Autocorrelation and Poisson Distribution

[mccormsa]

Hello! I am trying to figure out what procedure to use that will allow me to test and adjust for autocorrelation in non-normally distributed count data on which I am conducting an interrupted time series. Any advice is appreciated- thanks!

[Stu_SAS]

Take a look at PROC COUNTREG.

https://support.sas.com/rnd/app/ets/procedures/ets_countreg.html

If it doesn't do everything you need, check out this excellent SGF 2015 paper by Michael Leonard and Bruce Elsheimer: Count Series Forecasting.

https://support.sas.com/resources/papers/proceedings15/SAS1754-2015.pdf