From the "Is Using a Car Phone Like Drunk Driving Article" we explored how to compute the relative risk. On page 7 we are essentially told that the relative risk should be computed as
P(Accident | On Phone AND Driving)/P(Accident | Not on Phone AND Driving).
Answer the following carefully : Explain carefully why this ratio indeed captures the notion of relative better than the ratio
P(Accident | On Cell Phone)/P(Accident | Not on Cell Phone).
as computed on page 6 and in class.
Attempt the following (this is not going to be easy!): Using the information in the box on page 6 and the fact stated on page 7 that the
P(Driving During the Control Period)=.65
compute the above improved relative risk. Please explain each of your steps and assumptions as best as you can. Does your answer agree with the one in the article?