Mean Absolute Deviation (MAD)
What you'll learn: How to measure spread by calculating the average distance of data points from the mean, giving you a simple, intuitive measure of variability.
The Core Idea
Mean Absolute Deviation (MAD) answers a straightforward question: On average, how far away is each data point from the mean?
Unlike variance (which squares deviations) and standard deviation (which takes the square root of variance), MAD works directly with the actual distances. You calculate it in three steps:
- Find the mean of your dataset
- Calculate the absolute difference between each data point and the mean (absolute means we ignore whether it's above or below—distance is always positive)
- Take the mean of all those absolute differences
Why "Absolute"?
If you averaged raw deviations without taking absolute values, negatives and positives would cancel out, always giving you zero—not useful! The "absolute" part ensures every deviation contributes positively to your measure of spread.
An Analogy
Imagine you're standing at the mean location on a street, and your friends are scattered at different addresses. MAD tells you the average number of blocks each friend is away from you, regardless of direction. Some might be 3 blocks north, others 5 blocks south—MAD averages those distances.
MAD vs Standard Deviation
MAD is more robust to extreme values than standard deviation because it doesn't square deviations (which amplifies outliers). If your dataset has a few wildly unusual points, MAD gives a more stable picture of typical spread.
Key Takeaway: MAD measures variability by averaging the absolute distances from the mean—it's intuitive, resistant to outliers, and complements other spread measures like IQR and standard deviation.