Date:         Mon, 5 Aug 1996 17:19:46 +0200
Reply-To:     Hans-Peter Piepho <piepho@WIZ.UNI-KASSEL.DE>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Hans-Peter Piepho <piepho@WIZ.UNI-KASSEL.DE>
Subject:      Re: SAS PROC MIXED and random coefficients
Comments: To: shi@UMBC.EDU

>IMHO,
>
>SAS PROC MIXED is not able to deal with real random coefficients
>problem in a regression setting. It considers
>
>      y=X*beta +Z*gamma + error
>
>and beta -- fixed coefficients; gamma -- random coefficients.
>
>One of its key assumptions is
>
>       E(gamma)=0
>
>which is stupid if you consider a random coefficient regression.
 
I am a bit late, because our server was down. Also, for the same reason, I
do not know if this has already been answered by SAS-L.
 
How about this reparameterization:
 
y=X*beta + Z* (gamma-E(gamma) ) +  Z*E(gamma)  + error
 
 = X*beta + Z*delta              + intercept  + error
 
delta = gamma - E(gamma) has zero mean
 
intercept = Z*E(gamma)
 
This can be fitted with MIXED.
 
Hans-Peter
_______________________________________________________________________
Hans-Peter Piepho
Institut f. Nutzpflanzenkunde  WWW:   http://www.wiz.uni-kassel.de/fts/
Universitaet Kassel            Mail:  piepho@wiz.uni-kassel.de
Steinstrasse 19                Fax:   +49 5542 98 1230
37213 Witzenhausen, Germany    Phone: +49 5542 98 1248