Guidelines
Suggest edit

Stack bar segments to show part-whole structure

For part-whole reading in categorical composition displays, use stacking in bar charts to improve readability and mitigate separate-part interpretations for readers.

  • purpose:refine
  • basis:empirical
  • chart:bar
  • data:categorical
  • operator:part-whole
  • quality:readability:use
  • lever:layout-structure

advice

Stacking for composition

Stack bar segments when the message is that several values add up to one whole. For example, use a stacked bar instead of separate bars when the chart must make the part-to-whole relation explicit.

reason

Why stacking changes the meaning

Part-whole charts need a part-whole metaphor, not only a precise comparison channel. A stacked bar keeps the total visible and makes each component read as belonging to that total.

Mechanism: Stacking embeds each part inside one shared whole, while separate bars frame the same values as independent quantities.

Evidence: The paper explicitly contrasts pie charts, stacked bars, and separate bars, arguing that stacked bars and pies express part-to-whole structure while separate bars do not, even though separate bars can be more precise for value comparison (Bertini et al., 2020).

context

Use when the whole must stay visible

  • User Goal: Show that components belong to one total.
  • Task: Part-whole interpretation rather than only comparison among parts.
  • Data: Categorical components of a shared whole.
  • Chart Setting: A bar-based display where you can choose between stacking the parts or separating them.
  • Audience: Readers who need the composition relationship to be obvious.
  • Success Criterion: Readers immediately see each segment as part of one total.

exceptions

Do not use when exact comparison among parts is primary

Break it when: The main job is precise comparison between the parts themselves. Why: Separate bars give a more direct comparison frame than stacked segments.

costs

Costs of stacking the bars

Sacrifice: You give up some comparison precision between individual parts. Risk: Blind stacking can make exact differences between parts harder to judge. Mitigation: Return to separate bars when comparing the parts is more important than expressing the whole.

mistakes

Common composition mistake

Mistake: Leaving the parts as separate bars when the message is composition. Why it fails: The layout removes the part-to-whole metaphor and makes the values look independent.

check

Check whether stacking is needed

Failure Sign: Reviewers talk about separate values but not about one whole. Quick Check: Show a stacked and separate version of the same parts, then ask which one makes the total obvious at a glance. Stronger Test: Ask a reviewer to explain the relationship among the values; if they do not mention a shared whole, the separate layout is obscuring the message.

fix

Fix the part-whole structure

  • Combine the separate bars into one stacked total.
  • Keep each component inside the shared total so the whole remains visible.
  • Remove layouts that isolate the parts when the chart’s message is composition.

References

Bertini, E., Correll, M., & Franconeri, S. (2020). Why Shouldn’t All Charts Be Scatter Plots? Beyond Precision-Driven Visualizations. 2020 IEEE Visualization Conference (VIS), 206–210. https://doi.org/10.1109/VIS47514.2020.00048