Re: Re: st: STATA and survey data

[Ronald McDowell <McDowell-R3@email.ulster.ac.uk>]

That's helpful Steve thanks.

If I've read the paper correctly STATA isn't able to (currently) produce residuals with eg. the svy: logistic command, which is why I was getting an error message, and
so I have to look at other diagnostic tools. It also looks from reading that other packages don't either, such as SAS and SUDAAN, but if anyone knows otherwise do let me know.

Cheers

Ron





---------------------------------------------------------------------------
Ron-

The recent book by Heeringa et al. (2010) uses Stata (note program
spelling, Statalist FAQ 8.2) throughout.  For diagnostics, see Rick
Valliant's  Stata Conference presentation: http://econpapers.repec.org/paper/bocdcon09/15.htm

With logit models, you can estimate ROC curves with Roger Newson's -
senspec- ("findit") See: http://www.stata.com/statalist/archive/2009-01/msg00689.html
.  Archer and Lemeshow have generalized the Hosmer-Lemeshow goodness
of fit test for logistic regression to survey data, and implemented it
in -svylogitgof- ("findit"). The usual techniques of adding
interactions and testing augmented against reduced models also apply.

Reference:
Heeringa, Steven, Brady T. West, and Patricia A. Berglund. 2010.
Applied Survey Data Analysis. Boca Raton, FL: Chapman & Hall/CRC.


Steve
sjsamuels@gmail.com

On Feb 2, 2011, at 4:46 AM, Ronald McDowell wrote:

Can anyone recommend any materials for using STATA in the analysis of
survey data (bar the manual?). I'm particularly interested in the key
areas around model fit, postestimation, examining residuals and the
like.

Thanks
Ron McDowell
University of Ulster


*
*   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/