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: Interpretation of interaction term in log linear (non linear) model


From   David Hoaglin <dchoaglin@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Interpretation of interaction term in log linear (non linear) model
Date   Wed, 12 Jun 2013 21:21:16 -0400

Dear Suryadipta,

A key idea underlying Bill Gould's blog post is that -poisson- enables
you to use quasi-likelihood. That command does not require that the
data follow a Poisson distribution.

Please read carefully the part of the blog post that discusses zero
values.  I doubt that simply fitting your model by using -poisson-
will adequately handle your "many zeros."

Regards,

David Hoaglin

On Wed, Jun 12, 2013 at 12:06 PM, Suryadipta Roy <sroy2138@gmail.com> wrote:
> Dear David,
> I have used both log linear model (that apply only to country pairs
> with nonzero bilateral imports) as well as fixed effects Poisson to
> include the zero observations (based on some wonderful suggestions
> from Bill Gould in Statablog here
> http://blog.stata.com/2011/08/22/use-poisson-rather-than-regress-tell-a-friend/).
> Earlier, Dimitriy had suggested an intuitive interpretation of the
> coefficient terms in terms of semi-elasticity, and I can probably
> persist with that interpretation for Poisson regression.
>
> Best regards,
> Suryadipta.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index