Contigency table 4x4 - what test to use?

#1
Hi,

I am comparing the frequencies of different maturity stages (4 stages) between different treatments (4 treatments). If I am not wrong, to do that, I would need to do a 4x4 contingency table. As I have several zero in my results, I thought to use a fisher exact test instead of a chi square. Do you think this test would be more appropriate?

I am using SPSS; I only managed to do the fisher exact test if I compare two by two the results (so for example treatment A vs treatment B with the four different maturity stages). Is there a way to do a 4x4 fisher test directly? And if yes, how could I then know what treatments are responsible for the differences? (i.e., posthoc test?)

To be clearer, my results are presented to like this:

Treatment 1 Treatment 2 Treatment 3 Treatment 4
Stage 1/ 0 4 0 4
Stage 2/ 3 1 0 0
Stage 3/ 4 0 3 0
Stage 4/ 0 5 4 4

Thank you verymuch for your help,
Cheers,
N.
 

Karabiner

TS Contributor
#2
I am comparing the frequencies of different maturity stages (4 stages) between different treatments (4 treatments).
"Maturity stage" sounds like an ordinal scaled variable. Does your research question
refer to something like "in some treatments, stages tend to be later than in other
treatments"? In that case, you could perform a Kruskal-Wallis H-test.

With kind regards

Karabiner