Time reported the results of a telephone poll of 800 adult Americans (600 of them were nonsmokers) who were asked the following question:"should the federal tax on cigarettes be raised by dollars 1.25 to pay for health care reform?" let p1 and p2 are the proportions of nonsmokers and smokers, respectively, who would say yes to this question. We are given that 360 nonsmokers and 40 smokers said yes.
a) Is the difference between the sample proportions for positive answers of nonsmokers and smokers statistically significant? To answer this question, test H0:p1=p2 against H1:p1 not equal to p2 with 10 percents significance level.
b) Find the (approximate) 90 percents confidence interval for p1-p2. is this in agreements with the conclusion of part a?