Keep denominator size consistent across equivalent icon arrays

Keep the total number of icons consistent when comparing equivalent risks or treatment effects. For example, do not switch between 100-icon and 1,000-icon arrays for the same percentage, because the larger array makes the baseline risk and especially the treatment benefit look larger.

A larger denominator makes more people appear affected or helped even when the proportion is unchanged. That shifts perceived seriousness and helpfulness away from the underlying rate.

Mechanism: Readers react to the larger absolute count of marked icons, not only to the proportion they represent.

Evidence: In the experiment, 1,000-icon arrays produced somewhat higher seriousness ratings than 100-icon arrays for equivalent baseline risks and significantly higher helpfulness ratings for equivalent screening benefits (Galesic et al., 2009).

• User Goal: Compare risks or benefits fairly across displays.
• Task: Show equivalent rates without changing their perceived size.
• Data: The same proportion can be rendered with different possible denominators.
• Chart Setting: Icon arrays are used for untreated and treated outcomes.
• Audience: Readers interpreting baseline risk and treatment benefit.
• Success Criterion: Equivalent rates are not made to feel larger or smaller by denominator choice.

Break it when: The absolute number of people represented is itself the message rather than an interchangeable presentation of the same rate. Why: The source tests denominator size as an alternative display choice for equivalent risks.

Sacrifice: You give up a presentation lever that can make the same rate feel larger. Risk: Switching from 100 to 1,000 icons can make equal risks and especially equal risk reductions seem bigger for reasons unrelated to the data. Mitigation: Use one denominator size consistently for every comparable icon array in the set.

Mistake: Varying total icon count across otherwise equal icon arrays while treating them as directly comparable. Why it fails: Readers infer different seriousness and helpfulness from denominator size alone.

Failure Sign: The same percentage appears in matched views with different total icon counts. Quick Check: Scan all comparable icon arrays and verify that they use the same denominator. Stronger Test: A/B test 100-icon and 1,000-icon versions of the same rate and check whether seriousness or helpfulness ratings drift.

• Choose one denominator size for a comparison set and apply it to every icon array in that set.
• If you revise one view from 100 to 1,000 icons, revise all matched views.
• Do not enlarge the icon array just to make the same rate or benefit feel more important.
• Keep the untreated and treated icon arrays matched within each comparison.