Bar height is the frequency. The dark line is the mean and the shaded band covers mean ± MAD.
below meanabove mean± MAD
Where the total deviation comes from
Each segment is one row's f × |x − x̄|. The blue half comes from values below the mean, the orange half from values above. The two halves always match.
Step-by-step working
The full frequency table with every column filled in.
How mean absolute deviation table calculator works
The mean absolute deviation table calculator works in 5 steps: it adds the frequencies to get N, finds the weighted mean Σf·x ÷ N, measures each row's distance |x − x̄|, multiplies each distance by its frequency, and divides the total by N. For class intervals, the calculator uses each class midpoint as x.
RowsEnter value and frequency, or lower limit, upper limit and frequency.
NTotal frequency N = Σf.
Meanx̄ = Σf·x ÷ N.
Weightf · |x − x̄| for every row.
DivideMAD = Σf·|x − x̄| ÷ N.
The calculator reports 7 results: MAD, total frequency N, mean, rows used, Σf·x, Σf·|x − x̄| and MAD as a percentage of the mean. The example table below has 4 rows, N = 20 and MAD = 1.64.
Below the mean (5.3)Above the mean
Bar height is the frequency f. The 20 values balance at the mean, 5.3.
Mean absolute deviation table calculator formula
The mean absolute deviation table calculator formula is MAD = Σ fᵢ|xᵢ − x̄| ÷ Σ fᵢ: the frequency-weighted sum of absolute deviations from the mean, divided by the total frequency.
MAD=Σfᵢ|xᵢ − x̄|Σfᵢ
MAD: mean absolute deviation of the frequency table.
Σ (sigma): add over every row of the table.
fᵢ: the frequency of row i. Σfᵢ = N, the total count.
| |: absolute value, the distance without a sign.
xᵢ: the value of row i, or the class midpoint for grouped data.
The frequency formula and the ungrouped formula give identical results for the same data. The grouped formula gives an estimate, since each class stands in for its midpoint.
Mean absolute deviation table calculator technique
The standard mean absolute deviation table calculator technique is the 5-column frequency table: x, f, f·x, |x − x̄| and f·|x − x̄|, with the f, f·x and f·|x − x̄| columns totalled. The f·x total gives the mean, and the f·|x − x̄| total gives the MAD.
x
f
f·x
|x − x̄|
f·|x − x̄|
2
3
6
3.3
9.9
4
5
20
1.3
6.5
6
8
48
0.7
5.6
8
4
32
2.7
10.8
Σ
20
106 → x̄ 5.3
32.8 → MAD 1.64
1 check that catches mistakes
The weighted distances below the mean equal the weighted distances above the mean. The example gives 16.4 below and 16.4 above. Unequal totals mean the mean is wrong.
Blue rows (below 5.3) add to 16.4. Orange rows (above 5.3) add to 16.4. The 2 halves always match.
Grouped data: class midpoints
Class
m
f
f·m
|m − x̄|
f·|m − x̄|
0–10
5
4
20
14
56
10–20
15
7
105
4
28
20–30
25
6
150
6
36
30–40
35
3
105
16
48
Σ
20
380 → x̄ 19
168 → MAD 8.4
The grouped MAD of 8.4 is an estimate. Excel and Google Sheets compute the frequency MAD with =SUMPRODUCT(B2:B5,ABS(A2:A5-SUMPRODUCT(A2:A5,B2:B5)/SUM(B2:B5)))/SUM(B2:B5), where column A holds the values and column B the frequencies.
Frequently asked questions
What is the mean absolute deviation of 2, 4, 6, 8 with frequencies 3, 5, 8, 4?
The mean absolute deviation is 1.64. The total frequency is N = 20, Σf·x = 106 and the mean is 106 ÷ 20 = 5.3. The weighted distances f·|x − 5.3| are 9.9, 6.5, 5.6 and 10.8, they add to 32.8, and 32.8 ÷ 20 = 1.64.
What is the formula for mean absolute deviation of grouped data?
The formula for mean absolute deviation of grouped data is MAD = Σ fᵢ|mᵢ − x̄| ÷ Σ fᵢ, where mᵢ is the midpoint of class i. Classes 0–10, 10–20, 20–30 and 30–40 with frequencies 4, 7, 6 and 3 have midpoints 5, 15, 25 and 35, a mean of 19 and a MAD of 168 ÷ 20 = 8.4.
Is the mean absolute deviation of grouped data exact?
No, the mean absolute deviation of grouped data is an estimate. The calculation places every value at its class midpoint, so the true values inside each class are unknown. Narrower classes, such as 5 units instead of 10, bring the estimate closer to the exact MAD.
Does a frequency table give the same MAD as the full list of values?
Yes, a frequency table gives the same mean absolute deviation as the full list of values. Values 2, 4, 6, 8 with frequencies 3, 5, 8, 4 and the 20-value list 2, 2, 2, 4, 4, 4, 4, 4, 6 (8 times), 8 (4 times) both give a MAD of 1.64.
How do you find the mean absolute deviation of a frequency table?
Find the mean absolute deviation of a frequency table by dividing Σf·|x − x̄| by the total frequency Σf. Multiply each value by its frequency, add the products and divide by the total frequency to get the mean. Multiply each distance |x − x̄| by its frequency, add the results and divide by the total frequency again.
What do I do with class intervals like 10–20?
Use the midpoint of each class as its value: the class 10–20 has the midpoint (10 + 20) ÷ 2 = 15. Switch the calculator to Class intervals and the calculator uses the midpoints for you. The result is an estimate, since the real values inside each class are unknown.
Can I paste a table from Excel or Google Sheets?
Yes, you can paste a table from Excel or Google Sheets. Open Paste rows, paste 2 columns (value and frequency) or 3 columns (lower limit, upper limit and frequency), and press Use these rows.
Why are the blue and orange halves of the deviation bar always equal?
The blue and orange halves are always equal: the mean is the balance point of the data, so the total distance below the mean equals the total distance above the mean. The stacked bar shows the 2 totals side by side.