x̄ = Σwᵢxᵢ / Σwᵢ
Each value xᵢ is multiplied by its weight wᵢ, the products are added up, and the sum is divided by the
sum of the weights. A value with weight 3 counts three times as much as one with weight 1. The formula is
a definition.
When all the weights are equal, it becomes the ordinary mean. The weighted mean is not a
different kind of mean, then, but the ordinary one with room to let some values count for more.
Worked examples
Two tests that count differently. One test scored 80 and has weight 2, another scored 90 and has
weight 3. The numerator is 2 · 80 + 3 · 90 = 160 + 270 = 430, and the sum of the weights is 2 + 3 = 5.
The weighted mean is 430 / 5 = 86. The unweighted mean would be 85. The second test counts for more
and pulls the result up.
Weights as percentages. The same tests weighted 40% and 60%: 0.4 · 80 + 0.6 · 90 = 32 + 54 = 86.
The weights add up to 1, so the division does not show.
Common mistakes
- Forgetting to divide by the sum of the weights. The answer comes out as 430, which is not even
between 80 and 90.
- Pairing weights with the wrong values. The first weight belongs to the first value. Two lists of
different lengths are rejected before anything is worked out.
- Dividing by the number of values instead of the sum of the weights. 430 / 2 = 215 is just as wrong.
Common questions
What if all the weights are equal?
Then it is the ordinary mean. With weights 1 and 1 it comes to (80 + 90) / 2 = 85.
Can the weights be percentages?
Yes. If they add up to 100%, the sum of the weights is 1 and the formula is just the sum of the
products. If they do not add up to 100%, you still divide by their sum.
Why is it not 85?
85 is halfway between 80 and 90, which would be right if the tests counted equally. Here the second test
counts for more, so the mean lands closer to 90.