Hey,

I have some questions about which test I can or have to use. Which test do I use to only identify if there is a relationship between those variables? I only want that to know, for the identification of confounding factors.
  • Ordinal dependent categorical variable <-> ordinal independent categorical variable
  • Ordinal dependent categorical variable <-> nominal independent categorical variable

Can I use for both cases the Pearsons chi-square test?


I also have:
  • Dependent numerical variable <-> ordial independent categorical variable

Can I use for this the Kruskal Wallis H Test?
  • Dependent numerical variable <-> nominal independent categorical variable

Can I use for this the One-way Anova Test?

However, I don't think any of my variables is normal distributed and I am not sure about equal variance. Can I still use these test? If not, which ones?

BTW I am using R commander

Thanky you!