appropriate nominal-by-nominal in small samples


New Member
I was just reading your forum, and wondering if it is appropriate to do a chi-squared test with post-hoc if 44% of my cells have a count of <5.
I have 3 x 3 for my iv x dv.
All are categorical (personality group = iv, blood group = dv).
The chi squared p=0.015, but I get a warning about the low counts.
What would be the best way to analyze this, and other similar data?
Thank you,