Guidelines
Suggest edit

Use wrapped bars instead of treemaps for single-item rank lookup

For single-item rank lookup in dense record lists, use wrapped bars on ranked-list views instead of treemaps to improve fidelity and mitigate item-rank errors for readers making perceptual rank judgments.

  • purpose:select
  • basis:empirical
  • chart:bar:use
  • chart:treemap:avoid
  • quality:fidelity:use
  • lever:chart-family
  • operator:rank
  • density:dense

advice

Wrapped bars for rank lookup

Choose a wrapped bar layout when readers need to judge where one item falls in a long ranked list. For example, use wrapped bars rather than a treemap when a list of roughly 75-300 items must stay on one screen and the main readout is the selected item’s rank.

reason

Why wrapped bars work here

Length-encoded bars in wrapped columns keep the full list visible while preserving an ordered reading path for rank judgments. Area-coded treemap tiles stay compact, but they gave less accurate single-item ranking in this study.

Mechanism: Wrapped bars preserve length judgments and visible order cues for one marked item while avoiding scroll-driven interruption.

Evidence: In the collated results, wrapped bars had the lowest error on the single-item rank task and significantly outperformed treemaps in accuracy, while treemaps were among the faster conditions on the same task (Zeng & Battle, 2023; Mylavarapu et al., 2019).

Notes: The evidence comes from low-level perceptual tasks on unlabeled ranked-list charts.

context

Use when all of these are true

  • User Goal: Find where one highlighted item sits in the full ordering.
  • Data: A sorted ranked list with many items.
  • Chart Setting: The full list should fit within one screen without scrolling.
  • Audience: Readers are relying on visual marks more than text labels.
  • Success Criterion: Lower rank-estimation error matters more than raw speed.

exceptions

Do not use when any of these are true

Break it when: the main success criterion is faster single-item ranking rather than more accurate ranking. Why: treemaps completed faster than wrapped bars on this task.

costs

Tradeoffs of wrapped bars here

Sacrifice: You give up some speed relative to faster compact layouts. Risk: Multiple column baselines reduce horizontal resolution and can make across-column reading less direct. Mitigation: Use this choice specifically for single-item rank lookup, then validate it on representative long lists.

mistakes

Common failure mode

Mistake: Replacing a treemap with any compact ranked-list layout. Why it fails: packed bars, piled bars, and Zvinca plots were all less accurate than wrapped bars for this task.

check

How to check the choice

Failure Sign: Reviewers frequently misplace a highlighted item’s position in the ranking. Quick Check: A/B test a wrapped-bar version against a treemap version with one-item rank questions on representative long lists. Stronger Test: Compare normalized absolute rank error across several list sizes and keep the lower-error option.

fix

What to change

  • Replace the treemap with a wrapped bar layout.
  • Split the sorted bars into columns so the whole ranked list stays visible on one screen.
  • Re-test the design if speed becomes the main requirement for this task.

References

Mylavarapu, P., Yalcin, A., Gregg, X., & Elmqvist, N. (2019). Ranked-List Visualization: A Graphical Perception Study. Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems, 1–12. https://doi.org/10.1145/3290605.3300422
Zeng, Z., & Battle, L. (2023). A Review and Collation of Graphical Perception Knowledge for Visualization Recommendation. Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, 1–16. https://doi.org/10.1145/3544548.3581349