Mean Absolute Deviation Calculator

Mean Absolute Deviation Calculator: Find MAD With Steps, Formula and Examples

A mean absolute deviation calculator shows how far your numbers typically sit from their average. Paste a list of values and it returns the MAD, every absolute deviation, and the full working.

Students, teachers, analysts, planners and quality engineers all use it. It answers one plain question: how spread out is my data?

That matters because an average alone can mislead. Two classes can both average 75 on a test while one class scores almost the same and the other swings wildly. MAD exposes that difference in one number, in the same units as your data.

This guide covers how to use the calculator, how to find mean absolute deviation by hand, and how MAD compares with standard deviation. It also covers grouped data, forecasting, Excel, limits and mistakes to avoid.

What Is MAD and How Do You Find It?

Mean absolute deviation (MAD) is the average distance between each value in a data set and the data set’s center, usually the mean.

MAD = Σ|xᵢ − x̄| ÷ n

To find it:

1. Add all values and divide by the count to get the mean.
2. Subtract the mean from each value.
3. Drop any minus signs (take the absolute value).
4. Average those distances.

For the data set 12, 15, 18, 20, 25, the mean is 18 and the MAD is 3.6.

What Is Mean Absolute Deviation?

Mean absolute deviation measures variability. It tells you how tightly a data set clusters around its center.

A small MAD means the values sit close together. A large MAD means they spread far apart. Because MAD uses your original units, it is easy to explain. A MAD of 3.6 marks on a test means scores are typically 3.6 marks from the average.

The Mean Absolute Deviation Formula

Many people search for the “mean average deviation formula.” It is the same idea as MAD, and it looks like this:

MAD = (1 ÷ n) × Σ |xᵢ − m(X)|

• n is the number of values.
• xᵢ is each individual value.
• m(X) is the center. In the standard school definition, it is the mean (x̄).
• |…| means “absolute value,” which removes the sign.

Why Use Absolute Values?

Deviations from the mean always cancel out. Add them up and you get zero every time. For 12, 15, 18, 20, 25, the signed deviations are −6, −3, 0, +2 and +7, and they sum to 0.

Taking the absolute value turns each distance into a positive number. Now the distances add up instead of canceling.

Key Terms You Need to Know

• Mean: the sum of values divided by the count. See our mean calculator.
• Median: the middle value after sorting.
• Mode: the most frequent value.
• Deviation: the signed difference between a value and the center.
• Absolute deviation: that difference with the sign removed.
• Variability or dispersion: how spread out the data is.
• Outlier: a value far from the rest of the data.

Need the center first? Our mean, median and mode calculator and average calculator can help.

How to Use the Mean Absolute Deviation Calculator

The tool above is built from 12 connected calculators. Card 1 is the main MAD calculator. Its results feed most of the other cards, so you enter your data once.

Card 1 Inputs, Explained

Pro tip: Use value:frequency mode for repeated values. Typing 12:2, 15:3 means 12 appears twice and 15 appears three times.

Card 1 Results, Explained

The calculator also labels spread using relative MAD. Under 10% is low variability, 10% to 25% is moderate, and above 25% is high.

The Other 11 Cards

How to Find Mean Absolute Deviation by Hand

Anyone asking “how do you find mean absolute deviation” can follow four steps. The same steps answer “how to solve mean absolute deviation” and “how to get absolute deviation.”

1. Find the mean. Add the values and divide by n.
2. Find each deviation. Subtract the mean from every value.
3. Take absolute values. Remove all negative signs.
4. Average the absolute deviations. Add them and divide by n.

If you only need the distances, stop at step 3. Those are your absolute deviations.

Worked Examples With Step-by-Step Solutions

Every number below uses the same formulas as the calculator.

Example 1: A Simple Data Set

Calculate the mean absolute deviation of the following numbers: 12, 15, 18, 20, 25.

• Mean: (12 + 15 + 18 + 20 + 25) ÷ 5 = 90 ÷ 5 = 18
• Absolute deviations: 6, 3, 0, 2, 7
• Sum: 18
• MAD: 18 ÷ 5 = 3.6

Relative MAD is 3.6 ÷ 18 = 20%, which the calculator labels moderate variability.

Example 2: Test Scores

A teacher records eight scores: 72, 85, 90, 68, 77, 94, 81, 88.

The mean is 655 ÷ 8 = 81.875. The MAD is 59 ÷ 8 = 7.375 marks.

Relative MAD is about 9%, so this class has low variability. Most students land within roughly 7 marks of the average.

Example 3: The Outlier Experiment

We tested how one extreme value changes MAD. We took the data from Example 1 and added a single new value of 90.

One outlier raised the mean-based MAD from 3.6 to 20.0. That is an increase of about 456%. The median absolute deviation moved only from 3 to 5.

The lesson is that mean-based MAD is sensitive to outliers. If your data might include extreme values, check the median-based version in Card 7.

Example 4: Grouped Data

The “mean absolute deviation calculator for grouped data” (Card 8) works from a frequency table. Here is a table of class intervals with midpoints:

• Mean: 730 ÷ 30 = 24.33
• Grouped MAD formula: Σ f|x − x̄| ÷ N
• MAD: 266.67 ÷ 30 = 8.89

Limitation: grouped MAD is an estimate. It assumes every value sits at its class midpoint. Narrow classes give better estimates.

Example 5: Case Study Comparing Two Groups

Two teams each average 75 on a skills test.

• Team A: 70, 72, 75, 78, 80 → MAD = 3.2
• Team B: 60, 70, 75, 80, 90 → MAD = 8.0

The averages match, but Team B is 2.5 times more spread out. A manager who only reads the average would treat the teams as equal. MAD shows Team B needs more consistent training.

Card 12 automates this comparison and shows how much the two groups overlap.

Sample vs. Population: Which Divisor Should You Use?

The textbook definition divides by n. The calculator uses it by default.

You can switch to n−1 in Card 1. On the data from Example 1, that gives 18 ÷ 4 = 4.5 instead of 3.6.

Treat the n−1 version as a non-standard variant. It is not what most courses teach for MAD. Use it only if your instructor or organization requires it.

MAD in Forecasting

In forecasting, MAD measures forecast accuracy. It is the average size of your forecast errors, ignoring direction.

The MAD Forecasting Formula

MAD = Σ|Actual − Forecast| ÷ n

People also search for “how to calculate MAD forecasting” and “mean absolute deviation forecasting calculator.” Card 9 handles both.

Forecasting Example

• MAD = 19 ÷ 5 = 3.8 units
• Running sum of forecast errors (RSFE) = +5
• Tracking signal = RSFE ÷ MAD = 5 ÷ 3.8 = 1.32
• MAPE ≈ 3.24%
• RMSE ≈ 4.07

A tracking signal of 1.32 is well inside the common ±4 limit. The forecast shows no strong bias.

Safety Stock From MAD

The calculator estimates error spread as 1.2533 × MAD. That factor is √(π/2) and holds when errors are roughly normal.

Safety stock = z × (1.2533 × MAD) × √lead time

At a 95% service level (z = 1.6449) and a one-period lead time, safety stock is about 7.83 units. Planners use this to cover forecast error between orders. See our inventory turnover calculator for related stock planning.

Forecast Methods Card 9 Compares

• Naïve (last period)
• Simple moving average
• Weighted moving average
• Exponential smoothing

The card ranks them by MAD, RMSE and MAPE on the same periods.

Limitation: MAD is in your data’s units, so you cannot compare it across items of different scale. Use MAPE for that, but note MAPE is undefined when any actual value is 0. Forecasts and safety stock are decision aids, not guarantees.

How to Calculate MAD in Excel and Google Sheets

Searchers often ask “how to calculate MAD forecasting in Excel.” Card 10 generates the exact formula for your ranges. Here are the core versions, assuming data in A2:A6.

Test it: for 12, 15, 18, 20, 25, AVEDEV should return 3.6. For the forecast example, the forecast MAD should return 3.8.

Older Excel versions need Ctrl+Shift+Enter for the median-based array formula. Google Sheets needs ARRAYFORMULA() around array math. Use a semicolon instead of a comma if your region uses commas as decimals.

Common errors:

• #VALUE! means a cell contains text or the ranges differ in length.
• #DIV/0! means the range is empty.
• A result of 0 often means the formula was not array-entered in an older version.

MAD vs. Standard Deviation and Other Spread Measures

Standard deviation (SD) squares the deviations before averaging them. That makes large deviations count far more. MAD treats every distance equally.

For normal data, MAD ≈ 0.7979 × σ, and σ ≈ 1.2533 × MAD. The value 0.7979 is √(2/π).

For Example 1, SD is 4.427 and MAD is 3.6. Their ratio is 0.813, close to the normal benchmark. Card 11 shows this ratio for your data. A ratio well below 0.80 hints at heavy tails or outliers.

Expert Perspective

The debate is old. Eddington (1914) argued for the mean absolute deviation. Fisher (1920) showed that SD has lower sampling variability for normal data. Both are correct for different jobs.

The NIST/SEMATECH e-Handbook of Statistical Methods lists these among its standard measures of scale. A practical rule:

• Use SD for inference, such as confidence intervals and tests, on roughly normal data.
• Use MAD to explain spread to a general audience.
• Use median-based measures when outliers are possible.

Need a p-value after measuring spread? Try our p-value calculator. More tools are in our statistics calculators hub.

Mean vs. Median: The Different Meanings of “MAD”

“MAD” is not one fixed term. Courses and industries use it differently. Card 7 shows them side by side.

The constant 1.4826 equals 1 ÷ 0.6745. It makes the scaled value comparable to standard deviation for normal data.

Card 7 also flags outliers with the modified z-score: 0.6745 × (x − median) ÷ MAD. Values above 3.5 are flagged, following Iglewicz and Hoaglin. In the outlier experiment, 90 scores about 9.6, so it is clearly flagged.

Always check which definition your instructor or organization requires.

Assumptions, Limitations and Edge Cases

MAD is simple, but it has limits. Know them before you rely on the result.

Outliers. One extreme value can swing mean-based MAD sharply, as Example 3 shows. Use Card 7 or the outlier handling option in Card 1.

Small samples. With only a handful of values, MAD is noisy. Treat the MAD-to-SD ratio as a rough hint, not proof of a distribution.

Not a test. MAD describes spread. It does not test hypotheses. Use a method like a chi-square calculator or p-value for that.

Different centers give different answers. Deviation about the mean, median and mode produce different values. Always state the center you used.

No mode. If no value repeats, there is no unique mode. The calculator will tell you to choose the mean or median.

Zero actuals. MAPE is undefined when any actual value is 0. MAD is not affected.

Scale dependence. MAD uses your data’s units. A MAD of 5 means very different things for heights in cm and for salaries. Use relative MAD or the coefficient of variation to compare across scales.

Calculator limits. The tool accepts up to 5,000 values per data set. It is for information and learning, not professional advice. Check high-stakes decisions, such as inventory or quality limits, with a qualified professional.

Common Mistakes and Pro Tips

Common Mistakes

• Forgetting the absolute value. Signed deviations always sum to zero.
• Dividing by the wrong number. Standard MAD divides by n.
• Mixing up MAD definitions. Mean absolute deviation and median absolute deviation are different.
• Using the median as the center without saying so. State your center.
• Rounding too early. Keep extra decimals until the final step.
• Comparing MAD across different units. Use relative MAD instead.
• Treating grouped MAD as exact. It is an estimate based on midpoints.

Pro Tips

• Check your hand calculation with Card 3, the practice problem checker.
• Pair MAD with the tracking signal in forecasting. MAD gives the size of misses, and the signal shows whether they lean one way.
• Report MAD next to SD when your audience is mixed.
• Sort by “Largest Deviation First” to spot problem values fast.
• Test every spreadsheet formula on the five-row sample before using real data.

Where Is MAD Used in the Real World?

• Education: Common Core 6.SP.B.5c names mean absolute deviation as a measure of variability for grade 6 students. Teachers also use it to read exam results. Our grade calculator and test grade calculator pair well with it.
• Operations and supply chain: MAD measures forecast error and sets safety stock.
• Quality control: The scaled median absolute deviation monitors a process without being fooled by one bad reading.
• Finance: Median absolute deviation suits heavy-tailed returns. Simple dispersion reports often use the mean version.
• Data science: Robust scaling and outlier detection rely on median-based MAD.
• Health and fitness: MAD summarizes day-to-day swings. Seven daily step counts averaging 7,829 have a MAD of about 1,347, or 17.2% relative MAD (moderate).

Related Calculators

• Mean, median and mode calculator
• Mean calculator
• Average calculator
• Geometric mean calculator
• Percent error calculator
• Average percentage calculator
• Box and plot generator
• Square root calculator
• Statistics calculators hub

You can also read how we review our content on our editorial policies page.

Frequently Asked Questions

What is mean absolute deviation?

Mean absolute deviation is the average distance between each value in a data set and the data set’s center, usually the mean. It shows how spread out the data is, in the same units as the data.

How do you find mean absolute deviation?

Find the mean, subtract it from each value, take the absolute value of each difference, and average those distances. For 12, 15, 18, 20 and 25, the mean is 18 and the MAD is 3.6.

How do I find the mean absolute deviation using a calculator?

Paste your numbers into Card 1 and leave “Deviation Taken About” on Mean. The calculator returns the MAD, the full deviation table and the working. Use Card 2 if you want every step written out.

What is the mean absolute deviation formula?

The formula is MAD = Σ|xᵢ − x̄| ÷ n. Here xᵢ is each value, x̄ is the mean and n is the count. People also call it the mean average deviation formula.

How do you calculate MAD in math class?

Follow the same four steps: mean, deviations, absolute values, then the average of those. This is the version taught in grade 6 under Common Core 6.SP.B.5c.

What is the difference between MAD and standard deviation?

MAD averages the absolute distances, so all distances count equally. Standard deviation squares them first, so large distances count more. For normal data, MAD is about 0.7979 times the standard deviation.

Is MAD the same as average deviation?

Usually yes. “Average deviation” and “mean deviation” often mean the mean absolute deviation about the mean. Check your course, because some texts take the deviation about the median or mode.

What is the difference between mean absolute deviation and median absolute deviation?

Mean absolute deviation averages the distances from the mean. Median absolute deviation takes the median of the distances from the median. The median version is far more resistant to outliers.

How do you calculate MAD for forecasting?

Subtract each forecast from its actual value, take the absolute values, and average them. The forecast MAD formula is Σ|Actual − Forecast| ÷ n. A lower MAD means a more accurate forecast.

How do you calculate MAD in Excel?

Use =AVEDEV(range) for mean absolute deviation about the mean. For forecast MAD, use =SUMPRODUCT(ABS(actuals-forecasts))/COUNT(actuals). Card 10 builds the formula for your cell ranges.

Can mean absolute deviation be negative?

No. Every absolute deviation is zero or positive, so MAD is always zero or greater. A MAD of 0 means every value equals the center.

Does MAD divide by n or n−1?

The standard definition divides by n. Dividing by n−1 is a non-standard variant. The calculator lets you choose, but n is the default.

What is a good MAD value?

It depends on your units and goal. Relative MAD helps: under 10% is low variability, 10% to 25% is moderate, and above 25% is high. In forecasting, lower is better when compared on the same data.

Can I use MAD for grouped data?

Yes. Use class midpoints and frequencies, then compute Σ f|x − x̄| ÷ N. Card 8 does this and also draws a histogram. The result is an estimate because it assumes values sit at midpoints.

What does a high MAD mean?

A high MAD means values sit far from the center on average. Check for outliers, different subgroups or measurement errors before drawing conclusions.

Conclusion

Mean absolute deviation gives you a clear, honest view of spread. It is easy to compute, easy to explain, and expressed in your own units.

Key takeaways:

• MAD = Σ|x − mean| ÷ n. Find the mean, take absolute distances, then average them.
• The data set 12, 15, 18, 20, 25 has a MAD of 3.6.
• One outlier can inflate mean-based MAD sharply. Median-based versions resist it.
• For normal data, MAD ≈ 0.7979σ.
• In forecasting, MAD measures error size, and the tracking signal measures bias.
• Always state your center and divisor, because “MAD” has more than one definition.

Use our mean absolute deviation calculator to get the answer, the steps and the formula in seconds. Then explore the grouped data, forecasting, Excel and comparison cards to go deeper.