Re: st: subpop with svy commands

["L Allen" <indiananonprofit@hotmail.com>]

Ritesh,
When I have run into this problem, I have created a new dichotomous 
variable.  So, if you create a variable =1 for being in category 1 and =0 
for all others and use this in combination with your subpop command that 
removes categories 3&4 from the point estimates, but keeps their sampling 
information for the variance estimates, the only categories left to be 
tabulated in the "all others" value of your new dichotomous variables are 
category 2 respondents.
Rich Clerkin



> From: "Ritesh Mistry" <riteshm@ucla.edu>
> Reply-To: statalist@hsphsun2.harvard.edu
> To: "statalist" <statalist@hsphsun2.harvard.edu>
> Subject: st: subpop with svy commands
> Date: Tue, 30 Sep 2003 18:40:28 -0700
>
> hello statalisters,
>
> i am using survey procedures to analyze data, and need to select a 
> subsample
> on which to conduct all my analysis.  i understand that the subpop option
> needs to be used and i have to create a variable which identifies the
> subsample.  i have done this but something unexpected happens.
>
> one criteria for exclusion from the subsample are responses that i am not
> interested in for certain variables (e.g. for variable A there are 4 [1, 2,
> 3, 4] categories, but i am only interested in those that responded either 1
> and 2). Although the variable indicating the subsample correctly identifies
> only those that responded 1 or 2, when i run any analysis using the subpop
> command (e.g. in svytab), i get a table with cells for all four categories,
> and the cells for categories 3 & 4 are empty.  what i expected was to get a
> table with cells for only categories 1 and 2.
>
> any advice on how to deal with this.
>
> as an alternative i designated categories 3 and 4 as missing so they would
> not be included in the analysis, but this creates another problem: the
> sample size becomes small enough so that there are some strata with only 
> one
> PSU (this requires that i combine strata, which i really don't want to do
> but may have too).  is it appropriate to designate categories 3 and 4 as
> missing?
>
> i am sure i am missing something.
>
> thanks!!
>
> ritesh
>
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
_________________________________________________________________
High-speed Internet access as low as $29.95/month (depending on the local 
service providers in your area). Click here.   https://broadband.msn.com

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/