Guidelines
Suggest edit

Add icon arrays to numerical risk-reduction displays

For treatment-benefit comparison in non-temporal medical risk displays, use icon arrays alongside numerical risk information on paired baseline-versus-treatment views to improve fidelity and mitigate misunderstanding of risk reduction for mixed-numeracy audiences.

  • purpose:refine
  • basis:empirical
  • task:compare
  • quality:fidelity:use
  • lever:encoding
  • operator:difference
  • knowledge:mixed

advice

Add icon arrays to numeric risk reductions

Add icon arrays to numerical risk descriptions when readers must judge how much a treatment changes risk. For example, pair the numerical statement with one icon array for risk without treatment and one for risk with treatment, and do this even when the numbers already use absolute risk reduction.

reason

Why paired icon arrays improve accuracy

Paired icon arrays make the untreated and treated groups visible as counts out of a common whole. That helps readers recover the baseline risk, the treated risk, and the reduction more accurately from the display.

Mechanism: Icon arrays externalize the comparison between two risks so readers do less mental conversion from percentages or relative reductions.

Evidence: In controlled experiments with older adults and students, adding icon arrays to numerical risk descriptions increased correct understanding of risk reduction, with especially large gains for below-median numeracy participants and additional gains even when the numbers were already expressed as absolute risk reduction (Galesic et al., 2009).

context

Use when explaining treatment benefit from two risk values

  • User Goal: Explain how much a treatment or screening changes risk.
  • Task: Compare baseline risk with treated risk and infer the reduction.
  • Data: Two risk values for the same outcome, without treatment and with treatment.
  • Chart Setting: Numerical risk text is already present, and you can add paired icon arrays for untreated and treated outcomes.
  • Audience: Readers with mixed numeracy, including lower-numeracy adults.
  • Success Criterion: Readers can correctly estimate the size of the risk reduction.

exceptions

Do not use when urgency is the main communication goal

Break it when: The communication goal is to make the risk or treatment benefit feel more serious. Why: In this study, icon arrays made both baseline risks and treatment benefits seem less serious than numerical presentations.

costs

What you trade away

Sacrifice: You may reduce the perceived seriousness of the same risk or benefit. Risk: If you also change the total number of icons, equivalent risks can start to look larger or smaller for reasons unrelated to the data. Mitigation: Keep denominator size deliberate and consistent across comparable icon arrays.

mistakes

Common failure mode with icon arrays

Mistake: Adding icon arrays only to boost understanding while ignoring that they also change perceived seriousness. Why it fails: The display can improve accuracy and lower perceived seriousness at the same time.

check

How to test the revision

Failure Sign: Readers cannot correctly state the untreated risk, the treated risk, or the reduction from the display. Quick Check: Compare a numerical-only version against a numerical-plus-icon-array version and ask readers how many out of a fixed population are affected without and with treatment. Stronger Test: Split results by lower and higher numeracy and confirm that the icon-array version improves correct answers in both groups, especially the lower-numeracy group.

fix

What to change

  • Add one icon array for the baseline risk and a matched second icon array for the treated risk.
  • Keep the numerical statement and use the icon arrays as a supplement to it.
  • Add icon arrays even if the numerical text already uses absolute risk reduction.
  • Keep the same total number of icons across comparable views.

References

Galesic, M., Garcia-Retamero, R., & Gigerenzer, G. (2009). Using icon arrays to communicate medical risks: Overcoming low numeracy. Health Psychology, 28(2), 210–216. https://doi.org/10.1037/a0014474