Guidelines
Suggest edit

Use no more than four shapes in grouped scatterplots

For grouped scatterplots, prefer shape encoding on point marks to improve accessibility and mitigate color-only grouping, and avoid large shape sets that create visual clutter for readers with color-vision deficiency.

  • purpose:refine
  • basis:heuristic
  • chart:scatter
  • quality:accessibility
  • lever:encoding
  • channel:shape:use
  • group-cardinality:few
  • needs:color-vision-deficiency

advice

Replace color-only grouping with a small shape set

Use distinct point shapes in a scatterplot when color alone separates groups. For example, replace some circles with triangles, rectangles, crosses, or stars, but stop at three or four shapes instead of giving every group a different marker.

reason

Why a small shape set works

Shapes provide a noncolor way to separate groups, but too many shape types overwhelm the display. A small, repeated set preserves group identity without turning the plot into visual confetti.

Mechanism: Shape differences survive color confusion, while limiting the number of shapes preserves a readable point field.

Evidence: The article recommends using different shapes in scatterplots instead of just circles and warns that using all those shapes quickly looks like confetti, advising a limit of three or four. (Muth, 2020)

context

Use when a scatterplot has a few groups

  • User Goal: Distinguish grouped points in a scatterplot.
  • Task: Compare or identify categories in a point cloud.
  • Data: A few groups encoded on the same scatterplot.
  • Chart Setting: A scatterplot where color would otherwise be the only group marker.
  • Audience: Readers who may include people with color-vision deficiency.
  • Success Criterion: Groups remain distinguishable without the plot becoming cluttered.

exceptions

When shape encoding breaks down

Break it when: The scatterplot needs more than four distinct group markers. Why: The article warns that larger shape sets quickly make the plot look like confetti.

costs

Costs of adding point shapes

Sacrifice: You give up some visual uniformity in the point field. Risk: Too many shapes create clutter and slow scanning. Mitigation: Keep the shape set small and reuse only a few clearly distinct markers.

mistakes

Common scatterplot failure

Mistake: Encoding all groups with circles and asking readers to rely on color alone, or adding too many different shapes. Why it fails: The first choice depends entirely on color, and the second makes the plot visually noisy.

check

How to verify point shapes

Failure Sign: The plot either collapses into same-looking colored circles or turns into a field of many unrelated marker types. Quick Check: Count the distinct shapes in the plot and keep the total at four or fewer. Stronger Test: View the plot with color differences reduced and check whether groups still separate by marker shape.

fix

What to change

  • Replace some circle-only groups with clearly different marker shapes.
  • Limit the distinct marker set to three or four shapes.
  • Merge or split groups differently if the plot would need more shapes than that.

References

Muth, L. C. (2020). What to consider when visualizing data for colorblind readers. https://www.datawrapper.de/blog/colorblindness-part2