Mean vs Median vs Mode
Mean, median and mode are the three averages — three different answers to the question “what is a typical value?” The mean adds everything up and divides by the count, the median is the middle value once the data are sorted, and the mode is the value that appears most often. They agree on symmetric data and disagree on skewed data, and that disagreement is exactly what makes them useful.
In short, the mean is the average: add all the values and divide by how many there are. The median is the middle value after sorting the data. The mode is the most frequent value. On symmetric data all three agree; on skewed data the mean follows the tail and the median stays in the middle.
This guide works each average out by hand on the same datasets, from the perfectly symmetric 2, 4, 6, 8, 10 to a salary example in which one $500k pay packet moves the mean from $50k to $95k while the median moves only from $50k to $52.5k. Every example can be repeated in seconds on the Statistics Calculator, which computes the mean, median and mode — and the standard deviation — in one go.
What Are Mean, Median and Mode?
Three averages, three answers. The mean, the median and the mode each summarise a dataset with a single number, but each answers a different question. The mean asks: if the total were shared out evenly, what would each share be? The median asks: what value sits in the middle? The mode asks: what is the most common value?
Start with a simple example: the test scores 2, 4, 6, 8, 10. The mean is 6 — add them up (30) and divide by 5. The median is also 6 — the value in the middle after sorting. And the mode? There is none, because no value repeats. One dataset, three answers, and they do not have to agree.
The mean. Often called the arithmetic mean or simply the average, the mean uses every value in the data. It is the total divided by the count, and it is the only one of the three that can recover the total: total = mean × count. That is why the mean is the right average for money, revenue and totals — if the mean salary is $50,000 and there are 10 employees, the payroll is $500,000.
The median. The median is the middle value of the sorted data. Half the values lie below it and half above it. Because it is based on position rather than size, an extreme value can shift it by at most one place — which is why the median is the standard measure of a “typical” income, salary or house price.
The mode. The mode is the value that occurs most often. It is the only average that works for categories — shoe sizes, survey answers, the most common failure code — and it is the only one that can be plural: a dataset can have two or more modes, or none at all.
How to Calculate Each One
The formulas are short, but each has a step that is easy to get wrong.
Mean: x̄ = Σx / n. Add all the values (Σx) and divide by the number of values (n). For [2, 4, 6, 8, 10]: (2 + 4 + 6 + 8 + 10) / 5 = 30 / 5 = 6.
Median: find the (n + 1) / 2-th value after sorting. With an odd count, the median is the middle value: for [4, 8, 6, 8, 10], sorting gives [4, 6, 8, 8, 10], and the 3rd value is 8. With an even count, take the average of the two middle values — the n / 2-th and the (n / 2 + 1)-th. For [2, 4, 6, 8], the two middle values are 4 and 6, so the median is (4 + 6) / 2 = 5. Sort first: finding the median of unsorted data is the most common mistake in this guide.
Mode: the value with the highest frequency. Count how many times each value appears and take the most frequent. In [1, 2, 2, 3, 3, 3, 4, 5, 9], the value 3 appears three times — more than any other — so the mode is 3. If every value appears once, there is no mode.
All three are reported by the Statistics Calculator — enter the numbers separated by commas and it shows the mean, median and mode in one step, together with the range and the standard deviation.
Worked Example 1: All Three on One Dataset
Take the simplest possible dataset: [2, 4, 6, 8, 10]. It is perfectly symmetric, so the mean and the median agree — and the mode shows its limits.
| Measure | Value | How it is found |
|---|---|---|
| Mean | 6 | (2 + 4 + 6 + 8 + 10) / 5 = 30 / 5 |
| Median | 6 | Middle value of [2, 4, 6, 8, 10] |
| Mode | None | Every value appears once |
| Range | 8 | 10 – 2 |
| Population SD | 2.828 | √(40 / 5) |
| Sample SD | 3.162 | √(40 / 4) |
The mean and the median both say 6, but there is no mode at all. That is not a failure of the data — it is information: the dataset has no “most common” value, so the mode cannot summarise it. Run the same numbers through the Statistics Calculator and you will see “Mode: No mode” alongside the same figures.
Worked Example 2: A Right-Skewed Dataset
Now add some realism: [1, 2, 2, 3, 3, 3, 4, 5, 9]. Most values are small, and one value (9) sits far to the right. This is the same dataset used in the guide to skewness and kurtosis, where it scores an adjusted skewness of 1.728.
| Measure | Value | How it is found |
|---|---|---|
| Mean | 3.556 | (1 + 2 + 2 + 3 + 3 + 3 + 4 + 5 + 9) / 9 = 32 / 9 |
| Median | 3 | Middle value of the sorted list |
| Mode | 3 | 3 appears three times |
| Range | 8 | 9 – 1 |
| Population SD | 2.217 | √(44.222 / 9) |
| Sample SD | 2.351 | √(44.222 / 8) |
Here mean > median = mode: 3.556 > 3 = 3. The 9 pulls the mean up by more than half a point, while the median and the mode stay put at the cluster of 3s. This ordering — mode < median < mean — is the signature of right skew: whenever the mean sits clearly above the median, a long right-hand tail is doing the pulling.
Worked Example 3: The Income Outlier
Salaries are the classic case. Ten salaries in thousands of dollars: [30, 35, 40, 45, 50, 55, 60, 65, 70, 500] — nine ordinary salaries and one executive on $500,000.
| Measure | With the $500k salary | Without it |
|---|---|---|
| Mean | 95 ($95,000) | 50 ($50,000) |
| Median | 52.5 ($52,500) | 50 ($50,000) |
| Mode | None | None |
| Range | 470 | 40 |
One salary moves the mean from 50 to 95 — it nearly doubles — while the median moves only from 50 to 52.5. The mean of $95,000 describes nobody: nine of the ten people earn $70,000 or less. The median of $52,500 is the honest “typical” salary, which is exactly why the U.S. Census Bureau reports the median household income: its 2024 figure was $83,730, while the mean household income is considerably higher — the mean follows the high earners, the median follows the middle household. Try the same numbers on the Statistics Calculator: add the 500, then remove it, and watch the mean jump while the median barely moves.
Worked Example 4: Bimodal and No Mode
Not every dataset has a single mode. Take [1, 2, 2, 3, 3, 4, 5, 5, 6]: the values 2, 3 and 5 each appear twice, and nothing appears more often.
| Measure | Value | How it is found |
|---|---|---|
| Mean | 3.444 | (1 + 2 + 2 + 3 + 3 + 4 + 5 + 5 + 6) / 9 = 31 / 9 |
| Median | 3 | Middle value of the sorted list |
| Mode | 2, 3 and 5 | Each appears twice |
| Range | 5 | 6 – 1 |
| Population SD | 1.571 | √(22.222 / 9) |
| Sample SD | 1.667 | √(22.222 / 8) |
This dataset is trimodal — it has three modes (2, 3 and 5). A dataset with exactly two modes is called bimodal. Two peaks usually mean two groups have been mixed together, such as the heights of boys and girls measured in one class. And the no-mode case is just as common: the symmetric [2, 4, 6, 8, 10] from Example 1 has no mode at all, because every value appears exactly once. A dataset with no mode is not broken; the mode simply is not a useful summary for it. Check the bimodal data in the Statistics Calculator — it reports “Mode: 2, 3, 5”, and “Mode: No mode” for [2, 4, 6, 8, 10].
Skewness is read from the longest tail; kurtosis from how much of the data sits far from the centre — not from the height of the peak.
The Skewness Rule: How Mean, Median and Mode Move Apart
Compare the figures in the diagram above with the numbers from the worked examples and a rule emerges:
- Symmetric data: mean = median = mode. The distribution is a mirror image, so all three sit in the centre. A normal distribution is the classic case: its mean, median and mode are all equal.
- Right-skewed data: mode < median < mean. A long right-hand tail pulls the mean furthest right, the median sits in the middle, and the mode stays with the peak.
- Left-skewed data: mean < median < mode. A long left-hand tail pulls the mean furthest left.
The rule is a tendency, not a law — it can fail in unusual distributions — but it is the fastest way to read skew from three numbers. The NIST/SEMATECH e-Handbook makes the same point: extreme values in the tails distort the mean, while the median, which is based on ranks, is not distorted. The reference is the NIST Measures of Location page.
Three datasets from the skewness and kurtosis guide show the rule in action:
| Dataset | Shape | Mean | Median | Mode |
|---|---|---|---|---|
| [2, 4, 6, 8, 10, 12, 14, 16, 18, 20] | Symmetric | 11 | 11 | None |
| [1, 2, 2, 3, 3, 3, 4, 5, 9] | Right-skewed | 3.556 | 3 | 3 |
| [1, 5, 6, 7, 8, 8, 9, 9, 10] | Left-skewed | 7 | 8 | 8 and 9 |
In the symmetric set the mean and median are both 11 (and there is no mode — a reminder that “no mode” is common even in perfectly symmetric data). In the right-skewed set the mean (3.556) sits above the median (3) and the mode (3). In the left-skewed set the mean (7) sits below the median (8); the dataset is bimodal, with modes of 8 and 9, so the rule holds as mean < median ≤ mode. Verify the third row on the Statistics Calculator — it reports “Mode: 8, 9”. For the full skew story, with skewness coefficients and tail weight, use the site’s Skewness and Kurtosis Calculator.
What Is the Difference Between Mean, Median and Mode?
The difference is what each one measures. The mean is a balance point: it uses every value and is pulled by the extremes. The median is a position: the middle of the sorted data, untouched by how far the extremes lie. The mode is a frequency: the most common value, and the only one of the three that can describe categories.
| Mean | Median | Mode | |
|---|---|---|---|
| Definition | Sum of all values ÷ count | Middle value of the sorted data | Most frequent value |
| Formula | x̄ = Σx / n | (n + 1) / 2-th value (odd n) | Highest frequency |
| Uses all values? | Yes | No — only the middle one or two | No — only the most frequent |
| Outliers | Strongly affected | Barely affected | Not affected |
| Works for categories? | No | No | Yes — the only one |
| Gives totals? | Yes — total = mean × n | No | No |
| Example [1, 2, 2, 3, 3, 3, 4, 5, 9] | 3.556 | 3 | 3 |
Put another way: if you want the average as a balance point, take the mean; if you want the average as the middle household, salary or house, take the median; if you want the most popular answer, take the mode.
When Should You Use the Median Instead of the Mean?
Use the median instead of the mean whenever the data are skewed or contain outliers — income, house prices, salaries, waiting times and prices are the usual cases. The mean is dragged towards the tail, so it describes a value nobody actually has; the median sits in the middle of the real data. The U.S. Census Bureau’s median household income of $83,730 in 2024, cited in Worked Example 3, is the classic case: it is the income of the middle household, while the mean household income — lifted by the highest earners — is considerably higher. The practical test is simple: if the mean is far from the median, report the median and say why. If the data are roughly symmetric with no extremes, the mean is fine — and it has an advantage the median lacks: it lets you recover the total.
Which Average Should You Use?
Match the average to the question and to the shape of the data:
| Situation | Use | Why |
|---|---|---|
| Data are symmetric, no outliers | Mean | Uses every value and lets you recover the total |
| Data are skewed or have outliers | Median | Unmoved by extremes; the honest “typical” value |
| Data are categories (sizes, answers, codes) | Mode | The only average that works for non-numbers |
| You need a total (revenue, payroll, cost) | Mean | Total = mean × count |
| You want the most common value or size | Mode | Only the mode answers “what is most usual?” |
| You want the middle and the spread | Median, with quartiles | Median plus quartiles describe the centre and the spread together |
When in doubt, report the mean and the median together — and the mode when it exists. The three numbers take one line on a page, and any gap between them tells the reader about the shape of the data. The Quartile Calculator reports the median alongside the quartiles and the interquartile range, the natural next step after choosing the median. And to describe the spread around your averages, use the Standard Deviation Calculator and the Mean Absolute Deviation Calculator — the two companion measures of spread.
Categorical Data: Where Only the Mode Works
Some data have no numbers to average. Shoe sizes are a classic: if a shop sells sizes 7, 7, 8, 8, 8, 9, 10, the mean size is 8.14 — a size nobody buys — and the median is 8, but the useful number is the mode: 8, the size to stock most of. The same applies to survey answers (the most common rating), the most common failure code on a production line, and the most frequent word in a text. In each case the mode is the only average that makes sense, because you cannot add the categories together or order them in a meaningful way.
The mode is also the only average that can be plural or absent: a survey can have two joint-most-common answers, and an even spread of answers has no mode at all — in which case the median does not rescue you either, because categories have no middle. For categories, the mode is not one option among three; it is the only option.
Can a Dataset Have No Mode?
Yes — a dataset has no mode whenever no value occurs more often than any other. The most common case is a dataset in which every value appears exactly once, such as [2, 4, 6, 8, 10]: each value occurs once, so there is no “most frequent” value. The opposite case is a tie at the top: in [1, 2, 2, 3, 3, 4, 5, 5, 6] the values 2, 3 and 5 each appear twice, so the dataset has three modes rather than none. Both answers are correct — “no mode” and “three modes” — and both are reasons to look at the median as well: the median always exists, even when the mode does not.
Common Mistakes to Avoid
- “Mean, median and mode are always the same.” They agree only on symmetric data. On skewed data they part — mean > median = mode for right skew, mean < median for left skew — and that gap is information, not an error.
- Using the mean on income or house prices. A few very high values drag the mean far above what most people earn or pay; the median is the honest “typical” figure for skewed money data.
- Forgetting to sort before finding the median. The median is the middle value of the sorted data. In [4, 8, 6, 8, 10] the middle value is 8 only after sorting to [4, 6, 8, 8, 10]; in the unsorted list the third value is 6, which is wrong.
- Saying a dataset with all-unique values has a mode. If every value appears once, there is no mode. [2, 4, 6, 8, 10] has no mode — do not report the middle value (6) as the mode.
- Treating a median that is not in the data as an error. The median of an even count is the average of the two middle values, and it often does not appear in the data: the median of [2, 4, 6, 8] is 5. That is fine — an average does not have to be one of the observed values. Only the mode must be.
- Judging spread from the range alone. The range uses only the largest and smallest values, so one outlier can make it huge: 40 for nine salaries, 470 once the $500k salary is added. For a steadier picture of spread, use the standard deviation.
Frequently Asked Questions
Tap a question to see the answer.
What is mean, median and mode in simple words?
The mean is the average: add everything up and divide by how many numbers there are. The median is the middle number once the data are sorted. The mode is the number that appears most often. On symmetric data all three give the same value; on skewed data they differ.
What happens to the mean when you add an outlier?
A single extreme value can move the mean a long way, because the mean uses every value. In the salary example in this guide, adding one $500k salary to nine salaries of $30k to $70k moved the mean from $50k to $95k, while the median moved only from $50k to $52.5k. That is why the median is preferred for income and house prices.
Can the mean, median and mode all be the same?
Yes, on symmetric data they can all be equal. In a normal distribution the mean, median and mode are identical. For [2, 4, 6, 8, 10] the mean and median are both 6, but there is no mode, because no value repeats. A dataset can agree on the mean and median and still have no mode.
Can a dataset have more than one mode?
Yes. A dataset with two modes is bimodal, and [1, 2, 2, 3, 3, 4, 5, 5, 6] is trimodal, with modes of 2, 3 and 5. Two or more peaks usually mean two groups have been mixed together.
Which average should I use for income or house prices?
Use the median. Income and house prices are right-skewed: a few very high values pull the mean far above what most people earn or pay. The U.S. Census Bureau reports median household income ($83,730 in 2024) for this reason.
How do you find the median of an even number of values?
Sort the data and average the two middle values. For [2, 4, 6, 8], the two middle values are 4 and 6, so the median is (4 + 6) / 2 = 5. With an odd count the median is simply the middle value.
Which average is best for skewed data?
The median. It is based on position, not size, so a long tail cannot drag it far. For right-skewed data the order is mode < median < mean, and for left-skewed data it is mean < median < mode, so the median stays in the middle of the three.
Does the mode have to be a number?
No. The mode is the only average that works for categories, so it can be a shoe size, a survey answer or a failure code. It is the most frequent value or category, whether or not it is a number.
Conclusion
Mean, median and mode are three averages that answer three different questions. The mean balances the total, the median sits in the middle, and the mode marks the most common value. On symmetric data they agree; on skewed data they part — and reading that gap is the whole skill. If you remember one rule, make it this one: use the mean when the data are symmetric and you need totals, use the median when they are skewed or have outliers, and use the mode whenever the data are categories or you want the most common value. Check the numbers for yourself on the site’s Statistics Calculator: type in your data and compare the three side by side.
Check Tools
Statistics Calculator
Compute the mean, median and mode — plus the range, quartiles and standard deviation — for any dataset, with the full working shown.
Skewness and Kurtosis Calculator
See how skewed your data are, and whether the mean has been pulled away from the median.
Standard Deviation Calculator
Work out the spread around your averages, in population and sample form.
Explore All Statistical Tools
Browse the full set of calculators and visualisers on the site.