1

As I am browsing the resources online, I encounter as least three ways to detect overdispersion after fitting a Poisson regression model.

1) A regression approach proposed by Cameron and Trivedi, 1990

2) The Lagrange Multiplier test

3) A third test proposed by Greene found below

http://people.tamu.edu/~b-wood/Maximum%20Likelihood/RLesson%208.htm

I found literature references to 1) and 2), but am having a hard time finding anything related to 3). Since every test has it own assumptions and limitations, would anyone shine some light on what passing/failing each test entails?

Ye Tian
  • 619
  • 6
  • 19
  • Other possible dups: https://stats.stackexchange.com/questions/256327/overdispersion-tests-dependence-on-used-covariates-in-poisson-model, https://stats.stackexchange.com/questions/38177/how-to-assess-overdispersion-in-poisson-glmm-lmer – kjetil b halvorsen Mar 28 '19 at 22:02
  • Thank you @kjetilbhalvorsen! I have read the other two links, and I don't believe that my question is a duplicate of either. My question was very specific -- I was asking for literature references of the "Greene" approach to overdispersion tests, which is not mentioned by the other two questions, their answers or their references. I have edited the title of my question. Hopefully this would give potential answerers a clearer picture. Thanks! – Ye Tian Mar 30 '19 at 23:23
  • 1
    Answer must be somewhere in this search: https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=Greene+overdispersion+test&btnG= – kjetil b halvorsen Apr 05 '19 at 19:57

0 Answers0