AVEDEV in Calligra Sheets: Mean Absolute Deviation, Examples and Checks

AVEDEV computes the mean absolute deviation from the arithmetic mean of numeric observations. Unlike variance or standard deviation, it takes the absolute distance from the ordinary mean for each data value, then divides their sum by the number of numeric observations (not n-1). KDE officially documents the syntax AVEDEV(value;value;...) and the example =AVEDEV(11.4;17.3;21.3;25.9;40.1) returning 7.84. KDE AVEDEV entry; source implementation. Checked 2026-10-13.

This result is a measure of dispersion, not the data set’s average value. For the values 2, 4, 4, 4, 5, 5, 7, 9 the mean is 5, while AVEDEV is 1.5. The values are independently calculated expectations; no Calligra Sheets application was run.

=AVEDEV(value1; [value2]; ...)
=AVEDEV(B2:B9)

The official help’s AVEDEV(value;value;...) entry describes optional floating-point/range parameters; the 26.08.2 source registers AVEDEV with a minimum one argument and a variable upper count (setParamCount(1,-1)), and its body calls calc->avg(args), calc->arrayWalk(..., awAveDev, ...), and calc->count(args). These are the implementation’s mean, accumulated absolute deviations and count, respectively. Registration; body. The precise handling of errors, booleans and text in an installed application is not established by this calculation; test rather than assume cross-spreadsheet coercion.

For n observations x1 … xn, with arithmetic mean μ = (Σxi)/n, the relevant mathematical formula is:

AVEDEV(x1; ...; xn) = (|x1-μ|+...+|xn-μ|) / n

Zero or missing numeric observations make the mathematical denominator zero. The function wrapper does not show a standalone zero-count error branch; Calligra’s visible error text and how blanks are filtered require native verification, not inference from Excel.

Scope: Source inspection targets Calligra 26.08.2 released by KDE on 2026-10-08. The stable-KF6 handbook retains an old original revision and may not describe every current execution detail. KDE recommends Linux packages or Flatpak; Windows development support remains in progress. Check the installed version, function availability and locale rather than assuming catalogue presence equals runtime proof. KDE release, download page, formula entry; accessed 2026-10-13.

1. In a fresh worksheet, place heading Observation in B1. Enter these eight numeric values in order, one per row: B2=2, B3=4, B4=4, B5=4, B6=5, B7=5, B8=7, B9=9.
2. Set D1 to AVEDEV formula, E1 to Expected value. Enter the formulas from the table into D2:D9; the formulas are literal text here only for guidance, and should be actual formula entries in Calligra.
3. Compare your application’s numeric results with the independent expected values shown. The table includes overlapping ranges to test argument processing; each expected value was calculated separately from the specified B-cell contents.
4. For greater precision, format the result cells to show nine decimal places. The repeating third in D4 can round at different precision settings.

Fully worked D5 calculation for all eight rows: the mean is (2+4+4+4+5+5+7+9)/8 = 5. Each absolute difference from 5 is 3, 1, 1, 1, 0, 0, 2, 4. Their sum is 12, so 12/8 = 1.5. This independently validates the interpretation without claiming any native spreadsheet run.

Vendor handbook’s second test: Paste =AVEDEV(11.4;17.3;21.3;25.9;40.1) in an unused cell. The independently verified arithmetic mean is 23.2, the absolute distances are 11.8, 5.9, 1.9, 2.7, 16.9, and the average of these distances is 7.84. This example and result are also directly documented by KDE.

Build a parallel calculation to confirm the full-range result without relying on AVEDEV’s internals:

1. Put =AVERAGE(B2:B9) in F2; the mathematical result is 5. This is an ordinary KDE-documented AVERAGE call, not a native test.
2. Enter =ABS(B2-$F$2) in C2 and fill through C9. The eight independently expected C values are 3, 1, 1, 1, 0, 0, 2, 4.
3. In F3, enter =AVERAGE(C2:C9). It should yield 1.5. Compare F3 with =AVEDEV(B2:B9) in D5.
4. If there is a mismatch, check the input cells, range boundaries, locale-specific formula syntax and real precision before assuming either function is incorrect.

Difference from STDEV: This computation uses absolute values and divides by n. Variance and standard deviation square the deviations, and the sample STDEV convention uses n−1. Neither is a substitute for AVEDEV if your model specifically calls for mean absolute deviation from the mean. Also note that this is not the median absolute deviation (MAD) from the median, a separate statistic.

Why it matters: In a quality-control or pricing example, averaging absolute deviations provides an interpretable unit-scale measure of how far observations are from their average. For sample price values in Canadian dollars, the AVEDEV output would be in the same currency units, not a percentage or squared dollars. Do not treat it as a forecast, confidence interval, or test of normality.

• Wrong statistic: AVEDEV uses deviations from the arithmetic mean, not from the median. A sample calculation for the values 2,4,4,4,5,5,7,9 uses mean=5 and produces 1.5.
• Single numeric input: =AVEDEV(B2) is mathematically 0, since the only value equals its own mean. Whether the installed program renders 0 for every numeric type should be confirmed natively.
• No numeric inputs / blanks / text-only range: Mean and count are undefined without observations; the source wrapper delegates to avg and count. The exact Calligra UI error and its locale-specific coercions remain untested.
• Text or Boolean mixed with values: The vendor’s AVEDEV description specifies floating-point values; do not apply AVERAGEA rules to AVEDEV. If you include text/Boolean values, verify the installed software’s handling rather than importing assumptions from Excel.
• Accidentally including a header: B1:B9 includes text Observation; B2:B9 is the intended numeric test range. The effect of including an extra text cell depends on observed conversion and should be tested.
• Date/currency formatting: The function computes over underlying numeric values; date serial and formatted values may not represent the real data you intend. Recheck numeric input types.
• Not the sample denominator: Formula uses n, not n−1; reproducing the mean deviations with STDEV-like mathematics gives a different answer.
• Regional separators: The vendor documentation uses semicolons for arguments; values use dot decimals in the English examples. Installed locales may use different separators and number formats. KDE formulas guide.
• Incorrect arity / overloaded cells: The source requires at least one input and permits a list of arguments; native error labels for zero arguments, unexpected array shapes and errors in referenced cells have not been measured.

Suggested native test set: Capture actual outputs/errors for AVEDEV(B2:B9), the KDE five-literal example, a one-number range, a blank-only range, a text-only range, mixed text/Boolean values, negative numbers, duplicated values, a manually produced error cell and numeric text. Record application release/locale. No native Calligra program was executed.

The generic statistical formula is portable mathematics, but an identically named function in Excel, LibreOffice Calc or another program is not evidence that Calligra treats empty cells, text, logical values or imported formulas identically. Vendor documentation confirms the intended numeric operation; pinned Calligra code identifies its execution path. Special-cell handling, error labels and package availability must be compared with real tests. In particular, Calligra’s AVEDEV help defines the function over numerical values, while AVERAGEA expressly describes special text/logical counting; these are different Calligra functions. Official AVEDEV and AVERAGEA entries.

• AVERAGE: AVERAGE guide; KDE AVERAGE entry.
• ABS: ABS guide; KDE ABS entry. It is used in the independent-check calculation.
• COUNT: COUNT guide; KDE COUNT entry. It helps diagnose the number of usable numeric observations.
• AVEDEV documentation: KDE AVEDEV syntax and five-value example; viewed 2026-10-13.
• AVEDEV source registration: Pinned Calligra 26.08.2 statistical.cpp L107–113; verified 2026-10-13.
• AVEDEV execution body: Pinned statistical.cpp L797–803, with average, absolute-deviation walk and numeric count; verified 2026-10-13.
• Packaged help XML: Pinned statistical.xml, which defines the documented numeric argument types and 7.84 example; viewed 2026-10-13.
• Release/platform and formula syntax: KDE Gear 26.08.2 announcement (2026-10-08), Calligra downloads, formulas handbook; reviewed 2026-10-13.