st: interpreting results from perturb

[Lloyd Dumont <lloyddumont@yahoo.com>]

Hello.
I am running some estimates that include two right-hand
side variables that are EXTREMELY highly-correlated.  It also includes some interactions involving
these variables.  I end up getting
interesting results which are almost assuredly a product of instability
relating to colinearity.  But, the
dataset is pretty large (n > 30,000), so it is possible that there is enough
independent variation to be generating these results legitimately.  It seems like the -perturb package is a good
way of getting at this.
The problem is that I'm not sure what the -perturb-
output means, and I am having trouble finding a good explanation or good
examples for interpreting the output from -perturb-.  I see that someone posted this question to
the list a few years ago, but it does not appear it was ever answered.
This is the most useful thing I could find, but it is not
very detailed re: interpretation.
http://home.wanadoo.nl/john.hendrickx/statres/stata/perturb/expl.html
(I am also curious as to whether -perturb- is accounting
for the clustered nature of the SEs that I am asking for when estimating the
model, but this question might be easily answered once I understand the output
I'm looking at.)
Might anyone be able to point me to something useful in
this regard?
Thank you for your help.  Lloyd

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/