ONLYOFFICE AVERAGEA: Text, Logical Values and Reference Differences

• Introduced: The first introduction release and date are not established by the checked sources.
• Availability: Currently documented in ONLYOFFICE’s rolling Spreadsheet Editor help; no final supported-version endpoint is established.
• Scope: Ten scalar cases were executed in Desktop Editors Community 10.0.0.169 on Linux x64. Docs, older desktop builds and mobile clients were not tested.
• Checked: October 14, 2026, against the official AVERAGEA reference, source code and saved native results. Additional arithmetic reconstructions are independently calculated.

=AVERAGEA(value1, [value2], ...)

The first input is required, with further inputs optional. The official reference documents up to 255 values and describes text/FALSE as zero and TRUE as one. These examples use its English comma-separated syntax.

That short summary is not a universal rule for every argument form. In the tested desktop build, a referenced numeric string contributed zero, a directly supplied numeric string was converted, and a direct ordinary-text literal returned an error. The official implementation revision also separates reference, array and direct-argument branches. Do not assume they are interchangeable.

Create MixedMean. Put Input and Value in A5:B5. Enter these cells, leaving B11 genuinely empty:

For B12, set the cell to text before entering 2 and verify its stored type. Do not include an apostrophe as part of the stored string. Enter TRUE/FALSE as logical values; a text field spelling TRUE is a different input. The native fixture’s stored types were checked in the saved workbook, rather than inferred from their appearance.

Place these formulas in G6:G14. The observed results are from the named desktop build; repeating thirds are rounded below to ten decimal places.

G7 uses 10/3: the status adds zero but remains in the denominator. G8 uses 11/5. G9 excludes the genuine blank, while G14 includes the empty-text formula result as a third entry. Those two cells look empty but are not equivalent in this test.

G10 and G11 also differ despite both displaying a 2 somewhere. The referenced string is counted as zero, but the direct literal contributes two. Converting types is a modeling decision, not a cosmetic formatting change.

Create FruitExample, with no header inside A1:B4:

In D1 enter =AVERAGEA(A1:A4,B1:B4,TRUE,FALSE). The vendor image and the native reproduction both give 34.5. Four text labels contribute zero, four numbers total 344, and the direct Booleans contribute 1 and 0. There are ten included entries, so the mean is 345/10.

The labels change the denominator. If the question concerns only the four numeric observations, their mathematical mean is 86, from 344/4. Adding descriptive labels to a mean does not make the result more representative.

On MixedMean, enter the explicit numeric mapping 4,6,0,1,0 in J6:J10. In L6 enter =SUM(J6:J10); expect 11. In L7 enter =COUNT(J6:J10); expect 5. In L8 enter =L6/L7; expect 2.2, independently reconstructing G8.

If pending means missing rather than a zero score, use a separate numeric measurement column and a status column. The mapping is mathematically consistent, but it may answer the wrong business question. A low mixed-value mean does not prove that the actual recorded measurements were low.

The ten native cases establish these exact scalar inputs on Desktop 10.0.0.169. Other numeric-string grammars, array constants, older versions, Docs and mobile clients need their own checks. The one observed #VALUE! does not describe every possible error path, and an all-empty range was not tested.

AVERAGE addresses ordinary numeric means. SUM and COUNT reconstruct the explicit numeric mapping. MAXA and MINA have their own mixed-input rules; do not transfer AVERAGEA’s text-to-zero summary to them.