Statalist


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

Re: st: dropping one unit at a time


From   Matt Spittal <Matt.Spittal@cancervic.org.au>
To   <statalist@hsphsun2.harvard.edu>
Subject   Re: st: dropping one unit at a time
Date   Thu, 18 Jun 2009 12:35:36 +1000

Dear Carola,

I would have a look at some the regression diagnostics first.  The ones to
focus on are the DFBETAs, see - help regress postestimation-, which shows
how the regression slope changes as each observation is deleted one at a
time.  It can be quite helpful to plot this against the subject id number to
identify the observations with the largest DFBETAs.  A second option is
Cook's distance, which is better for multivariate models.  Again, plotting
this value for each subject against the id number is probably the best way
to spot potentially influential observation.

-- Matt
matt.spittal@cancervic.org.au

  


On 18/6/09 12:14 PM, "Carola Herrera" <herrera.carola@yahoo.com> wrote:

> Dear Statalist participants,
> 
> I wish to examine whether my OLS results are driven by any particular country
> in my panel dataset.  What is the Stata code to drop one country at the time
> and re-run the regression?  Is there a better way to test for this?
> 
> Thanks
> Carola
> 
> 
> 
>       
> 
> *
> *   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/

*
*   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