CHISQ.TEST
Quirk foundCategory: Statistical · Last tested 2026-09-01
Real compatibility results for the CHISQ.TEST function: executed in Excel for the web, Google Sheets and LibreOffice Calc, with desktop Excel behavior from Microsoft’s official documentation (we do not run desktop Excel — Excel for the web is a different application and is executed separately). Syntax and links to that documentation are below.
Support matrix
| Engine | Documented | Live-tested | Verdict |
|---|---|---|---|
| Excel (desktop) | Yes | No — documented only | n/a |
| Excel for the web | — | Yes (recalc, 2026-09-01) | Supported, behaves as documented |
| Google Sheets | Yes | Yes (Drive import, 2026-08-31) | Quirk found |
| LibreOffice Calc | Yes | Yes (25.8.7.3, 2026-08-31) | Quirk found |
LibreOffice version history
We executed the same test cases under each LibreOffice release to show exactly when CHISQ.TEST’s support changed — not documentation claims, real results.
| LibreOffice version | Verdict | Tested |
|---|---|---|
| 24.2.0.3 | Quirk found | 2026-08-31 |
| 24.8.7.2 | Quirk found | 2026-08-31 |
| 25.2.0.3 | Quirk found | 2026-08-31 |
| 25.8.7.3 | Quirk found | 2026-08-31 |
Why isn't CHISQ.TEST working in LibreOffice?
CHISQ.TEST exists in LibreOffice 25.8.7.3, but it is not a drop-in match for
Excel — our executed tests found real behavioral differences (detailed in the test results on this
page). If a formula that works in Excel or Google Sheets misbehaves in LibreOffice, compare your usage
against the failing cases above before assuming your data is wrong.
Why isn’t CHISQ.TEST working in Google Sheets?
CHISQ.TEST runs in Google Sheets, but our executed cases show it does not match
Excel’s documented behavior on every input (the failing cases are listed on this page). If a
formula that behaves one way in Excel gives you a different answer in Sheets, compare your usage
against those cases before assuming your data is wrong.
Discovered quirks
-
=CHISQ.TEST(A2,A6) on
Google Sheets returned
0.06031830659, but the documented/expected
result is #N/A.
Provenance
Excel documents: "r = c= 1 is not allowed and #N/A is returned." With one row and one column there are no degrees of freedom left after conditioning on the margins, so CHISQ.TEST has no distribution to evaluate; MISMATCH vs expected: expected '#N/A', got 0.06031830659
-
=CHISQ.TEST(A2:B4,A6:B7) on
LibreOffice Calc returned
#VALUE!, but the documented/expected
result is #N/A.
Provenance
Excel documents: "If actual_range and expected_range have a different number of data points, CHISQ.TEST returns the #N/A error value." Here the actual range holds 6 values and the expected range 4; MISMATCH vs expected: expected '#N/A', got '#VALUE!'
-
=CHISQ.TEST(A2,A6) on
LibreOffice Calc returned
#VALUE!, but the documented/expected
result is #N/A.
Provenance
Excel documents: "r = c= 1 is not allowed and #N/A is returned." With one row and one column there are no degrees of freedom left after conditioning on the margins, so CHISQ.TEST has no distribution to evaluate; MISMATCH vs expected: expected '#N/A', got '#VALUE!'
Executed test cases
Excel for the web (executed 2026-09-01 via OneDrive recalculation)
These values come from Excel for the web, not from desktop Excel. They are two different implementations of the calculation engine, and this run measured only the web one: the corpus was uploaded to OneDrive as .xlsx, recalculated by Excel for the web on open, and downloaded again for readback. Excel for the web is a rolling service with no pinnable version, so the run is identified by its date. Where a value here disagrees with the Expected column — which is Microsoft’s documentation of the desktop product — we cannot tell you whether the web engine diverges from the desktop one or the documentation is wrong about both, because we do not run desktop Excel.
| Formula | Description | Result | Expected | Verdict |
|---|---|---|---|---|
| =ROUND(CHISQ.TEST(A2:B4,A6:B8),7) | Microsoft's documented worked example: chi-squared test of independence on a 3x2 agree/neutral/disagree by men/women table | 0.0003082 | 0.0003082ProvenanceMicrosoft's page publishes this example's result as 0.0003082 and states the underlying statistic is 16.16957 with 2 degrees of freedom. Both figures were re-derived independently: sum over cells of (A-E)^2/E for the documented 3x2 tables gives 16.16957507479688, and with df = (r-1)*(c-1) = (3-1)*(2-1) = 2 the right-tail probability is scipy.stats.chi2.sf(16.16957507479688, 2) = 0.0003081920170083095 -> 0.0003082 at 7 dp |
Matched |
| =ROUND(CHISQ.TEST(A2:B4,A2:B4),9) | Actual and expected ranges identical, so the chi-squared statistic is exactly 0 | 1 | 1.0ProvenanceExcel documents that the statistic "is 0 only if Aij = Eij for every i,j". With a statistic of 0 on 2 degrees of freedom the right-tail probability is exactly 1: scipy.stats.chi2.sf(0, 2) = 1.0. This case checks the documented statistic definition at its degenerate point without depending on any published figure |
Matched |
| =CHISQ.TEST(A2:B4,A6:B7) | Actual and expected ranges with different numbers of data points | #N/A | #N/AProvenanceExcel documents: "If actual_range and expected_range have a different number of data points, CHISQ.TEST returns the #N/A error value." Here the actual range holds 6 values and the expected range 4 |
Matched |
| =CHISQ.TEST(A2,A6) | Single-cell actual and expected ranges, i.e. r = c = 1 | #N/A | #N/AProvenanceExcel documents: "r = c= 1 is not allowed and #N/A is returned." With one row and one column there are no degrees of freedom left after conditioning on the margins, so CHISQ.TEST has no distribution to evaluate |
Matched |
Google Sheets (executed 2026-08-31 via Drive import)
Google Sheets is a rolling service with no pinnable version, so this run is identified by its date. The corpus was imported to Drive as .xlsx, recalculated by Sheets, and exported back for readback.
| Formula | Description | Result | Expected | Verdict |
|---|---|---|---|---|
| =ROUND(CHISQ.TEST(A2:B4,A6:B8),7) | Microsoft's documented worked example: chi-squared test of independence on a 3x2 agree/neutral/disagree by men/women table | 0.0003082 | 0.0003082ProvenanceMicrosoft's page publishes this example's result as 0.0003082 and states the underlying statistic is 16.16957 with 2 degrees of freedom. Both figures were re-derived independently: sum over cells of (A-E)^2/E for the documented 3x2 tables gives 16.16957507479688, and with df = (r-1)*(c-1) = (3-1)*(2-1) = 2 the right-tail probability is scipy.stats.chi2.sf(16.16957507479688, 2) = 0.0003081920170083095 -> 0.0003082 at 7 dp |
Matched |
| =ROUND(CHISQ.TEST(A2:B4,A2:B4),9) | Actual and expected ranges identical, so the chi-squared statistic is exactly 0 | 1 | 1.0ProvenanceExcel documents that the statistic "is 0 only if Aij = Eij for every i,j". With a statistic of 0 on 2 degrees of freedom the right-tail probability is exactly 1: scipy.stats.chi2.sf(0, 2) = 1.0. This case checks the documented statistic definition at its degenerate point without depending on any published figure |
Matched |
| =CHISQ.TEST(A2:B4,A6:B7) | Actual and expected ranges with different numbers of data points | #N/A | #N/AProvenanceExcel documents: "If actual_range and expected_range have a different number of data points, CHISQ.TEST returns the #N/A error value." Here the actual range holds 6 values and the expected range 4 |
Matched |
| =CHISQ.TEST(A2,A6) | Single-cell actual and expected ranges, i.e. r = c = 1 | 0.06031830659 | #N/AProvenanceExcel documents: "r = c= 1 is not allowed and #N/A is returned." With one row and one column there are no degrees of freedom left after conditioning on the margins, so CHISQ.TEST has no distribution to evaluate |
Mismatch |
LibreOffice Calc 25.8.7.3 (tested 2026-08-31)
| Formula | Description | Result | Expected | Verdict |
|---|---|---|---|---|
| =ROUND(CHISQ.TEST(A2:B4,A6:B8),7) | Microsoft's documented worked example: chi-squared test of independence on a 3x2 agree/neutral/disagree by men/women table | 0.0003082 | 0.0003082ProvenanceMicrosoft's page publishes this example's result as 0.0003082 and states the underlying statistic is 16.16957 with 2 degrees of freedom. Both figures were re-derived independently: sum over cells of (A-E)^2/E for the documented 3x2 tables gives 16.16957507479688, and with df = (r-1)*(c-1) = (3-1)*(2-1) = 2 the right-tail probability is scipy.stats.chi2.sf(16.16957507479688, 2) = 0.0003081920170083095 -> 0.0003082 at 7 dp |
Matched |
| =ROUND(CHISQ.TEST(A2:B4,A2:B4),9) | Actual and expected ranges identical, so the chi-squared statistic is exactly 0 | 1 | 1.0ProvenanceExcel documents that the statistic "is 0 only if Aij = Eij for every i,j". With a statistic of 0 on 2 degrees of freedom the right-tail probability is exactly 1: scipy.stats.chi2.sf(0, 2) = 1.0. This case checks the documented statistic definition at its degenerate point without depending on any published figure |
Matched |
| =CHISQ.TEST(A2:B4,A6:B7) | Actual and expected ranges with different numbers of data points | #VALUE! | #N/AProvenanceExcel documents: "If actual_range and expected_range have a different number of data points, CHISQ.TEST returns the #N/A error value." Here the actual range holds 6 values and the expected range 4 |
Mismatch |
| =CHISQ.TEST(A2,A6) | Single-cell actual and expected ranges, i.e. r = c = 1 | #VALUE! | #N/AProvenanceExcel documents: "r = c= 1 is not allowed and #N/A is returned." With one row and one column there are no degrees of freedom left after conditioning on the margins, so CHISQ.TEST has no distribution to evaluate |
Mismatch |
Docs & syntax
- Excel (desktop): official documentation
- Google Sheets: official documentation
- LibreOffice Calc: official documentation
Where CHISQ.TEST behaves differently
- Every error code LibreOffice reports as #VALUE!
Across our 2,334-case executed corpus, 272 cases in 135 functions return #VALUE! in LibreOffice Calc 25.8.7.3 where Microsoft documents #NUM! (244), #N/A (15), #DIV/0! (10) or #REF! (3). Google Sheets returns the documented code on 239 of the 249 it has a function for. Identical in all four LibreOffice builds tested.