Statalist The Stata Listserver


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

Re: st: Bootstrap with strata and sample sizes less than N


From   "Austin Nichols" <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Bootstrap with strata and sample sizes less than N
Date   Sat, 8 Apr 2006 15:38:07 -0400

Looks like doing your own sampling routine is easier than it might
have been yesterday:
. ssc install gsample
. gen nfrstr=10 if stratum==1
. replace nfrstr=20 if stratum==2
. gsample nfrstr, str(stratum)

On 4/8/06, Mike Lacy <Michael.Lacy@colostate.edu> wrote:
> I'm interested in getting a bootstrap distribution of the difference
> in the value of a statistic across two groups, with the
> sample size for each group being less than what prevails in the full
> sample, and the statistic being computed by a user-written program.
> Let's say, for example, that the total sample is 1000, with N1 = 400
> in group 1 and N2 = 600 in group 2, and let's say that I want
> bootstrap samples with n1 =10 from group 1 and n2 = 20 from group 2,
> in order to compute the difference in ranges across groups. Now, if
> n1 = N1 and n2 = N2, the -strata- option on
> -bootstrap- would do it, but I'm interested in n1 < N1 and n2< N2.  I
> could imagine using -simulate-, and doing my own sampling routine
> within the simulated program, but I was hoping for a simpler solution.

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index