I have the rate of tests performed by month over a time period, and I want to determine if the increase in rate of tests run after an intervention is statistically significant. Our hypothesis is that the rate of test will increase after the intervention. We are thinking that a segmented Poisson regression would be appropriate here, but are not positive. Would another type of analysis be more appropriate? Any thoughts and suggestions would be very much appreciated.
It is impossible to determine if a model is adequate without knowing what it is about. But whatever model you choose, make sure that:
1) It includes an overall time trend, so that you don't confuse the effect of an intervention (before/after) with a simple unrelated time trend.
2) It accounts for overdispersion. Most counting processes in the real world are not pure Poisson processes. Assuming they are leads to false conclusions.
PG
PG
Ready to join fellow brilliant minds for the SAS Hackathon?
Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.
ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.