Hi,
Hope someone could help me. I’m analysing school effect on many performance
variables among students, I’m using a longitudinal (5 years) mixed model with
student nested in a school to explain motivation over years versus school
effect.
I face the problem that many students (20%) change of school at year 2 or 3. My
question is, how can I perform a mixed model with a school code changing over
time in SAS.
Here’s the syntax I’m using in SAS.9:
proc mixed data=data METHOD = REML covtest ;
class year ;
model motiv = year schoolef year*schoolef / SOLUTION DDFM=bw ;
random intercept / SUBJECT=id;
repeated year/ subject=id(school) type=ar(1);
run;
Id: is the student longitudinal code
Motiv : Motivation (motivation scale of students) continuous variable
Schoolef: school effect. Continuous variable
School: code of school
Thank you in advance
Khalid Chenguiti
|