For each of the above models, state the form of the
estimated model in logit terms.
Model 1: logit P(X)¼
Model 2: logit P(X)¼
Using Model 1, compute the estimated risk for CVD
death (i.e., CVD¼1) for a high social class (SOC¼1)
smoker (SMK¼1) with SBP¼150. (You will need a
calculator to answer this. If you do not have one, just
state the computational formula that is required, with
appropriate variable values plugged in.)
Using Model 2, compute the estimated risk for CVD
death for the following two persons:
Person 1: SOC¼1, SMK¼1, SBP¼150.
Person 2: SOC¼0, SMK¼1, SBP¼150.
(As with the previous question, if you do not have a
calculator, you may just state the computations that are
required.)
Person 1:
Person 2:
Compare the estimated risk obtained in Exercise 3 with
that for person 1 in Exercise 4. Why are not the two
risks exactly the same?
Using Model 2 results, compute the risk ratio that
compares person 1 with person 2. Interpret your
answer.
If the study design had been either case-control or
cross-sectional, could you have legitimately computed
risk estimates as you did in the previous exercises?
Explain.
