- Foundation
- About 14 min
- 5 practice questions
The median and the effect of outliers
Find the median of odd and even sized data, spot outliers, and decide whether the mean or the median better represents data.
If something looks wrong, please tell us.
By the end of this lesson you can
- Find the median by sorting the data and taking the middle value (or the mean of the two middle values).
- Identify an outlier and describe how it pulls the mean.
- Use the positions of the mean and median to describe data.
The mean is useful, but one unusual value can drag it far from the rest of the data. Then a different representative value, the median, often tells the truer story.
When the mean misleads
A family's heights are 168, 172, 160, 175 and 95 cm (the youngest child is still small). Mean = (168 + 172 + 160 + 175 + 95) ÷ 5 = 770 ÷ 5 = 154 cm. But four of the five members are taller than 154 cm. The one very small value has pulled the mean down.
The median
To find the median, sort the data and take the middle value. Sorted heights: 95, 160, 168, 172, 175. The middle (3rd of 5) is 168 cm, which describes most of this family much better.
Median: sort the values in order. With an odd number of values it is the middle one; with an even number it is the mean of the two middle ones. Half the data is at or below the median and half at or above.
Worked example 1: an even number of values
Heights in another family: 150, 158, 162, 165, 170, 171 cm. There are 6 values, so there is no single middle. The two middle values are the 3rd and 4th: 162 and 165. Median = (162 + 165) ÷ 2 = 327 ÷ 2 = 163.5 cm. The mean is 976 ÷ 6 ≈ 162.7 cm, very close to the median, because this data has no unusual value.
Outliers
A value that is very far from the rest of the data is an outlier. The 95 cm above is a low outlier. Outliers pull the mean towards themselves, but hardly move the median.
Worked example 2: a high outlier
Seven students read these numbers of stories in the holidays: 3, 4, 5, 5, 6, 7, 40.
- Mean = (3 + 4 + 5 + 5 + 6 + 7 + 40) ÷ 7 = 70 ÷ 7 = 10.
- Median: the data is already sorted; the 4th of 7 values is 5.
- The mean (10) is twice the median (5). Six of the seven students read 7 or fewer, so 10 is not typical. The outlier 40 has pulled the mean up.
- Without 40: mean = 30 ÷ 6 = 5 and median = (5 + 5) ÷ 2 = 5. Removing the outlier changed the mean a lot and the median not at all.
Mean and median
-
Mean
- Uses every value in the sum
- Pulled strongly by outliers
- Good when data is balanced
-
Median
- Uses only the middle position
- Hardly moved by outliers
- Good when there are outliers
| What the data looks like | Mean compared with median |
|---|---|
| Balanced, no outliers | Mean and median are close |
| A low outlier | Mean is less than median |
| A high outlier | Mean is more than median |
The mean and the median are called measures of central tendency: they show where the data tends to gather.
Key takeaways: the median is the middle of the sorted data. Outliers pull the mean towards them but barely move the median, so with outliers the median is often the better summary. If the mean is well below the median, look for a low outlier; well above, a high outlier.
Check your understanding
-
1. What is the median of 7, 3, 9, 5, 11?
- 5
- 35
- 9
- 7
Show answer
Answer: D. Sorted: 3, 5, 7, 9, 11. The middle (3rd of 5) value is 7.
-
2. What is the median of 12, 18, 10, 16?
- 14
- 12
- 16
- 56
Show answer
Answer: A. Sorted: 10, 12, 16, 18. Two middle values: 12 and 16. Median = (12 + 16) ÷ 2 = 14.
-
3. For the data 4, 5, 5, 6, 30, how do the mean and median compare?
- They are equal
- The mean is greater than the median
- The mean is less than the median
- There is no median
Show answer
Answer: B. Mean = 50 ÷ 5 = 10; median = 5 (the 3rd value). The high outlier 30 pulls the mean above the median.
-
4. The weekly pocket money of seven friends is ₹50, ₹60, ₹55, ₹65, ₹50, ₹60 and ₹500. Which value better describes a typical friend?
- The mean, ₹120
- The maximum, ₹500
- The median, ₹60
- The range, ₹450
Show answer
Answer: C. Sorted: 50, 50, 55, 60, 60, 65, 500. Median = 60. The mean is 840 ÷ 7 = 120, which is more than six of the seven friends get, because of the outlier 500.
-
5. A data set has one very low outlier and no other unusual values. What do you expect?
- Mean less than median
- Mean greater than median
- Mean equal to the maximum
- Median equal to the outlier
Show answer
Answer: A. A low outlier pulls the mean down but barely moves the median, so the mean ends up below the median.
Sign in to keep track of the lessons you finish.