Contact
12 programs
Explore the site

STDEV Function in Calligra Sheets: Sample Standard Deviation

Statistical Intermediate Calligra Sheets

Calculate sample standard deviation with Calligra Sheets STDEV using reproducible inputs, correct n−1 denominator, outlier comparisons and documented limitations.

Verification status: Official documentation referenced; example outcomes independently reasoned or arithmetically checked where stated. Not executed in Calligra Sheets. How formula examples are checked

Quick answer: STDEV calculates sample standard deviation, a measure of how far numeric observations typically vary from their sample mean. KDE documents =STDEV(6;7;8) → 1. Unlike a population standard deviation, the sample calculation divides the sum of squared deviations by n−1, not n. Both the official documentation and Calligra 26.08.2 calculation source support this distinction. [K1, K2, K3, I1]

Validation (2026-10-13): KDE’s handbook, function registration, and version-pinned engine implementation were examined. All numerical examples are independently calculated expected outputs; no native Calligra Sheets application run was performed, and error displays, input coercion, and locale behavior are not asserted as tested. [K1–K5, I1]

Exact syntax, parameters, types, and defaults

=STDEV(value1; [value2]; ...)
=STDEV(B2:B6)
  • value1 (required): a numeric observation, cell reference, or numeric range. The official 26.08.2 registration uses setParamCount(1, -1) and setAcceptArray(): at least one function argument and array/range acceptance. The source does not encode a fixed numeric cap here, but it does not establish unlimited practical arguments. [K2]
  • value2 and later values (optional): more numeric observations or ranges. Although the function accepts one syntactic argument, a mathematically meaningful sample standard deviation requires at least two accepted numeric observations. This is a data requirement, not a second mandatory formula argument. [K2, K3, I1]
  • No separate sample/population flag: this registration always selects the sample form. Use the distinct registered function STDEVP when your observations represent the entire population. [K1, K2, K3]
  • Result: a nonnegative numeric spread for valid numeric samples. It is 0 when all observed numeric values are equal. It is in the same units as the input values. [I1]
  • Argument order: reordering the same valid numeric observations does not change standard deviation mathematically. [I1]
  • Locale punctuation: the KDE English handbook separates explicit arguments using ;, uses . as the decimal separator, and writes contiguous ranges with :. Only the English examples are asserted, not every installation’s locale parser. [K1]

For n valid observations x1…xn, let mean = (x1+…+xn)/n. Then:

sample standard deviation = SQRT( SUM((xi - mean)^2) / (n - 1) )

This is a mathematical expression, not a single formula with an undocumented array-entry convention. The pinned Calligra implementation func_stddev calls ValueCalc::stddev(args,false); its engine computes a mean, sums the squared deviations, and divides by count−1 before taking the square root. [K3, K4, I1]

Reproduce the sample workbook

  1. Open Calligra Sheets with a blank worksheet. Check whether STDEV is available in your installed build; the official 26.08.2 source registration alone cannot prove installed execution. [K2]
  2. Enter these numeric inputs exactly. B7 contains one deliberate outlier; keep it out of the shorter B2:B6 range.
  3. Enter each formula in column E, not in the B input range; begin with = and use the shown separators for the handbook-style example.
  4. Display enough decimal places to compare expected fractional results, and distinguish a mathematical expectation from an observed native test. [I1]
Cell Value Cell Value
A1 Trial B1 Measurement
A2 1 B2 6
A3 2 B3 7
A4 3 B4 8
A5 4 B5 9
A6 5 B6 10
A7 Outlier B7 100
Enter in Formula Independently expected result What it demonstrates
E2 =STDEV(B2:B6) 1.581138830084 Five-number sample uses divisor 4.
E3 =STDEVP(B2:B6) 1.414213562373 Same data treated as the complete population; divisor 5.
E4 =STDEV(6;7;8) 1 Official KDE handbook example.
E5 =STDEV(6;6;6) 0 All values identical.
E6 =STDEV(B2:B7) 37.585458180880 Outlier substantially increases spread.
E7 =AVERAGE(B2:B6) 8 Mean required for a manual check.
E8 =STDEV(B2:B6)/AVERAGE(B2:B6) 0.197642353761 Coefficient of variation: approximately 19.7642%.
E9 =(B6-AVERAGE(B2:B6))/STDEV(B2:B6) 1.264911064067 Sample z-score of B6=10 relative to the five-number mean.
E10 =STDEV(2;4) 1.414213562373 Two-value sample; denominator 1.

Arithmetic audit: For B2:B6 = 6, 7, 8, 9, 10, the mean is 8 and deviations are −2, −1, 0, 1, 2. Their squares total 4+1+0+1+4 = 10. Therefore STDEV = √(10/4) = √2.5 ≈ 1.581138830084. The population counterpart is STDEVP = √(10/5) = √2 ≈ 1.414213562373. [I1, K3]

The outlier case B2:B7 has a mean of 140/6 and an independently calculated sample deviation of 37.585458180880, illustrating why sample dispersion can change dramatically with extremes. This is not an anomaly in the formula. [I1]

Advanced technique 1: independently reconstruct STDEV without array formulas

You can audit the function row by row, avoiding any unverified dynamic-array behavior:

  1. In G1 type Mean, and in G2 enter =AVERAGE(B2:B6) (expected 8).
  2. In C1 type Squared deviation. Enter =(B2-$G$2)^2 in C2 and copy it through C6. Expected numeric results: 4, 1, 0, 1, 4; $G$2 holds the mean cell fixed. [I1]
  3. In G3, enter =SUM(C2:C6) (expected 10).
  4. In G4, enter =COUNT(B2:B6) (expected 5 for this numeric-only range).
  5. In G5, enter =SQRT(G3/(G4-1)) (expected 1.581138830084). Its result should agree numerically with E2, within normal floating-point rounding. [I1, K3, K4]

Why this matters: the helper calculation makes the n−1 divisor explicit, so it is easier to detect accidental use of a population formula or an incorrect range. This step-by-step reconstruction uses established scalar formulas, not an assumed spill or CSE array syntax. [I1]

Advanced technique 2: standardized values and unit-free spread

Coefficient of variation (CV): with a nonzero, meaningful mean, =STDEV(B2:B6)/AVERAGE(B2:B6) provides sample spread relative to the magnitude of that mean. In this example the result is 0.197642353761 = 19.7642%. Interpret CV cautiously if the mean is near zero, when the variable is on an interval (rather than a ratio) scale, or when negative values make the ratio ambiguous. These are statistical interpretation limits, not a particular Calligra error. [I1]

Sample z-score: =(B6-AVERAGE(B2:B6))/STDEV(B2:B6) is 1.264911064067 for the value 10. It measures how many sample standard deviations the observation is above the sample mean. If every value is identical, STDEV is 0, and the z-score formula divides by zero; do not claim a specific cell-rendered Calligra error without an application test. [I1, K3]

Errors, boundaries, and troubleshooting

Case Practical recommendation Verified status
Unknown function Check installed Calligra Sheets version/module; the statistical module in release 26.08.2 registers STDEV. Source registration, not runtime proof. [K2]
Formula appears as text Begin the entry with = and inspect cell format. General diagnostic; not application-tested.
Wrong separator or decimal mark Try the syntax in the installed KDE function help; ; and . are the documented English example convention. Handbook; locale variations untested. [K1]
Only one numeric observation The sample equation has denominator n−1=0; the pinned helper divides by cnt−1. No specific visible error string has been verified. Source and mathematics; native boundary test pending. [K3, K4, I1]
All empty or all text observations The count and calculation helpers determine which values contribute; do not assume Excel’s text/blank coercion applies. Native test required for exact result. [K4]
Booleans, string numbers, direct text parameters ValueCalc::stddev(..., false) chooses the numeric/non-A traversal, but upstream conversion and error propagation still need application tests. Source path verified, per-input outcome not observed. [K3, K4]
STDEV differs from STDEVP The sample form divides by n−1; the population form divides by n. They should differ for nonconstant data. Distinct registered functions and source helpers. [K2–K4]
Results jump after adding a 100 Verify whether your range includes the B7 outlier. Independent numeric demonstration. [I1]
A helper z-score divides by zero Check that sample spread is not zero before dividing. Mathematical boundary, no native error message asserted. [I1]
Very large numbers or many values Numerical precision, overflow and performance depend on installed engine and hardware. Not benchmarked on an installed Calligra build.

Native test matrix still outstanding: exactly one valid sample value; zero values; blank cells; text and Boolean cells within ranges; direct text arguments; error-valued cells; multi-range parameters; localized separators; very large/small magnitudes; comparison with STDEVP in a running 26.08.2 installation. Observe actual output strings before approving any error-specific claims. [K3, K4]

Version, platform and compatibility notes

  • Calligra 26.08.2 (released October 8, 2026): official source registers STDEV as accepting range input and calls the sample standard-deviation helper. The KDE handbook documents the sample interpretation and the STDEVP relationship. The handbook’s displayed revision is older than the source release. No running Calligra application was tested. [K1–K5]
  • Other software: Excel, LibreOffice Calc, Google Sheets and other spreadsheets have related statistical functions, but function naming across editions, separators, treatment of text and Booleans, floating-point precision and error propagation are not guaranteed to match Calligra. No other application’s behavior is offered as Calligra runtime evidence. [K3, K4]
  • Platforms: KDE’s download guidance lists Linux distro packages and Flatpak; Windows availability is described as in progress. This review ran no application on Linux, Windows or macOS. [K5]
  • Aliases: the pinned official release’s registered primary-function inventory lists STDEV with no explicit alias. The separately registered STDEVP is a related different mathematical operation, not an alias rather than a spelling variant. [K2]

Verified related KDE documentation: STDEVP — population standard deviation, AVERAGE — sample mean, COUNT — numeric observations, SUM, SQRT, and MEDIAN — outlier-robust centre. [K1]

Dated claim-level references — checked 2026-10-13