Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: bootstrap and model building


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: bootstrap and model building
Date   Fri, 14 Oct 2011 14:20:45 +0100

You don't give any references for this approach. It sounds like a very
roundabout way of getting the information that a direct bootstrap
gives you any way, albeit in different form. It has the extra
disadvantage that it wires in decision-making with fixed significance
levels. I could easily be missing some subtle differences here, but I
would be grateful to know what they are.

Bootstrapping stepwise is the stuff of nightmares.

Nick

On Fri, Oct 14, 2011 at 2:02 PM, Sunday Clark <sunday.clark@gmail.com> wrote:
> I saw model building approach using bootstrapping and I am wondering
> if it is possible to do it in Stata. I would like to bootstrap either
> a stepwise or a manual model building procedure to determine which of
> the factors considered for inclusion in the model should be retained
> in the final model. Ideally I would like to determine which factors
> are significant at p<=0.05 in more than 50% of the bootstrap samples.
> For example, stating that var1 was retained because if was significant
> in 90 of 100 bootstrap samples and var2 was not retained because it
> was only significant in 30 of 100 bootstrap samples. Is it possible to
> do this in Stata? If so, do you have any suggestions for how to code
> this? Thank you.
>

*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index