Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: problems with ztnb and ztp


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: problems with ztnb and ztp
Date   Thu, 13 Mar 2008 17:43:22 -0000

That doesn't necessarily sound surprising. For example, if you run the
example
given in the -ztp- help the smallest predicted value is 1.022. As you
want at least 
some predictions pass near one, it isn't surprising that some go below. 
You wouldn't think much of a model that always predicted much more than
one when 
the observed was one. 

There isn't anything here about sample size, merit of model or
characteristics
of data, which presumably experts would find instructive. 

Nick 
n.j.cox@durham.ac.uk 

Gillian Holloway

I'm analyzing zero-truncated bird count data using the standard code for

zero-truncated poisson and zero-truncated negative binomial regression.

However, when I calculate the predicted values I'm getting some 
estimates below 1, even though 1 is the minimum observed value in the 
data. Any ideas what the problem could be?

Example code:

ztp Count SLSoft if Count>0 & Sppcode==38, iterate(100)
predict GRJA

*
*   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/



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