if I were to
.regress Y X1 X2.Xn [pweight=weight]
would my standard error estimates be compromised as Winship and Radbill
noted back then, or does pweight now correct for any problems with these
estimates brought about by the act of weighting?
There was a lengthy discussion of weighting on Statalist several months ago
and I think my part of it ended with much the same question: have the
issues raised by Winship and Radbill been addressed by Stata today? I
would think so, but if not a lot of effort has been put into things like
pweights and the -svy- commands that don't work! Do any long-time Stata
users happen to know when pweights were introduced? The Winship-Radbill
article is very good and I'd love to see them do an update of it that
discusses modern software and techniques.
