Reeliy

Why the Average Can Mislead: Outliers, Skew and Weights

By Reeliy Editorial Team · · 3 min read

When someone quotes "the average", they almost always mean the arithmetic mean: add up the values and divide by how many there are. It is the right tool for many jobs, but it has a weakness. It uses every value, so a single extreme one can pull it a long way from what most of the data looks like.

This article covers the three most common ways an average misleads (outliers, skewed data and unequal group sizes), what to use instead in each case, and a quick checklist before you trust a figure.

What each average actually measures

  • Mean: the sum divided by the count. It is the balance point of the data.
  • Median: the middle value once the data is sorted (or the mean of the two middle values if the count is even). Half the data lies on each side.
  • Mode: the most frequent value. It is the only one of the three that also works for categories, such as the most common shoe size.

For perfectly symmetric data, the mean and median land in the same place. They drift apart when the data is lopsided, and the size of that gap is itself a useful warning sign.

How one outlier moves the mean

Worked example: seven employees at a small firm earn (in thousands) 32, 34, 35, 36, 38, 40 and 250, where the last figure is the owner. The total is 465, so the mean is 465 ÷ 7 = 66.4. Yet six of the seven people earn between 32 and 40.

The median is the fourth value in sorted order, which is 36. It sits in the middle of the group that most people belong to. Remove the owner and the mean of the other six falls to 35.8, almost identical to the median. The median barely noticed the outlier, while the mean moved by more than 30.

An outlier is not always an error. A genuine extreme value belongs in the data, but you should report it separately or choose a summary that is not dominated by it.

You can check any list yourself with the mean, median and mode calculator, which shows all three side by side so you can see the gap.

Skewed data and the long tail

Some quantities have a long tail on one side. Incomes and house prices are typical examples: most values cluster at the lower end, while a few very large ones stretch the right-hand side. In data like this the mean is usually higher than the median, because the tail pulls it upward.

That is why reports on pay or property often quote the median. It answers the question most readers have, which is what a typical case looks like. The mean answers a different question: what would each person receive if the total were shared out equally.

A simple rule of thumb: if the mean and median differ by a lot, look at the spread of the data before deciding which one to quote. The median calculator is quick for this.

Averaging averages: weights matter

A second trap is combining group averages without allowing for group size. Worked example: class A has 10 students with a mean score of 80, and class B has 30 students with a mean of 60. Averaging the two means gives 70, but that treats the classes as equal in size.

weighted mean = (n₁ × m₁ + n₂ × m₂) ÷ (n₁ + n₂)Each group's mean is weighted by its size.

The correct figure is (10 × 80 + 30 × 60) ÷ 40 = 2,600 ÷ 40 = 65. The larger class counts for more. A weighted average calculator handles this for any number of groups, and it is the same method used for course grades with different weightings.

Rates need extra care. If you drive out at 60 km/h and back at 40 km/h over the same distance, your average speed is not 50 km/h. It is the harmonic mean, 2 ÷ (1/60 + 1/40) = 48 km/h, because you spend more time at the slower speed.

A short checklist before you trust an average

  1. Ask how many values it is based on. An average of three numbers is fragile.
  2. Compare the mean with the median. A large gap means skew or an outlier.
  3. Look for the minimum and maximum, or a measure of spread such as the standard deviation.
  4. If groups of different sizes were combined, check that the average was weighted.
  5. For rates and ratios, ask whether an arithmetic mean is the right kind of average at all.
Try the tools
  • Mean, Median & Mode Calculator: Type or paste a list of numbers, separated by commas, spaces or new lines, and press Calculate. You get the mean, median, mode, range and the number of values at once; the guide below explains how each one is worked out.
  • Weighted Average Calculator: Enter one value and its weight per line and the calculator returns the weighted mean. The guide shows the formula and how to use it for course grades, average prices and similar problems.
  • Median Calculator: Enter your numbers, separated by commas, spaces or new lines, and press the button to find the median: the value that splits the sorted list into a lower half and an upper half.

This article is general information, not professional advice. See our Disclaimer and Editorial Policy.

More articles