Proc Mixed procedure: Standard error of the level-1 residual is 0?

[hsuhy0914]

Hi, SAS experts,

I have used the following codes to analyze my data:

proc mixed data=ABC covtest noclprint noinfo noitprint method=reml;
title3 'TOEP time and time2';
class id ;
model fam_rou=time_c time2_c/ solution notest ddfm=kr ;
random intercept time_c time2_c/ subject=id ;
repeated / type=TOEP subject=ID r;
run;

In this analysis, the with-individual residuals covariance matrix was modeled as the toeplitz structure. However, the SE of the residual (sigma2) was 0, leading no z-value and p value. 

Does anyone know the reason why I got SE=0 and have any suggestions for addressing this issue? I appreciate it.

[Rick_SAS]

In general, an STDDEV or STDERR of 0 indicates that data are constant or constant for some level or strata. For your REPEATED statement, a SE of 0 could indicate that each subject has a constant response. (Or there is only one obs per subject.)