Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: RE: RE: Using command svy glm to obtain risk ratios
From
Steve Samuels <[email protected]>
To
[email protected]
Subject
Re: st: RE: RE: Using command svy glm to obtain risk ratios
Date
Tue, 18 Feb 2014 17:28:47 -0500
Jose Maria asked:
"would it be possible to use *predict pr, pr* instead of *margins*?
If there are only a few discrete covariates, then you can graph or table
all the predicted probabilities. Otherwise -margins- is preferable, I think,
especially when followed -marginsplot-. My own strategy, when the choice is mine, is
to fit -svy: logistic-,This will (almost) always converge with no
difficulties. Then I look at -marginsplot- of predicted probabilities; I
add two-factor interactions and see what happens to the plot
For most studies, I prefer to describe effects in terms of linear
differences in predicted probabilities, even if the underlying model is
-logistic- with no or few interactions.
Steve
[email protected]
On Feb 18, 2014, at 8:49 AM, jose maria pacheco de souza <[email protected]> wrote:
Em 18/02/2014 03:41, Steve Samuels escreveu:
> I replied to Jose Maria privately with copies of the do files. The paper
> was:
> Waltermaurer, E, M Butsashvili, N Avaliani, S Samuels, and LA McNutt.
> 2013. An examination of domestic partner violence and its justification
> in the Republic of Georgia. BMC Womens Health 13, no. 1: 44.
> and is at:
> http://www.biomedcentral.com/1472-6874/13/44
> It contained only the -svy: glm- analyses.
>
> To correct an error: In some instances of -svy: poisson-, the CI for a
> predicted *risk*, not the estimated RR, exceeded 1.0.
>
> log relative risk regression was the choice of my co-authors. My
> personal choice would have been -svy: logistic-, followed by -margins-
> for interpretation of risks.
>
> Steve
> [email protected]
>
Steve was very kind.
Thank you.
Now, maybe following the thread: for the interpretation of risks after *svy logistic*, would it be possible to use *predict pr, pr* instead of *margins*?
josé maria
--
Jose Maria Pacheco de Souza
Professor Titular (aposentado), Professor Senior
Departamento de Epidemiologia/Faculdade de Saude Publica, USP
Av. Dr. Arnaldo, 715
01246-904 - S. Paulo/SP - Brasil
fone (11)3768-8612
www.fsp.usp.br/~jmpsouza
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/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/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/