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: random sampling matching the characteristics of the sample
From
David Hoaglin <[email protected]>
To
[email protected]
Subject
Re: st: random sampling matching the characteristics of the sample
Date
Tue, 1 May 2012 09:54:10 -0400
Dear Andrea,
Did the reviewer explain the reason for wanting a sample of the x=0
observations that is comparable in size to the x=1 observations? As
others have commented, you would mainly sacrifice precision, so that
looks like a mistake. It is not unheard of for a reviewer to have a
misunderstanding. The consensus is that you should not subsample.
Samples that are "representative" of the x=0 data and the x=1 data may
well not match on age and size. Those differences are part of what
your regression model is adjusting for. You should consider whether
the coefficients of age and size are similar in the x=0 data and the
x=1 data (e.g., by fitting a model that allows separate coefficients
in the two subgroups --- that is, an interaction between x and age and
an interaction between x and size). Might that be what the reviewer
is actually concerned about?
David Hoaglin
On Tue, May 1, 2012 at 4:22 AM, Andrea Rispoli <[email protected]> wrote:
> Dear Stan,
> Thank you. This is the request of a reviewer. Would you recommend that
> I simply chose a random sample?
*
* 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/