Hi Michael: Thanks! Yes, this is clearly the way Stata treats it.
Are you also saying, however, that this is theoretically the way it
has to be? Supposing I have no strata (err, one big strata) and 5
post-strata, wouldn't it be conceivable to correct for the number of
known population units in each post-strata? If I keep working w/ this
kind of data, I guess it'll be time for me to pick up a book that
carefully derives each of the formulas behind svy.
Peter
On Thu, Sep 24, 2009 at 7:55 AM, Michael I. Lichter
<[email protected]> wrote:
> Peter,
>
> The FPC derives from sampling and has nothing to do with any
> poststratification adjustments you apply. If you have a single stage,
> clustered, unstratified design, you should have only one FPC value, equal to
> the sampling fraction. You do not want to calculate poststratum-specific
> FPCs.
>
> Michael
>
*
* 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/
