ONLYOFFICE STANDARDIZE: Reference Scores, Units and Standard Deviations
Calculate standardized scores in ONLYOFFICE with fixed references, sample and population spreads, unit conversions and complete worksheet examples.
Verification status: See the evidence and test scope described in this guide. How formula examples are checked
STANDARDIZE expresses a value as a signed distance from a specified mean, measured in specified standard deviations. You supply both reference parameters; the function does not estimate them automatically from nearby cells.
Availability and version history
- Introduced: The first introduction release and date are not established by the checked sources.
- Availability: Currently documented in ONLYOFFICE’s rolling function help; no final supported-version endpoint is established.
- Scope: Shared Spreadsheet Editor documentation, with formulas corroborated by a pinned official source-code revision. Individual client/build parity is unverified.
- Checked: October 14, 2026, against the official STANDARDIZE reference, source and vendor screenshot. Example outputs were independently calculated; these cases were not tested in the native application.
Syntax and parameter requirements
=STANDARDIZE(x, mean, standard_dev)
All three arguments are required. The official reference defines x as the value to normalize, mean as the arithmetic mean and standard_dev as a standard deviation greater than zero.
The mathematical calculation is:
=(x-mean)/standard_dev
ONLYOFFICE’s source at a pinned revision corroborates that expression and checks for nonpositive standard deviation. This establishes a source-scoped check, not native behavior in every deployed version. The examples use English names and commas.
Example: compare measurements with a fixed reference
Create ReferenceScores. Put Measurement, Standardized value and Direct check in A1:C1. Enter 80,90,100,110,120 in A2:A6. Put Reference mean in E2 and Reference standard deviation in E3, with 100 in F2 and 10 in F3.
In B2 enter =STANDARDIZE(A2,$F$2,$F$3). In C2 enter =(A2-$F$2)/$F$3. Fill B2:C2 through row 6.
| Row | Measurement | Expected B and C |
|---|---|---|
| 2 | 80 | -2 |
| 3 | 90 | -1 |
| 4 | 100 | 0 |
| 5 | 110 | 1 |
| 6 | 120 | 2 |
A result of −2 means two reference standard deviations below the reference mean. A positive result is above the mean; it does not automatically mean better performance. The result is unitless because the difference and standard deviation have matching units.
Put Reconstructed measurement in D1, enter =$F$2+B2*$F$3 in D2 and fill to D6. The expected results are the original 80,90,100,110,120. This reverses the arithmetic while keeping the supplied reference parameters fixed.
Supplying an estimated standard deviation changes the question
Keep F2:F3 unchanged. Put =AVERAGE(A2:A6) in H2, =STDEV.S(A2:A6) in H3 and =STDEV.P(A2:A6) in H4. For these five numeric observations:
| Cell | Meaning | Expected result, approximately |
|---|---|---|
| H2 | Observed mean | 100 |
| H3 | Sample standard deviation | 15.8113883008 |
| H4 | Population standard deviation | 14.1421356237 |
The squared deviations sum to 1000. The sample calculation uses 1000/(5−1) before taking the square root; the population calculation uses 1000/5.
Put Using sample SD and Using population SD in J1:K1. Enter =STANDARDIZE(A2,$H$2,$H$3) in J2 and =STANDARDIZE(A2,$H$2,$H$4) in K2. Fill to row 6.
| Row | Expected J, approximately | Expected K, approximately |
|---|---|---|
| 2 | -1.2649110641 | -1.4142135624 |
| 3 | -0.6324555320 | -0.7071067812 |
| 4 | 0 | 0 |
| 5 | 0.6324555320 | 0.7071067812 |
| 6 | 1.2649110641 | 1.4142135624 |
The function has not selected one convention over another. You changed its inputs. Use a reference distribution, a sample estimate or a population description deliberately, and report which parameters were used.
Change units consistently
Create ConvertedUnits. Enter 800,900,1000,1100,1200 in A2:A6, representing the first fixture’s measurements multiplied by ten. Put Converted measurement in A1 and Standardized value in B1. Enter =STANDARDIZE(A2,1000,100) in B2 and fill to B6.
The targets remain −2,−1,0,1,2. Multiplying the measurements, mean and standard deviation by the same positive factor preserves standardized distances. In D2, =STANDARDIZE(1200,100,10) instead gives 110, because the input was converted but the reference parameters were not.
For another parameter comparison, put =STANDARDIZE(110,100,10) in D4 and =STANDARDIZE(90,80,5) in D5. The results are 1 and 2. A lower raw measurement can be farther above its own reference mean; meaningful comparisons still require compatible measurement purposes and reference populations.
Reproduce the official example
Create VendorStandardize with x = 5 in A1, mean = −2 in A2 and standard deviation = 10 in A3. In C3 enter =STANDARDIZE(A1,A2,A3); expect 0.7, agreeing with the official screenshot. The direct calculation is (5-(-2))/10.
Boundaries and interpretation
A standard deviation of zero or less is outside the documented input requirement. The short help does not identify every native error token or coercion case, so no unobserved diagnostic is promised.
Standardization does not make an arbitrary data distribution normal. A result of 2 is not itself a percentile or probability, and it does not establish that a measurement is an outlier under an unspecified policy. Those conclusions need assumptions beyond this three-argument calculation.