2
$\begingroup$

I need to model a certain response variable in a GLM model. The response variable is a count (amount of insurance claims over a year). So it would be natural to assume a Poisson distribution for this response variable. However, using a goodness of fit test in R, how can I validate this assumption?

Improve this question
$\endgroup$
4
  • $\begingroup$ This document has a good run down of most distribution goodness-of-fit measures in R. I would particularly look at fitdistr() in the MASS package and goodfit() in the vcd package. $\endgroup$
    – NatWH
    Commented May 14, 2018 at 12:59
  • 1
    $\begingroup$ Sorry, if your response variable depends on covariates (which it should if you want to model with a GLM) it probably is not Poisson distributed. However, that does not matter for regression. What matters is the residual distribution. $\endgroup$
    – Roland
    Commented May 14, 2018 at 13:03
  • $\begingroup$ You will need poisson regression, and then this post will be useful: stats.stackexchange.com/questions/331086/… $\endgroup$
    – kjetil b halvorsen ♦
    Commented May 14, 2018 at 13:20
  • $\begingroup$ Possible duplicate of Diagnostic plots for count regression $\endgroup$
    – kjetil b halvorsen ♦
    Commented Jun 27, 2018 at 11:19

0

Reset to default