When you read that the “average” income in a city is ₹80,000 a month, that number can be misleading. A handful of extremely high earners can pull the average upward, leaving most people earning far less than the figure suggests. This is exactly where the median becomes useful. It cuts through distortion and tells you the value that sits right in the middle of your data, giving a more honest picture of what is typical.
Table of Contents
What is the median?
The median is the middle value of a dataset when all the values are arranged in order, either from smallest to largest or largest to smallest. It is one of the three main measures of central tendency, alongside the mean and the mode. While the mean adds up every value and divides by the count, the median simply locates the point that splits the data into two equal halves.
Half of all observations fall below the median, and half fall above it. This makes the median a “robust” statistic, meaning it is not pulled around by extreme values the way the mean is.
Why the median matters when data is not evenly distributed
Real-world data is often skewed. Income, property prices, and household expenditure rarely spread out neatly around a central point. Instead, a small number of very high values stretch the data to one side. In such situations, the mean gets dragged toward the tail and stops representing the typical case.
The median does not have this problem. The median is less affected by outliers and skewed data than the mean and is usually the preferred measure when a distribution is not symmetrical. A classic example is salary data, where a few top earners create a false impression of the typical income if you only look at the mean. The median ignores how extreme those top values are and reports the genuine middle.
There is also a practical limit worth knowing. The median is also a good choice for ordinal data, such as satisfaction ratings, but it cannot be calculated for nominal categories like blood group or religion, because those cannot be logically ordered.
How to calculate the median
The method changes depending on whether your dataset is small and listed individually, or large and grouped into class intervals. Both follow the same underlying idea: find the middle.
Finding the median in a small dataset
For a short list of values, the process is straightforward.
Step 1: Arrange all the values in ascending order.
Step 2: Count the number of observations, written as n.
Step 3: Decide whether n is odd or even.
If n is odd, the median is the value at the position (n + 1) / 2. Consider the marks 12, 18, 21, 24, 30. Here n = 5, so the median is the (5 + 1) / 2 = 3rd value, which is 21.
If n is even, there are two middle values, so the median is their average. Take the values 12, 18, 21, 24, 30, 36. Now n = 6, and the two middle values are the 3rd and 4th items, 21 and 24. The median is (21 + 24) / 2 = 22.5. Notice that the median itself need not be a value that actually appears in the dataset.
Finding the median in a large dataset
When data is extensive, it is usually organised into a frequency distribution table with class intervals such as 0-10, 10-20, and so on. Grouped data is useful for large datasets where we study patterns across ranges rather than individual points. Here the exact middle value is hidden inside a class interval, so we estimate it using a formula.
The formula for the median of grouped data is:
Median = l + [(n/2 − cf) / f] × h
Each symbol has a plain meaning:
l is the lower boundary of the median class. n is the total of all frequencies. cf is the cumulative frequency of the class just before the median class. f is the frequency of the median class itself. h is the class width or size.
The steps to apply it are as follows. First, build the frequency distribution table and add a cumulative frequency column by running a total down the frequencies. Next, calculate n/2. Then locate the median class, which is the first class whose cumulative frequency is greater than or equal to n/2. Finally, read off l, cf, f, and h for that class and substitute them into the formula.
One caution applies to large datasets with intervals like 10-19, 20-29. Before calculating, convert these into continuous boundaries such as 9.5-19.5 and 19.5-29.5 by adjusting 0.5 on each side, so the classes touch without gaps.
A practical example using expenditure data
Suppose a researcher records the monthly expenditure of 50 households in a locality and groups the figures into intervals. The aim is to find the median expenditure, the point below which half the households spend less and above which half spend more.
Consider this distribution of monthly expenditure in rupees:
1000-2000: 6 households | 2000-3000: 10 households | 3000-4000: 16 households | 4000-5000: 12 households | 5000-6000: 6 households
The cumulative frequencies build up as 6, then 16, then 32, then 44, then 50. The total n = 50, so n/2 = 25.
The first class whose cumulative frequency reaches or exceeds 25 is the 3000-4000 class, since its cumulative frequency is 32. This is the median class. For it, l = 3000, the cumulative frequency before it is cf = 16, the frequency is f = 16, and the class width is h = 1000.
Substituting into the formula gives Median = 3000 + [(25 − 16) / 16] × 1000 = 3000 + (9 / 16) × 1000 = 3000 + 562.5 = ₹3562.5. So the median monthly expenditure is about ₹3,562. Half the households spend below this amount and half spend above it, regardless of how high the spending of the wealthiest few households reaches.
Why this matters beyond the classroom
This is not just a textbook exercise. National statistical agencies rely on the median exactly for this reason. The Ministry of Statistics and Programme Implementation reports Monthly Per Capita Consumption Expenditure (MPCE) through the Household Consumption Expenditure Survey, which in 2023-24 collected data from over 2.6 lakh households across all states and union territories. When consumption or income data is summarised, the median gives policymakers a clearer sense of how a typical household lives, because it is not distorted by the small number of very high spenders.
The same logic explains why median figures appear in economics and real estate, where distributions tend to be skewed. Median house prices, median wages, and median rents all describe the centre more faithfully than averages do.
A limitation to keep in mind
The median is powerful, but it is not perfect for every purpose. Because it depends only on position and not on the actual size of each value, the median cannot easily be used in further mathematical calculations the way the mean can. Statistical models, formulas, and many tests are built around the mean. So the choice of measure always depends on the data and the question being asked. For symmetrical data, the mean works well. For skewed data, the median earns its place.
What do you think? If you were analysing the monthly earnings of workers in your own town, would the mean or the median give a fairer picture of a typical income? And can you think of a situation where reporting only the median might actually hide something important about the data?
References
- https://www.abs.gov.au/statistics/understanding-statistics/statistical-terms-and-concepts/measures-central-tendency
- https://statisticsbyjim.com/basics/measures-central-tendency-mean-median-mode/
- https://statistics.laerd.com/statistical-guides/measures-central-tendency-mean-mode-median.php
- https://statistics.laerd.com/statistical-guides/measures-central-tendency-mean-mode-median-faqs.php
- https://www.geeksforgeeks.org/maths/median-of-grouped-data/
- https://www.cuemath.com/data/median-of-grouped-data/
- https://brightchamps.com/en-us/math/data/median-of-grouped-data
- https://www.pib.gov.in/PressReleasePage.aspx?PRID=2088390
- https://medium.com/@jaberi.mohamedhabib/measures-of-central-tendency-understanding-mode-median-and-mean-eab80ada8868
- https://yourcx.io/en/blog/2024/08/median-definition-and-examples-understanding-the-central-tendency-in-statistics/
Leave a Reply