Re: st: nonparametric regression with sampling weights

["Newson, Roger B" <r.newson@imperial.ac.uk>]

I don't know what questions you intend to answer. However, a rank stats 
package that allows sampling probability weights is the -somersd- 
package, downloadable from SSC.
The -somersd- package estimates a wide range of rank statistics, which 
are versions of Kendall's tau-a, Somers' D, or Theil-Sen median slopes, 
allowing the possibility of sampling probability weights and/or 
restriction to comparisons within strata. Apart from the .pdf manuals 
distributed with -somersd-, there are 3 Stata Journal references 
referred to in the on-line help, which can be downloaded from
http://www.stata-journal.com/archives.html

There is also an example presented at a Stata User Meeting (Newson, 
2006), which can be downloaded from
http://ideas.repec.org/s/boc/usug06.html

I hope this helps. BEst wishes

Roger


References

Newson R. 2006. On the central role of Somers’ D. Presented at the 12th 
UK Stata Users’ Group Meeting on 11-12 September, 2006.
Download from
http://ideas.repec.org/s/boc/usug06.html

Roger B Newson BSc MSc DPhil
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton Campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
UNITED KINGDOM
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: r.newson@imperial.ac.uk
Web page: http://www.imperial.ac.uk/nhli/r.newson/
Departmental Web page:
http://www1.imperial.ac.uk/medicine/about/divisions/nhli/respiration/popgenetics/reph/


On 17/02/2010 21:18, Andaleeb Alam wrote:
> Hi,
>
> I am working with Pakistan Standard Living Measurement Survey (it's a
> household level survey with a two-stage stratified sample design). I am
> interested in running a kernel weighted linear regression. From what I have
> tried, Stata does not allow the use of analytic weights if I am using
> svy:reg. What do you recommend I should do? Transform my data by my kernel
> weights and then do svy:reg? I have come across papers that do non
> parametric estimations using survey data so I am assumig that there must be
> some stata command that lets you do this?! If there is no such Stata command
> or my suggestion to transforms the data is incorrect, please recommend what
> can I do?
>
> Thanks
>
> Andaleeb
>
> *
> *   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/