1. Two groups of people with coronary heart disease receive two different drugs. The patients are 

followed over time. Among the patients in the first group, 1.7% eventually suffered a heart 

attack. 0.9% of the patients in the second group eventually suffered a heart attack. 

(a) Does this imply that the second drug works significantly better? 

(b) If I provide you the additional information that both groups had 500 patients, how will your 

answer change?

2. Assuming that the population proportions are same as the sample proportions obtained in #1

above, what is the sample size that you actually need to detect that difference across the two 

groups? Let's use an 80% power when calculating the sample size.

You must explain what test you are using, the hypotheses you are testing, and the values you 

are using as input in the sample size calculator.