varaince estimates for interaction of fixed and random effects in mixed models

[Ruhi]

Hi All

I am working on a project involving understanding interday and intraday varainces on a device and I found interday varaince to be less than intraday, which is our biggest concern.

So we took measurements on humans on 5 days, and 5 masurements on each day and each measurement gives us 12 outputs of the same measure at any given time. So I created a completely nested random effect model as follows to get all varaince estimates :

proc mixed data=required covtest cl;
class pt day scan  ;
model var= ;
random pt day(pt) scan(day*pt) ;
estimate "Mean" intercept 1/cl;
ods output covparms=Covparms estimates=EstimatedMean;
run;

To investigate why intraday varaince is higher than interday, I want to test if subgroups (say Male and female)  have equal interday and intarday varaince or how their variances are. Now I am not quite sure how can I get varaince estimates in these subgroups. Any help is greatly appreciated. Any other solution/trick to investigate  is welcome too.

Thanks soooo much for any help!!

[SteveDenham]

Why not introduce the subgrouping as a fixed effect, and include it in a group= option to the random statement.  For example, if sex were the factor, you might try:

proc mixed data=required covtest cl;
class pt day scan  sex;
model var= sex;
random pt day(pt) scan(day*pt)/group=sex ;
estimate "Mean" intercept 1/cl;
ods output covparms=Covparms estimates=EstimatedMean;
run;

Steve Denham