Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

Re: st: one stage cluster with preliliminary stratification


From   Steve Samuels <[email protected]>
To   [email protected]
Subject   Re: st: one stage cluster with preliliminary stratification
Date   Wed, 30 Jan 2013 18:07:34 -0500

Subir, I have some other observations after rereading your post.

Unless all clusters are of similar size, the design you have
described-equal probability sampling of clusters- is not recommended.

In many populations, clusters range in size from large to small, with a
few larger ones and a greater number of smaller ones. In such
situations, equal probability sampling of clusters will over-represent
people who live in smaller clusters. The preferred design with unequally
sized clusters is sampling with probability proportional to size (PPS).

If you find yourself in this situation, then only post-stratification of
the population by approximate cluster size will permit less-biased
estimates of population characteristics.

Note that the FAQ ask for full bibliographic references, including book,
article, and journal titles.

Steve



> Subir
> 
> You told us what what analysis you wish to do, but if it's descriptive,
> it will be easy to add the fpc() option, if appropriate, to Stas's
> -svyset- statements. See the help for -svyset-.
> 
> 
> Steve
> 

> On Jan 30, 2013, at 9:30 AM, Stas Kolenikov wrote:
> 
> *** assuming n1, n2, m1, m2, N are contained in the identically named scalars
> 
> gen wgt = scalar(n1)/10 if stratum==1
> replace wgt = scalar(n2)/8 if stratum==2
> assert !missing(wgt)
> * cluster size does not matter
> 
> *** option 1: poststrata
> gen pstrata = 1
> gen popsize = scalar(N)
> svyset cluster [pw=wgt], strata( stratum ) poststrata( pstrata )
> postweight( popsize )
> 
> *** option 2: rescale weights
> sum wgt
> generate wgt2 = wgt*scalar(N)/r(sum)
> svyset cluster [pw=wgt2], strata( stratum )
> 
> -- 
> -- Stas Kolenikov, PhD, PStat (SSC)  ::  http://stas.kolenikov.name
> -- Senior Survey Statistician, Abt SRBI  ::  work email kolenikovs at
> srbi dot com
> -- Opinions stated in this email are mine only, and do not reflect the
> position of my employer
> 
> 

On Wed, Jan 30, 2013 at 6:32 AM, Subir Mitra <[email protected]> wrote:
> ONE STAGE CLUSTER WITH PRELIMINARY STRATIFICATION -I stratify population N (members living in clusters , which is known) into 2 strata and  randomly pick up 10 clusters from 1st stratum and 8 clusters from 2nd stratum (stratum population n1 & n2  and total clusters m1 & m2 in both stratum also known) and all members of the clusters are sampled.
> 
> Any guidance to me to find the svyset command in this case, assuming N, m1,m2,n1 and n2 known and I want to make use of it? (The problem is from Schaeffer et al 1996-328 problem 8.19)
> 
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/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/faqs/resources/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/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index