On this concept: Formula

Not sure which lesson to do first? Find your level and get a plan →

Statistics · Grade 6-8

Mean Absolute Deviation (MAD)

Use Mean Absolute Deviation when the question asks how far the data values typically fall from the mean — a measure of spread, not center.

MAD=∑∣x−xˉ∣n

Quick Answer

Use Mean Absolute Deviation when the question asks how far the data values typically fall from the mean — a measure of spread, not center. Find the mean, measure how far each value is from it (ignoring whether it is above or below), and average those distances. For the data 3, 5, 9, 11 the mean is 7, the distances are 4, 2, 2, 4, and the MAD is 3. A small MAD means the values stay close to the mean; a large one means they wander far from it.

Why This Matters

The MAD turns "how consistent?" into a number in the same units as the data. If two pizza shops both deliver in 30 minutes on average, but one has a MAD of 2 minutes and the other a MAD of 12, you know whose promise to trust. It is also the stepping stone to standard deviation, which is built on the same idea of distance from the mean.

Intuitive Explanation

Five students score 12, 14, 15, 16 and 18 on a 20-point quiz. The mean is 75÷5=15. Now ask how far each student is from 15: 3, 1, 0, 1 and 3 points. Some are above and some below, but a distance is a distance. The average distance is (3+1+0+1+3)÷5=8÷5=1.6 points. That is the MAD: a typical score is about 1.6 points from the mean.

Why ignore the signs? If you kept them, the deviations would be −3,−1,0,+1,+3, which add to zero. They always add to zero, for any data set, because the mean is exactly the balance point. Taking absolute values stops the distances above and below the mean from cancelling. A second group scoring 5, 10, 15, 20 and 25 has the same mean, 15, but distances of 10, 5, 0, 5 and 10, so its MAD is 6 points, showing at once that it is far less consistent.

The MAD is not a center, even though it uses the mean. A class with a MAD of 2 points could have a mean of 50 or of 90; the MAD only says how tightly the scores cluster around whatever the mean is. It is also not the standard deviation, which squares the distances before averaging and so gives extra weight to far-away values. For 12, 14, 15, 16, 18 the standard deviation is 2.0 points, a little more than the MAD of 1.6.

Think of it like this. The MAD is like asking everyone in a class how far they live from school, ignoring direction, and averaging the answers: a 2 km trip from the north counts the same as a 2 km trip from the south. The analogy stops at the school itself, because for the MAD the center is the mean of the data, not a fixed building.

Core idea. Mean Absolute Deviation (MAD) asks how far, on average, the data values sit from the center - it measures spread, not where the center is.


When to Use

Use the MAD when a question asks how consistent a set of values is, how far values typically sit from the mean, or which of two groups with similar means varies more. Find the mean, the absolute distance of each value from it, and the average of those distances. If the question names standard deviation or squares the deviations, use standard deviation; if the data have large outliers and you are reporting the median, the IQR is the better spread.

Tip. Ask: Is the spread found by averaging each value's plain distance from the mean, with no squaring?

How to Recognize It

Before you start, check that the question wants a typical distance from the mean, and how those distances are meant to be combined.

  1. Does the question want how far the values typically sit from the average, not the average itself?

    Yes is a spread around the mean, such as the MAD. If it wants the average value itself, use the Mean as Fair Share.

  2. Are the distances used as they are, with no squaring and no square root?

    Yes is the MAD. If the distances are squared, averaged and then square-rooted, it is the Standard Deviation.

  3. Are the distances measured from the mean, rather than taken between the quartiles?

    Yes fits the MAD. If the spread is the gap from Q1 to Q3, it is the Interquartile Range (IQR).

  4. Do you need an actual number for the spread, not just a statement that one group is more scattered?

    Yes means calculate the MAD. A comparison only in words, such as more or less scattered, is Data Variability.

Mean Absolute Deviation (MAD) vs Standard Deviation vs Interquartile Range (IQR) vs Mean as Fair Share

All four start from the same data, and the MAD and the standard deviation both measure from the mean, so they look like twins. Ask how the distances are combined: plain, squared, between quartiles, or not at all.

Mean Absolute Deviation (MAD)

Meaning
The MAD is the average of the plain distances between each value and the mean.
Key test
Distances kept at their true size, signs dropped, then averaged.
Formula
MAD=1n∑i=1n∣xi−xˉ∣
Example
Heights 140, 145, 150, 165 cm: mean 150, distances 10, 5, 0, 15, so the MAD is 7.5 cm.

Standard Deviation

Meaning
The standard deviation squares each distance from the mean, averages the squares, then takes the square root.
Key test
Squares first, so far values count extra; it is never smaller than the MAD.
Formula
∑(x−μ)2n
Example
For the same four heights the standard deviation is about 9.4 cm, above the MAD.

Interquartile Range (IQR)

Meaning
The IQR is the width of the middle half of the sorted data, from Q1 to Q3.
Key test
Built from positions in sorted data, not from distances to the mean.
Formula
IQR=Q3−Q1
Example
Heights 140, 145, 150, 165 cm have Q1=142.5 and Q3=157.5, so the IQR is 15 cm.

Mean as Fair Share

Meaning
The mean is the center that the MAD is measured from, not a spread itself.
Key test
One location value; it is the first step of the MAD, not the answer.
Formula
xˉ=∑xn
Example
Heights 140, 145, 150, 165 cm have a mean of 150 cm, which says nothing about spread.

Formula & Notation

MAD=∑∣x−xˉ∣n
For a dataset {x1,x2,…,xn} with mean xˉ, the Mean Absolute Deviation is MAD=1n∑i=1n∣xi−xˉ∣.

How to read it: MAD stands for Mean Absolute Deviation. ∣xi−xˉ∣ is the absolute deviation of the ith data point from the mean xˉ.

Worked Examples

Example 1 — MAD of a week's sleep

Easy

Jo slept 7, 9, 8, 6, 10 and 8 hours on six nights. Find the mean absolute deviation.
Show the steps and the answer
  1. Find the mean: 7+9+8+6+10+8=48, and 48÷6=8 hours.

    Every distance is measured from the mean, so it comes first.

  2. Find each distance from 8: 1, 1, 0, 2, 2, 0.

    Absolute value turns a night 2 hours short and a night 2 hours long into the same distance, 2.

  3. Add the distances: 1+1+0+2+2+0=6.

    This is the total distance of all six nights from the mean.

  4. Divide by the number of values: 6÷6=1 hour.

    Averaging the distances gives the typical distance from the mean.

Answer: MAD=1 hour.

Takeaway: Mean first, then distances without signs, then average the distances.

Example 2 — When one value dominates the MAD

Standard

Six friends get weekly allowances of $5, $6, $6, $7, $8 and $40. Find the MAD. Then find the median and IQR, and say which pair describes the group better.
Show the steps and the answer
  1. Mean: 5+6+6+7+8+40=72, and 72÷6=12 dollars.

    The $40 lifts the mean above every other allowance.

  2. Distances from 12: 7, 6, 6, 5, 4, 28; total 56, so MAD=56÷6≈9.33 dollars.

    The $40 allowance alone contributes 28 of the 56.

  3. Median: (6+7)÷2=6.5. Quartiles: lower half 5, 6, 6 gives Q1=6; upper half 7, 8, 40 gives Q3=8, so IQR=2.

    The median and quartiles depend on positions, so the $40 cannot stretch them.

  4. Compare: five of the six friends get between $5 and $8, which the median of $6.50 and IQR of $2 describe well.

    A mean of $12 with a MAD of about $9 suggests a spread-out group, which is not true of five of the six friends.

Answer: The MAD is about $9.33 around a mean of $12; the median of $6.50 with an IQR of $2 describes the group better.

Takeaway: The MAD is built on the mean, so a large outlier inflates both; with outliers, describe the data with the median and IQR.

Example 3 — Find the mistake: forgetting the absolute value

Application

For the data 3, 5, 6, 10, Dev writes: "The mean is 6, the deviations are −3,−1,0,4, and their average is 0÷4=0, so the data have no spread." Find the mistake and correct it.
Show the steps and the answer
  1. Check the mean: 3+5+6+10=24, and 24÷4=6, so Dev's mean is right.

    The error comes after the mean.

  2. Explain the zero: signed deviations from the mean always add to zero, so their average says nothing about spread.

    The values below the mean exactly balance the values above it.

  3. Use distances instead: 3, 1, 0, 4, which add to 8.

    The absolute value keeps every distance positive, so nothing cancels.

  4. Divide: MAD=8÷4=2.

    A typical value is 2 away from the mean, which matches data that run from 3 to 10.

Answer: MAD=2.

Takeaway: Take absolute values before averaging; signed deviations from the mean always cancel to zero.

Where You'll See It

  • How off a weather forecast usually is

    Take each day's forecast error (ignore whether it was too high or too low), then average those distances; that MAD tells you how many degrees the forecast is typically off.

  • How much your game scores wander from your average

    If your average score is 80, find how far each game sat from 80, drop the plus/minus signs, and average them; a small MAD means you play consistently.

  • Comparing two buses' on-time records

    Both buses average a 10-minute wait, but computing the MAD of each shows which bus's waits typically stray further from 10 minutes, so you know which is more reliable.

Common Mistakes

  1. The slip: You can skip the absolute value and just average the signed differences from the mean.

    Why it's tempting: Averaging deviations seems natural, but the positive and negative deviations from the mean always cancel to zero, giving no information.

    Instead: Take the absolute value of each deviation first; the signed deviations always sum to zero, so MAD requires the distances.

  2. The slip: MAD is just another name for the standard deviation.

    Why it's tempting: Both measure average distance from the mean and look similar, so they blur together.

    Instead: MAD averages the absolute deviations; standard deviation squares the deviations, averages, then square-roots - so the two give different numbers.

  3. The slip: MAD is a measure of the center of the data.

    Why it's tempting: MAD uses the mean in its formula, so it feels tied to the center, but it reports distance from the center, not the center itself.

    Instead: MAD is a spread (variability) measure: the average distance of values from the mean.


Mini Practice

Try these on your own. Tap Reveal when you want to check.

  1. Find the MAD of 3, 5, 5, 7, 10.

    Hint: Find the mean, then the distance of each value from it.

  2. A week's daily high temperatures are 66, 68, 70, 70, 72, 72 and 72 °F, with mean 70 °F. On average, how far is a day's high from 70 °F?

    Hint: Find each distance from 70, ignoring signs, then average them.

  3. Class A's scores are 78, 79, 80, 81, 82 and Class B's are 60, 70, 80, 90, 100. Both means are 80. Find each class's MAD. Which class is more consistent?

    Hint: List each score's distance from 80.

  4. Two dart players have MADs of 0.8 points and 3.5 points around the same mean. What does each MAD tell you, and which player is steadier?

    Hint: MAD is the typical distance of a throw's score from the mean.

  5. Five quiz scores are 12, 15, 18, 20, 20. The question asks for the average score. Do you need the MAD or the mean?

    Hint: Is the question about the center or about distances from it?

  6. For 4, 6, 8, 10, 12 (mean 8), a student adds the deviations −4, −2, 0, 2, 4, gets 0, and writes "MAD = 0". What is the MAD?

    Hint: Distances are never negative.

Want the full set?

50 practice questions for this concept — free to try, every one with a complete worked solution showing the why, not just the answer.

Frequently Asked Questions

If the MAD is easier, why do statisticians usually use the standard deviation?

Mostly because the standard deviation fits the rest of statistics better. It is the spread built into the bell curve, and squared spreads have tidy rules, such as adding up neatly when two independent amounts are combined. The MAD is easier to work out and to explain, and it is less upset by one far-off value. For describing a single data set, both are honest choices, as long as you say which one you used.

How can the MAD tell me whether two groups are really different?

Compare the gap between the two means with the MAD. Suppose Class A averages 82 points and Class B averages 70, and both classes have a MAD of about 4 points. The gap of 12 points is 3 MADs, so the groups clearly differ. If the MADs were 15 points instead, a 12-point gap would be smaller than the usual wobble inside each class, and the groups would overlap a lot.

Can the MAD ever be bigger than the range?

No, and in fact it is never more than half the range. Every value lies between the smallest and the largest, and so does the mean, so no distance from the mean can be longer than the range. The biggest possible MAD comes when half the values sit at each end. For 0, 0, 10, 10 the mean is 5, every distance is 5, and the MAD is 5, exactly half the range of 10.

Learning Path

← Before

Mean as Fair Share
Mean Absolute Deviation (MAD)

You are here

You need Mean as Fair Share first, because every distance is measured from the mean. The MAD adds one number for spread: the average distance of the values from that mean. Next, Standard Deviation measures spread in a similar way but squares the distances, and Data Variability uses measures like the MAD to compare how consistent groups are.

See Also