[Bug]: two-tailed Fisher's exact test

[Diesfeldt]

JASP Version

0.18.1

Commit ID

No response

JASP Module

Frequencies

What analysis are you seeing the problem on?

Contingency Tables

What OS are you seeing the problem on?

Windows 11

Bug Description

I try to reproduce in JASP the results of a two-tailed Fisher’s exact test, as discussed on page 134 in Pandit JJ, Picton P, Mashour G. Penehcyclidine and awareness during anaesthesia: caution with zero numerators. Anaesthesia 2013; 68: 121–35. JASP gives the same p-value for a one-tailed and a two-tailed test. However, the 95%-CI of the log odds ratio does not contain zero in the one-tailed test, but contains zero in the two-tailed test. What is going on?
Pandit2013.zip

Expected Behaviour

JASP should give different p-value for a one-tailed and a two-tailed test.

Steps to Reproduce

...
JASP should give different p-value for a one-tailed and a two-tailed test.

Log (if any)

No response

More Debug Information

No response

Final Checklist

• I have included a screenshot showcasing the issue, if possible.
• I have included a JASP file (zipped) or data file that causes the crash/bug, if applicable.
• I have accurately described the bug, and steps to reproduce it.

[tomtomme]

Thx for the Report. That certainly does not sound correct. Will try confirming later on my system

[Diesfeldt]

Thank you very much, Thomas. I am looking forward to the results of your enquiry. Best, Han Diesfeldt

…

________________________________ Van: Thomas Langkamp ***@***.***> Verzonden: dinsdag 19 november 2024 09:23 Aan: jasp-stats/jasp-issues ***@***.***> CC: Han Diesfeldt ***@***.***>; Author ***@***.***> Onderwerp: Re: [jasp-stats/jasp-issues] [Bug]: two-tailed Fisher's exact test (Issue #3021) Thx for the Report. That certainly does not sound correct. Will try confirming later on my system — Reply to this email directly, view it on GitHub<#3021 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AKXOHB7TAYZQ4HHOE24DWQL2BLYPFAVCNFSM6AAAAABR77RXVCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIOBVGAYDQNRTG4>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[tomtomme]

@Diesfeldt
So I tried this in jasp 0.19.1 flatpak and cannot confirm:

Also the jasp file - remove the .zip from the file name to open:
Dancing Cats.jasp.zip

The file you attached sadly did not contain any data, only two analysis:

Could you provide your zipped jasp file so that I can check if something with your data is special?

[Diesfeldt]

Thanks again, the analysis for the dancing cats and dogs is fine, indeed. I enclose my JASP-file. I now see I sent a different but irrelevant file to Git-hub, sorry for that. I hope you will find the data that are enclosed in the file. If not, let me know and I send you an Excess data file (it is just a small table). Best, Han

…

________________________________ Van: Thomas Langkamp ***@***.***> Verzonden: dinsdag 19 november 2024 10:08 Aan: jasp-stats/jasp-issues ***@***.***> CC: Han Diesfeldt ***@***.***>; Mention ***@***.***> Onderwerp: Re: [jasp-stats/jasp-issues] [Bug]: two-tailed Fisher's exact test (Issue #3021) @Diesfeldt<https://github.com/Diesfeldt> So I tried this in jasp 0.19.1 flatpak and cannot confirm: grafik.png (view on web)<https://github.com/user-attachments/assets/6ec28fc1-5d0e-4ed0-b420-72731b4b2272> Also the jasp file - remove the .zip from the file name to open: Dancing Cats.jasp.zip<https://github.com/user-attachments/files/17812457/Dancing.Cats.jasp.zip> The file you attached sadly did not contain any data, only two analysis: grafik.png (view on web)<https://github.com/user-attachments/assets/2c9e4f3e-1331-4136-a467-1e9e43be7908> Could you provide your zipped jasp file so that I can check if something with your data is special? — Reply to this email directly, view it on GitHub<#3021 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AKXOHB3MO5ZPYZZD33UBSVT2BL5XFAVCNFSM6AAAAABR77RXVCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIOBVGEYDKMBXGA>. You are receiving this because you were mentioned.Message ID: ***@***.***>