700 real-world data narratives.
TIC spans data, analysis, visual encoding, text, reasoning, and audience interpretation. A communication-process framework situates these issues across analysis, narrative construction, and consumption. The annotated corpus and interactive explorer let readers examine cases, coding rationales, and sources. A single narrative can contain multiple interacting issues.
Team: Yu Fu, Jiawei Zhou, Sichen Jin, Munmun De Choudhury, Cindy Xiong Bearfield, and John Stasko.
The paper is available as a public preprint.
Explore the TIC taxonomy
Six dimensions of issues in data-driven communication.
Hover or select an issue type for a definition condensed from the paper, §IV-A. Sub-types are non-exhaustive.
Issues that originate before analysis or visualization and compromise the underlying dataset's integrity, completeness, or credibility.
- Measurement Error; Missing/
Incomplete Data - Sampling Bias; Outdated Data
- No Supporting Data; Unreliable Data Source