PHD Discussions Logo

Ask, Learn and Accelerate in your PhD Research

Question Icon Post Your Answer

Question Icon

7 years ago in Discussion By Aanchal S

Fisher’s exact test

I am using Chi-square in my study, but there are some variables which have their expected value less than 5. I was thinking of using Fisher’s exact test but I know it is used for  2X2 table only. So, is there any other option?
 

All Answers (3 Answers In All)

By Ritu Gupta Answered 6 years ago

Fisher exact test is for r x c contingency table. You must make the comparison on chi-square statistic not to an asymptotic chi-square distribution but the exact permutation distribution of the chi-square statistic for the accurate results.

By Nancy Trivedi Answered 7 years ago

If you have small samples, then, optimal data analysis is ideal for it. It provides alternatives for chi-square. You can refer to APA book with software: Yarnold and Solystik (2005). Optimal data analysis. American Psychological Association. 
Also, look into the journal whose link is attached below: http://odajournal.com/.

By Amrita Goyal Answered 7 years ago

You can go to Procedures Applicable to Categorical Frequency Data. If the Fisher Exact Probability Test for Tables is Larger than 2×2
2×3 2×4 3×3

Your Answer