statcompute

Charts

Box Plot Maker

Paste one group or several and get a box and whisker plot, with the numbers behind it.

Updated August 2026Runs in your browser — nothing is uploadedVerified against R
One value per line, or separated by commas, tabs or spaces. A header row is ignored.

The calculation runs in your browser, so this box needs JavaScript. The formula, the worked example and the interpretation below do not.

01

How to read this result

The box is the middle half of the data. Its edges are the first and third quartiles, and the line inside is the median — not the mean, which is drawn as a hollow diamond so the two can be compared.

The whiskers are not the minimum and maximum. They reach the furthest value within 1.5 × IQR of the box; anything beyond that is drawn as a separate point. This is the single most common misreading of a box plot.

A median off-centre in its box means skew. If the line sits nearer the bottom edge, the upper half of the middle 50% is more spread out — the distribution leans right. The gap between the median line and the mean diamond says the same thing.

Box plots hide multimodality. Two distinct clusters and one even spread can produce identical boxes. When the shape matters, draw a histogram as well.

Comparing groups is what they are for. Several boxes on one scale make differences in centre, spread and skew visible at once, which no table of summary statistics does as quickly.

minimum insidethe fenceQ1medianQ3maximum insidethe fenceoutlierinterquartile range — the middle 50%
Every part of a box plot is one of the five numbers. The whiskers stop at the last value within 1.5 × IQR of the box, so they are not the minimum and maximum whenever an outlier is drawn separately.
02

The formula

IQR=Q3Q1whiskersreachthelastpointwithinQ11.5IQRandQ3+1.5IQR

Tukey chose 1.5 so that roughly 0.7% of normal data — about 1 observation in 150 — is drawn separately. It is a convention, not a test.

03

Worked example

Two treatment groups side by side

Five plots under a control and five under a treatment. A table of means would say the treatment is higher; the box plot says that and also shows that the two groups do not overlap at all.

Control: 23, 25, 21, 24, 22 · Treated: 28, 30, 27, 29, 31

  1. Sort each group and find its median.
    control: 21 22 23 24 25 → 23 treated: 27 28 29 30 31 → 29
  2. The hinges are the medians of each half, including the median itself.
    control Q1 = 22, Q3 = 24 treated Q1 = 28, Q3 = 30
  3. The interquartile ranges.
    both are 2
  4. The fences at 1.5 × IQR.
    control: 19 to 27 treated: 25 to 33
  5. No value falls outside its own group's fences.
    no outliers
control median 23treated median 29both IQRs 2outliers none

The two boxes do not overlap, which is a strong visual signal — and a t-test on these numbers returns p = 0.0002, confirming it. Non-overlapping boxes usually mean a significant difference; overlapping ones do not mean the opposite.

Checked against R's fivenum and boxplot.stats.

The calculator above is loaded with these numbers by the Load the worked example button.

04

Assumptions, and when to use something else

None about the distribution — box plots work on skewed, ordinal and outlier-ridden data, which is what they were designed for.

Two things to state when you publish one: which quartile convention you used (the calculator names it), and how many observations each box represents. A box plot of four points looks exactly as authoritative as a box plot of four hundred, which is why n belongs in the caption.

  • You want to see the shape, not just the quartilesHistogram makerBox plots hide multimodality; histograms do not.
  • You want the numbers aloneFive-number summaryThe five values the plot is built from.
  • You want to test whether the groups differT-testTwo groups; use ANOVA for three or more.
  • The data set is small enough to keep every digitStem-and-leaf plotA distribution plot that does not discard the values.
05

Questions people ask

What do the whiskers on a box plot mean?

They reach the most extreme value still within 1.5 × IQR of the box. Points beyond that are drawn individually. So the whisker end is not the minimum or maximum whenever an outlier exists.

Can I make a box plot with several groups?

Yes — use the second mode and put one group per line, or separate them with blank lines. All groups are drawn on a shared scale so they can be compared directly.

Why does my box plot differ from Excel's?

Almost certainly the quartile convention. Box plots traditionally use Tukey's hinges; Excel's QUARTILE interpolates. The method selector above lets you match either, and the result names which was used.

How do I download the chart?

Every chart draws to a canvas in your browser. Use the download button for a PNG, or right-click the image and save it. Nothing is uploaded to produce it.