Fisher's exact test

#1

Hi all :smile:

I am trying to differentiate between participants' (from different groups) ability to classify pictures into categories. Each participant watched the same pictures and was asked to categorize them into categories according to their perceived similarities.

My data looks like this:

Categorization (pictures groupings) Number of participants using this categorization in Group 1 Number of participants using this categorization in Group 2 Number of participants using this categorization in Group 3
A 10 4 1
B 10 2 1
C 10 4 0
D 1 3 2

How do I conduct a Fisher's exact test (and pairwise comparisons) on this data?

Your suggestions and help are much appreciated. Thank you!

#2

Is this related to R in some way?
And if so what steps did you already do to solve your problem?
Did you google for "Fisher exact test R" ?

#3

Thank you.
Yes - I have. I am still not sure how to conduct the pairwise comparisons.
Thank you.

#4

Have you got your data in R?

#5

Thank you. Yes.
I get the following error:
Error in fisher.test(data) :
all entries of 'x' must be nonnegative and finite.
Thanks in advance for trying to help.

#6

in R you can access function documentation with the question mark symbol in the console like this
?fisher.test
I read:
If x is a matrix, it is taken as a two-dimensional contingency table, and hence its entries should be nonnegative integers.

Is your data of this form? if not you should transform it.
For a resource on how to transform data in R, I recommend

#7

You might want to check out this page:

https://stats.idre.ucla.edu/r/whatstat/what-statistical-analysis-should-i-usestatistical-analyses-using-r/#exact

It will have some more information that you might find helpful.

#8

Thank you!!!! :grin: