Statalist


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

Re: st: Stata11 cannot estimate xthtaylor with bootstrap SE


From   tmb@stata.com
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Stata11 cannot estimate xthtaylor with bootstrap SE
Date   Mon, 24 Aug 2009 12:41:35 -0500

In my previous response to Mike's question, I forgot to answer two questions. 

> 1. What does "xx...x..xx" mean in Stata10?

By default, -bootstrap- displays dots as a means of showing its progress
throught the replications.  Occasionally you'll run into an "x" in the list of
dots.  The "x" means that the prefixed command (-xthtaylor- in Mike's case)
failed for some reason.  Results from failed replications are not used in
variance estimation.

We suspect that the source for Mike's failed regression are due to
collinearity, -xthtaylor- is omitting different variables in some of the
replications compared to the original model fit.

Mike can use -bootstrap-'s -noisily- option to see what is causing the
failures in these replications.

> 3. What is the difference between "dropping" and "omitting" a variable due 
> to collinearity?

Substantively there is no difference.  Stata 11 omits collinear variables from
the model instead of dropping them from the list of covariates. This is the
same as constraining the coefficient to zero.  The "o." operator identifies
the omitted variables, and the regressiong table reports them as '(omitted)'
by default.

This change of behavior is largely motivated by the new factor variables
features in Stata 11.  The base levels of factor variables remain part of the
list of covariates.

--Theresa Boswell			--Jeff Pitblado
  <tmb@stata.com>			  jpitblado@stata.com
*
*   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   |   What's new   |   Site index