I am going to use Poisson regression in my model, but I don't know what assumptions I need to check to verify adequacy of the model. In a normal regression, I had to check Gauss Markov assumptions for my estimates to be BLUE. Do I have to check same assumptions for Poisson? Can I apply tests like White's test, Breusch Pagan test, etc., to Poisson regression? I googled these questions but unfortunately didn't find anything helpful, so if you know good source please share.
Asked
Active
Viewed 598 times
1

kjetil b halvorsen
- 63,378
- 26
- 142
- 467

John
- 21
- 1
-
Related: [Diagnostic plots for count regression](http://stats.stackexchange.com/q/70558/). – gung - Reinstate Monica Nov 07 '16 at 19:35
-
1There is a great answer about count data regression diagnostics by Momo [here](http://stats.stackexchange.com/questions/70558/diagnostic-plots-for-count-regression). – Simon Nov 09 '16 at 17:23
-
Relevant: https://stats.stackexchange.com/questions/142338/goodness-of-fit-and-which-model-to-choose-linear-regression-or-poisson/142353#142353 – kjetil b halvorsen Sep 08 '17 at 15:32