6. Imagine another group of investigators were examining the association caffeine intake and the incidence rate of Parkinson’s disease in a prospective cohort whether individuals with low and high caffeine intake were matched on age and smoking status. Could these investigators evaluate effect measure modification by smoking status in their matched cohort? Yes Could the investigators evaluate the overall association between smoking status and the incidence rate of Parkinson’s disease in their cohort? Yes The investigators can evaluate both the main effect of smoking status on the incidence rate of Parkinson’s disease as well as possible effect modification by smoking status for the association between caffeine intake and Parkinson’s disease. The investigators matched individuals with high and low caffeine intake on smoking status, so the distribution of smoking is the same within an arrow from smoking status to the incidence of Parkinson’s disease so the main effect of smoking can still be evaluated in this cohort.
- Summer '14