groups comparaison

#1
Hello,

As part of an article review, a reviewer asks me to check a statistical problem but I don't know how to do it. I am doing group comparisons with widely varying numbers (10 - clinical group to 230 - control group). For continuous data, I did Kruskal-Wallis tests with Dwass-Steel-Critcholow-Fligner posthoc tests. How can I control for group size?

Also, for categorical data, I am told that Chi2 is not the most appropriate test for 6-group comparisons. What test can I use?
 

Karabiner

TS Contributor
#2
How can I control for group size?
Why should that matter? The respective group sizes are included in the formula,
and there is no indication that unequal sizes matter. What exactly did the reviewer
aks you?
Also, for categorical data, I am told that Chi2 is not the most appropriate test for 6-group comparisons. What test can I use?
Did he tell you that it not appropriate in general, or did he tell you that in your
particular case it is not appropriate? And what do you mean by "categorical data",
how many categories?

With kind regards

Karabiner