When the Median Belongs in a Descriptive Statistics Summary

You have just calculated a median of $52,000 for a list of 15 salaries and are about to report it alone. Stop. A median without its supporting context is like a book cover without pages. The question of when the median belongs in a descriptive statistics summary is answered by what else you must report alongside it: the interquartile range, the sample size, and a shape indicator. Those three pieces let a reader see spread, sample size, and shape from your summary alone.

Report the interquartile range (IQR = Q3, Q1) to show spread. Report the count n to show sample size. Report which way the distribution leans, either the mean versus median comparison or a skewness coefficient, to show shape. Together these turn a single number into a summary that a reader can interpret without the raw data.

OpenStax Introductory Statistics defines the median and IQR as part of a five-number summary. That set is the minimum needed. Without n, a reader cannot judge whether the median is stable (n under 10 is shaky). Without shape, a reader cannot tell if the median sits near the mean (symmetric) or far from it (skewed). Without IQR, a reader cannot see how spread out the middle 50 percent is.

Take the dataset {2, 4, 6, 8, 10}. The median is 6. Q1 is 4, Q3 is 8, so IQR = 4. n = 5. The mean is also 6, so mean equals median: symmetric distribution. A reader sees spread (IQR = 4), sample size (5), and shape (symmetric). Now consider {1, 2, 3, 100, 200}. The median is 3. Q1 is 2, Q3 is 100, IQR = 98. n = 5. The mean is 61.2. Mean is far greater than median, so the distribution is right‑skewed. The IQR of 98 shows the middle 50 percent spans a wide range, but it still understates the total spread because the outliers at 100 and 200 are in the tails. Reporting range (max, min = 199) alongside IQR helps a reader see that the IQR alone masks the full variability.

You need the quartiles to compute IQR. For the dataset {1, 2, 3, 100, 200}, the standard spreadsheet method (R‑7, used by Excel's QUARTILE.INC and Google Sheets' QUARTILE) gives Q1 = 2 and Q3 = 100. The statistics calculator uses the same R‑7 method. To compute Q1 and Q3 by hand: sort the data, find the median (3), then find the median of the lower half {1, 2} = 1.5 and the upper half {100, 200} = 150. That gives Q1 = 1.5 and Q3 = 150, IQR = 148.5. The difference between 98 and 148.5 is the difference between the R‑7 method (which interpolates) and the hinge method (which takes medians of halves). Report which method you use, because the IQR changes.

For grouped data, calculate the median class using cumulative frequencies. The standard interpolation formula from NCERT Class 10 Mathematics is: Median = L + [(n/2, cf)/f] × w, where L is the lower boundary of the median class, cf is the cumulative frequency before the median class, f is the frequency of the median class, and w is the class width. Report n first, then compute.

The median belongs in a descriptive statistics summary when your data is skewed, ordinal, or contains extreme values that would pull the mean away from the typical observation. For household incomes, the U.S. Census Bureau reports the median exactly because it resists the influence of a few very high earners. But the median alone tells nothing about the gap between rich and poor. That gap is revealed by the IQR and by comparing mean to median.

Failure Case: Two Identical Medians, Different Worlds

Dataset A: {1, 1, 1, 50, 100}. Median = 1. IQR = (50-1) = 49. n = 5. Mean = 30.6, right‑skewed. Dataset B: {1, 1, 1, 2, 3}. Median = 1. IQR = (2-1) = 1. n = 5. Mean = 1.6, near‑symmetric. Same median, but A has huge spread and severe skew. Reporting only the median would mislead a reader into thinking the datasets are similar. The IQR and skewness prevent that error.

Limit of IQR

The IQR ignores the tails. Two datasets can have the same IQR but completely different tail behavior. For example, {1, 2, 3, 4, 100} and {1, 2, 3, 4, 5} both have IQR = 2 (Q1 = 2, Q3 = 4), but the first has an outlier at 100. Always report n and a shape indicator to catch cases where the IQR is misleading.

Component	What It Shows	How to Report
Interquartile range (IQR)	Spread of the middle 50%	IQR = Q3 – Q1, with method stated
Sample size (n)	Reliability of the median	n < 10 means the median is unstable
Shape indicator	Symmetry or skew	Mean vs. median, or skewness coefficient

The median suits middle‑school and high‑school students learning the sort‑and‑average rule. It suits introductory college statistics students who need the formal definition and the mean‑median‑skew relationship. It suits data analysts handling income or price data where outliers are common. It suits business and economics readers who want a resistant summary for skewed distributions like salaries or home prices.

Who should skip the median as the sole summary? Anyone needing the mode or the range should look elsewhere. Advanced statisticians needing quantile regression are beyond this topic. People working with medical survival data need survival analysis, not a median calculator. And never report the median for nominal categories, they have no order, so the median is undefined.