Hi all. I'm working with some survey data employing noninteger inverse-
probability weights and I need to do an mvreg as I'm interested in the
covariance matrix of the coefficients across equations. I was surprised to
see that mvreg doesn't accept pweights, only fweights and aweights. Work
done by my client in the past on this same dataset employed aweight in the
mvreg call, but that's not strictly correct.
I can't think of any theoretical reason why mvreg shouldn't accept pweights.
Am I insufficiently imaginative about theory problems, or did it just prove
impracticable from a programming standpoint?
STS
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
