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 on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: Selecting a sample to compromise between significant size and geographical dispersion


From   Steven Samuels <sjsamuels@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Selecting a sample to compromise between significant size and geographical dispersion
Date   Fri, 16 Sep 2011 09:20:23 -0400

Partha,  Your problem is a fairly standard one, with some wrinkles, and the standard solution begins with optimal stratum formation (which can include take-none strata) and optimal allocation of observations to strata.  The calculations minimize estimation error for a bound on cost or vice-versa. Your iterative scheme ignores estimation error and so will be non-optimal. I strongly suggest that you consult a good sampling text or a sampling statistician instead of spending your  time on an inferior, possibly disastrous, design.  

Steve

On Sep 16, 2011, at 3:34 AM, Partho Sarkar wrote:

Thank you, Steven.  The program looks promising at first glance-I will
look at it in detail later.  For the time being, I think I can put
together an iteration method within Stata.  I will report back here on
this.

Regards
Partho

On Fri, Sep 16, 2011 at 3:17 AM, Steven Samuels <sjsamuels@gmail.com> wrote:
> 
> Partho-
> 
> The "stratification" package in R might meet your needs.
> 
> Steve
> 
> On Sep 15, 2011, at 4:42 AM, Partho Sarkar wrote:
> 
> To clarify further, this is not really a case of defining a population
> by the size of units.  What I am trying to do is to get a country
> level price-index as an alternative to the official index (which is
> also obtained from a sample of markets across the country, selected
> with similar considerations).  The only reason to impose a cut-off is
> that I have found (as one might expect)  that including EVERY market
> reporting a price at every date results in an index which is much more
> volatile compared to the official index.
> 
> Partho
> *

*
*   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/


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