These three "averages" describe data in different ways — and using the wrong one can be seriously misleading. Here's what each one means and which to use in real situations.
When someone says "the average salary is £35,000," which average do they mean? The answer matters more than most people realise — and the choice of mean, median, or mode can completely change the story a dataset tells.
The mean is what most people think of as "the average." Add all values together and divide by how many there are.
When to use it: When data is evenly distributed with no extreme outliers. Test scores in a consistent class. Daily temperatures over a month. Product weights from a manufacturing line.
When it misleads: When outliers exist. Five people earn £20k, £22k, £24k, £25k, and £500k. The mean is £118k — but that describes nobody in the group accurately.
The median is the middle value when all data is sorted in order. If there's an even number of values, it's the mean of the two middle values.
When to use it: Any dataset with outliers or skewed distribution. Income, house prices, wait times — any real-world data that has a long tail on one end. The median UK salary (around £34k) is more meaningful than the mean (pulled up by high earners).
The UK housing market example: Average house price is often reported as "mean" by estate agents trying to make the market look stronger, and as "median" by journalists trying to show how unaffordable housing is. Both can be technically correct — but describe very different realities.
The mode is simply the value that appears most often. A dataset can have no mode (all values unique), one mode, or multiple modes.
When to use it: Categorical data and discrete counts. The most common shoe size stocked by a retailer. The most frequent customer complaint category. The most popular product variant. The mode tells you what the typical case is when "typical" means most common, not mathematical middle.
10 employees earn: £18k, £19k, £20k, £21k, £22k, £23k, £24k, £25k, £28k, £200k (the owner).
The owner could truthfully advertise "average pay of £42k" — which is technically the mean — while all but one employee earns less than half that. This is why whenever you see an "average" quoted, the first question should be: which average?
Averages only tell half the story. Two datasets can have identical means but very different distributions. This is why averages are almost always more useful alongside a measure of spread — standard deviation, range, or interquartile range — that tells you how clustered or dispersed the values are.