Statistics and data
Read a box plot: quartiles, spread, and what it cannot show
Find the median and middle half, compare two groups carefully, and avoid treating a box plot as a picture of every observation.
A box plot compresses a whole set of values into a small picture. That makes it useful when a statistics assignment asks which group is more typical, which has greater spread, or where the middle half of observations lies. It also makes the plot easy to overread. The box does not display every value, and its central line is the median, not automatically the mean. Learning to read a box plot means knowing both what the marks encode and what they leave hidden.
You can work through a box plot without an app or paid tool. Start with the axis and the plotting convention, locate the five landmarks, calculate the interquartile range, and turn any comparison into a sentence tied to the data. This guide uses concrete numbers and a short practice example so you can check your interpretation rather than memorize the shape. If a teacher uses a different quartile or whisker rule, follow that rule and say which one you used.
1. Identify the scale before reading the box
First read the title, measured quantity, units, and scale. A horizontal box plot might show minutes of travel, while a vertical one shows scores. The method is the same, but the numerical direction changes. Trace the axis from low to high and note whether tick marks are evenly spaced. A box that looks twice as long can only support a numerical comparison when both groups share a compatible scale.
Find out how the whiskers are defined. In a basic school plot they often reach the observed minimum and maximum. In another common convention they stop at the most extreme values that are not classed as outliers; separate points show outliers beyond them. You cannot identify the minimum from a whisker tip in that second convention if there is a farther plotted point. The legend, textbook, or task instruction settles the question.
Before calculating, write one plain sentence about what the data represent: for example, ‘Each value is the number of minutes one student spent on a journey.’ That prevents you from calling a range of 12 minutes a number of students. OpenStax gives a clear reference for the five-number summary and box-plot reading; compare its convention with the one in your course.
2. Locate the five landmarks
In the simple minimum-to-maximum version, read the left whisker tip, left box edge, line inside the box, right box edge, and right whisker tip. They are minimum, first quartile Q1, median Q2, third quartile Q3, and maximum. If the plot is vertical, read from bottom to top. Mark each value against the axis instead of estimating from the size of the drawing alone.
Imagine a group with minimum 4, Q1 8, median 10, Q3 13, and maximum 18 minutes. The middle line says that the ordered observations split around 10 minutes. The interval from 8 to 13 contains the middle half of the observations, approximately 50 percent. The interval from 4 to 18 describes the full observed span under the simple whisker convention.
Quartiles are positions in ordered data, not four equal-width stretches of the axis. A long interval from Q3 to the maximum can contain roughly a quarter of the observations just as a short interval from Q1 to the median can. Its greater length shows that those observations are more spread out numerically. It does not tell you how many students are present without a separate sample-size label.
3. Reconstruct quartiles from a short list
When a task supplies raw values rather than a finished plot, sort them first. For the eight values 2, 4, 6, 8, 10, 12, 14, and 18, the median is halfway between 8 and 10, so it is 9. The lower half is 2, 4, 6, 8; its median is halfway between 4 and 6, giving Q1 = 5. The upper half is 10, 12, 14, 18; its median is halfway between 12 and 14, giving Q3 = 13.
This produces the five-number summary 2, 5, 9, 13, 18. The arithmetic is short, but every step has a purpose: sort, locate the centre, split the lower and upper halves, then locate their centres. When there is an odd number of values, curricula may disagree about whether to include the overall median in both halves. Software can also use interpolation. State the rule you followed instead of treating a one-unit difference as a mysterious error.
A box plot drawn from the summary would place the box edges at 5 and 13 and the middle line at 9. The whiskers would reach 2 and 18 if your class uses minimum and maximum. Sketch the scale before drawing the marks, then verify that each value is in increasing order. If the median falls outside the box, either the values were copied incorrectly or the marks were placed in the wrong positions. For a refresher on mean, median, mode, and range, compare what those summaries answer before you return to quartiles.
4. Calculate two different kinds of spread
The interquartile range, or IQR, is Q3 minus Q1. In the first imagined group, 13 − 8 = 5 minutes. It describes the width of the middle half, not the average deviation from the median. The full range is maximum minus minimum, here 18 − 4 = 14 minutes. These two measures answer different questions: middle consistency and total observed span.
Use the units in the data. If test times are measured in minutes, both IQR and range are in minutes. If the data are scores on a 100-point test, the differences are score points. Do not say the IQR is ‘five students’ because 5 is a distance along the value axis, not a frequency. A useful verbal reading is ‘the central 50 percent spans five minutes.’
5. Compare two box plots without overstating
Suppose group A has minimum 4, Q1 8, median 10, Q3 13, maximum 18. Group B has minimum 2, Q1 7, median 11, Q3 17, maximum 23. Start with the middle: B has a median one minute higher. Then compare the boxes: A has IQR 5 minutes, while B has IQR 10 minutes. The middle half of B is therefore more spread out. The full ranges are 14 and 21 minutes respectively.
Write a comparison that respects both location and spread: ‘Group B has a slightly higher median journey time, but its central half varies more.’ A larger median does not prove every B observation exceeds every A observation; the ranges overlap substantially. Nor does the larger IQR tell you which group is better. That judgment depends on the measured outcome and the question being asked.
Check whether the groups have similar sample sizes and whether the same axis and whisker rule are used. Box widths sometimes encode sample size, but ordinary textbook boxes often do not. If no sample sizes are given, avoid claiming that one group contains more people. If the plots use different units, convert them first or do not compare their lengths at all. A defensible sentence names the measured quantity, the relevant statistic, and the limit of the inference.
6. Read outliers and asymmetric whiskers carefully
In a modified box plot, a point beyond a whisker may mark an outlier under a specified rule. A common rule marks values below Q1 − 1.5 × IQR or above Q3 + 1.5 × IQR. This is a screening convention, not a declaration that a value is wrong. A very long commute, for example, may be genuine and important to the question. Inspect the source and measurement before excluding it.
Do not calculate a mean from the five marks. Many different datasets share the same five-number summary while having different averages. Also avoid turning a single distant dot into a story about a particular person. The plot may represent anonymized measurements, and an unusual value needs context. Your assignment may ask for a description of distribution, not a cause; separate what you can observe from what you merely suspect.
7. Resolve common misreadings before seeking help
The most common slip is labelling the centre line as the mean. Say ‘median’ unless the plot explicitly adds a mean marker. A second slip is treating a wider box as a larger number of observations; under a standard box plot, it means greater numerical spread for the middle half. A third is reading the whisker ends as minimum and maximum without checking for separately plotted outliers.
If you are stuck, work from the paper first: copy the scale, annotate the five positions, and calculate Q3 − Q1. You can compare the result with the source material or use a free manual practice question before opening any tool. Lirno is free to download or start, while some AI usage levels or advanced features may require Premium. It may be useful for a narrow explanation or for turning your own notes into practice questions, but the box-plot method in this guide works on its own.
If you do use AI, it can misread a small axis mark, mistake an outlier dot for a whisker end, or reason incorrectly. Check its interpretation against the original graph and your course convention. School rules still apply to assessed work; Lirno does not guarantee correctness, grades, mastery, or permission. Keep the explanation you submit in your own words and be ready to point to the marks that support it.
8. Try a short comparison without looking back
Create two simple five-number summaries on paper: C = 3, 6, 9, 12, 15 and D = 1, 5, 9, 16, 20. Draw both on the same numbered line using the minimum-to-maximum whisker convention. Pause before reading on. Which group has the greater median? Which has the wider middle half? Which has the larger full range? Give a reason for each answer, not just a letter.
Both medians are 9, so neither group has the greater median. C has IQR 12 − 6 = 6; D has IQR 16 − 5 = 11, so D's central half is wider. C's range is 15 − 3 = 12; D's is 20 − 1 = 19. D also has the larger total span. These claims are supported by the supplied five numbers. Neither summary tells you the sample mean or how many observations lie at exactly 9.
Now change only D's maximum from 20 to 30 and redraw it. The median and IQR stay the same, while the total range grows to 29. That one change demonstrates why a statement about the middle half is not interchangeable with a statement about endpoints. Create one more version where only Q1 changes, and predict which measures must move. For a repeatable self-test, see how to make a quiz from notes using your own five-number summaries.
9. Finish with an evidence-based reading
A reliable final answer has four parts. Name the measured variable and units. Report the relevant location, usually the median. Report spread using the IQR, range, or both as the question requires. Finally qualify the conclusion: mention overlap, possible outliers, missing sample sizes, or the plotting convention when they affect the claim. This structure is more useful than saying that one box simply ‘looks bigger.’
For a comparison, choose the statistic that answers the actual prompt. ‘Which group has a more consistent middle half?’ calls for IQR. ‘Which group has the higher typical value?’ often calls for median. ‘Which includes the most extreme observation?’ calls for endpoints and the outlier display. If the question asks about every data value or the mean, explain why the box plot alone cannot settle it and request the underlying data.