Statalist


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

st: Zero-truncated Negative Binomial convergence


From   Emily Wilson <ECWILSON@DAL.CA>
To   statalist@hsphsun2.harvard.edu
Subject   st: Zero-truncated Negative Binomial convergence
Date   Mon, 16 Mar 2009 13:40:37 -0300

Hello,

I am having trouble running a zero-truncated negative binomial regression.
The dependent variable is: # of visits to a specialist physician in the past
year, and the distribution is something like:
0 visits ~= 93,000
1 visit ~= 15,000
2 visits ~= 1,000
3 visits ~= 500
..
.
.
300 visits = 1

I have a number of covariates that I eventually want to include in the model,
but beginning only with x=sex, I can not get the model to converge.

We ran the same regression using # of visits to a general practitioner
0 visits ~= 30,000
1 visit ~= 25,000
2 visits ~= 25,000
3 visits ~= 15,000
and it ran without any problems with all of the covariates.

Any ideas of why the specialist model will not converge? Is it the skewed
distribution? What are some alternatives?

Thanks for your assistance!!

Emily


*
*   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   |   What's new   |   Site index