Bookmark and Share

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: After logistic regression, can I determine ORs and CIs for nonsignificant variables


From   Fernando Furquim <ffurquim@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: After logistic regression, can I determine ORs and CIs for nonsignificant variables
Date   Tue, 3 Dec 2013 10:22:59 -0600

Rachel,

I'm not sure why you are using a stepwise selection, nor am I very
familiar with when it's a good idea, but my understanding is that
Stata would iteratively add a new independent variable to the model if
it is statistically significant at the level specified. So if a
variable doesn't reach that threshold, it is not included in the model
and thus has no p/CIs/ORs. They are not in the model at all - and if
they were, the p/CIs/ORs of your other variables may be different than
what the stepwise selection is reporting in the end.

If you need to know how each variable is contributing to the model,
I'd just fully specify the model rather than rely on stepwise
selection. Then you can use whatever criteria make sense to you in
deciding what to include or exclude.

Fernando



On Mon, Dec 2, 2013 at 7:19 PM, Rachel Meserole
<rachel.meserole@jhmi.edu> wrote:
> Statalist,
>
>
> I am using logistic regression for predictive modeling.  I want to report the outcome in terms of ORs and CIs for all variables studied.
>
> stepwise, pe(0.05): logit outcome indepvar1 indepvar2 indepvar3 indepvar4..., or
>
> Because of the method of variable selection, only variables with a significant p value are included in the output.  The other variables must have nonsignificant p values and therefore CIs that cross 1.  I would like to know what those nonsignificant p values, CIs, and ORs are.  Is it possible to get this output?  Thank you so much!
>
>
>
>
> Rachel
>
> *
> *   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index