Use wrapped bars instead of Zvinca plots for two-item comparison
For two-item comparison in dense record lists, use wrapped bars on ranked-list views instead of Zvinca plots to improve fidelity and mitigate pairwise size-estimation errors for readers making perceptual comparisons.
- purpose:select
- basis:empirical
- task:compare
- chart:bar:use
- chart:dotplot:avoid
- quality:fidelity:use
- lever:chart-family
- density:dense
advice
Wrapped bars for pairwise comparison
Choose a wrapped bar layout when readers must decide which of two items is larger and by how much in a long ranked list. For example, use wrapped bars rather than a Zvinca plot when both items need to stay visible on one screen and comparison accuracy matters more than the fastest possible response.
reason
Why wrapped bars work here
Bar length supported more accurate two-item size judgments than the dot-based Zvinca layout in this study. The wrapped layout kept the list compact without dropping to the lowest-accuracy condition.
Mechanism: Wrapped bars preserve length comparisons for two marked items while keeping all items visible, which helps pairwise magnitude judgments.
Evidence: For two-item comparison, wrapped bars were in the top accuracy tier and significantly outperformed Zvinca plots, while Zvinca plots had the fastest completion times (Zeng & Battle, 2023; Mylavarapu et al., 2019).
Notes: The evidence comes from low-level perceptual comparison tasks on unlabeled ranked-list charts.
context
Use when all of these are true
- User Goal: Compare two highlighted items and estimate the difference between them.
- Task: Pairwise magnitude comparison inside one ranked list.
- Data: A sorted ranked list with many items.
- Chart Setting: The full list should remain visible within one display.
- Success Criterion: Lower comparison error matters more than minimum response time.
exceptions
Do not use when any of these are true
Break it when: the main success criterion is fastest completion time for two-item comparison. Why: Zvinca plots were the fastest condition on this task.
costs
Tradeoffs of wrapped bars here
Sacrifice: You give up speed relative to Zvinca plots. Risk: Per-column baselines can still make some across-column judgments less direct. Mitigation: Use this choice when comparison accuracy is the primary requirement, then test it on representative long lists.
mistakes
Common failure mode
Mistake: Choosing the fastest dot-based layout by default for pairwise comparison. Why it fails: the fastest condition here also had the lowest comparison accuracy.
check
How to check the choice
Failure Sign: Readers answer quickly but often miss which of two highlighted items is larger. Quick Check: A/B test a wrapped-bar version against a Zvinca version using two-item comparison questions. Stronger Test: Compare normalized absolute comparison error and completion time across representative list sizes.
fix
What to change
- Replace the Zvinca plot with a wrapped bar layout for the comparison view.
- Keep value encoded by bar length rather than dot position alone.
- Keep the list sorted and wrapped into columns so the full set remains visible without scrolling.