Mean vs median: which average should you use?

The mean shares the total equally. The median finds the middle case. In skewed data, those can tell very different stories.

Mean and median are both measures of center, but the mean is pulled by extreme values while the median tracks the middle observation.

By the Math Says Yes editorial team

Human-reviewed under our source and correction standards.

How we review content

THE TRAP

Calling both numbers averages and forgetting that one follows the total while the other follows the middle case.

Bars rising from short to tall, a pointer on the middle bar and a dashed line sitting high above most of them.

The short difference

The answers a sharing question: if the total were spread equally across every observation, what would each one get? The answers a position question: what value sits in the middle after all observations are ordered? When data are balanced and roughly symmetric, the two are often close. When data are skewed, a few very large or very small values can pull the mean away from what most observations look like.

Quick Comparison

Mean

Median

What it answers

What would everyone get if the total were shared equally?

What value sits in the middle after sorting?

Common mistake

Using it for skewed data as if outliers did not matter.

Treating it as if it described totals or budgets.

Use when

Totals matter or the distribution is roughly symmetric.

You want the typical case in skewed data.

What the Numbers Show

one high outlier

Nine house prices are 250,000 and one is 2.5 million. The mean is 475,000, but the median stays 250,000.

Worked example

Take ten house prices on one street: nine at 250,000 and one at 2.5 million. The is 475,000 because the expensive house adds a lot to the total. The is 250,000 because the middle of the ordered list still sits among the nine ordinary homes. Both numbers are correct. Only one describes what a typical house on that street costs.

What people get wrong

Asking for “the ” assumes there is only one sensible center. A mean can be useful for totals, budgets, and quantities that really do add up. A is often better for typical income, house prices, waiting times, and other skewed quantities. Another mistake is treating a gap between mean and median as a problem. That gap is information: it says the data have a tail or outliers.

When it applies

Use the when values are fairly symmetric, when every unit contributes to a total, or when the total itself matters. Use the when you want the typical case and the data have outliers or a long tail. For public claims, the best answer is often to show both. If the mean and median are close, one number may be enough. If they are far apart, the needs more explanation.

Source note

The Census household income tables show a domain where medians are central, because income distributions are skewed. Our World in Data is included for broader context on income inequality and shape. The page uses a small illustrative salary list rather than a live national estimate, because the goal is to explain the mechanical difference between and .

Try It

Mean vs. median income

Add high earners and watch the average climb.

0 high earners added

38k

average, k

38k

median, k

With no high earners added, the mean and median agree at 38k — the average still describes a typical person. Add a few and watch them split.

FAQ

Is median the same as average?

Median is one kind of average in the loose sense of a measure of center, but it differs from the arithmetic mean: the median is the middle value after sorting the data.