I have data from an infection study that I did and I'm trying to figure out if the differences I have observed are significant or not.
I have infections with 4 different bacterial strains, with 10 mice infected per strain (40 mice total). Then a colleague blindly scores the inflammation induced by the bacteria on a scale of inflamed, mildly inflamed, and not inflamed. I understand that this creates data that is qualitative and ordinal, since the inflammation is ranked.
I'm trying to figure out if I should be using a Fisher's exact test or a chi-squared test (or something else?), as well as what the minimum group size should be. I already observe big differences comparing the wild-type infected mice to my mutant infected mice with only 10 mice per group.
When I run a chi-squared test in Graphpad it tells me that there are significant differences but the calculations are only valid when all expected values are greater than 1.0 and at least 20% of the expected values are greater than 5. Can anyone suggest what to do? I'd like to use the least number of mice possible.