Team
statistics

Mean Absolute Deviation Calculator

Mean absolute deviation quantifies how far individual data points typically stray from a chosen central value. Unlike standard deviation, MAD uses linear distances rather than squared differences, making it more intuitive and robust to outliers. Data analysts, statisticians, and quality control engineers rely on MAD to assess variability in distributions without the computational complexity of variance-based measures.

Last updated:

Creators Wojciech Sas, PhD
5,604 people find this calculator helpful

Understanding Mean Absolute Deviation

Mean absolute deviation measures the average distance between each observation and a reference point—usually the mean, median, or mode. The term breaks down naturally: deviation refers to how far a value strays from normal; absolute means we take the positive distance regardless of direction; and mean is simply the average of those distances.

This metric reveals data dispersion in the original units, rather than squared units like variance does. For example, if your dataset measures test scores out of 100, the MAD result is also in points, making it immediately interpretable. MAD is particularly valuable when outliers could skew analysis, since it doesn't amplify extreme values the way standard deviation does.

You can anchor your calculation to any reference point: the arithmetic mean for typical central tendency, the median for robustness, or a theoretical target value for quality assessment.

Mean Absolute Deviation Formula

The mathematical definition sums the absolute differences between each data point and the central reference, then divides by the count of observations:

MAD = (1/n) × Σ|xᵢ − m|

where:

n = number of observations

xᵢ = the i-th data point

m = central reference point (mean, median, or custom value)

|...| = absolute value (ignore sign)

  • n — The total count of values in your dataset
  • xᵢ — Each individual value in the dataset
  • m — The central point—typically mean, median, or a target reference value
  • Σ — Summation symbol: add all the absolute differences together

Step-by-Step Calculation Example

Consider daily rainfall measurements (in mm) over a week: 12, 8, 15, 9, 14, 11, 10.

First, calculate the mean: (12 + 8 + 15 + 9 + 14 + 11 + 10) ÷ 7 = 11.29 mm.

Next, find each deviation from 11.29:

  • 12 − 11.29 = 0.71
  • 8 − 11.29 = −3.29
  • 15 − 11.29 = 3.71
  • 9 − 11.29 = −2.29
  • 14 − 11.29 = 2.71
  • 11 − 11.29 = −0.29
  • 10 − 11.29 = −1.29

Take the absolute value of each (remove negative signs): 0.71, 3.29, 3.71, 2.29, 2.71, 0.29, 1.29.

Finally, average these absolute deviations: (0.71 + 3.29 + 3.71 + 2.29 + 2.71 + 0.29 + 1.29) ÷ 7 = 1.90 mm. This means daily rainfall typically varies by about 1.9 mm from the weekly average.

Common Pitfalls and Considerations

When computing MAD, watch for these frequent mistakes and interpretation challenges:

  1. Choosing the wrong central point — Your MAD result depends entirely on which reference value you select. Using the mean works well for normally distributed data, but the median is more reliable when outliers are present. Always verify your choice aligns with your analysis goal.
  2. Confusing MAD with median absolute deviation — These are distinct measures. Mean absolute deviation uses the arithmetic mean of distances; median absolute deviation uses the median of those distances. Some fields refer to the latter as 'MAD' too, causing confusion. Check context carefully.
  3. Forgetting to take absolute values — The most common computational error is omitting the absolute value step. Negative deviations will cancel positive ones, giving zero or misleadingly small results. Every distance must be treated as positive.
  4. Misinterpreting units — MAD is expressed in the same units as your original data. A MAD of 5 cm for length measurements means deviations average 5 cm, not 5 squared centimeters. This directness is MAD's advantage over variance-based metrics.

MAD vs. Standard Deviation

Both metrics quantify spread, but they differ in calculation and sensitivity. Standard deviation squares each deviation before averaging, then takes the square root. This amplifies the impact of outliers and produces results in squared units conceptually (though the root brings it back to original units).

Mean absolute deviation avoids squaring, so extreme values don't dominate the calculation. If your data contains legitimate outliers that shouldn't be weighted more heavily, MAD is often the better choice. For normally distributed data with few extremes, standard deviation's mathematical properties enable further statistical inference like confidence intervals and hypothesis tests.

In practice, MAD appears frequently in quality control, weather analysis, and financial risk assessment, where robustness and interpretability matter more than traditional inference tools.

Frequently Asked Questions

When should I use mean absolute deviation instead of standard deviation?

Choose MAD when your dataset contains outliers you want to treat equally with other points, or when you need a dispersion measure in the original data units for easy interpretation. Standard deviation is better suited for normally distributed data where you plan further statistical modeling, such as calculating confidence intervals or conducting parametric hypothesis tests.

Can I use a value other than the mean or median as my central point?

Yes. You can calculate MAD around any reference value—a theoretical target, industry benchmark, or specific threshold. This flexibility is one of MAD's strengths. For example, in manufacturing, you might measure deviations from a specification rather than from the sample mean, directly assessing whether production meets standards.

What does a low vs. high mean absolute deviation indicate?

A low MAD suggests data points cluster closely around your chosen central point, indicating consistency and low variability. A high MAD means values are scattered widely from the center. The interpretation depends on context: in manufacturing, low MAD is desirable for quality; in exploratory analysis, high MAD reveals which datasets are more volatile.

How does sample size affect the mean absolute deviation?

Sample size doesn't inherently inflate or deflate MAD. The metric is calculated the same way regardless of how many observations you have. However, larger samples generally provide more reliable estimates of true population variation. Very small samples may not represent the full range of your data.

Is mean absolute deviation affected by extreme values?

Less so than standard deviation. Because MAD uses absolute values rather than squared differences, an extreme outlier contributes proportionally to the final result without disproportionate amplification. However, it still does affect the central point itself—if you're using the mean as reference, a strong outlier will shift the mean and thus change your MAD.

What's the relationship between mean absolute deviation and variance?

Both measure spread, but variance squares deviations before averaging. This makes variance harder to interpret (it's in squared units) but mathematically convenient for further analysis. MAD is approximately 0.8 times the standard deviation for normally distributed data. They rank datasets similarly in terms of variability, but MAD is more intuitive for practitioners who need direct interpretation.

statistics
Sum of Squares Calculator
statistics
Midrange Calculator
statistics
Coefficient of Variation Calculator

More statistics calculators (see all)

Sample Size Calculator OR Probability Calculator Central Limit Theorem Calculator Polynomial Regression Calculator SMp(x) Distribution Calculator Relative Risk Calculator Cubic Regression Calculator Decimal Odds Calculator