Chi Squared test?

#1
I am designing a research project to determine what bodily sensations individuals experience during emotions. Participants will indicate on a silhouette of a body locations where they feel sensations for a particular emotion (for example love might be felt in the chest area) and this will produce frequency data for body regions. I have 13 body regions in total.

I want to compare the data for different emotions to see whether any two emotions have bodily sensations which occur in the same regions and whether this is larger than would be expected by pure chance. The data is all category data with simple counts (frequency).

This is not two different populations, i.e. Male / Female, it is within a single population (although later I may break it down further).

Is a Chi Squared test the right way to do this? And if so is it a test of independence or homogeneity?

At present I am assuming an equal distribution of responses for all regions for all emotions. So I will have a data table with emotions in on direction and body regions in the other.

I then need to work out the next step, but can't seem to get Chi Squared results which make sense. When I have high numbers of results which occur in the same body region, I get a Chi Squared close to 0. My understanding is a 0 would support the null hypothesis that there is no relationship, but with imbalanced numbers I get higher Chi Squared, which seems counter intuitive?

I'm relatively new to statistics so any help gratefully received. Happy for you to assume no knowledge on my part, I won't be offended!

Thanks

David