0

I'm rather new to R and poisson regression and have a few questions.

I have count data of incidents per month, a couple years prior to an intervention and 1 year after. I want to use poisson regression to test if the intervention had a significant impact on the number of incidents. This is a quasi-experimental study in which the intervention was applied in a non-random way. There is only 1 zero in the data. The data has a seasonal pattern so I am wondering how to model this in the poisson, do I need to create a dummy variable? How best do I go about modeling the dummy variables (month, quarter)?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
indubitable19
  • 1
  • 1
  • 1
    It may be useful to provide some example data – OliverFishCode Mar 01 '19 at 06:56
  • Possible dups: https://stats.stackexchange.com/questions/175584/count-data-time-series-with-multiple-seasonality, https://stats.stackexchange.com/questions/7579/how-to-model-zero-inflated-over-dispersed-poisson-time-series, https://stats.stackexchange.com/questions/144745/is-this-an-appropriate-method-to-test-for-seasonal-effects-in-suicide-count-data – kjetil b halvorsen Sep 13 '19 at 09:14

0 Answers0