Krysty,
First, be careful of templates without studying the assumptions of the
procedure and what other tests may be available to you. In this case,
Dunnett assumes independent observations, equal variances, etc..
It's also a good idea to plot data and print summary stats for each
group level.
Robin High
University of Oregon
* simple "balanced" example;
DATA indat;
input Type $ y;
datalines;
a 2.7
a 1.0
a 1.5
..
d 4.2
d 3.8
d 2.5
;
proc print; run;
* check sample sizes and variances;
proc tabulate data=indat noseps;
class type; var y;
table type, y*(n*f=3.0 mean*f=5.2 var*f=8.3) / rts=11;
run;
goptions reset=all cback=white;
symbol i=hilo value=dot h=1 color=blue;
proc gplot data=indat;
plot y*type / noframe haxis = 0 to 8 by 1;
run;
proc glm data=indat;
class Type;
model y = Type / solution;
means Type / dunnett('d');
lsmeans Type / pdiff=control('d') cl adjust=dunnett;
run; quit;
-----Original Message-----
From: SAS(r) Discussion [mailto:SAS-***@LISTSERV.UGA.EDU] On Behalf Of
Munnster
Sent: Monday, October 22, 2007 8:40 AM
To: SAS-***@LISTSERV.UGA.EDU
Subject: Major SAS help - Dunnetts Test
Hello!
Im very very new to this SAS program and I need to run a Dunnett's
Test. I was wondering if it were at all possible for someone to
provide with a basic template of SAS program where I could potentially
add my data points for this test.
Thanks so much,
Krysty