I have read quite a bit about this procedure and I am now trying to estimate my first mixed model.
Here's the data that I have, and would like to get some help to estimate a model.
1. Each physician has 5 patients/records in the data (100 physicans for a total of 500 records/rows)
2. Physician level attributes are Phy_ID phy1 phy2 phy3
3. Patient level attributes are Pat_ID pat1 pat2 pat3 DEP
I would like to estimate a model that would help me understand the impact of the various physician and patient level attribute on DEP (continuous and norma).
Would this be a proper way to specify my model?
/model DEP = phy1 phy2 phy3 pat1 pat2 pat3;
/random = int phy1 phy2 phy3 pat1 pat2 pat3 / type = un sub=phy_ID;