Statalisters,
Is there a hard and fast rule as to how small your mean and variance
need to be when modeling count data via negative binomial regression? I
know about the assumption of overdispersion but have never seen anything
on the relative "size" of these two measures. I am assuming that the
larger the mean the more your model looks like a normal distribution...
Thanks in advance!
Carter Rees
*
* 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/
