Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: hypothesis testing/R2 with svy commands


From   "Nick Cox" <[email protected]>
To   <[email protected]>
Subject   RE: st: hypothesis testing/R2 with svy commands
Date   Sun, 7 Mar 2004 15:47:00 -0000

Richard Williams replied to Karen Matsuoka

> >So I am using the svylogit command for all my multivariate 
> analyses on
> >dichotomous outcome variables. But svylogit does not produce 
> an R2 value. Is
> >there another command that will produce this R2 value for 
> me? Or does Stata
> >have another measure--other than R2--that measures how much 
> variance a model
> >explains when using complex survey data? (I believe the 
> logit command *does*
> >return an R2 result, so I'm not sure why svylogit doesn't. Is there
> >something about complex survey data that makes R2 values 
> inappropriate?)
> 
> Another thing that struck me as really weird at first is that 
> svylogit 
> reports F and t statistics rather than chi-square and z 
> statistics.  But, 
> Stata generally has a reason for what it does, even if I don't always 
> understand it.  I don't see a specific discussion of R^2 in 
> the Stata 8 
> Survey Data manual, but on p. 28 it does say that most of the svy 
> estimation commands use "pseudo-maximum likelihood methods."  
> These are not 
> "true likelihoods" and hence "standard LR tests are no longer 
> valid".  Since Pseudo R^2 is computed from the likelihoods, 
> my guess is 
> that it is not considered valid in this case either.  As for 
> alternatives?  I don't know.

<exchange of SPSS nostalgia> 

I guess the guess about likelihood is likely right on target.

For more discussion, see 
http://www.stata.com/support/faqs/stat/rsquared.html

 
Nick 
[email protected] 

*
*   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 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index