Use uncertainty-revealing rendering when certainty is not warranted

Add rendering that visibly signals tentativeness when the data should be questioned. For example, use sketchy drawing styles or explicit uncertainty-conveying renderings instead of a clean minimalist presentation when uncertainty, missingness, or bias is part of the message.

A polished, exact-looking chart can make uncertain data feel more authoritative than it is. Visual signals of tentativeness can slow overconfident reading and make viewers more willing to question the evidence.

Mechanism: Clean minimalist rendering can project objectivity and certainty, while sketchy or uncertainty-revealing styling makes uncertainty part of what the reader perceives.

Evidence: The paper argues that clean layouts can grant unwarranted authority and objectivity to data, while sketchy and uncertainty-conveying renderings can increase willingness to critique and appreciation of uncertainty (Bertini et al., 2020).

• User Goal: Communicate uncertainty, tentativeness, or the need for critique.
• Task: Uncertainty communication rather than only fast value extraction.
• Data: Uncertain, missing, biased, or otherwise not fully settled.
• Chart Setting: A quantitative display that might otherwise look overly clean and authoritative.
• Audience: Readers who might otherwise take the chart as objective truth.
• Success Criterion: Readers notice uncertainty and remain willing to question the display.

Break it when: The display is meant for fast, exact reading of settled values and uncertainty is not part of the message. Why: The extra visual difficulty can slow decoding without adding needed information.

Sacrifice: You give up some speed and precision in immediate value reading. Risk: Blind use can make already-clear data harder to decode. Mitigation: Apply the treatment when the chart should explicitly communicate uncertainty or invite critique.

Mistake: Keeping a clean minimalist rendering when uncertainty is central to the story. Why it fails: The chart can look more objective and settled than the data deserves.

Failure Sign: Uncertain data looks fully settled and authoritative. Quick Check: Remove the title and caption and ask whether the marks themselves still signal uncertainty. Stronger Test: Compare a clean version and an uncertainty-revealing version; if the clean version looks more certain than the data warrants, the original is masking uncertainty.

• Apply sketchy styling to the uncertain marks.
• Add explicit uncertainty-conveying rendering where values are missing or uncertain.
• Replace a purely clean minimalist treatment that hides the tentativeness of the data.