Use the mean when your data is fairly balanced and every value should count equally; reach for the median instead when one or two values are unusually large or small. That single rule covers most of the choice between these two averages.
The mean is the ordinary average: add all the values and divide by how many there are. For numbers ,
Add the values, count how many you used, then divide the total by that count.
Mean vs. Median Side by Side
Mean Median
What it uses every value in the set only the middle position after sorting
How it is found sum of values divided by n middle value of the sorted list
Sensitivity pulled by outliers stable against outliers
Best when data is balanced, equal data has strong outliers or skew
weighting makes sense
The mean uses every value, so it reflects the full set, not just the middle position. The median only looks at the middle position after sorting, which is why strong outliers move it much less.
When to Use Which
Use the mean when you want one number to summarize a numerical data set and equal weighting makes sense, such as test scores, daily temperatures over a short period, or the average number of items sold per day. It is a basic tool in statistics because it is simple to compute and easy to compare across groups.
Switch to the median when the data has strong outliers. The mean income of a group can rise because of one very high income, even if most people earn much less; the median describes that center more clearly. And if some values should genuinely count more than others, neither plain average fits and you need a weighted mean.
Worked Example
Suppose four quiz scores are , , , and .
Add the scores:
Count how many there are:
Divide the total by the count:
So the mean score is . The pattern is always total first, count second, division last.
Where the Mean and Median Get Confused
The most common mix-up is treating every "average" as the mean. In everyday language that happens constantly, but in math the mean, median, and mode are different ideas, and the right one depends on the data.
A few other traps follow from that confusion:
- Dividing by the wrong count. Divide by the number of values, not the largest value or a guess.
- Forgetting a number when adding. One missing value changes the result.
- Using the mean when values should not count equally. Grades with different weights are a common example.
To see the difference for yourself, take , , , , and and find the mean step by step. Then change one number to and watch how far the mean moves. That single swap shows exactly when the mean is useful and when an outlier pulls it off-center, which is the moment to prefer the median.
Frequently Asked Questions
- How do you calculate the mean?
- Add all the values, count how many values there are, then divide the total by that count. For example, quiz scores of 6, 8, 7, and 9 add up to 30, and dividing by 4 gives a mean of 7.5. The same pattern works for any list: total first, count second, division last.
- When can the mean be misleading?
- When one value is much larger or smaller than the rest, it can pull the mean away from what feels typical. For example, the mean income of a group can rise because of one very high income, even if most people earn much less. In cases like that, the median may describe the center more clearly.
- When should you use the mean?
- Use the mean when you want one number to summarize a numerical data set and equal weighting makes sense, such as test scores or daily temperatures over a short period. If some values should count more than others, like grades with different weights, you need a weighted mean instead of the ordinary average.
- What are common mistakes when finding the mean?
- Dividing by the wrong count instead of the actual number of values, forgetting a number during the addition, using the mean when values should not count equally, and calling every average a mean. In everyday language average is used loosely, but in math, the mean, median, and mode are different ideas.
Need help with a problem?
Upload your question and get a verified, step-by-step solution in seconds.
Open GPAI Solver →