Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: -svy- commands with a pps sample vs. a simple random sample


From   Stas Kolenikov <skolenik@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: -svy- commands with a pps sample vs. a simple random sample
Date   Sat, 14 May 2011 08:44:16 -0500

On Sat, May 14, 2011 at 3:33 AM, Nick Cox <njcoxstata@gmail.com> wrote:
> Stas' initial statement is incorrect if the sense of "black box" is
> that you can't see the detailed code. The source code for
> -mm_sample()- is accessible once you have installed the -moremata-
> package. Example code below. There are calls to other functions.
>
> The user-written commands referred to, -moremata- and -gsample-, are from SSC.

Oh, I see (-help moremata_source-). The code in -mm_ups*()- is only
good for drawing one unit from with probability proportional to a
measure of size. If you are drawing more than one unit, you won't get
PPS, although your actual probabilities of selection will be
correlated with the measure of size. I suspect that the more you draw,
the farther you'll be from PPS. But this is a tangential issue; the
more relevant one is the more subtle explanation about what it is that
PPS is able to bring down: the largest residuals in the case when the
magnitude of residuals is correlated with size.

-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.
*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index