Outliers and Their Effect on Mean and Median

The mean is a total split evenly: add every value, divide by the count. Change one value and the total changes by exactly that amount, so the mean shifts by that change divided by the count. The median is different. It is whatever value sits in the middle once everything is lined up in order. It does not care how big the extreme values are, only where they rank. So a value can grow enormous and the median will not move, as long as it stays in the same position in the order.

This is why one billionaire walking into a room full of ordinary incomes drags the average income way up, but the median income barely budges. The billionaire was already the richest person before walking in, now they are just a more extreme richest person; their rank does not change, so the middle value of the group does not change. An outlier is not just a value that looks a little bigger than its neighbors. It is a value far enough from the pattern of the rest of the data that it behaves differently under the mean, pulling it noticeably off center. A mildly higher value barely moves the mean either.

This matters for reading skewed distributions. A right-skewed data set, like household incomes, has most values clustered low with a thin tail of very large ones stretching right. That tail pulls the mean above the median, since the mean absorbs the size of every value while the median stays anchored where the bulk of the data sits. Comparing mean and median this way is a quick test for skew, and it connects to why an unusual measurement should not always be discarded: sometimes it is the real signal, not an error.