STANDARDIZE Function (Gnumeric)

STANDARDIZE expresses how far a value lies above or below a chosen mean in units of standard deviation. This standardized value is often called a z-score: z = (x − mean) / stddev. A z-score of +2 means the value lies two standard deviations above the supplied mean; −1 means one below. Unlike AVERAGE or STDEV, STANDARDIZE takes the mean and standard deviation as inputs—it does not calculate either parameter from a range automatically (Gnumeric Statistics manual, STANDARDIZE, accessed 2026-10-13).

Use this when comparing scores measured on different scales (for example, exam results from classes with different means and spreads), detecting unusually high/low observations, or feeding normalized values to another analysis. The interpretation assumes the mean and standard deviation were chosen appropriately for the dataset.

=STANDARDIZE(x,mean,stddev)

All three arguments are required in the published synopsis and no defaults are documented. There is a documentation/source discrepancy: the Gnumeric manual says zero stddev produces #DIV/0!, but the checksum-verified 1.12.62 callback returns #NUM! for every stddev <= 0, including zero. No native execution was performed, so the manual’s token must not be presented as a confirmed 1.12.62 runtime result (manual; indexed 1.12.62 implementation).

Version/platform: The upstream 1.12.62 Gnumeric desktop release is the target; project catalogue registration lists STANDARDIZE as Complete / Basic testing, not as evidence of execution here. Actual packaged availability on your operating system was not checked (Gnumeric catalogue, accessed 2026-10-13). Locale: Examples use English names, periods for decimals and commas between three arguments. Some decimal-comma locales use semicolons; use Gnumeric’s function dialog for the right syntax (official mailing-list guidance).

Set up an empty sheet as follows:

Expected D2 = 1.5 because (85 − 70)/10 = 1.5. The value is 1.5 standard deviations above the supplied mean. The formula does not establish whether the score is statistically significant, a percentile or normally distributed; those are separate questions.

The positive numeric outputs were checked by independent subtraction/division, and the last output by Gnumeric documentation only. Display formatting may round the result.

Enter raw values 10, 12, 14, 16, 18 into B2:B6. In C2 enter:

=STANDARDIZE(B2,AVERAGE($B$2:$B$6),STDEV($B$2:$B$6))

Copy C2 down to C6. The absolute $ references keep the comparison population fixed while the B2 observation changes by row. AVERAGE(B2:B6) is expected to be 14; sample STDEV(B2:B6) is expected to be √10 ≈ 3.162277660. (Gnumeric’s official Statistics manual identifies STDEV as its N−1 sample standard-deviation function.)

Independent verification: the deviations are −4,−2,0,2,4; their squares sum to 40; the sample variance is 40/(5−1)=10. The z-scores above follow by division by √10. These results were computed independently, not generated by executing the worksheet in Gnumeric.

Suppose those five scores are the entire population of interest, not a sample. The Gnumeric manual identifies STDEVP as the population standard-deviation function, in contrast to STDEV’s N−1 sample interpretation. For these five values, population variance is 40/5 = 8, so standard deviation √8 ≈ 2.828427125.

=STANDARDIZE(B2,AVERAGE($B$2:$B$6),STDEVP($B$2:$B$6))

Expected result for B2=10: about -1.41421356, compared with -1.26491106 under sample STDEV. Neither parameter is intrinsically always correct—choose the one matching the question and do not mix standard-deviation definitions across groups. This distinction and the STDEV/STDEVP functions are in the Gnumeric Statistics manual.

An optional guard in a sheet where C2 already holds a positive standard deviation is:

=IF(C2>0,STANDARDIZE(A2,B2,C2),"Check standard deviation")

This guards the documented zero-denominator problem for a known numeric value. It is a formula design suggestion, not a claim that every Gnumeric invalid-value branch has been tested.

The official Gnumeric function help describes STANDARDIZE as Excel-compatible; that is not a blanket promise of matching coercion, all error cases, platform availability or locale. The official function catalogue classifies it Complete / Basic; Basic describes upstream tests, not validation of this page.

• AVERAGE supplies a mean from a range.
• STDEV supplies the sample standard deviation.
• STDEVP supplies the population standard deviation.
• IF can guard invalid denominator entries.
• COUNT helps check that numeric data exist.

The official 1.12.62 source archive was checked against GNOME’s published SHA-256 checksum; plugins/fn-stat/functions.c supplied the branch inspection described above.

1. Gnumeric Statistics manual — STANDARDIZE, STDEV and STDEVP, accessed 2026-10-13 — exact argument order, zero-stddev error, sample-versus-population semantics, Excel-compatible annotation.
2. Gnumeric function catalogue, accessed 2026-10-13 — registered/Complete/Basic project-testing status.
3. Gnumeric official 1.12.62 release announcement, accessed 2026-10-13 — version/platform release context, not independent testing.
4. GNOME source release series, accessed 2026-10-13 — identifies 1.12.62 release source; the official source archive was checksum-verified and inspected without compilation.
5. Indexed Gnumeric 1.12.62 statistical-function source, accessed 2026-10-13 — symbol help_standardize and function implementation metadata, not Gnumeric runtime output.
6. GNOME Gnumeric locale discussion, accessed 2026-10-13 — locale-specific argument separators.

Validation limits: Official documentation and checksum-verified Gnumeric 1.12.62 source were inspected, and examples were independently calculated. No native Gnumeric execution, screenshots, locale regression or workbook round-trip was performed.