Guidelines
Suggest edit

Use a histogram instead of a box plot for novice distribution reading

For distribution reading, prefer a histogram over a box plot on quantitative data to improve readability and mitigate dependence on quartile-based statistical knowledge for novice readers.

  • purpose:select
  • basis:empirical
  • task:distribute
  • chart:histogram:use
  • chart:box-violin:avoid
  • lever:chart-family
  • quality:readability
  • literacy:novice

advice

Choose the chart family for distribution reading

Use a histogram instead of a box plot when non-expert readers need to read a quantitative distribution. For example, show the distribution with bins and counts in a histogram, and avoid a box plot when the reader would need percentile, quartile, or interquartile-range knowledge before they can interpret the chart.

reason

Why the chart choice works

A histogram lets the reader inspect the displayed distribution directly. A box plot shifts the task toward statistical terminology that non-expert readers may not already know.

Mechanism: The histogram keeps the reading task on visible distribution shape and binned frequency. The box plot adds a prerequisite layer of percentile and quartile knowledge before the reader can interpret the marks.

Evidence: The VLAT was designed for non-expert users and explicitly excluded box plots because understanding them requires specific statistical knowledge such as percentile, quartile, and interquartile range, while histograms were retained in the test blueprint (https://doi.org/10.1109/TVCG.2016.2598920">Lee et al., 2017).

context

Use when the audience is reading a distribution

  • User Goal: Understand the shape or spread of a quantitative distribution.
  • Task: Distribution reading by a non-expert audience.
  • Data: One quantitative variable.
  • Chart Setting: A static chart where the reader should interpret the display without prior statistical instruction.
  • Audience: Novice or non-expert readers.
  • Success Criterion: Readers can interpret the distribution without learning quartile-based terms first.

exceptions

Do not use when quartile knowledge is already part of the task

Break it when: The audience already knows percentile, quartile, and interquartile-range concepts, and those summaries are part of the intended message. Why: Then the extra statistical abstraction of a box plot is not a barrier.

costs

Tradeoffs of the chart swap

Sacrifice: You give up the compact quartile summary of a box plot.
Risk: A histogram depends on binning choices, so the displayed shape reflects those bins.
Mitigation: Use the histogram when the main goal is readable distribution interpretation by novices, not compact statistical summarization.

mistakes

Common failure mode

Mistake: Using a box plot for a novice audience and expecting immediate distribution reading. Why it fails: The reader must first understand percentile and quartile concepts before the display becomes interpretable.

check

How to test the choice

Failure Sign: Readers ask what the box, whiskers, or quartiles mean before they can answer the distribution question.
Quick Check: Compare a histogram and a box plot for the same data with a novice reviewer; keep the histogram if the box plot needs concept explanation first.
Stronger Test: If the reader can answer the distribution question from the histogram but not from the box plot without teaching quartiles or percentiles, choose the histogram.

fix

What to change

  • Replace the box plot with a histogram for the novice-facing version.
  • Phrase the task around the visible distribution rather than quartile-based summaries.
  • Remove explanations that are only needed to decode percentile or quartile terminology.

References

Lee, S., Kim, S.-H., & Kwon, B. C. (2017). VLAT: Development of a Visualization Literacy Assessment Test. IEEE Transactions on Visualization and Computer Graphics, 23(1), 551–560. https://doi.org/10.1109/TVCG.2016.2598920