My question is about how to analyse very small samples on contingency tables. I have a contingency table with observation for 2 groups (first variable) answering questions with multiple choices (second variable), the values in the contingency table are as low as 2, 1 and 0. and my total sample was as low as 2, 3.

Is there any chance I can statistically investigate if my answers are different between my groups?

I know that for small sample sizes we should use Fisher's exact test, but with this small sample the test never reject the null hypothesis.


Group/answer:  YES YES/SOLVED  NO 
group 1:        0       2      0
group 2:        3       1      0
  • 1
    $\begingroup$ Using Fisher's exact test is correct, but with a sample size so small, it seems that you are lacking statistical power to find significant differences. $\endgroup$ – Ertxiem Apr 30 at 15:16

Your Answer

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

Browse other questions tagged or ask your own question.