What Are Mean, Median, and Mode? (Simple Definitions)
Median is the middle value of a dataset when all values are arranged in ascending or descending order. If the count is even, it is the average of the two middle values.
Mode is the value that appears most frequently in a dataset. A dataset can be unimodal, bimodal, multimodal, or have no mode at all.
All three are measures of central tendency — statistics that summarize the center or "typical value" of a dataset. They answer the same question ("what is a representative value?") but in three fundamentally different ways. According to the NIST/SEMATECH e-Handbook of Statistical Methods, selecting the appropriate measure of location (central tendency) is one of the foundational decisions in descriptive statistics. The team at Statistics Fundamentals has built this reference page to make that decision clear and permanent.
- Mean: x̄ = Σx / n | sensitive to outliers | best for symmetric data
- Median: middle value after sorting | resistant to outliers | best for skewed data
- Mode: most frequent value | unaffected by outliers | only option for categorical data
- Normal distribution: mean = median = mode (all equal at the center)
- Right-skewed data: mode < median < mean (mean pulled toward the long right tail)
- Left-skewed data: mean < median < mode (mean pulled toward the long left tail)
Master Comparison Table: Mean vs Median vs Mode
This table is the single most useful reference for exams, teaching, and data analysis decisions. Every row addresses a question that comes up repeatedly in statistics — from which formula to apply to which measure to report in a research paper.
| Property | 📊 Mean | 📍 Median | 🎯 Mode |
|---|---|---|---|
| Definition | Arithmetic average — sum of all values divided by count | Middle value when data is arranged in order | Value that appears most frequently in the dataset |
| Formula / Method | x̄ = Σx / n | Sort data; pick value at position (n+1)/2 (odd n), or average of positions n/2 and (n/2)+1 (even n) | Find value(s) with the highest frequency count |
| Uses all data values? | Yes — every value contributes to the result | No — only the middle position(s) matter | No — only frequency of each value matters |
| Outlier sensitivity | High — one extreme value shifts the mean significantly | Low — outliers do not change the middle position | None — frequency counts are unaffected by outliers |
| Best data type | Continuous numerical (interval or ratio scale) | Ordinal, skewed numerical, or data with outliers | Categorical, nominal, or discrete numerical data |
| How many results? | Always exactly one | Always exactly one | Can be zero, one, or multiple (bimodal / multimodal) |
| Used in normal distribution? | Yes — mean = center of the bell curve | Yes — median = center (equals mean) | Yes — mode = peak (equals mean and median) |
| Used in skewed distribution? | Misleading — pulled toward the tail | Best choice — stays near the true center | Useful — identifies most common value |
| Real-world example | Average exam score in a class of 30 | Median household income (U.S. Census Bureau) | Most popular shoe size in a store's inventory |
| Notation | x̄ (sample), μ (population) | M or Md | Mo or simply "mode" |
Sources: NIST/SEMATECH e-Handbook of Statistical Methods §1.3.5.1; Penn State STAT 200 — Lesson 2: Summarizing Data; OpenIntro Statistics, 4th Edition, §2.1 (Diez, Çetinkaya-Rundel, Barr).
Formulas for Mean, Median, and Mode
Mean Formula
x̄ = sample mean (x-bar)
μ = population mean (mu)
Σx = sum of all data values
n = number of values in the sample
N = number of values in the population
The mean is the most commonly used average in everyday contexts. Its weakness is that it treats every value as equally important — including extreme ones. A single data point of $1,000,000 in a dataset of five $30,000 salaries shifts the mean to $203,333. The other four salaries average $30,000 each. The mean no longer describes any of them accurately. That is not a flaw in the formula — it is what the formula does. Knowing when to not use the mean is as important as knowing how to calculate it.
When values carry different weights (e.g., a final exam worth 50% and a quiz worth 10%), use the weighted mean: x̄w = Σ(wᵢ × xᵢ) / Σwᵢ. This appears on grade transcripts, economic indices, and financial portfolios. For ungrouped data with equal weights, this reduces back to the standard mean formula. See the full mean guide for grouped data methods.
How to Calculate the Median (Step-by-Step)
If n is even: average of positions n/2 and (n/2) + 1
n = number of values in the dataset
Sorted = data arranged in ascending order
The median's calculation does not use arithmetic on all the values — it only uses their order and position. That is exactly why it is unaffected by outliers. Whether the highest salary in a dataset is $100,000 or $10,000,000, if it sits at position 8 of 9 values, it changes nothing about the median. Penn State's STAT 200 curriculum describes the median as the preferred measure for any distribution with a "long tail" precisely for this reason (Penn State STAT 200 §2.2).
How to Find the Mode
The mode is the only measure of central tendency that works with categorical data. You cannot average colors or calculate the median political party — but you can find the most common one. In retail, the mode of clothing sizes sold in a week directly informs restocking decisions. In survey analysis, the mode of a 5-point Likert scale response identifies the most prevalent opinion. For numerical data, a frequency distribution chart or histogram makes the mode immediately visible as the tallest bar.
Worked Examples: Finding Mean, Median, and Mode
The same dataset runs through all five examples so you can compare how each measure is found and how they relate to each other.
A teacher records quiz scores for 9 students: 4, 7, 13, 2, 7, 9, 4, 7, 1
Example 1 — Finding the Mean
Dataset: 4, 7, 13, 2, 7, 9, 4, 7, 1 — Find the mean quiz score.
Sum all values: 4 + 7 + 13 + 2 + 7 + 9 + 4 + 7 + 1 = 54
Count the values (n): There are 9 values.
Divide: x̄ = 54 ÷ 9 = 6
✓ Mean = 6. The average quiz score is 6 out of a possible 13. Notice that no student actually scored exactly 6 — the mean describes the group, not any individual.
Example 2 — Finding the Median (Odd Dataset)
Dataset: 4, 7, 13, 2, 7, 9, 4, 7, 1 — Find the median quiz score.
Sort in ascending order: 1, 2, 4, 4, 7, 7, 7, 9, 13
Count the values: n = 9 (odd)
Find the middle position: (9 + 1) / 2 = position 5
Read the 5th value: 1, 2, 4, 4, 7, 7, 7, 9, 13 → 5th value = 7
✓ Median = 7. Exactly four students scored below 7 and four scored above. Compare to the mean of 6 — the median is slightly higher because the outlier (score of 13) pulled the mean down relative to the true center.
Example 3 — Finding the Median (Even Dataset)
Remove the score of 1 from the dataset. New dataset: 4, 7, 13, 2, 7, 9, 4, 7 — Find the median.
Sort: 2, 4, 4, 7, 7, 7, 9, 13
Count: n = 8 (even)
Find the two middle positions: n/2 = 4th and (n/2)+1 = 5th
Read those values: 2, 4, 4, 7, 7, 7, 9, 13 → 4th = 7, 5th = 7
Average them: (7 + 7) / 2 = 7
✓ Median = 7. When both middle values are equal, the median equals that value exactly. If the 4th had been 6 and the 5th had been 8, the median would be (6 + 8) / 2 = 7 — the same. Always average the two middle values for an even-count dataset.
Example 4 — Finding the Mode
Original dataset: 4, 7, 13, 2, 7, 9, 4, 7, 1 — Find the mode.
Tally each value's frequency:
| Value | Frequency (count) | Mode? |
|---|---|---|
| 1 | 1 | — |
| 2 | 1 | — |
| 4 | 2 | — |
| 7 | 3 | ✓ Highest |
| 9 | 1 | — |
| 13 | 1 | — |
✓ Mode = 7. The value 7 appears 3 times — more than any other value. The dataset is unimodal. Notice that the mode (7) happens to equal the median here, but they coincide for different mathematical reasons.
Example 5 — Summary: All Three Measures from One Dataset
The mean (6) is lower than the median and mode (both 7) because the three low scores (1, 2, 4) pull the mean downward — a mild example of left-side influence. The median and mode coincidentally agree here, but that is not always the case. Understanding why they diverge is more important than noticing that they match.
When to Use Mean, Median, or Mode
No single measure is universally correct. The right choice depends on the shape of your data, the presence of outliers, and the type of variable you are working with. Reporting the wrong measure gives a technically accurate but fundamentally misleading result.
| Situation | Best Measure | Reason |
|---|---|---|
| Symmetric distribution, no outliers | Mean | Uses every data point; gives maximum statistical precision |
| Skewed distribution (income, house prices, wait times) | Median | Unaffected by the long tail; accurately represents the typical value |
| Categorical or nominal data (colors, brands, political parties) | Mode | Mean and median are mathematically meaningless for non-numeric categories |
| Dataset with clear outliers | Median | Outliers pull the mean; median stays stable at the true center |
| Finding the most popular item or response | Mode | Mode directly identifies the most common value — mean and median cannot |
| Normally distributed exam scores (no extreme outliers) | Mean | All three measures are approximately equal; mean is most useful for further calculations (e.g., standard deviation) |
| Ordinal scale (e.g., satisfaction ratings 1–5) | Median | Ordinal data has rank but unequal intervals; mean assumes equal spacing between values |
| Small dataset, discrete whole numbers | Mode | Mode is immediately visible and requires no calculation; easy to communicate |
| Reporting government statistics (income, wealth) | Median | The U.S. Census Bureau reports median household income to avoid distortion from high-earner outliers |
| Quality control and process consistency checks | Mean | Control charts and Six Sigma processes use the mean as the process target — deviations from the mean are what matter |
Draw a histogram of your data. If it looks like a symmetric bell → use mean. If one side has a long tail → use median. If you want to know the most common answer, or your data is categorical → use mode. When in doubt between mean and median, check for outliers first.
How Outliers Affect Mean, Median, and Mode
This is the most practically important concept in the entire mean-median-mode discussion. An outlier is a value that sits unusually far from the rest of the dataset. A single outlier can silently destroy the usefulness of the mean while leaving the median and mode completely intact.
The Salary Demonstration
Six people work at a small company. Their annual salaries are: $30,000 / $31,000 / $32,000 / $33,000 / $34,000 / $500,000 (the CEO).
| Measure | Without CEO Salary ($500K) | With CEO Salary ($500K) | Change |
|---|---|---|---|
| Mean | $32,000 | $110,000 | +$78,000 ↑↑↑ |
| Median | $31,500 | $32,500 | +$1,000 (negligible) |
| Mode | No mode | No mode | Unchanged |
Outlier Effect — One CEO Salary vs. Five Employee Salaries
The green line (median) barely moves. The orange/red marker (mean) shifts dramatically to the right — past four of the five actual employees' salaries.
This is not an abstract scenario. This is precisely why the U.S. Census Bureau reports median household income in its official publications. If they used the mean, the existence of a small number of ultra-high earners would make "average income" appear far higher than what the typical American household actually earns. The median — by definition — tells you what a person in the middle of the distribution earns, regardless of how extreme the ends are.
When you see "average salary in [city] is $X," that is almost certainly a mean — and likely an overestimate of what most workers actually earn. Check whether the source reported mean or median. In most cases, the median paints a more honest picture of what a typical person in that group actually experiences.
Mean, Median, and Mode in Skewed Distributions
The relationship between the three measures changes depending on how the data is distributed. This is a frequently tested concept in statistics exams and a critical consideration in data science and research.
How Mean, Median, and Mode Relate in Different Distribution Shapes
In a right-skewed distribution — common in income, house prices, population sizes, and insurance claims — the long tail extends to the right. High-value outliers pull the mean upward, past both the median and mode. The mode sits at the peak (most common value), the median sits in the true middle, and the mean is pulled furthest right. For a deeper treatment of distribution shapes, see the normal distribution guide and the statistics and probability overview.
In a left-skewed distribution — seen in exam scores on easy tests (most students score high with a few scoring very low), age at death in populations with good healthcare, and reaction times — the tail extends left. The mean is pulled below both the median and mode.
A useful rough approximation for moderate skew: Skewness ≈ 3 × (Mean − Median) / Standard Deviation. Positive result → right skew; negative → left skew; near zero → roughly symmetric. This is a quick diagnostic, not a formal test. For formal skewness measurement, see the third standardized moment in the descriptive statistics section.
Real-World Applications of Mean, Median, and Mode
These three measures are not textbook abstractions. They appear in every government report, business dashboard, sports broadcast, and medical paper that summarizes data. The examples below show exactly which measure each field relies on and why the choice matters.
Why the Government Reports Median Household Income
The U.S. Census Bureau's Current Population Survey reports median household income — not mean. In 2023, U.S. median household income was approximately $77,000. The mean would have been notably higher because a small percentage of ultra-high earners (hedge fund managers, tech executives, inherited wealth holders) pull the mean upward dramatically. If policymakers used the mean to design income support programs, they would systematically underserve the majority of households whose income sits well below the mean. The median tells you what a household at the 50th percentile actually earns.
Mean Scores in Standardized Tests
When exam scores follow a roughly normal distribution — as they often do for well-designed standardized tests — the mean, median, and mode are close to equal, and the mean is the preferred summary because it enables standard deviation, confidence intervals, and t-tests. The National Center for Education Statistics (NCES) reports mean scale scores for the National Assessment of Educational Progress (NAEP). However, when a test is very easy or very hard, scores skew — and the median becomes a more reliable indicator of typical performance. A class where most students scored 90–95 but three students scored 20 has a mean dragged down to, say, 80 — a number no student actually represents.
Mode Drives Product Restocking
A shoe retailer sells sizes 6 through 13. The mean shoe size sold in a week might calculate to 9.3 — a size that does not exist in whole number form. The median might be 9. But the mode — say, size 10 — is what the buyer uses to decide how many units of each size to restock. No matter how mathematically elegant the mean is, you cannot order 9.3 shoes. In fashion, food service, and manufacturing, mode directly drives operational decisions: the most common clothing size, most frequently ordered dish, most commonly produced component dimension. Mean and median cannot substitute for mode in these contexts.
Why Medical Studies Use Median Survival
In clinical oncology research, survival time after treatment follows a highly right-skewed distribution. A few patients survive many decades; most patients have shorter survival windows. Using the mean survival time would make a treatment appear more effective than it is for the typical patient. Oncologists and journal editors specifically require median survival time for this reason. The National Cancer Institute documents this convention in clinical trial reporting standards. A treatment with median survival of 18 months tells patients far more about their likely experience than a mean of 24 months pulled up by a handful of exceptional long-term survivors.
Median Home Price as the Standard
The National Association of Realtors and most real estate market reports publish median home prices, not mean prices. One $20 million luxury property sale in a neighborhood of $350,000 homes raises the mean price substantially while the median barely moves. Homebuyers, sellers, and mortgage lenders are best served by the median because it accurately describes what a typical transaction in that market looks like. The same logic applies to commercial real estate, rent prices, and land valuations globally.
Interactive Mean, Median & Mode Calculator
Enter your dataset below and the calculator will compute all three measures instantly with a step-by-step breakdown. For a dedicated calculator with additional statistics, see the full mean/median/mode calculator and the complete descriptive statistics calculator.
🧮 Mean, Median & Mode Calculator
Enter numbers separated by commas or spaces. Works for any size dataset.
For more statistical measures — variance, standard deviation, quartiles, and range — use the descriptive statistics calculator. For individual measures, try the mean calculator or median calculator.
Mean, Median, and Mode from Grouped Data (Advanced)
When data is given in a frequency table with class intervals — as is common in exam questions and real survey data — the calculation method changes. The exact values are unknown; only the classes and their frequencies are given.
Mean from a Grouped Frequency Table
Use the midpoint of each class interval as a representative value, multiply by the frequency, sum the products, then divide by the total frequency.
fᵢ = frequency of class i
mᵢ = midpoint of class i
Σfᵢ = total frequency (n)
Test scores grouped into intervals. Find the estimated mean.
| Score Interval | Midpoint (mᵢ) | Frequency (fᵢ) | fᵢ × mᵢ |
|---|---|---|---|
| 10–19 | 14.5 | 3 | 43.5 |
| 20–29 | 24.5 | 7 | 171.5 |
| 30–39 | 34.5 | 12 | 414.0 |
| 40–49 | 44.5 | 8 | 356.0 |
| 50–59 | 54.5 | 5 | 272.5 |
| Total | — | Σf = 35 | Σ(fm) = 1,257.5 |
✓ Estimated Mean = 1,257.5 / 35 ≈ 35.9. Note this is an estimate — the actual mean could differ because the true values within each interval are unknown. For an explanation of computed vs actual mean, see computed mean vs actual mean.
Modal Class in Grouped Data
The modal class is the class interval with the highest frequency. In the example above, the modal class is 30–39 (frequency = 12). The mode is estimated as the midpoint: 34.5. For more precise estimation, the modal class method uses the frequencies of neighboring classes — covered in advanced statistics textbooks and curricula such as MIT OpenCourseWare STAT 18.650.
Exam Revision: Quick Notes & Formula Cheatsheet
- Mean = total ÷ count; best for symmetric, clean numerical data; sensitive to outliers.
- Median = middle value after sorting; best for skewed data or data with outliers; robust to extreme values.
- Mode = most frequent value; best for categorical data and finding popularity; can be bimodal or have no mode.
- Outlier effect: outliers pull the mean significantly but leave the median and mode largely unchanged.
- Distribution pattern: symmetric → mean = median = mode; right-skewed → mode < median < mean; left-skewed → mean < median < mode.
| Measure | Formula | Key Step | Common Mistake |
|---|---|---|---|
| Mean | x̄ = Σx / n | Add all values, divide by count | Forgetting to divide by n (just summing) |
| Median (odd n) | Position (n+1)/2 | Sort first, then pick the middle value | Not sorting before finding the middle |
| Median (even n) | Average positions n/2 and (n/2)+1 | Sort, average the two central values | Picking just one middle value instead of averaging |
| Mode | Highest frequency | Tally or sort to count repeats | Stating "no mode" when values appear once (correct: no mode is valid) |
1) Calculating the median without sorting first — always sort. 2) Averaging the middle two values when n is odd — only do this for even n. 3) Saying there is no mode when a value appears once — if all values appear once equally, then there is no mode; if one value appears more than once (even just twice), that is the mode. 4) Confusing the modal class (grouped data) with a specific mode value. 5) Choosing mean for salary or house price questions — always use median for skewed economic data in exams.
Formula & Concept Glossary
Each term here connects directly to mean, median, or mode. This glossary is designed for rapid exam review and AI-parseable reference.
| Term | Formula / Notation | Plain-English Definition | Connection to MMM |
|---|---|---|---|
| Mean | x̄ = Σx / n | Sum of all values divided by count — the arithmetic average | First and most common measure of central tendency |
| Median | Position (n+1)/2 | The value that splits sorted data into two equal halves | Best for skewed data; unaffected by outliers |
| Mode | Max frequency | The most frequently occurring value in a dataset | Only MMM measure that works for categorical data |
| Central Tendency | (concept) | A single value that describes the center or typical value of a dataset | The category containing mean, median, and mode |
| Outlier | (no formula) | A value significantly distant from the rest of the dataset | Distorts the mean; does not affect median or mode |
| Skewness | γ₁ = Σ[(xᵢ−x̄)³/n] / σ³ | Asymmetry in a distribution — positive (right) or negative (left) | Determines the ordering of mean, median, and mode |
| Frequency Distribution | (table format) | A table or chart showing how often each value or range appears | Used to calculate mode and estimated mean from grouped data |
| Normal Distribution | f(x)=e^−(x−μ)²/2σ² / σ√2π | Symmetric bell-shaped distribution defined by μ and σ | The only distribution where mean = median = mode exactly |
| Bimodal | (two equal max frequencies) | A dataset with exactly two modes (two values tied for highest frequency) | Unique to mode — mean and median cannot be bimodal |
| Standard Deviation | σ = √[Σ(xᵢ−μ)²/N] | Average distance of data points from the mean | Calculated from the mean; used in skewness approximation |
For related measures of spread, see the standard deviation guide and the variance guide. For the full terminology reference, visit the Statistics Fundamentals glossary.
Common Mistakes with Mean, Median, and Mode
| Mistake | What People Do Wrong | What's Correct |
|---|---|---|
| Not sorting before median | Finding the middle value of unsorted data | Always sort ascending first — the median is a positional measure |
| Averaging for even n — wrong values | Averaging the wrong two positions (e.g., position 3 and 4 when n=10) | For n=10 (even): average positions 10/2=5 and 10/2+1=6 |
| Using mean for skewed data | Reporting "average income" as the mean in a right-skewed salary dataset | Use median for income, house prices, and any skewed distribution |
| Saying no mode when values repeat | "This dataset: 3, 3, 5, 7 has no mode" (wrong — 3 is the mode) | No mode only when every value appears the same number of times |
| Reporting mean for categorical data | Computing "average eye color" or "mean political party" | Categorical data requires mode; mean and median are meaningless |
| Ignoring bimodal distributions | Reporting one mode when two values tie for highest frequency | Report both modes — bimodal distributions convey important structural information about the data |
Frequently Asked Questions — Mean, Median & Mode
Sources cited in this guide: NIST/SEMATECH e-Handbook of Statistical Methods, §1.3.5.1 — Measures of Location · Penn State STAT 200, Lesson 2.2 — Measures of Central Tendency · U.S. Census Bureau — Income and Poverty Reports · National Center for Education Statistics — NAEP · National Cancer Institute — Clinical Trial Reporting Standards · MIT OpenCourseWare, 18.650 Statistics for Applications · OpenIntro Statistics, 4th Ed. (Diez, Çetinkaya-Rundel, Barr)