I'm almost sure about separate models since the test strongly reject the null of equality between the coef.
my coefs refer to geographical partitions (groups of region) of residence. These are binary variablesintroduced into my model for the duration of unemploymnet. basically the test confirms the presence of heterogeneity of this phenomena acrooss different partitions of my country.
ah, sorry: the model is a piecewise constant hazard
-----Messaggio originale-----
Da: [email protected] per conto di Richard Williams
Inviato: lun 03/03/2008 17.21
A: [email protected]; [email protected]
Cc:
Oggetto: Re: RIF: RIF: st: chow test for more than 2 coef
At 10:45 AM 3/3/2008, Mussida Chiara wrote:
>Sure, that is what I want to test: done with the command
>testparm...I can now proceed with separate models
I'm unclear on the separate models part. Just because the effects of
x1-x5 are equal does not mean that you would run separate models, one
using only x1, one using only x2, etc. All 5 variables (or perhaps
an additive scale created from them) should probably still be in the
model. But again, I don't know what the x's are, nor do I know what
you are trying to accomplish. Perhaps you could describe your goals
in a bit more detail.
-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME: (574)289-5227
EMAIL: [email protected]
WWW: http://www.nd.edu/~rwilliam
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
<<winmail.dat>>
