Mean vs Median vs Mode: How Averages Work

When someone says "the average is 88," that is not one fact — it is one of three different facts wearing the same coat. Mean, median, and mode each answer a different question about a set of numbers, and picking the wrong one is how an honest dataset gets used to tell a misleading story.

The mean: add up, then divide

The mean is what most people picture: add every value and divide by how many there are.

mean = (sum of all values) / (number of values)

For test scores 72, 84, 90, 78, 96: the sum is 420, and 420 / 5 = 84. The mean uses every value in the calculation, which makes it efficient — and is also its weakness. Change one score from 90 to 500 and the mean leaps to 166 even though nothing else moved. The mean is hostage to its most extreme member.

The median: the middle of the sorted list

The median sorts the values and picks the middle one — the person in the center, not the totals. With an odd number of values that is simply the middle entry; with an even number it is the average of the two middle ones. From 2, 4, 6, 8 the median is (4 + 6) / 2 = 5.

Now the salary example: five people earn 40k, 45k, 50k, 55k, and 250k. The mean is 440k / 5 = 88k — a number nobody in the room actually earns. The median is 50k: the real middle person. When one extreme value would drag the mean away from typical reality, the median is the more honest summary.

The mode: the most common value

The mode is the value that appears most often. In 3, 5, 5, 7, 9 the mode is 5. Two special cases: if every value is unique there is no mode, and if two values tie for most frequent the data is bimodal — 1, 2, 2, 3, 3, 4 has modes 2 and 3.

The mode's real home is categorical data, where a mean makes no sense: the most common shoe size sold, the most frequent complaint category, the most popular shipping option. You cannot average "medium" and "large," but you can count which one appears most.

How skew pulls the mean and median apart

In a symmetric distribution the mean and median nearly coincide. When the data is skewed — with a long tail on one side — the mean gets pulled toward the tail while the median stays put. Income and house prices are the classic cases: a few very large values stretch the right tail, so mean income sits well above median income. That is why official statistics report median household income and median sale prices.

A practical reading rule: when the mean is clearly higher than the median, the data has a long high tail and "the average" flatters the typical case. Neither number is wrong — they answer different questions. The mean is the right tool for totals (average revenue per customer × number of customers = total revenue); the median is the right tool for describing a typical individual.

Weighted averages: when values don't count equally

Sometimes each value carries a different weight, and the plain mean is simply wrong. A GPA is a weighted average: grade points multiplied by credit hours, divided by total credits. Take A (4.0) in a 3-credit course, B (3.0) in a 4-credit course, and A− (3.7) in a 3-credit course:

GPA = (3×4.0 + 4×3.0 + 3×3.7) / (3 + 4 + 3)
    = (12.0 + 12.0 + 11.1) / 10
    = 35.1 / 10 = 3.51

The unweighted mean of 4.0, 3.0, 3.7 would be 3.57 — close here, but the gap widens as the weights diverge. The same trap hides behind "averages of averages": averaging a 90% across 10 tasks and a 60% across 1,000 tasks is not 75%; weighted by size it is (60 × 1000 + 90 × 10) / 1010 ≈ 60.3%. Whenever groups have different sizes, weight by size.

The average alone isn't the story

Two datasets can share a mean and still be completely different: 4, 5, 6 and 0, 5, 10 both average 5, but the second has roughly five times the standard deviation. An average describes a center; it says nothing about variability. That is what the standard deviation adds — a measure of how far values typically sit from the mean, turning "the average is 5" into "the average is 5, give or take 3." Center plus spread is the minimum honest summary of any dataset.

Try it yourself

Paste any list of numbers into the Average Calculator — it returns the mean, the median, and the mode side by side, and flags the no-mode case explicitly. Seeing all three at once on your own data is the fastest way to feel how they diverge. When you are ready to go beyond the center, the Standard Deviation Calculator adds the spread. And for a weighted average you may already be computing, the GPA Calculator handles the credit-weighted math on the 4.0 scale.