Use proc GLIMMIX, this is a new procedure in SAS version 9.13, if you don't have 9.13, you can download this procedure and its user guide from support.sas.com and search for 'GLIMMIX'. The syntax is very similar to proc MIXED, or proc GLM, but the graphs are much better.

the link for Download:

http://www.sas.com/apps/demosdownloads/sasstatglimmix_PROD__sysdep.jsp?packageID=000353&jmpflag=N

[此贴子已经被作者于2007-5-16 1:52:00编辑过]

*This is the data:

data origdata;
input ttime1 ttime2 ttime3 choice @@;
datalines;
16.481 16.196 23.89 2 15.123 11.373 14.182 2
19.469 8.822 20.819 2 18.847 15.649 21.28 2
12.578 10.671 18.335 2 11.513 20.582 27.838 1

...;

RUN;

*This is how you create the array dataset.

data newdata(keep=pid decision mode ttime);
set origdata;
array tvec{3} ttime1 - ttime3;
retain pid 0;
pid + 1;
do i = 1 to 3;
mode = i;
ttime = tvec{i};
decision = ( choice = i );
output;
end;
run;

*This is PROC MDC for mixed logit.

*You are saying the coefficient is normally distributed. You can also use UNIFORMPARM= or LOGNORMALPARM= as alternatives.

proc mdc data=one type=mixedlogit;
model decision = ttime / nchoice=3
mixed=(normalparm=ttime);
id pid;
run;

GOOD LUCK !!!