Multiple responses/variables in marco magree

[rola13]

Hi there,

I successfully followed this description for calculating a multiple rater Fleiss kappa value: https://support.sas.com/kb/25/006.html

My question now: is it possible to have more than one collumn with variables?

Collumns x,y,z instead of only y?

In the data I would like to process the subjects (s) were observed by the raters (r) 1 to 5 not only in one category "y" but severeal more (category "x" and "z").

Do I have to enter the data for each category individually? Or is there a way of calculating the Fleiss kappa per category in one go?

My individual code loks like this:

      data ratings;
        do s=1 to 1122;
        do r=1 to 3;
          input y;
output;
        end; end;
        datalines;
0
1
etc.
1
1
;
	  %inc "C:--------\magree.sas";
	  %magree(data=ratings, items=s, raters=r, response=y,	stat=nominal, options=counts)

Thank you!

[PaigeMiller]

Since the documentation speaks about a response variable (singular), it would appear that you will have to run the macro several times, one time for each response. Then you could merge the results into a single table if you want.

[rola13]

Oh, I will be busy for a while copying in 75 response variable collumns 😅 Thank you anyway for your answer!

[ballardw]

---

@rola13 wrote:

Oh, I will be busy for a while copying in 75 response variable collumns 😅 Thank you anyway for your answer!

---

You may want to search this forum and the documentation for the SAS Call Execute function. You can place information, such as variable names into a data set and create syntax using that function to call the macro once for each variable and providing other parameters as needed.

[PaigeMiller]

---

@rola13 wrote:

Oh, I will be busy for a while copying in 75 response variable collumns 😅 Thank you anyway for your answer!

---

Did you even try it with more than one response variable? Sometimes the documentation can be wrong.

[rola13]

Yes I treid but it results in a lot of errors.