Mean Absolute Deviation Calculator

Mean Absolute Deviation Calculator

Paste your numbers below, separated by commas, spaces or line breaks. The calculator shows MAD from the mean, MAD from the median and the median absolute deviation, with every step shown.

MAD from the mean uses the mean as its centre; MAD from the median and the median absolute deviation use the median. They are three different numbers, so this tool reports all three and you can compare them side by side.

MAD from the mean
2.40

MAD from the median = 2.4 · Median absolute deviation = 2

Mean
6
Median
6
MAD from the median
2.4
Median absolute deviation
2
Sample SD (s)
3.162
Population SD (σ)
2.828
σ estimate: 1.2533 × MAD (mean)
3.008
σ estimate: 1.4826 × median abs. dev.
2.965

Absolute deviations from the mean

Each bar is |x − mean| for one value. The dashed line is the MAD from the mean.

Step-by-step calculation

#Value (x)x − mean|x − mean|x − median|x − median|

    Mean absolute deviation (MAD) measures how far values sit from a typical value, on average. This calculator finds all three versions at once: MAD from the mean, MAD from the median and the median absolute deviation, plus the mean, median, standard deviation and two normal-distribution sigma estimates β€” paste your data and get every step shown.

    What is mean absolute deviation?

    Mean absolute deviation (MAD) is the average distance between each value in a data set and a central value. Instead of squaring the differences, as standard deviation does, MAD takes their absolute values, so the result stays in the same units as the data β€” dollars, points, centimetres β€” and is easy to explain: it is roughly how far a typical value sits from the centre.

    The confusion starts with the name. “MAD” is used for three different statistics, and they can give three different answers for the same data. MAD from the mean averages |x βˆ’ mean|; MAD from the median averages |x βˆ’ median|; and the median absolute deviation is the median of |x βˆ’ median|. Always check which version a source means before you compare numbers. The formal definitions are set out in the NIST/SEMATECH e-Handbook of Statistical Methods.

    Formulas: MAD from the mean = Ξ£|x βˆ’ mean| Γ· n Β· MAD from the median = Ξ£|x βˆ’ median| Γ· n Β· median absolute deviation = median(|x βˆ’ median|). Here n is the number of values, and |…| means absolute value.

    What is the difference between MAD-from-mean, MAD-from-median and median absolute deviation?

    The three MADs differ in two choices: which centre you measure from β€” the mean or the median β€” and how you combine the distances β€” by their mean or by their median. That is the whole difference, and it matters: for the same data, the three answers can be far apart.

    MeasureCentreHow distances are combinedFor 2, 4, 6, 8, 10For 1, 2, 3, 4, 100Robust to outliers?
    MAD from the meanMeanMean of |x βˆ’ mean|2.431.2No β€” the mean moves with outliers
    MAD from the medianMedianMean of |x βˆ’ median|2.420.2Partly β€” the median centre helps, but the distances are still averaged
    Median absolute deviationMedianMedian of |x βˆ’ median|21Yes β€” the most robust of the three

    For the outlier data 1, 2, 3, 4, 100, the mean is dragged to 22, so MAD from the mean jumps to 31.2. The median stays at 3, so MAD from the median is 20.2 and the median absolute deviation is just 1 β€” the single value 100 barely moves it. That is why the median absolute deviation is the standard robust scale in statistics, and why it pairs naturally with the quartile calculator and the interquartile range: both describe spread without letting one extreme value dominate.

    Which one should you report? If your audience expects the classic definition, give MAD from the mean. If outliers are present, give MAD from the median or the median absolute deviation β€” and always name the version you used.

    How to calculate mean absolute deviation

    Calculating MAD from the mean takes five steps. Work through them once by hand and the idea sticks:

    1. Add all the values and divide by how many there are to get the mean.
    2. Subtract the mean from each value. (For MAD from the median, use the median instead.)
    3. Take the absolute value of each difference, so negatives become positives.
    4. Add the absolute differences together.
    5. Divide by the number of values. The result is the MAD from the mean.

    Worked example β€” 2, 4, 6, 8, 10. The mean is (2 + 4 + 6 + 8 + 10) Γ· 5 = 6. The deviations are βˆ’4, βˆ’2, 0, 2, 4; their absolute values are 4, 2, 0, 2, 4; the sum is 12, and 12 Γ· 5 = 2.4. So the MAD from the mean is 2.4.

    Value (x)x βˆ’ mean|x βˆ’ mean|
    2βˆ’44
    4βˆ’22
    600
    822
    1044
    Total012

    For the median versions, first order the data and find the median β€” 6 here β€” then repeat steps 2–5 from the median. For the median absolute deviation, take the median of the absolute differences instead of their mean: the absolute differences are 4, 2, 0, 2, 4, and their median is 2. Try the same numbers in the Standard Deviation Calculator to compare the two measures side by side.

    MAD vs standard deviation β€” which should you use?

    Use MAD when you want a robust, easy-to-explain summary of spread, especially with skewed data or outliers. Use standard deviation when you need the measure that underpins the normal distribution, confidence intervals and most statistical tests. In practice, report both: they answer slightly different questions.

    Β MADStandard deviation
    What it measuresAverage absolute distance from the centreSquare root of the average squared distance from the mean
    UnitsSame as the dataSame as the data
    OutliersResistant in its median versionsSensitive β€” squaring magnifies extremes
    Equals zero whenAll values are equalAll values are equal
    Best forRobust summaries, skewed dataNormal models, confidence intervals, tests

    The outlier example makes the gap concrete: for 1, 2, 3, 4, 100, the sample standard deviation is about 43.62, while the median absolute deviation is 1. The standard deviation is dominated by the single value 100; the median absolute deviation describes the other four values. If you need a spread that every value contributes to, use the standard deviation; if you need one that an outlier cannot hijack, use a median-based MAD.

    Mean absolute deviation and the normal distribution

    For data that follow a normal distribution, MAD can be converted into an estimate of the standard deviation (Οƒ). The links are fixed constants: Οƒ β‰ˆ 1.2533 Γ— MAD from the mean, and Οƒ β‰ˆ 1.4826 Γ— median absolute deviation. The second constant is the one used most often, because the median absolute deviation is the robust estimator β€” multiply it by 1.4826 and you get a standard-deviation estimate that outliers cannot drag around.

    For 2, 4, 6, 8, 10, MAD from the mean is 2.4, so Οƒ β‰ˆ 1.2533 Γ— 2.4 β‰ˆ 3.008. The true sample standard deviation is 3.162, so the estimate is close. For the outlier data 1, 2, 3, 4, 100, the median absolute deviation is 1, giving Οƒ β‰ˆ 1.4826 β€” far below the sample standard deviation of 43.62. That is exactly the point: the estimate describes the bulk of the data, not the outlier. Turn a mean and standard deviation into probabilities with the normal distribution calculator, or see the 68-95-99.7 shares with the empirical rule probability finder.

    These conversions assume a roughly normal shape. For skewed data, report the MAD itself rather than converting it.

    Mean absolute deviation in Excel and Google Sheets

    Excel and Google Sheets have a built-in function for MAD from the mean only: =AVEDEV(A1:A5) returns the average of the absolute deviations from the mean. For 2, 4, 6, 8 and 10 it returns 2.4.

    There is no built-in function for the median versions. For MAD from the median, use =AVERAGE(ABS(A1:A5-MEDIAN(A1:A5))) as an array formula β€” in older Excel, confirm it with Ctrl+Shift+Enter; in Google Sheets, wrap it in =ARRAYFORMULA(…). For the median absolute deviation, use =MEDIAN(ABS(A1:A5-MEDIAN(A1:A5))) the same way. When your data are skewed, a shape check helps too β€” try the skewness and kurtosis calculator.

    Worked examples

    Example 1 β€” an outlier: 1, 2, 3, 4, 100. The mean is 22 and the median is 3. MAD from the mean is 31.2, MAD from the median is 20.2, the median absolute deviation is 1 and the sample standard deviation is about 43.62. The contrast is the lesson: one value of 100 multiplies the mean-based measures but barely touches the median-based ones.

    Example 2 β€” no spread at all: 5, 5, 5, 5. Every value equals the mean and the median, so every absolute deviation is 0. All three MADs are 0 and both standard deviations are 0. When every value is identical there is no spread to measure.

    Example 3 β€” 10, 20, 30. The mean and the median are both 20. MAD from the mean is about 6.667, the median absolute deviation is 10, the sample standard deviation is 10 and the population standard deviation is about 8.165. Here the median absolute deviation (10) is larger than MAD from the mean (6.667) β€” with only three values, the middle absolute deviation is the largest one. It is a reminder that the three MADs are different numbers, not three names for one thing. For a robust companion to MAD, compare the interquartile range, which also ignores extreme values.

    Example 4 β€” a single value: 7. With n = 1, the mean and median are both 7, every MAD is 0 and the population standard deviation is 0. The sample standard deviation is not defined, because it divides by n βˆ’ 1 = 0 β€” the calculator shows a dash for it instead of a number.

    Keep exploring: try the standard deviation calculator on the same data to compare the two measures, or browse all our free statistics tools.

    Frequently asked questions

    What is mean absolute deviation?

    Mean absolute deviation (MAD) is the average distance of the values in a data set from a central value, using absolute values instead of squares. MAD from the mean is Ξ£|x βˆ’ mean| Γ· n. It is measured in the same units as the data, so a MAD of 2.4 for test scores means a typical score sits 2.4 points from the mean.

    What is the difference between MAD-from-mean, MAD-from-median and median absolute deviation?

    They measure from different centres and average the distances differently. MAD from the mean averages |x βˆ’ mean|; MAD from the median averages |x βˆ’ median|; the median absolute deviation is the median of |x βˆ’ median|. For 1, 2, 3, 4, 100 they are 31.2, 20.2 and 1 β€” three different numbers, so always name the version you used.

    How do you calculate mean absolute deviation step by step?

    Find the mean, subtract it from each value, take the absolute value of each difference, add them up and divide by the number of values. For 2, 4, 6, 8, 10: the mean is 6, the absolute deviations are 4, 2, 0, 2, 4, their sum is 12, and 12 Γ· 5 = 2.4.

    Is MAD more robust than standard deviation?

    The median-based versions are. The median absolute deviation barely moves when an outlier is added, while the standard deviation is inflated because it squares large deviations. For 1, 2, 3, 4, 100, the sample standard deviation is about 43.62 but the median absolute deviation is 1. MAD from the mean is less robust, because the mean itself moves with outliers.

    Can mean absolute deviation be zero or negative?

    It can be zero but never negative. MAD is an average of absolute values, so it is always zero or positive. It is zero only when every value is identical β€” for example, 5, 5, 5, 5 has a MAD of 0.

    What is the median absolute deviation used for?

    It is used as a robust measure of spread and to spot outliers: values more than a few scaled median absolute deviations from the median are flagged as unusual. It is also converted to a standard-deviation estimate for roughly normal data with Οƒ β‰ˆ 1.4826 Γ— median absolute deviation. See the empirical rule to turn that standard deviation into probabilities.

    How do you calculate mean absolute deviation in Excel?

    Use =AVEDEV(range) for MAD from the mean; for 2, 4, 6, 8 and 10 it returns 2.4. There is no built-in function for the median versions β€” use =MEDIAN(ABS(range-MEDIAN(range))) as an array formula for the median absolute deviation.

    What is the MAD of 2, 4, 6, 8 and 10?

    The mean and the median are both 6. MAD from the mean is 2.4, MAD from the median is 2.4, and the median absolute deviation is 2. The sample standard deviation is about 3.162 and the population standard deviation is about 2.828.

    More Tools to Explore

    Found your MAD? Keep exploring the spread of your data with these free tools β€” no sign-up, instant results.

    ● LIVE

    Z-Score Calculator

    Turn any value into a z-score and find its percentile and probability on the bell curve.

    Try Calculator β†’

    ● LIVE

    Normal Distribution Calculator

    Find probabilities, percentiles and critical values for any normal distribution.

    Try Calculator β†’

    ● LIVE

    Standard Deviation Calculator

    Calculate the mean, variance and standard deviation of any data set.

    Try Calculator β†’

    ● LIVE

    Empirical Rule Probability Finder

    Find the 68-95-99.7 probabilities for any mean and standard deviation.

    Try Calculator β†’

    ● ALL TOOLS

    Explore All Statistical Tools

    Browse all 58 free statistics tools in one place β€” no sign-up, instant results.

    Browse All Tools β†’