RE: st: AW: confidence intervals, saved results, pweights

["Martin Weiss" <martin.weiss1@gmx.de>]

<>

For the "manual" approach, also see Maarten`s
http://www.stata-journal.com/article.html?article=st0137



HTH
Martin

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Steve Samuels
Sent: Samstag, 17. Juli 2010 18:49
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: AW: confidence intervals, saved results, pweights

That's true of matrices e(b) and e(V) which are returned by every
Stata estimation command, not just -svy- commands.  It's irritating.

You can:
matrix list e(V)
matrix v = e(V)
di el(v,1,1)

As to your first question, -parmest- is the best choice, as Martin
pointed out, but you can do it from first principles.  You also don't
need to refer to the matrices, because Stata builds in _b[] and _se[]
system variables. See the -help- for "system variables".

*******************
sysuse auto, clear
svyset _n
svy: mean weight trunk
matrix v =e(V)
matrix list e(V)
di sqrt(el(v,1,1))
di _se[weight]
local bound_wt = _se[weight]*invttail(e(df_r),.025)
gen llim_wt = _b[weight]-`bound_wt'
gen ulim_wt = _b[weight]+`bound_wt'
di llim "   " ulim
*********************





On Sat, Jul 17, 2010 at 11:29 AM, Martin Weiss <martin.weiss1@gmx.de> wrote:
>
> <>
>
> Try Roger`s -ssc d parmest-:
>
>
> *************
> capture which parmest
> if _rc ssc install parmest
>
> webuse fpc, clear
> svyset [pw=weight]
> svy: mean x
> parmest, saving(myfile, replace)
>
> u myfile, clear
> l
> *************
>
>
>
> HTH
> Martin
>
>
> -----Ursprüngliche Nachricht-----
> Von: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Ben Zipperer
> Gesendet: Samstag, 17. Juli 2010 17:24
> An: statalist@hsphsun2.harvard.edu
> Betreff: st: confidence intervals, saved results, pweights
>
> confidence intervals of variable means in survey data with pweights.
>
> As I understand it, -svy: mean- will show the confidence interval, but
> only returns the variance and mean in its e() results. Is it
> recommended that I simply use these e() results to calculate my own
> confidence interval (to match the one reported but not saved by -svy:
> mean-), or is there a command like -ci- that saves as macros the
> confidence interval results with pweighted survey data?
>
> Additionally, after -svy: mean- why can I display the elements of some
> variance matrices but not others? For example,
>
> . webuse fpc, clear
>
> . svyset [pw=weight]
>
>      pweight: weight
>          VCE: linearized
>  Single unit: missing
>     Strata 1: <one>
>         SU 1: <observations>
>        FPC 1: <zero>
>
> . svy: mean x
> (running mean on estimation sample)
>
> Survey: Mean estimation
>
> Number of strata =       1          Number of obs    =       8
> Number of PSUs   =       8          Population size  =      27
>                                    Design df        =       7
>
> --------------------------------------------------------------
>             |             Linearized
>             |       Mean   Std. Err.     [95% Conf. Interval]
> -------------+------------------------------------------------
>           x |   5.448148    .723725       3.73681    7.159486
> --------------------------------------------------------------
>
> . di el(e(V_srs),1,1)
> .54734078
>
> . di el(e(V),1,1)
> matrix operators that return matrices not allowed in this context
>
> thanks,
> Ben
> *
> *   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/
>
>
> *
> *   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/
>



-- 
Steven Samuels
sjsamuels@gmail.com
18 Cantine's Island
Saugerties NY 12477
USA
Voice: 845-246-0774
Fax:    206-202-4783

*
*   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/


*
*   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/