Ken, thank you for your comments regarding use of MANOVA. I read some
primary literature on the two approaches and decided there were enough
pros and cons on both sides. But, given the matsize limits might be
sidestepped with MANOVA, I'll go find a book on the topic and see if I
can't make better use of Stata's "MANOVA" command. I called SAS sales
regarding their limitations. They don't have a matsize limit per se,
and in version 9 some of their code is multi-threaded (potentially one
could take advantage of the Vanderbilt "VAMPIRE" parallel computing
system with this feature), but for PC use, the issue with computational
time and memory need (that is their limitation) was not trivial for SAS
either. In other words my planned data set would likely choke a 64 bit
workstation as well. So it's best I search for alternative analyses and
or experimental designs.
