Proc Mixed

[agoerke]

Using proc mixed to analyze price prediction data and seeking some assistance in the repeated clause.

given that

repeated id=

   specifies the defining variable of the (in my case CS) matrices

we are interested in the effect of thr group statement... it is our understanding that in a situation where the group statement defines a between subject variable such as

group = gender

   both male and female subjects would now be allotted their own R

the issue is that we are also interested in group on a within subject effect (lets say trials) and are looking to evaluate matrices within the subject matrix

As all of a single respondent's answers are related wedo not want 0's assumed in the off diagonals within the ID matrix

My co-authors and I are considering something of the following

   group = subjectid*trials

OR

   group = trials(subjectid)

Anyone out there familiar with this problem/aware of a solution?

[SteveDenham]

You are definitely on the right track.  The last two group= statements are equivalent, when it comes to constructing the R matrix.  It will be block diagonal, with each block being a CS submatrix referring to the subject-trial combination.  I am curious about the full design--it may be that a doubly repeated measures using type=UN@CS might be an approach to consider.  However, the best specification is all dependent on the design.  More info would probably clear this up pretty quickly.

Steve Denham

[SteveDenham]

Overnight I thought of something.  Does the variable trials index repeated observations?  If so, then perhaps it is as simple as changing the type= from CS to CSH.  Still, more information on the design would make this much easier.

Steve Denham