Every dataset has a center. The problem is that "center" means three different things depending on what you ask. The mean asks where the data balances. The median asks what value splits it in half. The mode asks what value appears most. On well-behaved data these three answers are close enough to ignore. On real data they diverge — and that divergence is not a problem to resolve. It is the most important thing the data is telling you.
The mode is the only measure of central tendency that works on nominal data. You cannot average nationalities. You cannot find the median shoe brand. But you can find the most common one. That's the mode's entire advantage — and it comes with two serious weaknesses: it sometimes doesn't exist, and it sometimes isn't unique.
The True Power: Categorical Data
The mode's limitation is also its strength. It is the only measure that doesn't require numbers — which makes it the only measure that works on categorical data.
The median doesn't care about your outliers. It only cares about position — which value sits exactly in the middle when everything is sorted. You can replace the largest value with a number a thousand times bigger and the median won't move. That robustness is its defining property, and it's also its limitation: by ignoring the extremes, it ignores real information those extremes contain.
The median's resistance to outliers is mathematical, not accidental. Position is immune to magnitude.
The arithmetic mean is just a weighted mean where every observation gets equal weight. Once you see it that way, the weighted mean stops being a special formula and becomes the general case — and the arithmetic mean becomes its simplest instance. Change the weights and the center moves. Set them all equal and you're back where you started.
The Grand Mean
When combining two groups, simply averaging their means assumes both groups are exactly the same size. Real life rarely works that way.
The arithmetic mean assumes everyone is equally important. The weighted mean lets you say otherwise.
This is the moment the three measures stop being abstract. Give them a dataset with an outlier and watch them respond differently. The mode doesn't notice. The median barely moves. The mean chases the extreme value across the number line. None of them is wrong. They are answering different questions about the same data.
The mean uses all the data. That's its strength when data is well-behaved. That's its weakness when it isn't.
When a distribution is symmetric, mean, median, and mode are identical. When it skews, they separate — and the direction they separate tells you which tail is pulling the data. The mean always chases the tail. The median follows reluctantly. The mode stays at the peak. The gap between them is not measurement error. It is the shape of the distribution, expressed as a number.
Symmetric Distribution
All three measures agree. The curve is balanced. This is the only situation where the mean is unambiguously the right choice for the 'center'.
The distance between mean and median is a measure of skewness. The mean always chases the longest tail.
The distance between mean and median is a measure of skewness. The direction tells you which tail is longer.
It's time to test your intuition. Given a specific scenario or a set of values, identify which measure of central tendency is the most appropriate or what shape the distribution must have.
Part 1: Which Measure?
Select the most appropriate measure of central tendency for each situation.
You want to know the most popular major at FEPS.
You want to describe average income in Egypt without being misled by billionaires.
You want to describe the average exam score in a normally distributed class.
You want to find the central tendency of students' satisfaction ratings: Poor, Fair, Good, Very Good, Excellent.