Is it right to use kwallis here?

[Tallys Feldens]

Dear community,

I have a big dataset from the same population. This population is divided in four exclusive groups A, B, C and D. The amount of people in each group is somewhat balanced. All observations have an ordinal categorical variable W very right-skewed that has 4 categories 1, 2, 3 and 4. The histograms for each group seem practically the same.
I wanted to test whether the distribution into the categories 1-4 for the variable W is different according to the groups A-D. It is not clear to me if kwallis is right to my case. What are your opinions on this subject?

[Felix Bittmann]

I think you can use kwallis or ologit, the difference between the commands are probably small. If they agree with your graphical assessment you have solid evidence.

[Carlo Lazzaro]

Tallys:
welcome to this forum.
As an aside to Felix's helpful reply, you do not tell is if you are dealing with a regression (where the dependent variable is a multinomial ordered one) or else.

[Tallys Feldens]

Thank you Carlo for the kind answer. It was just a statistical test for distribution; I was not meant to regress anything in this context. Thanks again