I am a little new to mixed models and proc mixed and was hoping someone could tell me if I am coding my Proc Mixed statement correctly. It converges and the results appear reasonable, but I would greatly appreciate input from someone a little more versed on the subject
Here is the intercepts only model
Yij = Y00 + Y10(gradeband)+ Y20(Ethnic) Y01(SES) + Y11gradeband (SES) (Ethnic) + uoj + rij
Gradeband is a level 1 variable that can take one of three values: 0 1 or 2
Ethinc is a dichotomous level-1 variable: Hispanic or non-Hispanic
SES is dichotomous level 2 variable: lowSESschool or highSESschool
schoolcode is just the identifier variable for the individual schools. Students are nested within schools (obviously)
Here is the Proc Mixed Statement
proc mixed data=schooldata covtest;
class schoolcode gradeband ethnic SES ;
model testscore = gradeband ethnic SES gradeband*ethnic*SES/solution
ddfm=contain;
random intercept /type=un sub=ethnic(schoolcode);
run;
Does this proc mixed statement look right given what I am trying to do? Thanks.