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]

AW: st: Industry dummies - inflated standard errors


From   "Seliger Florian" <seliger@kof.ethz.ch>
To   "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Subject   AW: st: Industry dummies - inflated standard errors
Date   Mon, 9 Jul 2012 14:38:57 +0000

Thank you Maarten.

Using factor variables solves the problem.
However, I don't understand why? What is the differenc between using dummy_1-dummy_33 and i.dummy?

Best,
Florian


-----Ursprüngliche Nachricht-----
Von: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Maarten Buis
Gesendet: Montag, 9. Juli 2012 11:21
An: statalist@hsphsun2.harvard.edu
Betreff: Re: st: Industry dummies - inflated standard errors

On Mon, Jul 9 2012 at 10:28, Seliger Florian wrote:
>     After entering industry dummies into probit or multinomial logit 
> models  (or other models, I haven't tried...), I sometimes experience 
> the problem that standard errors of the coefficients of the dummies 
> tend to infinity and p-values tend to one.

On Mon, Jul 9, 2012 at 10:42 AM, Stefano Lugo wrote:
> Sometimes those models have an hard time when fixed effects are 
> included; here are a couple of heuristic solution that in my experience might work:
>
> 1) Try to directly exclude collinear indicators instead of letting 
> stata dropping them

Or better yet, use Stata's factor variable notation, see: -help fvvarlist-.

I would also look for small industries, and combine them with similar industries such that each industry is large enough to get a reliable estimate.

Moreover, I assume you know that adding indicator variables to non-linear models will not result in a fixed effects model.
<http://www.stata.com/statalist/archive/2003-09/msg00103.html>

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------
*
*   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   |   Site index