I have a Poisson GLM in R that is overdispersed, so I fit a quasipoisson GLM, however the residual deviance nor the degree of freedom change. Can that happen? What does it mean in that case?
$\begingroup$
$\endgroup$
1
-
$\begingroup$ Same Q: stats.stackexchange.com/questions/176918 $\endgroup$– amoebaCommented Nov 17, 2017 at 10:56
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
Yes, it can and is perfectly fine. Actually in this case, the estimated coefficients should be the same as well as the residual deviance and the degrees of freedom! But have a look at the standard errors of the estimated coefficients. You will see larger standard errors when fitting a quasipoisson GLM. It means that you are fitting a more flexible model that allows an over-dispersion. Therefore, you have larger confidence intervals.
