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 at the end of May, and its replacement, statalist.org is already up and running.


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

Re: st: Cluster robust se in logit fixed effects model


From   "Justina Fischer" <JAVFischer@gmx.de>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Cluster robust se in logit fixed effects model
Date   Thu, 26 Jul 2012 13:42:49 +0200

Dear David,

depending on how the question is formulated you could treat it as a count-variable, applying negative binomial estimation methods.

Hope this helps

Justina
-------- Original-Nachricht --------
> Datum: Thu, 26 Jul 2012 11:34:45 +0000
> Von: "Deller, David R D" <drddel@essex.ac.uk>
> An: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
> Betreff: st: Cluster robust se in logit fixed effects model

> Dear statalist,
> 
> I am currently using a short and highly unbalanced panel data set to
> estimate a logit model of individuals committing property crime. Having run
> basic logit and probit regressions successfully I'm now running a fixed effects
> logit model using xtlogit, fe. As well as my dependent variable being
> binary most of my independent variables are dummies as I am using survey data
> composed largely of categorical questions. I want to obtain cluster robust
> standard errors as each individual represents a cluster who has errors which
> are likely to correlated through time.
> 
> To get cluster robust standard errors when using xtlogit, fe Cameron and
> Trivedi (2010) advises to use the option vce(bootstrap). However, given that
> relatively few individuals commit property crime, when I run the code I
> get lots of "red x's" which indicate stata has posted a missing value. I
> believe these missing values are due to stata dropping variables in the
> resampling process for which no one commits a crime.
> 
> My question is how does stata treat these missing values when it finally
> reports the standard errors? I know there are issues about the averaging
> process if one simply uses the bootstrap command and there are missing values.
> See:
> 
> http://www.stata.com/support/faqs/statistics/resampling-and-missing-values/
> Does this same problem apply to the command vce(bootstrap)?
> 
> I appreciate that even if this averaging problem doesn't occur with
> vce(bootstrap) the situation is not ideal as some of the standard error figures
> will be based on fewer replications than others.
> 
> Any help would be much appreciated.
> 
> Regards
> 
> David Deller
> University of Essex
> 
> *
> *   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/

-- 
Justina AV Fischer, PhD
COFIT Fellow
World Trade Institute
University of Bern

homepage: http://www.justinaavfischer.de/
e-mail: javfischer@gmx.de. justina.fischer@wti.org
papers: http://ideas.repec.org/e/pfi55.html


*
*   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