Question: A prospective cohort study was conducted to investigate the association between smoking at study entry and cardiovascular death during the follow-up period. Participants, categorized at the start of the study as smokers (exposed – yes) and nonsmokers (exposed – no), were followed over time to see if they ultimately died from a cardiovascular cause. The data are summarized in the following table:

Death from

a. Compare the proportion of those who died from a cardiovascular cause between the two groups (where group is based on smoking exposure status) using a 98% confidence interval. Provide a written conclusion of the interval.

b. Compare the proportion of those who died from a cardiovascular cause between the two groups using a hypothesis test with significance level .01. Provide a written conclusion of the test.

c. Generate a relative risk point estimate. Also, calculate a relative risk interval estimate using a 95% confidence interval. Provide a written conclusion of the interval.

d. Generate an odds ratio point estimate. Comment on the magnitude of this estimated odds ratio as compared to the estimated relative risk. Also, calculate an odds ratio interval estimate using a 95% confidence interval. Provide a written conclusion of the interval.

e. Conduct a chi-square test of independence for this data at the .02 significance level. Provide a written conclusion of the test.

f. Comment on your findings in total from (a) through (e) on specifically whether it appears there is an association between smoking status and future death from cardiovascular causes.