Mean Absolute Deviation Calculator
Find the average distance of your data from the mean, the median, or any value you choose—with every step shown and a plot of each deviation.
Paste or type up to 5,000 numbers
Separate values with commas, spaces, semicolons, or new lines—a column copied from Excel or Google Sheets works as is. Most classes use the mean as the center; switch to the median when the data have outliers.
—
—
Count (n)
—
—
Mean
—
Average value
Median
—
Middle value
Range
—
—
Distance of Each Value from the Center
—
How it was calculated
Other measures of spread
Absolute Deviations
| # | Value xᵢ | xᵢ − m | |xᵢ − m| |
|---|
How to Use This Calculator
- Type or paste your numbers into the data box. The counter below it confirms how many values were read.
- Choose the center: Mean for the standard school definition, Median for data with outliers, or Other value to measure distance from a target or forecast.
- Pick the number of decimals and press Calculate MAD.
- Read the MAD, check each deviation in the table, and compare it with the median absolute deviation and standard deviation shown alongside.
What Is Mean Absolute Deviation?
A mean absolute deviation calculator measures the average distance between each value in a dataset and the dataset's mean. It calculates each value's absolute deviation from the mean and averages these deviations. A lower mean absolute deviation indicates values cluster more closely around the mean, while a higher value indicates greater dispersion.
MAD answers a simple question: how far is a typical value from the center? The center is usually the mean, but the median or another reference value can be used too. A MAD of 9 test points means scores are, on average, 9 points above or below the class average.
Distances are taken as absolute values, so a value 5 below the mean counts the same as one 5 above it. Without the absolute value, the positive and negative deviations from the mean would always cancel out to exactly zero.
Same units
MAD is in the units of the data—points, kg, minutes—so it is easy to explain.
Zero means identical
MAD is 0 only when every value is equal. Larger values mean more spread.
Taught in grade 6
MAD is the first measure of variability many students learn.
Curriculum reference: Common Core State Standards — Grade 6, 6.SP.B.5.c.
Mean Absolute Deviation Formula
MAD = (|x₁ − m| + |x₂ − m| + … + |xₙ − m|) ÷ n = Σ|xᵢ − m| ÷ n
xᵢ are the data values, n is how many there are, and m is the central point—the mean x̄ unless stated otherwise.
Four steps by hand
- Find the center. For the mean, add the values and divide by n.
- Subtract the center from each value.
- Drop the minus signs to get the absolute deviations.
- Add the absolute deviations and divide by n.
Definition reference: NIST/SEMATECH e-Handbook — Measures of Scale.
Worked Example: Eight Test Scores
Scores: 72, 85, 90, 64, 78, 88, 95, 70. The sum is 642, so the mean is 642 ÷ 8 = 80.25.
| Score | 72 | 85 | 90 | 64 | 78 | 88 | 95 | 70 |
|---|---|---|---|---|---|---|---|---|
| |x − 80.25| | 8.25 | 4.75 | 9.75 | 16.25 | 2.25 | 7.75 | 14.75 | 10.25 |
The absolute deviations add up to 74, so MAD = 74 ÷ 8 = 9.25. A typical score is about 9 points from the class average. For comparison, the population standard deviation is 10.23—a little larger, as it always is.
Mean Absolute Deviation vs. Other Measures of Spread
All describe how spread out data are, but they treat large deviations differently.
| Measure | Formula | Effect of outliers | Best for |
|---|---|---|---|
| Mean absolute deviation | Σ|x − x̄| ÷ n | Moderate | Explaining a typical distance in plain terms |
| Standard deviation | √(Σ(x − x̄)² ÷ n) | Strong—squaring magnifies large gaps | Normal-distribution work, inference, finance |
| Median absolute deviation | median |x − median| | Very weak—up to half the data can be extreme | Robust statistics, outlier detection |
| Interquartile range | Q3 − Q1 | Weak | Box plots, skewed data |
| Range | max − min | Extreme—uses only two values | Quick first look |
One outlier, four answers
For 3, 4, 4, 5, 6, 40: SD = 13.30, MAD around the mean = 9.89, MAD around the median = 6.67, and median absolute deviation = 1.0. The single value 40 dominates the first two.
MAD ≤ SD, always
For any data set, MAD around the mean is never larger than the population standard deviation. For normally distributed data, MAD ≈ 0.80 × SD, so SD ≈ 1.25 × MAD.
Mean vs. Median as the Center
The abbreviation MAD is used for three different things. Check which one your textbook, teacher, or software means.
Mean absolute deviation around the mean
The school default. Average of |x − mean|.
Mean absolute deviation around the median
Average of |x − median|. It is the smallest possible MAD for that data set—no other center gives a lower value.
Median absolute deviation
Median of |x − median|. Multiply by 1.4826 to estimate the SD of normal data while ignoring outliers.
In forecasting, "MAD" usually means the mean absolute deviation of forecast errors, the same idea as mean absolute error. Use Other value only when every value is compared with one fixed target; for a separate forecast per period, calculate each error first and enter the errors with a center of 0.
Learning reference: Khan Academy — Mean absolute deviation (MAD) review.
Frequently Asked Questions
What does the MAD tell you?
How far a typical value is from the center, in the same units as the data. If two classes both average 80 points but one has a MAD of 3 and the other a MAD of 12, the first class scored with much more consistency. Where a measure of central tendency such as the mean tells you where the data sit, MAD tells you how much variability there is around that point.
Can the mean absolute deviation be negative?
No. Every absolute deviation is zero or positive, so their average is too. A MAD of 0 means every observation in the dataset is identical, so there is no dispersion at all.
Why use absolute values instead of plain differences?
The plain differences from the arithmetic mean always sum to zero, because values above and below the mean balance each other. Taking absolute values keeps every distance positive so they can be averaged. Standard deviation solves the same problem a different way, by squaring each difference.
Is MAD the same as standard deviation?
No. Standard deviation squares each deviation before averaging (that average is the variance) and takes a square root at the end, which gives large deviations more weight. As a result, the standard deviation is always at least as large as the MAD around the mean. For a normal distribution, MAD is about 80% of the standard deviation.
Should I divide by n or n − 1?
Divide by n, the number of data points. The n − 1 correction belongs to the sample variance and standard deviation, where it adjusts for estimating a population value from a sample; the standard MAD formula, and functions such as Excel's AVEDEV, divide by n whether the data are a sample or a whole population.
How do I calculate MAD in Excel or Google Sheets?
Use =AVEDEV(A1:A20) for the mean absolute deviation around the mean. For the median absolute deviation, use =MEDIAN(ABS(A1:A20−MEDIAN(A1:A20))), entered as an array formula in older Excel versions. If your data come as a frequency table, expand it into individual values first, or paste the full list into this calculator as input—the output includes both versions.
When should I use the median instead of the mean?
When the data are skewed or contain outliers, such as incomes, house prices, or response times. One extreme value pulls the mean toward it and inflates every deviation; the median stays put. A quick check is the range: if the gap between the minimum and maximum is driven by one or two values, the median is the safer center.
How do I compare spread between data sets with different units or sizes?
Divide MAD by the mean to get a relative MAD, shown in the calculator as a percentage. A MAD of 5 g on a 50 g item (10%) is far more variable than 5 g on a 1,000 g item (0.5%). The same idea is used in statistics to judge forecast accuracy: the mean absolute deviation of the errors, or residuals, between predicted values and each actual measurement shows how far off a forecast typically is.
Other Useful Calculators
Empirical Rule Calculator
Find the 68–95–99.7% ranges of a normal distribution.
Process Capability Index Calculator
Cp, Cpk, Cpm, and expected defects.
Poisson Distribution Calculator
Probability of a number of events in an interval.
Pie Chart Calculator
Turn values into percentages, angles, and a chart.
Lognormal Distribution Calculator
Probabilities, quantiles, and measures for skewed data.
Negative Binomial Distribution Calculator
Probability of the r-th success on a given trial.