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: Restricting sample size
From
Maarten Buis <[email protected]>
To
[email protected]
Subject
Re: st: Restricting sample size
Date
Fri, 29 Nov 2013 11:48:17 +0100
On Fri, Nov 29, 2013 at 10:50 AM, Amal Khanolkar wrote:
> The Swedish group is very large relative to the two other categories. As this might bias results for potential differences in outcomes of interest between the three ethnic groups (non-Europeans being very much smaller then the Swedish group), I would also like to run the analysis by reducing the Swedish group from 394,050 to say just 50,000 or 100,000 subjects. How would I tell Stata to run the above analysis restricting to a random sample when ethncity_bi2x==1 (i.e. the Swedish group) ?
Why would the Swedish case be biased? The large sample size means that
your Swedish estimates are a lot better then the other estimates.
Reducing the quality of one estimate such that all estimates are
equally bad seems like a very unproductive strategy. Below is how you
could do that, but please think again if this is really what you want
to do.
gen weight = .
bsample 50000, weight(weight)
eststo: regress syst bwtgestage_sd age byear i.conscript_office bmi
height_cons ib2.education2 ib3.gestwk_catx2 [fweight=weight] if
multibirth==1 & sex==1 & ethnicity_bi2x==1, cluster (mor_lopnr)
-- Maarten
---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany
http://www.maartenbuis.nl
---------------------------------
*
* 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/