What Do You Do If There Are Two Medians

Introduction

When you first encounter the term median in a statistics class, the definition sounds simple: it is the middle value that separates the higher half of a data set from the lower half. The intuition works perfectly for odd‑sized lists, where a single observation sits exactly in the centre. Things become a little more nuanced, however, when the data set contains an even number of observations. In that case there are two middle numbers, and the concept of “the median” seems to split into two possibilities.

This article explains exactly what to do when you face two medians, why the convention exists, and how the rule fits into the broader theory of central tendency. By the end you will be able to compute the median for any data set, recognize common pitfalls, and appreciate why statisticians have settled on the averaging solution.

Detailed Explanation

What the median really measures

The median is a measure of central tendency that identifies a typical value in a distribution. Unlike the mean, which adds all values and divides by the count, the median depends only on the order of the data. To find it, you first sort the observations from smallest to largest and then locate the point that splits the ordered list into two equal halves.

When the total number of observations, denoted n, is odd, there is a unique observation that has exactly ((n-1)/2) values below it and ((n-1)/2) values above it. That observation is the median.

When n is even, no single observation can satisfy that exact split because half of the data would require a non‑integer count on each side. Instead, the two central observations—call them (x_{(n/2)}) and (x_{(n/2+1)})—each have the same number of data points on one side and one more on the other. Because neither alone perfectly bisects the set, statisticians adopt a convention: the median is the arithmetic mean of those two middle values.

Why averaging makes sense Averaging the two central numbers yields a value that lies exactly halfway between them. This choice preserves several desirable properties of the median:

1. Robustness to outliers – The median still ignores extreme values because it depends only on the middle of the ordered list.
2. Equivariance under monotone transformations – If you apply any strictly increasing function (e.g., taking logarithms) to all data points, the median of the transformed set is the same function applied to the original median. Averaging the two middle points respects this property.
3. Minimisation of absolute deviations – The median is defined as any point that minimises the sum of absolute deviations (\sum |x_i - m|). For an even‑sized data set, every point between the two middle observations (including the endpoints) yields the same minimal sum. Picking the midpoint is a natural, single‑value representative of that whole interval.

Thus, when you encounter two medians, the correct action is to calculate their average; the result is the conventional median used in virtually all statistical software and textbooks.

Step‑by‑Step or Concept Breakdown Below is a clear, repeatable procedure for finding the median when the data set size is even.

1. Collect and list the raw data.
   Example: ({12, 7, 9, 15, 20, 4}). 2. Sort the observations in ascending order.
   Sorted list: ({4, 7, 9, 12, 15, 20}).

2. Determine the positions of the two middle numbers.
   For n = 6, the middle positions are (n/2 = 3) and (n/2 + 1 = 4).

3. Identify the values at those positions.
   The 3rd value is 9; the 4th value is 12.

4. Compute the average of those two values.
   (\text{Median} = \frac{9 + 12}{2} = \frac{21}{2} = 10.5).

5. Report the result.
   The median of the original data set is 10.5. If the two middle numbers happen to be identical (e.g., both are 9), the average is simply that same number, and the median coincides with an actual data point.

Real Examples

Example 1: Exam Scores

A teacher records the scores of eight students on a quiz:

[ {78, 85, 92, 67, 74, 88, 91, 80} ]

1. Sort: ({67, 74, 78, 80, 85, 88, 91, 92}). 2. Middle positions: 4th and 5th → 80 and 85.
2. Median: ((80 + 85)/2 = 82.5).

Even though no student scored exactly 82.5, this value best represents the centre of the distribution and is unaffected by the unusually high score of 92 or the low score of 67.

Example 2: Household Incomes (in thousands of dollars)

A small survey of ten households yields:

[ {35,

Continuing seamlessly from the provided text:

Real Examples (Continued):

Example 2: Household Incomes (in thousands of dollars)

A small survey of ten households yields:

[ {35, 40, 45, 50, 55, 60, 65, 70, 75, 80} ]

1. Sort the data:
   ({35, 40, 45, 50, 55, 60, 65, 70, 75, 80})

2. Identify the middle positions:
   For n = 10, the middle positions are (n/2 = 5) and (n/2 + 1 = 6).

3. Values at those positions:
   The 5th value is 55, the 6th value is 60.

4. Calculate the median:
   (\text{Median} = \frac{55 + 60}{2} = \frac{115}{2} = 57.5)

Interpretation:
The median household income is $57,500. This value is robust against the extreme values present (e.g., the $35,000 and $80,000 households). Unlike the mean ($56,000), which is slightly pulled downward by the lower income, the median remains centered among the bulk of the data. It provides a more representative picture of the typical household income in this sample, unaffected by the skew introduced by the lowest and highest values.

Why the Median Matters in Practice

The median’s unique properties make it indispensable across diverse fields:

1. Income and Wealth Analysis:
   As demonstrated, medians avoid distortion from extreme wealth or poverty, offering a clearer view of economic well-being than averages.

2. Real Estate Pricing:
   Median home prices reflect typical market value, ignoring the impact of a few ultra-luxury or distressed properties.

3. Health Statistics:
   Median survival times or patient ages are preferred over means when data is skewed (e.g., rare long-term survivors).

4. Education:
   Median test scores provide a balanced assessment of class performance, less influenced by a few exceptionally high or low performers.

Conclusion:
The median transcends its role as a simple measure of central tendency. Its resilience to outliers, invariance under monotonic transformations, and optimality for minimizing absolute deviations make it a cornerstone of robust statistics. Whether analyzing income distributions, evaluating medical outcomes, or assessing educational achievement, the median delivers a stable, interpretable summary that resists distortion from extreme values. By averaging the two central values in even-sized datasets, it transforms raw data into a meaningful indicator of the "typical" observation—a principle consistently applied in scientific research, policy-making, and everyday decision-making. Its enduring utility lies in its ability to distill complex datasets into a single, reliable value that faithfully represents the data’s center.