Weibull is a model for durations. If your outcome variable is binary, I'd
try logit or probit.
Gary
On Mon, 5 Mar 2007, Israel Marques II wrote:
Hello,
I am conducting a project which seeks to analyze survival amongst Communist regimes and
am having a little bit of trouble getting Clarify to work with my data. My variables are
as follows:
Country: Name of the country (15 different countries)
Year: varies from 1978-1991
Transition: coded 0 for non-event and 1 for event
Electrical_Production: varies from 1 to 1000 (billions of kWh)
Whenever I try to get Clarify to use the weibull model in order to estimate a model where
transition is the dependent variable, I always end up with an error message that says
"Transition <= 0 in some observations". I've checked the data itself to
make sure there is nothing in it (stringed variables, etc.) that State might not like and
there is nothing there. Clarify runs perfectly if I try using a non-weibull model. I
assume I'm making some sort of coding error. How do I code a non-transition? I think
my use of zero is a problem here. Any help would be vastly appreciated.
-IM
--
Clarify mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive:
http://lists.gking.harvard.edu/?info=clarify
--
Clarify mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive:
http://lists.gking.harvard.edu/?info=clarify