ProMathTools
ProMathTools

Mean Absolute Deviation Calculator

Average the distances from the mean, or switch to the median absolute deviation when a single outlier would stretch that average.

Commas, spaces, tabs, semicolons, or new lines. Example: 2, 4, 6, 8, 10

MAD = 2.4000

mean

Population standard deviation is 2.8284. It is never smaller than the MAD, because squaring distances gives extra weight to the far points before the square root brings the unit back.

Show steps

MAD = Σ|x − mean| ÷ n = 12.0000 ÷ 5 = 2.4000.

xx − mean|x − mean|
2.0000-4.00004.0000
4.0000-2.00002.0000
6.00000.00000.0000
8.00002.00002.0000
10.00004.00004.0000

What mean absolute deviation tells you

Mean absolute deviation, often shortened to MAD, is the average distance between the numbers in a list and their arithmetic mean. Picture five quiz scores. The mean says where the class sat overall. The MAD says how far a typical score wandered from that center. A small MAD means the scores clustered. A large MAD means they were scattered, even if the average looks the same as another class.

The word absolute matters. Subtracting the mean from each score produces negatives for scores below the average and positives for scores above it. Those signed gaps cancel, and the total is always zero. Taking the absolute value keeps every gap as a distance, so the average describes spread instead of collapsing to nothing.

Formula

For a plain list, MAD = Σ|x − mean| ÷ n. Add the values, divide by how many there are to get the mean, subtract that mean from each value, drop the sign, add those distances, and divide by n again. When a value repeats, weight it: MAD = Σ f|x − mean| ÷ Σ f. The frequency f counts how many times that value occurred, and Σ f replaces n. This calculator keeps full precision while it works and rounds only the numbers you see. Use the decimal-places control to show 0 through 6 digits.

How to use this calculator

Paste a list in the data box. Commas, spaces, tabs, semicolons, and line breaks all count as separators, so a column copied from a spreadsheet is fine. Or switch to the value-and-frequency table when a textbook already grouped repeats: type each distinct number, type how often it appears, and add or remove rows as needed. Choose mean absolute deviation for the school definition, or median absolute deviation to see the middle distance from the median. Copy result puts n, the mean, the sum of absolute deviations, both deviation measures, and the population standard deviation on the clipboard. Reset restores the sample list 2, 4, 6, 8, 10.

An empty box asks you to enter a number. A token such as abc is named together with its position in the list, so you can find it quickly. A frequency of 0, a negative frequency, or a frequency that is not a whole number is rejected with the row number. A single number is valid: its MAD is 0, because that number is the mean and the only distance is zero.

MAD, standard deviation, and the interquartile range

All three describe spread, and they answer slightly different questions. MAD stays in the original unit and treats every distance evenly, which is why it is usually the first spread measure taught. Population standard deviation squares each distance before averaging, then takes a square root. Squaring makes a point that is 10 units away count much more than ten points that are 1 unit away. For any data set the population standard deviation is at least as large as the MAD; they match only in special cases such as every value being identical, where both are 0. The interquartile range ignores the outer quarter of the data on each side and reports the width of the middle half. It does not use the mean at all.

Outliers move these measures by different amounts. One huge value shifts the mean toward itself, and every absolute deviation is then measured from that shifted mean, so the MAD rises sharply. Squared deviations rise even faster, so the standard deviation usually jumps more than the MAD. The interquartile range often barely moves, because the extreme value sits outside the middle half. The median absolute deviation is the robust cousin of MAD: find the median, measure each distance from it, and take the median of those distances. A single spike becomes one large distance, and the median of the distances can ignore it.

Try 1, 2, 3, 4, and 100 in the list. The mean is 22 and the MAD is 31.2, even though four of the five numbers are tiny. The median absolute deviation stays at 1, which describes the tight cluster. The population standard deviation is about 39.01, larger than the MAD, as the inequality promises. For a balanced list such as 2, 4, 6, 8, 10 the mean is 6, the distances add to 12, and the MAD is 2.4, with a population standard deviation near 2.83.

Compare this result with the standard deviation calculator, the variance calculator, the interquartile range calculator, the mean calculator, and the mean, median, and mode calculator. The longer walk-through, including a frequency example and the mistakes that make the signed deviations sum to zero, is in the guide below.

Read the full guide

A MAD of 2 means the typical value sits 2 units away from the average. It is an average of distances, so the original units stay intact: scores, centimeters, dollars, and so on.

No. MAD averages the plain distances from the mean. Population standard deviation averages the squared distances and then takes a square root, so large gaps pull it up more. For the same data, population standard deviation is at least as large as the MAD.

Use the mean absolute deviation when the question asks for MAD, or when the data has no extreme spikes. Use the median absolute deviation when one or two unusual values would otherwise drag the average distance up.

No. Each term is an absolute distance, and the count of values is positive, so MAD is 0 or greater. It is 0 only when every value is identical.

Related Calculators

Standard deviationVarianceInterquartile rangeMeanMean, median, and modeHow to find MADCorrelation CoefficientLinear RegressionSample Size CalculatorT-Test CalculatorMedian CalculatorMode Calculator

Related Guides

About UsBlogChrome ExtensionFAQsTerms and ConditionsPrivacy PolicySitemap
EmailLinkedInFacebook

© 2025-2026 PromathTools. All rights reserved.

Built by KruskalCode – SaaS & Automation Experts · EverydayDevTools