Roger Newson <[email protected]> asks why -e(b)- and -e(V)- are not
explicitly mentioned in the 'Saved results' section of -[SVY] svy-:
> I was investigating the -svy- part of Stata 10 yesterday, and checked
> the results returned by -svy-, both in the on-line help for -svy- and in
> the manuals under -[SVY] svy-. I was alarmed to notice that, in the
> "Matrices" subsection of the Returned Results, there was no mention of
> either -e(b)- (the row vector of parameter estimates) or -e(V)- (the
> matrix of covariance estimates). Both of these returned results are
> mentioned in the corresponding section of the Stata 9 manuals, although,
> of course, the Stata 9 on-line help does not mention returned results.
> However, on running some Examples from the manuals, I found that -e(b)-
> and -e(V)- are still returned. This was reassuring, as both of these
> results are assumed to exist by my SSC package -parmest-.
>
> Is there any particular reason why -e(b)- and -e(V)- are no longer
> mentioned in the documentation? Has there been some kind of oversight
> that nobody else has noticed? Or is there a hidden long-term agenda, in
> which these returned results have been de-publicized and de-documented
> in preparation for withdrawing them at some time in the future?
No hidden agenda, just trying to be efficient with our descriptions.
The following is mentioned at the end of the 'Saved results' section:
svy also carries forward most of the results already in e() from
command.
The results -e(b)- and -e(V)- are implied here.
After reading my response a few times, we've decided to add them back.
This will only affect the on-line help file for -svy- for Stata 10.
--Jeff
[email protected]
*
* 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/
