Does z-test work?

The statistical z-test returns #DIV/0! for the z-statistic and P-values. I notice that even the help page on the z-test has these #DIV/0! results in the screenshots.

For the Data > Statistics > Z-test function

Tested in:

Version: 7.0.0.3 (x64)

Build ID: 8061b3e9204bef6b321a21033174034a5e2ea88e

CPU threads: 2; OS: Windows 10.0 Build 18363; UI render: Skia/Raster; VCL: win

Locale: en-US (fr_FR); UI: en-US

Calc: CL

The help page also has this in 7.1.

I notice that even the help page on the z-test has these #DIV/0! results in the screenshots

Can you share where? The help pages seem to not have screenshots, and guides seem to not have any #DIV/0! - but maybe I just didn’t look at proper topic.

Ah - likely it’s this help page.

@mikekaganski yes, there. Thank you for sharing those links. It looks from the Chapter 9 Data Analysis link that it requires a known variance, which makes sense.

The guide describes this:

For the Z-test tool to work properly, a known variance for each sample must be inserted in the related cell. In the example shown in Figure 58, the variances (125.076923 and 94.435897) were inserted using the formula =VAR(A1:A13) into cell E5 and the formula =VAR(B1:B13) into cell F5. The subsequent z and P values will be updated automatically.

It seems that this function requires one to enter the “Known Variance” for the variables. This makes sense, as the z-test relies on having a population variance. Additional information on this function is available at this link:

https://books.libreoffice.org/en/CG71/CG7109-DataAnalysis.html#toc48

Thank you, @mikekaganski !