Date:         Fri, 13 Feb 2004 14:19:06 -0500
Reply-To:     Paul Kononoff <pjk2@MAPLE.UNH.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Paul Kononoff <pjk2@MAPLE.UNH.EDU>
Subject:      stuck in GLIMMIX, count or %?

Hello all:

I'm stuck and am in need of some advice.

My question is as follows:

Because my dataset is not balanced (unequal observations per treatment), I
do not want to test the actual counts between treatment but rather the
proportion of counts between treatments using GLIMMIX.

I am not sure if the treatment (denoted GT in model) P-values I see in the
in OUTPUT are testing the number of counts or testing the percentage of
counts within treatment.

My code is as follows:

data one;

INPUT CASE GT $ SEX $ KLOT CARWT YG QG $ QGDET $;

data onea;

set one;

QGa = QG;

expt =1;

run;

%glimmix(data=onea,

stmts=%str(

class GT YG KLOT SEX;

model QGa = GT YG SEX SEX*GT / solution;

random KLOT;

lsmeans GT / cl diff;

),

error=binomial

)

run;