Mathematics & Logic · Numbers & patterns
When the average is the wrong summary
When one number stands in for a whole set, which number you pick changes the story. Mean, median and mode tend to pull apart when the shape is lopsided.

What you will learn
- Calculate a mean and locate a median in an ordered list
- Explain why an outlier moves the mean and leaves the median where it was
- Predict the order of mean, median and mode from the direction of a skew
- Use the gap between a reported mean and median as evidence about the shape
Lesson 1 of 4 · 5 min
Mean, median and mode give three different middles
Picture a town of 50 people where the average income is 129,400 dollars a year, and nobody earns anything close to that. The arithmetic is right. The word "average" has three meanings, and they can sit far apart.
The usual one is the arithmetic mean. You add the values and divide by how many there are.
The median is the value that splits an ordered set into two equal halves. To find it, you sort the data first and then locate the middle. The location is n plus 1 divided by 2, where n is how many values there are.
With 97 values that gives 49, so the median is the 49th value. With 100 values it gives 50.5, so the median sits midway between the 50th and the 51st. The formula returns a position in the sorted list, so keep the location of the median separate from its value.
The mode is simply the most frequent value. A set with two modes is called bimodal. The mode is also the only one of the three that works on categories, such as colors, which can't be added or put in order.
Back to the town. One of its 50 people earns 5,000,000 dollars a year, and the other 49 each earn 30,000. Add the incomes and divide by 50, and the mean is 129,400 dollars. Sort them and find the middle, and the median is 30,000. Both numbers are correct.
The mean moved because it uses the precise value of every income. One enormous number drags the total upward, and the division spreads that pull across everybody. The median only cares about the ordering. Replace the 5,000,000 with 50,000,000 and the mean rises almost eightfold, to 1,029,400 dollars. The median doesn't budge by a single dollar.
That's why the median is generally the better measure of the center when there are extreme values, called outliers. Quoting the mean as the typical income here would describe a person who doesn't exist. The mean is still correct. It answers a question about the total, not about the typical resident.
Try to recallIn a town of 50 people, one person earns 5,000,000 dollars and the other 49 earn 30,000 each. Give the mean and the median, and say which describes a typical resident.
The mean is 129,400 dollars, found by adding 5,000,000 to 49 times 30,000 and dividing by 50. The median is 30,000 dollars, because 49 of the 50 values are 30,000 and the middle of the sorted list falls among them. The median describes a typical resident, since no one earns anything near 129,400. The mean was pulled upward because it uses the outlier's exact value, while the median depends only on position in the order.
The mean uses every value, so one outlier can drag it far from what anyone earns. The median uses only the order, so it stays put. When you see the word "average", ask which one it is.
Lessons in this topic
Mean, median and mode give three different middles
Mean, median and mode all claim to be the center. A town of 50 people shows how far apart two of them can sit, and why.
What the gap between mean and median tells you
Mean, median and mode coincide when data are symmetrical and separate in a predictable order when the data lean. Learn to read the gap between a reported mean and median.
Which average answers which question
The question you're asking decides which center to use. See when the mean, the median and the often-ignored mode are each the right one.
Quartiles show the spread an average hides
Quartiles and the interquartile range describe the spread that any center discards, and they also give a stated rule for flagging an outlier.
Three quick stories from this topic
One town, two averages
Add one 5-million-dollar earner to 49 people on 30,000 and the average income is 129,400
That average is the mean, and nobody in the town earns anything close to it. The median, the income of the person in the middle, is still 30,000. One extreme value moves the mean a long way and leaves the median where it was.
Two correct averages
The mean is 129,400 dollars. The median is 30,000 dollars. Nobody in the town earns anything close to the mean.
Why they split
The mean uses the precise value of every observation, so one enormous number drags it upward. The median depends only on position in the sorted order, so the outlier doesn't move it at all.
The six-pound average
An ad can truthfully claim an average weight loss of six pounds when most people lose two
The six is the mean, pulled up by a few people who lost a lot. The two is the mode, the most common result. Both are honest averages of the same group, and the advertiser chooses which one to print.
Three honest averages
The mean adds every value and divides by the count. The median is the middle value of the sorted list. The mode is the most frequent one.
Which one to look at
Someone deciding whether to join wants to know what's likely to happen to them. The mode or the median answers that. The mean answers a question about the group's total.
Percentile isn't percent
Scoring in the 90th percentile on a test doesn't mean you got 90% of it right
It means 90% of the scores were the same as yours or lower. On a hard exam, 60% correct could put you in the 90th percentile. On an easy one, 90% correct might leave you well below it.
A rank, not a score
Percentiles divide ordered data into hundredths. Your percentile says where you stand among everyone who took the test.
Quartiles work the same way
Quartiles cut ordered data into quarters. The distance from the first quartile to the third, the interquartile range, covers the middle half of the data.
Sources
- Measures of the Center of the Data, OpenStax, Rice University
- 2.3 Measures of the Location of the Data - Introductory Statistics 2e, OpenStax, Rice University





