I have used Fisher's Exact rather than Chi-square due to having some cells with small sample sizes, but I have no clue how to report this? In the output, I am given the Fisher's test value (14.510) and the statistical significance (.042) That's it - no df etc.

As I would usually report a chi square test as x2 (df, n=participants) = 14.510, p=.042 along with a phi or cramer's v value, how would you suggest I report the results for a Fisher's Exact Test?

Finally, would I still report Phi/Cramer's V?


I am not familiar with the Fisher's test output in SPSS, but in other computing software's you will receive things like test-statistics, p-values, confidence intervals, and odds ratios.

I think odds ratios are a good thing to report because it will show the difference between the groups you are looking at (and maybe the confidence interval of the odds ratio). I think generally, it depends on who you are reporting to. If you are reporting to more math literate people, the more information about the test (even your SPSS output) is satisfactory. However, otherwise I would simply tell them that based on the data you were given (and the low p-value) that there is significant evidence that the groups (or whatever you are testing) differ from one another -- and you can mention in what direction.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.