Need help with Fisher's exact test

General help and assistance with jamovi. Bug reports can be made at our issues page: https://github.com/jamovi/jamovi/issues . (If you're unsure feel free to discuss it here)
Post Reply
NewAtThis
Posts: 4
Joined: Tue Feb 14, 2023 12:27 pm

Need help with Fisher's exact test

Post by NewAtThis »

Hi everyone, I am a med student with hardly any knowledge of statistics, so apologies for the trivial questions.

1. I wanted to do a chi-squared test on some data. However, the data didn't meet the requirements of observed and expected counts. Therefore, I changed it to Fisher's exact test. With this, I did get a significant p, but the "Value" space is blank. What could be the reason for this? Is the result of the test still valid?

2. Concerning some other data, I want to test different locations of a tumor depending on whether it is a recurrent or primary tumor. So, for each row in my sheet, there is a column with "recurrent or primary tumor" and a column with the category "tumor location", i.e. "upper eyelid, lower eyelid, eyebrow" etc. The chi-squared test yielded a significant p-value, so my clinical supervisor told me to now test all the individual items, in order to figure out where exactly the significance came from. How do I get the chi-squared test to individually consider each category of location?

Any help would be greatly appreciated!
User avatar
reason180
Posts: 278
Joined: Mon Jul 24, 2017 4:56 pm

Re: Need help with Fisher's exact test

Post by reason180 »

Hi. FYI: Typically, I think questions get answered more effectively when the questioner provides a jamovi omv file containing either the data or some very similar made-up data.

(1) Customarily, no test statistic is reported for Fisher's exact test. That's why there's no "value."

(2) To conduct a different chi square contingency test for each of several subsets of the data, you would need to either:
  • Filter-out all data that's not part of the subset in question, run the chi relevant chi square test, filter the data differently to focus on different subset of the data, run the relevant chi square test, etc., or
  • Create multiple pairs of columns, based on the original two, with each new pair containing only the data-rows relevant to testing a particular subset of the data.
Post Reply