# Thread: I need advice on Fisher's exact test

1. ## I need advice on Fisher's exact test

I recently submitted a paper and in my study, I wanted to see if there is an association between a gene mutation and the absence/presence of the expression of another gene in 14 cancer cell lines. I used the Fisher's exact test because to my knowledge, it's used for the relationship between two categorical variables but I was told by the reviewer that this test is not appropriate. I wonder if anyone could help me with this. Thank you.

2. ## Re: I need advice on Fisher's exact test

Well, fisher's test is used with 2 categorical variables like you mentioned. So what are we missing? Can you post their comment in its entirety, plus any more details about your variables.

So n= 14; variable 1 (gene present): yes/no; variable 2 (expression): yes/no.

Can you post your 2x2 table, so we can see your cell counts, etc.

Could their critique be related to the content not the appropriateness or that the test or complaints about the statistical power of the test?

3. ## Re: I need advice on Fisher's exact test

Originally Posted by hlsmith
Well, fisher's test is used with 2 categorical variables like you mentioned. So what are we missing? Can you post their comment in its entirety, plus any more details about your variables.

So n= 14; variable 1 (gene present): yes/no; variable 2 (expression): yes/no.

Can you post your 2x2 table, so we can see your cell counts, etc.

Could their critique be related to the content not the appropriateness or that the test or complaints about the statistical power of the test?

'Authors claim that there is a significant association between NF2 status and CD44v protein expression based on the results of the literature study and the Fisher's exact test. I don't think Fisher's exact test can be used to determine the relationship between NF2 status and CD44v expression.'
This is the comment from the reviewer.
And my 2x2 table is

NF2 mutation Without NF2 mutation
CD44v expression 9 1
No CD44v expression 1 3

The number of cell lines without NF2 mutation is indeed too few. So in this case, is Fisher's exact test still valid?

4. ## Re: I need advice on Fisher's exact test

Hi,
I believe this statement needs a lot of clarification as to why the reviewer THINKS the Fisher exact test is not applicable. It definitely can not be the small sample size.

"Although in practice it is employed when sample sizes are small, it is valid for all sample sizes. "

https://en.wikipedia.org/wiki/Fisher%27s_exact_test

5. ## Re: I need advice on Fisher's exact test

I agree with rogojel. I believe that the reviewer should do a lot more than claiming that a test cannot be performed: (s)he should strongly motivate his/her claim.
I would reply to the reviewer in a gentle way, giving solid bases (i.e., literature) to the use of that test for that particular ami, in that particular context.

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts