I finally figured out how to get my version of SPSS to report Fisher's Exact numbers...that took several days! Most places on the internet either give you a Fisher's Exact calculator (that don't have enough cells for my data) or tell you that you need the Exact module...this is not true!!
My question is this. If I have a cell that has a frequency of zero, will the Fisher's Exact test still work? I know that the Fisher's Exact can report on cells with frequencies <5, but would zero be acceptable? Reason I'm asking is I have a 2x7 contingency table. The seven rows are to show how many people from my population responded 1-7. The two columns are for yes/no.
Some cells have zeros in one of the columns but numbers in the other, so I was wondering if I put '0' in would it be ok.
I would appreciate your help on this!
Last edited by August 15; 08-23-2010 at 10:52 PM.
Advertise on Talk Stats