The median is another measure of central tendency representing the “middle score” in a set. It’s the point in a score distribution in which 50% of the scores fall below it and 50% fall above it. The median is usually represented by Md. but can sometimes be represented by M.

Computing the median is a two-step process:

1. List all the scores in order, starting with the highest value and progressing to the lowest. For example: 100, 88, 76, 65, 40, 25, 16
2. Find the score that is in the middle of this list. It is the 4th score in the list (there are 3 scores above it and 3 scores below it) and the median would be 65.

If there are an even number of scores, the median is the average of the two middle scores. For example, in this set, there are six scores: 76, 65, 51, 49, 39, 23. The two middle scores (the third and fourth in the list) are 51 and 49, so the average of these two scores is 50 and 50 would be the median score of the set.

The median can be very useful in describing central tendency when a data set has extreme scores; when the mean has been affected by extreme scores, the median will still reflect the exact middle score of the distribution. We want our central tendency measure to reflect the best score estimate that represents all the scores and the median will still do that in all situations. For example, examine the following data set: 78, 78, 76, 75, 74, 73, 2. The mean of these scores is 65, yet most scores are in the 70s, except for the one extreme score of 2. Meanwhile, the median is 75, which is right in the middle of the non-extreme scores and represents the entire set much more accurately. This demonstrates the medians’ ability to mitigate the effect of extreme scores.

Also, the median is not affected by the values of the scores themselves, only by the number of scores. The median can also effectively indicate the middle of rank order data points in ways that the mean typically cannot. A prominent example of how the median is regularly used is in describing the central tendency of annual income and home prices in the United States.  There are enough very high salaries and very expensive homes that push the mean of these values higher than the true middle score in these distributions. Due to the extreme nature of some of the data points in these distributions, median income and median home price more accurately describe the central tendency of these distributions.

Watch this video for a quick description of finding the median using Excel.