COLLEGE STATISTICS • DESCRIPTIVE STATISTICS

Measures of Center

Quantifying the typical value of a distribution through mean, median, and mode.

Historical Context & Motivation

The impulse to summarize a collection of observations with a single representative number is as old as quantitative reasoning itself. Ancient astronomers in Babylon and Egypt routinely averaged multiple measurements of celestial positions to arrive at more reliable estimates—an intuitive acknowledgment that individual observations carry error, but their collective center carries signal. The formal development of measures of central tendency into a rigorous statistical framework, however, unfolded over several centuries and involved contributions from astronomers, mathematicians, and social scientists who each recognized different facets of what it means for a value to be "typical."

The problem these measures solve is fundamental: raw data, even modest datasets, overwhelm human cognition. A list of 50 exam scores or 10,000 patient blood-pressure readings conveys almost nothing until we distill it into digestible summaries. Measures of center provide the first and most essential such summary—a single number that communicates where the bulk of the data resides. Understanding the historical trajectory of these measures illuminates why we have multiple competing definitions of "center" and when each is most appropriate.

c. 1500 BCE
Babylonian Astronomical Averaging
Babylonian astronomers computed midpoint values of repeated observations of planetary positions, an early form of averaging that improved the accuracy of ephemerides used for calendrical and astrological purposes.
1669
Christiaan Huygens & the Expected Value
Huygens formalized the notion of expectation in his treatise on games of chance, providing a probability-weighted average that would later underpin the arithmetic mean in statistical theory.
1809
Gauss & Least Squares
Carl Friedrich Gauss demonstrated that the arithmetic mean minimizes the sum of squared deviations—a result that anchored the mean as the optimal estimator under normal error assumptions and gave rise to the method of least squares.
1882
Francis Galton & the Median
Galton promoted the median as a robust alternative to the mean, noting its resistance to extreme observations—an insight that would become central to modern robust statistics.
1895
Karl Pearson & the Mode
Pearson coined the term mode and systematically studied skewness, articulating how the relative positions of mean, median, and mode reveal the shape of a distribution.

This historical arc reveals a recurring theme: no single measure of center is universally optimal. The mean, median, and mode each answer a subtly different question about where data concentrate, and the choice among them depends on the distribution's shape, the presence of outliers, and the inferential goals of the analyst. The remainder of this lesson develops each measure formally, examines their geometric and algebraic properties, and establishes the criteria for selecting the most appropriate one in a given context.

Core Principles & Definitions

A measure of center (also called a measure of central tendency) is a single value that attempts to describe a dataset by identifying a central position within the data. The three primary measures—mean, median, and mode—each capture a different aspect of centrality. Understanding the conceptual foundation of each is essential before examining their mathematical formulations and comparative behavior.

1

Arithmetic Mean (x̄)

The sum of all observed values divided by the number of observations. The mean is the balance point of the distribution: it is the unique value at which the sum of signed deviations equals zero. It uses every data point, making it sensitive to outliers.
2

Median (M or x̃)

The middle value when all observations are arranged in ascending order. For an even number of observations, the median is the average of the two central values. The median minimizes the sum of absolute deviations and is highly resistant to extreme values.
3

Mode (Mo)

The value that occurs most frequently in a dataset. A distribution may be unimodal, bimodal, or multimodal. The mode is the only measure of center applicable to nominal (categorical) data.
4

Sensitivity vs. Robustness

A central trade-off in choosing a measure of center: the mean incorporates all data points (high sensitivity, low robustness), while the median ignores magnitude of extreme values (lower sensitivity, high robustness). The mode may not even be unique.
5

Symmetry Criterion

For perfectly symmetric, unimodal distributions, the mean, median, and mode coincide at the axis of symmetry. Skewness pulls the mean toward the tail, causing the three measures to diverge—a diagnostic signal for distributional shape.
KEY TAKEAWAY
Think of a dataset as a collection of weights placed on a number line—a physical beam. The mean is the fulcrum point where the beam balances perfectly; move a single weight far to the right and the fulcrum shifts with it. The median is the position that divides the beam so that equal numbers of weights sit on each side—regardless of how far any individual weight is from center. The mode is simply the location on the beam where the most weights are stacked. Each tells you something different about where the "center" of mass, count, or frequency lies.

Visual Explanation: Mean, Median & Mode on a Distribution

The relationship among the three measures of center becomes most transparent when visualized on a frequency distribution. In a symmetric distribution all three coincide, but in skewed distributions they separate in a predictable order. The following diagram illustrates a right-skewed distribution—the type commonly encountered with income data, hospital lengths of stay, and reaction times—and marks the positions of the mode, median, and mean.

In a right-skewed distribution, the long right tail pulls the mean to the right of the median, which in turn lies to the right of the mode. The ordering is reversed for left-skewed distributions: mean < median < mode.

The diagram encapsulates a crucial diagnostic principle. When you compute a mean and a median for a dataset and find that the mean exceeds the median appreciably, you can infer right skewness without plotting the data at all. Conversely, when mean < median, left skewness is indicated. For symmetric distributions—the normal distribution being the canonical example—the three measures converge to a single value, and any one of them adequately characterizes the center. This convergence property is one reason the normal distribution enjoys its privileged status in statistical theory.

Mathematical Framework

Each measure of center can be defined both operationally (how to compute it) and variationally (what quantity it minimizes). The variational perspective is especially illuminating because it connects descriptive statistics to optimization theory and motivates the choice of one measure over another in different loss-function contexts.

The Arithmetic Mean

SAMPLE MEAN
x̄ = (1/n) Σᵢ₌₁ⁿ xᵢ
Where is the sample mean, n is the number of observations, and xᵢ denotes the i-th observation. The population analogue replaces x̄ with μ and n with N.

The mean possesses a key variational property: it is the unique value c that minimizes the sum of squared deviations, Σ(xᵢ − c)². This connects the mean to the principle of least squares and explains its centrality in regression analysis. Additionally, the mean is an unbiased estimator of the population mean μ, meaning E[x̄] = μ regardless of the underlying distribution.

VARIATIONAL PROPERTY OF THE MEAN
x̄ = argmin_c Σᵢ₌₁ⁿ (xᵢ − c)²
The mean minimizes the sum of squared deviations—equivalently, it minimizes the L₂ loss function.

The Median

SAMPLE MEDIAN
M = x₍₍ₙ₊₁₎/₂₎ if n is odd; M = (x₍ₙ/₂₎ + x₍ₙ/₂₊₁₎) / 2 if n is even
Where x₍ₖ₎ denotes the k-th order statistic (the k-th smallest value after sorting). The median is the value at the 50th percentile of the empirical distribution.

The median's variational characterization is equally elegant: it minimizes the sum of absolute deviations, Σ|xᵢ − c|. This L₁ loss criterion explains the median's robustness—absolute deviations grow linearly with distance rather than quadratically, so outliers exert far less influence. The breakdown point of the median is 50%, meaning that up to half the data can be arbitrarily corrupted before the median becomes unbounded—compared to a breakdown point of 0% for the mean, where a single extreme value can drag it arbitrarily far.

VARIATIONAL PROPERTY OF THE MEDIAN
M = argmin_c Σᵢ₌₁ⁿ |xᵢ − c|
The median minimizes the sum of absolute deviations—equivalently, the L₁ loss function. This is the foundation of least absolute deviations (LAD) regression.

The Mode

The mode is defined as the value (or values) with the highest frequency. For continuous distributions, the mode is the value at which the probability density function attains its maximum. Unlike the mean and median, the mode is not guaranteed to be unique—a distribution may be bimodal or multimodal. From a variational standpoint, the mode can be connected to the L₀ loss (zero-one loss), as it minimizes the expected misclassification rate when predicting a single value for the entire dataset. This makes the mode the optimal point prediction under a discrete classification loss function.

Behavior Under Different Distributions

The relative positions of the mean, median, and mode shift systematically with the shape of the distribution. Understanding this relationship is not merely academic—it directly informs which summary statistic to report and how to interpret it. The following diagram presents three canonical distributional shapes side by side, marking the position of each measure of center to illustrate their divergence under skewness.

Three panels show how the relative ordering of mean, median, and mode shifts with skewness. In a left-skewed distribution, the mean is pulled toward the left tail. In a symmetric distribution all three coincide. In a right-skewed distribution the mean is pulled toward the right tail.

The mnemonic ordering for right-skewed data—mode < median < mean—follows from the fact that the mean, being a sum-based statistic, is dragged disproportionately by the extreme values in the long tail. The median, depending only on ranks, shifts modestly. The mode, fixed at the peak of the density, is least affected. Pearson's empirical rule of thumb, mean − mode ≈ 3(mean − median), quantifies this relationship for moderately skewed, unimodal distributions, though it is only approximate and can fail for strongly skewed or multimodal data.

💡 Practical Guideline
When reporting a "typical" value to a general audience, consider the distribution shape. For skewed distributions (e.g., household income, home prices), the median is usually the more informative and less misleading measure. The mean is preferred when the distribution is approximately symmetric or when subsequent algebraic operations on the summary are needed (since the mean of sums equals the sum of means).

Worked Example

Consider a random sample of 11 commute times (in minutes) recorded by employees at a mid-sized company: 12, 15, 18, 18, 22, 25, 27, 30, 35, 42, 95. We will compute all three measures of center, compare them, and interpret the results in context.

Computing Mean, Median, and Mode for Commute Times
1
Step 1 — Organize the DataSort the data in ascending order (already done): 12, 15, 18, 18, 22, 25, 27, 30, 35, 42, 95. We have n = 11 observations.
2
Step 2 — Compute the Arithmetic MeanSum all values: 12 + 15 + 18 + 18 + 22 + 25 + 27 + 30 + 35 + 42 + 95 = 339. Divide by n: x̄ = 339 / 11 ≈ 30.82 minutes.
x̄ ≈ 30.82 minutes
3
Step 3 — Identify the MedianSince n = 11 (odd), the median is the value in position (11 + 1)/2 = 6th. Counting from the sorted list: the 6th value is 25.
M = 25 minutes
4
Step 4 — Identify the ModeExamine frequencies: 18 appears twice; all other values appear once. Therefore, 18 is the mode.
Mode = 18 minutes
5
Step 5 — Interpret and CompareWe observe Mode (18) < Median (25) < Mean (30.82), which is the characteristic ordering for a right-skewed distribution. The mean is inflated by the outlier commute time of 95 minutes. In this context, the median of 25 minutes provides the most representative summary of a "typical" commute, as it is not distorted by the extreme value. If the outlier (95) were removed, the mean would drop to (339 − 95)/10 = 24.4, much closer to the median—confirming that the single extreme observation substantially inflated the original mean.

Strengths, Limitations & Selection Criteria

No single measure of center dominates in all settings. Each has strengths that make it the preferred choice under certain conditions, and limitations that render it misleading in others. The following table provides a systematic comparison across several criteria that are relevant to data analysts, researchers, and practitioners.

Comparative properties of the three primary measures of center
CriterionMean (x̄)Median (M)Mode (Mo)
Uses all data pointsYes — every observation contributesNo — depends only on positional ranksNo — depends only on frequencies
Sensitivity to outliersHigh — a single extreme value shifts the mean substantiallyLow — 50% breakdown pointNone — completely unaffected by outlier magnitude
UniquenessAlways uniqueUnique (though conventionally averaged for even n)May not be unique — distributions can be bimodal or have no mode
Algebraic tractabilityExcellent — supports additivity (mean of sums = sum of means)Poor — median of sums ≠ sum of medians in generalPoor — no algebraic closure properties
Applicable data levelsInterval and ratio scales onlyOrdinal, interval, and ratio scalesAll levels including nominal
Best use caseSymmetric distributions; algebraic operations neededSkewed distributions; ordinal data; robust summaryCategorical data; identifying most common category
KEY TAKEAWAY
Choosing a measure of center is analogous to choosing a loss function in machine learning. If you penalize errors quadratically (L₂ loss), you're implicitly selecting the mean. If you penalize errors linearly (L₁ loss), you choose the median. If you use zero-one loss (correct vs. incorrect classification), you choose the mode. The "best" measure depends entirely on how you define "error" in your particular application.

Connections to Advanced Statistical Theory

The measures of center introduced in this lesson serve as the foundation for a wide array of advanced topics. The arithmetic mean generalizes naturally into the expected value E[X] of a random variable, which is the cornerstone of probability theory, moment-generating functions, and maximum likelihood estimation. The median's robustness leads to entire subfields—robust statistics and nonparametric methods—where estimators with high breakdown points are preferred. The mode connects to maximum a posteriori (MAP) estimation in Bayesian inference.

How each measure of center connects to advanced statistical methods
Descriptive MeasureAdvanced GeneralizationContext
Arithmetic mean (x̄)Expected value E[X]; weighted mean; trimmed meanProbability theory, regression, ANOVA, moment estimation
Median (M)Population median; quantile regression; Hodges–Lehmann estimatorRobust statistics, nonparametric tests (Wilcoxon, sign test)
Mode (Mo)MAP estimate; kernel density peak; mixture component centersBayesian inference, clustering, density estimation
Geometric meanexp(E[ln X]); mean of log-normal dataGrowth rates, financial returns, log-transformed data
Harmonic mean1 / E[1/X]; F-score in classificationRates, ratios, precision-recall trade-offs

Beyond the arithmetic mean, two other Pythagorean means deserve brief mention. The geometric mean, defined as (∏xᵢ)1/n, is the appropriate measure of center for multiplicative processes such as compound interest rates or population growth rates. The harmonic mean, defined as n / Σ(1/xᵢ), is used when averaging rates (e.g., speeds, throughput). A key inequality governs them: harmonic mean ≤ geometric mean ≤ arithmetic mean, with equality if and only if all observations are identical. These extensions reinforce the lesson's central thesis: the notion of "average" is not monolithic but context-dependent.

Practice Problems

PROBLEM 1CONCEPTUAL
A researcher reports that the mean income in a city is $92,000 while the median income is $58,000. What does the large gap between these two values tell you about the shape of the income distribution? Which measure would you recommend reporting to represent a "typical" resident's income, and why?
PROBLEM 2BASIC CALCULATION
A sample of 8 test scores is: 72, 85, 88, 90, 90, 93, 95, 97. Compute the mean, median, and mode of this dataset.
PROBLEM 3INTERMEDIATE
A frequency distribution reports the following data for the number of daily customer complaints at a service center over 30 days: 0 complaints (5 days), 1 complaint (8 days), 2 complaints (9 days), 3 complaints (5 days), 4 complaints (2 days), 10 complaints (1 day). Compute the mean and identify the median and mode. Then explain which measure would be most appropriate for setting staffing levels.
PROBLEM 4APPLIED
An environmental scientist collects dissolved oxygen (DO) readings (in mg/L) at 10 sampling sites along a river: 7.2, 7.5, 7.8, 8.0, 8.1, 8.3, 8.5, 8.7, 9.0, 14.6. A regulatory standard requires that the "central tendency" of DO exceed 8.0 mg/L. Compute both the mean and median. Does the site pass the regulatory standard under each measure? Discuss which measure is more scientifically defensible and why.
PROBLEM 5CRITICAL THINKING
Prove that for any dataset {x₁, x₂, …, xₙ}, the sum of deviations from the mean equals zero: Σᵢ₌₁ⁿ (xᵢ − x̄) = 0. Then explain why this property, while algebraically elegant, is actually a limitation when assessing spread—and how it motivates the use of squared deviations in the variance formula.

Summary

Measures of center distill a dataset into a single representative value. The arithmetic mean (x̄ = Σxᵢ/n) is the balance point that minimizes squared deviations (L₂ loss), uses every observation, and supports algebraic operations—but is sensitive to outliers. The median is the middle-ranked value that minimizes absolute deviations (L₁ loss) and possesses a 50% breakdown point, making it the preferred measure for skewed distributions. The mode identifies the most frequent value and is the only measure applicable to nominal data.

For symmetric, unimodal distributions all three measures coincide. Under right skewness the ordering is mode < median < mean, and under left skewness it reverses. Selecting the appropriate measure requires considering the data's level of measurement, distributional shape, the presence of outliers, and the analytical goals. The variational perspective—mean minimizes L₂, median minimizes L₁, mode minimizes L₀—provides a unified framework that extends naturally into regression, robust estimation, and Bayesian inference.

Varsity Tutors • College Statistics • Measures of Center