Statistical test for multiple raters

#1
Hi all, I'm having trouble finding the right statistical test to use to analyze the following situation. I was wondering if anyone has an idea!

There are two independent cohorts, 1 and 2, who received different surgical procedure. I want to compare cosmetic outcomes of these cohorts. To do this, there are 4 independent raters who are rating all cohort participants on cosmetics using an ordinal scale from 0-15. I have done an intraclass correlation to test for inter-rater reliability, but now I want to compare cohorts 1 and 2 for cosmesis and utilize all the data from these 4 raters. I don't think a Mann Whitney U test is accurate because that treats the 4 ratings for each person in a cohort as a separate independent sample. Anyone have ideas on what I can do?

Let me know if I can provide more information on the scenario. Thanks!
 

Karabiner

TS Contributor
#2
there are 4 independent raters who are rating all cohort participants on cosmetics using an ordinal scale from 0-15.
Are there any experiences / discussions / references studies which indicate that perhaps these variable can be treated as interval scaled?
I have done an intraclass correlation to test for inter-rater reliability,
What were its results?
but now I want to compare cohorts 1 and 2 for cosmesis and utilize all the data from these 4 raters.
What exactely do you mean by this? Had there been any plans concernig this beforehhand?
What about aggretation of those 4 ratings?

With kind regrads

Karabiner