Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: stratify question


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: stratify question
Date   Fri, 12 Sep 2008 07:10:25 +0100 (BST)

That sounds a bit like propensity score matching, type -findit
propensity score- and you'll find several programs that can help you.

-- Maarten

--- Sim.Oertel@t-online.de wrote:

> Dear Stata Users,
> 
> I have a quick question. I am working with two datasets. The first
> one
> (dataset "A") contains 3,500 cases, the second one (dataset "B")
> contains
> 150,000 cases. 
> 
> In dataset "A" variable "x" equals 1, in dataset "B" 0. For a
> pre-study, I
> would like to analyze how variable "x" influences the survival
> chances of a
> subject. Each dataset contains further variables, while the
> distribution of
> the values of these variables differs between them. For example,
> welfare can
> take values between 0 and 10. In sample "B" welfare is generally
> higher than
> in sample "A". Since I am not interested in the effect of welfare, I
> would
> like to draw a stratified sample of dataset "B" containing subjects
> with the
> similar quota of welfare than the subjects included in sample "A".
> The
> datasets contain many other variables. Therefore, I would like to
> enlarge
> this process on more than one variable.
> 
> Put differently, I would like to tell Stata: "Generate a sample "C"
> based on
> the dataset "B" which holds the same proportion of variable values
> for
> "welfare", "age", "location", . as sample "A"." 
> 
> Finally, I would combine the stratified sample of "B" and the full
> sample of
> "A" and run a logit model.   
> 
> Is there a simple way of doing this in Stata? How can I generate a
> stratified sample with Stata?
> 
> Thanks for your help
> Simon
> 
> *
> *   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/
> 


-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room N515

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------


      
*
*   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   |   What's new   |   Site index