VAR in Collabora Office: Estimate Sample Variance with the n Minus One Denominator
Use Collabora VAR for sample variance, verify centered squares and distinguish missing cells from actual zeros.
Verification status: See the evidence and test scope described in this guide. How formula examples are checked
Quick answer: for 2, 4, 6 and 8 in A2:A5, =VAR(A2:A5) has independently expected result 20/3, approximately 6.666666667.
Syntax and sample interpretation
=VAR(Argument1 [; Argument2 ...])
The dedicated Collabora help describes VAR as a sample-variance estimate, permits up to 255 arguments and requires at least two usable numeric values. Arguments can include references and ranges. Ordinary text and empty cells inside a data range are ignored.
The current source divides the sum of squared deviations from the mean by n-1. This denominator distinguishes VAR from population variance, which divides by n. VAR is retained here under its exact documented legacy name rather than silently replacing its page identity with another spelling.
Variance is measured in squared input units. A variance of 6.666666667 for measurements in meters is in square meters, not meters.
Example 1: check the centered-square calculation
Enter 2,4,6,8 in A2:A5. Their mean is 5. In B2 enter =A2-AVERAGE($A$2:$A$5) and fill down. Deviations are -3, -1, 1 and 3.
In C2 enter =B2^2 and fill down. Squared deviations are 9, 1, 1 and 9, totaling 20. Dividing by 4-1 gives the sample variance 20/3.
The expression =SUM(C2:C5)/(COUNT(A2:A5)-1) provides a transparent reference calculation for this complete numeric fixture. Use the unrounded mean and stored values when checking a general dataset.
Example 2: compare the population denominator
=VARP(A2:A5) has independently expected result 5, because it divides the same centered-square total twenty by four.
The two results answer different questions. Choose the sample estimator when the observations are treated as a sample under the intended statistical procedure; choose the population calculation when describing the complete population of interest. The function does not infer that context from the number of rows.
For n>1 on the same numeric observations, VAR=VARPn/(n-1). Here 54/3 recovers 20/3.
Advanced application: interpret scale and shift
Adding ten to every observation gives 12,14,16,18 and the same sample variance 20/3. Centering removes the common shift.
Multiplying each original observation by two gives 4,8,12,16 and sample variance 80/3, approximately 26.666666667. Variance scales with the square of the multiplier, not the multiplier itself.
Taking the square root of 20/3 gives a sample standard deviation approximately 2.581988897 in the original measurement units. Do not compare a variance directly with a standard deviation as though their units matched.
Advanced application: audit ignored cells versus numeric zeros
A referenced range containing 2,4,“note”,blank,6,8 still has four usable numeric observations and the same independent sample variance 20/3 under the documented range rule.
If the text and blank are deliberately replaced with actual zeros, the six values become 2,4,0,0,6,8. Their independent sample variance is 32/3, approximately 10.666666667. Inserting zeros is a data change, not a neutral way to fill missing cells.
Check COUNT alongside the variance when importing numeric-looking text. The number of selected cells need not equal the number used in the denominator.
Errors and troubleshooting
The current source uses a division-by-zero result when fewer than two usable numeric values remain. Two or more identical numeric observations have valid variance zero; that is different from having too few observations to estimate sample variance.
The help’s ignored-text statement concerns referenced ranges. This guide does not certify every direct string, Boolean or array coercion. Investigate upstream errors and missing-data policy rather than replacing every failed calculation with zero.
Large magnitudes, subtraction near a large mean and extreme floating-point values require numerical care. These ordinary independent examples establish the calculation and units, not universal native accuracy or a statistical conclusion about the population.
Availability and verification
The exact function is documented in Collabora Office 26.04 desktop help. The cited help snapshot was retrieved on 2026-10-13 and hash-verified October 15, 2026. Desktop documentation does not establish Online or mobile support. First introduction and last functioning release are unknown in this evidence.
No native Collabora Office execution occurred. All numerical results are independent mathematical expectations, with English names and semicolon argument separators. Native error displays, all coercions and extreme numerical behavior are not certified.
Implementation statements are corroborated by Collabora current engine source, retrieved October 15, 2026. This is moving-main source evidence, not a pinned desktop 26.04 binary or execution.