The Bakery Where the Average Misleads
Imagine a small bakery with nine people on its payroll. Eight workers each earn $30,000 a year. The owner earns $120,000. A newspaper writer wants to describe how much people at the bakery make, and she has to choose which kind of average to use.
The mean is what most people think of first. To find it, add up all the values and divide by how many there are. The eight workers add up to $240,000, and with the owner's pay the total is $360,000. Dividing by nine people gives a mean of $40,000. That number sounds nice, but not one worker actually earns it. The single large paycheck pulled the mean upward.
The median is a different average. To find it, put the values in order from least to greatest and pick the one in the middle. With nine people, the fifth value is in the middle, and it is $30,000. The median describes the typical worker much better, because it is not dragged around by one unusual value.
Statisticians call a value far from the rest an outlier. When a data set has an outlier, the mean can be misleading, while the median stays steady. Home prices and salaries are often reported with the median for this reason. A careful reader asks which average was used, and a careful writer sometimes reports both so that the full picture is clear.