Help using Base SAS procedures

Mixed model repeated measures: difference between lsmeans and estimate statements

Reply
Highlighted
Learner
Posts: 1

Mixed model repeated measures: difference between lsmeans and estimate statements

Hello,

 

I have noticed that when doing a MMRM using baseline value as a covariate, I find differences between the lsmeans and the estimate statements. Am I doing wrong the estimate parameters?? In the example below, using lsmeans the estimate at month 3 would be 7.4 while using estimate, the estimate is 19.4

 

data test;
input
subjid $ month $ baseline  postbaseline;
datalines;
1 1 10 11 
2 1 11 14
3 1 12 16
4 1 13 17
5 1 14 15
1 2 10 13
2 2 11 16
3 2 12 14
4 2 13 17
5 2 14 17
1 3 10 18
2 3 11 19
3 3 12 20
4 3 13 21
5 3 14 19
;
run;

data test2;
set test;
change= postbaseline - baseline;
run;

proc mixed data=test2 method=reml;
class  change month subjid;
model change=month baseline / ddfm=satterthwaite;
repeated month / subject=subjid type=un;
lsmeans month;
estimate "month 3"  intercept 1 month 0 0 1;
run;

Thanks a lot for the help!

 

 

 

Ask a Question
Discussion stats
  • 0 replies
  • 241 views
  • 0 likes
  • 1 in conversation