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: Binary Variables


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: Binary Variables
Date   Wed, 2 Jun 2010 12:20:02 +0100

Another alternative is just to aggregate some regions. If the response is identical, you lose nothing, except that some thought needs to be given to the combination of predictor values for those regions. For example, is population weighting appropriate? 

Nick 
n.j.cox@durham.ac.uk 

Maarten buis

--- On Wed, 2/6/10, Natalie Trapp wrote:
> I use a Dummy Variable for 150 regions within the EU27.
> When I regress the model, I use one region as a reference
> group, but Stata still automatically omits four to five more
> regions. Is it maybe because the regions are too similar in
> their characteristics so that I have to build groups of
> similar regions? Or is there another way how I can do the
> regression with all regions of interest?

This is a fixed effects regression (assuming you are using 
linear regression, -regress-).  This type of regression can 
only make use of variation within a region, so if the dependent
variable is constant within a region, the region will be dropped.
Your alternative is to use random effects regression (see:
-help xtreg-), but that has disadvantages of its own. It is up 
to you to decide which disadvantages you think are least bad... 


*
*   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