SEs for fixed effects affected by solution in random statement in proc mixed

[ehmst3]

Hello,

The inclusion of solution option in random statement in proc mixed affects the SEs of the fixed effects, which is very surprising to me. Can anyone provide insight?

General code:

proc mixed data=data;
class sub;
model AVAL = BL TIME TIMEK /solution cl ddfm=KR;
random int time timek / subject=sub type=un;

run;

Depending on whether solution option is included in random statement, fixed effects results (SEs) are different. Effect is small, but I don't see how explicit estimation of subject specific random effects (EBLUP) would change the estimation of fixed effect variance. In the parameterization in SAS help, this would appear to affect the estimate of V. 

[SteveDenham]

Good question.  I have a hunch you will need to take this up with Tech Support.  When you get an answer, please post it here.  This is the kind of unexpected behavior that some of us need to consider.

SteveDenham

[ehmst3]

Thanks for the suggestion. I have reached out to Tech Support and I will post the response.

[lsandell]

Hi there, did you ever receive an answer from Tech Support? I had the same question, and don't want to post again if you already found the solution. Thanks!