Did you know that the middle score in a class may be a score nobody earned? A median can be mathematically correct and still describe the class poorly if the marks fall into two separated groups. That is an important lesson in what education is for: using a number without losing the people and patterns behind it.
The median is the middle value of sorted data, or the average of the two middle values when the number of observations is even. If the scores form two clusters with a large gap, the median may fall in that gap. The immediate answer is to report the median accurately and describe the two clusters instead of calling the median the typical pupil’s result.
A small diagnostic: does the middle represent any student?
Ask a learner to write down the following ten illustrative quiz marks, each out of 20: 3, 4, 4, 5, 5, 15, 16, 16, 17, 17. Before calculating anything, ask: Are there many marks near 10? What different stories could produce two groups? The purpose is not to label children or assume a reason. It is to see how much information a single summary loses.
Many learners immediately select the fifth score (5) as the median because they forget that with ten values there are two middle positions. Others say that a median of 10 means most pupils scored near 10. Both errors are useful diagnostic signals; they call for a different explanation, not simply more arithmetic drills.
Worked example: a correct answer with a misleading description
- Count the data: n = 10. They are already in ascending order.
- Find the two middle positions: position 5 is 5 marks; position 6 is 15 marks.
- Calculate the median: (5 + 15) ÷ 2 = 10 marks out of 20.
- Check the mean independently: (3 + 4 + 4 + 5 + 5 + 15 + 16 + 16 + 17 + 17) ÷ 10 = 102 ÷ 10 = 10.2 marks.
- Read the pattern: five marks lie from 3 to 5; five lie from 15 to 17. No mark lies from 6 through 14.
Both 10 and 10.2 are valid summaries, but neither is a score earned by anyone in this invented example. This is a two-cluster distribution: most of its observations sit in two separated regions. It is not evidence that the class is divided into permanent ability groups; one quiz and ten marks cannot establish that.
A helpful report would say: ‘The median score was 10/20, although the ten marks formed two clusters, 3–5 and 15–17, with no marks in between.’ That sentence preserves a valid calculation and describes what the calculation does not show.
Why the mechanism matters
A median identifies position in an ordered collection. It does not automatically identify the most frequent value or the most representative observation. The mode describes the most frequent value; a dot plot or histogram shows gaps and clusters. With these data there is no single unique mode: 4, 5, 16 and 17 each occur twice. The distribution shape is a separate fact from its centre.
The US National Institute of Standards and Technology explains that histograms can reveal bimodal patterns; noticing the two peaks invites further investigation of the underlying process. The fact that a centre lies between peaks does not make the median formula wrong. It means the centre alone cannot tell the complete story.
Teach it progressively: from guided checking to independent judgement
First: recognise the structure
Give the child two ordered sets: A = 7, 8, 8, 9, 10 and B = 2, 3, 4, 18, 19. Both sets contain five values, and both have their middle at position 3. Learners calculate medians of 8 and 4, then sketch each set as dots on a number line. Ask why the position is easy to find but the shape can differ.
Next: test a claim
Return to the ten-mark example. Say, ‘The typical pupil got 10/20 because the median is 10.’ Ask the learner to challenge that sentence with an observation, not an opinion. One possible answer: ‘None of the ten pupils scored between 6 and 14, so 10 is the central statistic, not a commonly achieved mark.’
Finally: work independently
Use the new hypothetical set 4, 5, 5, 6, 14, 15, 15, 16. The median is (6 + 14) ÷ 2 = 10, yet no one earned 10. Ask the learner to calculate the mean [80 ÷ 8 = 10], show the two clusters, and write one careful sentence without claiming causes. Check that the explanation gives evidence from the actual values.
A practical plan for pupils, parents and teachers
- Pupils: always count and sort first; calculate the statistic; then make a quick dot plot before using words such as ‘usual’ or ‘typical’.
- Parents: ask ‘Who actually scored near the median?’ rather than just ‘Is the answer right?’ This helps a child interpret information instead of memorising formulas.
- Teachers: compare a compact group and a split group with the same median, then ask learners to write the limitation of a one-number summary.
An observable sign of progress is that a student can find the correct median, notice an empty interval, and independently qualify an overconfident statement. Accuracy and interpretation are both necessary. For further practice on data calculations, see eduKate’s guide to counting visible spreadsheet rows, which asks a similar question: what did we actually count?
What this does not tell us
Two visible clusters do not, by themselves, diagnose teaching quality, pupil background, cheating, or the difficulty of an assessment. Even a histogram can suggest or conceal features depending on grouping; small samples are particularly sensitive to individual values. Investigate possible reasons only with more appropriate evidence and with care for pupil privacy.
Frequently asked questions
Is the median wrong if no pupil earned that score?
No. For an even number of observations, the median may be the average of two observed values. It need not occur in the data.
Should we always replace the median with the mean?
No. The mean can also land in the gap; in the example it is 10.2/20. Describe the distribution as well as its centre.
Is a two-cluster pattern always a bimodal distribution?
Two clear separated piles in a small dot plot can be described as two clusters. In statistical modelling, the exact number of modes depends on how data are represented and estimated. For school work, describe what the plotted data actually show without overstating a technical classification.
The learning goal
Mathematics becomes more useful when pupils can ask both ‘Did I calculate this correctly?’ and ‘What does this answer really mean?’ That’s the purpose of diagnosis-first, explanation-first teaching reflected in eduKate’s Secondary 1 Mathematics learning approach. Explore related skills in the Education Hub or the Core Aim of Education learning hub. The best answer is both true and well interpreted.