Understanding Statistics: Mean, Median, Mode and When Each Matters

Why “average” is more complicated than you think

You have probably heard someone say “the average person” does this or that. But which average are they talking about? In statistics, there is more than one way to summarise a set of numbers, and picking the wrong summary can paint a completely misleading picture.

When I first started tutoring undergrads in Bangalore, I noticed the same confusion every semester: students could calculate a mean but had no intuition for when the mean was the wrong tool. This guide walks you through the three most common measures of central tendency — mean, median, and mode — and then we will look at two ideas that tell you how spread out your data really is.

The mean: adding up and dividing

The arithmetic mean is what most people picture when they hear “average.” You add up every value and divide by the count. If five friends spend 200, 250, 300, 275, and 225 on dinner, their mean spend is 250.

Simple enough. But now imagine one friend orders an extravagant tasting menu and spends 2,000. The mean jumps to 650 — a number that does not describe anyone in the group accurately. This is the classic weakness of the mean: it is sensitive to outliers. A single extreme value can drag it far away from what is typical.

When to use the mean: it works best when your data is roughly symmetric and free of wild outliers. Think of repeated measurements in a lab, or daily temperatures across a month.

The median: the middle ground

The median is the value that sits right in the middle when you sort your data from smallest to largest. Half the values fall below it, half above. In our dinner example the sorted values are 200, 225, 250, 275, 300 — the median is 250, the same as the mean. But once that 2,000 outlier enters the picture, the median barely moves while the mean rockets upward.

This is exactly why salary reports and house price statistics almost always quote the median rather than the mean. A handful of extremely high earners can inflate the mean salary of a city, but the median tells you what a person right in the middle actually earns.

When to use the median: whenever your data is skewed or contains outliers — incomes, property values, hospital wait times, or anything where a few extreme cases could distort the picture.

The mode: the most popular value

The mode is simply the value that appears most often. In the set 4, 7, 2, 7, 9, 3, 7, the mode is 7 because it shows up three times. A dataset can have no mode (if every value is unique), one mode, or multiple modes.

The mode shines in situations where you care about frequency rather than magnitude. If a shoe store wants to know which size to stock most heavily, the mode of their sales data is far more useful than the mean shoe size.

When to use the mode: categorical data (favourite colour, most-selected survey option) or any context where you want to know the most common outcome.

Try plugging in your own numbers below and watch how the mean, median, and mode respond differently to the same dataset. Use the Statistics Calculator to experiment:

Beyond the centre: why spread matters

Knowing the centre of your data is only half the story. Consider two classes that both score a mean of 72 on an exam. In Class A, every student scores between 68 and 76. In Class B, scores range from 30 to 100. The averages are identical, yet these are very different classrooms. To capture that difference you need a measure of spread.

Standard deviation

Standard deviation tells you, on average, how far each data point sits from the mean. A small standard deviation means the values cluster tightly; a large one means they are scattered.

Here is the intuition I like to share: imagine the mean is a tent pole and each data point is a rope staked to the ground. Standard deviation is the average length of those ropes. Short ropes, tight tent. Long ropes, floppy tent.

In practice, standard deviation powers everything from quality control in manufacturing (are our widgets consistently the right size?) to finance (how volatile is this stock?). When a report says a metric is “within one standard deviation,” it roughly means the value is not unusual.

Use the Statistics Calculator to see how adding or removing values changes the spread:

Percentiles: where do you stand?

Percentiles answer the question “what percentage of values fall below this point?” If your exam score is at the 90th percentile, you scored higher than 90 percent of test-takers. Paediatricians use percentile charts to track a child’s height and weight relative to other children of the same age — a toddler at the 75th percentile for height is taller than 75 percent of peers.

Percentiles are closely related to the median: the median is simply the 50th percentile. The 25th and 75th percentiles (called the first and third quartiles) form the boundaries of the interquartile range, another handy measure of spread that resists outliers.

A useful mental model: imagine lining up 100 people by height. The person at position 25 marks the 25th percentile, the person at position 50 is the median, and the person at position 75 marks the 75th percentile. Percentiles give you a precise sense of relative standing that a raw score alone cannot provide.

Explore how different data distributions affect percentile rankings with the Percentile Calculator:

Choosing the right measure

There is no single best statistic — it depends on your data and your question. Here is a quick guide:

• Symmetric data, no outliers — the mean is your best friend.
• Skewed data or outliers present — report the median (and mention the mean alongside it so readers can gauge the skew).
• Categorical data or “most common” questions — use the mode.
• Need to describe spread — pair your central measure with standard deviation or the interquartile range.
• Comparing individuals against a group — percentiles give the clearest picture.

The next time you see a headline claiming “the average household” does something, pause and ask: is that the mean or the median? That one question can completely change the story the data is telling. Statistics is not about memorising formulas — it is about choosing the right lens so the numbers actually make sense.