From sjhsamuels@earthlink.net To statalist@hsphsun2.harvard.edu Subject Re: st: sample adjustment by substitution instead of weighting Date Fri, 24 Apr 2009 17:47:43 -0400 (GMT-04:00)

Ways of randomly selecting a sample so that estimates from it match population characteristics are covered in Tille' (2006, Chapter 8).  The method you describe is not one of them.  How do the people who advocate it propose to construct the probability weights?

-Steve
Tillé, Y. (2006). Sampling algorithms (Springer series in statistics). New York: Springer.

-----Original Message-----
>From: Dirk Enzmann <dirk.enzmann@uni-hamburg.de>
>Sent: Apr 24, 2009 4:08 PM
>To: statalist-digest@hsphsun2.harvard.edu
>
>Did anyone of the StatList community ever encounter a method to adjust a
>sample to the population (according to some criteria such as the age or
>gender distribution) that uses a method that might be called "adjustment
>by substitution" (or similar)? The basic idea is to delete some cases
>and replacing them randomly with duplicates *of the same sample* (!)
>such that the number of cases of the original sample remains unchanged
>and that the sample is adjusted to the distribution of certain variables
>of the population (I don't mean that you try to substitute a
>missing/lacking case by resampling from the population)?
>
>I ask this because I have serious doubts whether this procedure is
>reasonable (creating populations weights seems to be much better), but I
>would nevertheless like to know whether this procedure is/has been
>actually used and/or whether there exists a publication that discusses
>this method.
>
>Yours,
>Dirk
>
>*************************************************
>Dr. Dirk Enzmann
>Institute of Criminal Sciences
>Dept. of Criminology
>Schlueterstr. 28
>D-20146 Hamburg
>Germany
>
>phone: +49-(0)40-42838.7498 (office)
>        +49-(0)40-42838.4591 (Mrs Billon)
>fax:   +49-(0)40-42838.2344
>email: dirk.enzmann@uni-hamburg.de
>www:
>http://www2.jura.uni-hamburg.de/instkrim/kriminologie/Mitarbeiter/Enzmann/Enzmann.html
>*************************************************
>*
>*   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/

*
*   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/