1. A study is conducted to assess the relationship between cholesterol levels and coronary heart disease. 1329 adult aged subject's hospital records are randomly sampled. The cholesterol levels and chronic heart disease status of each subject is used. Cholesterol levels are classied into the ranges (0199), (199-219], (220-259], and (260,300] mg/ 100cc, which correspond to levels Normal, Above Normal, High, and Very High respectively. The 4x2 table of the 1329 observations is below: Cholesterol CHD no CHD yes Normal 307 12 Above Normal 246 8 High 439 31 Very High 245 41 a. Say we are only interested in the risk of CHD comparing two groups, the Normal Cholesterol group and the Very High cholesterol group. Create the two way table for this scenario (and state it) and compute the odds of CHD yes for the Normal group and compute the odds of CHD yes for the Very High group. b. Using part a., compute the odds ratio for comparing the Very High group (numerator) and the Normal group (denominator). Interpret this estimated odds ratio in context of the problem. c. Say you want to test if the odds of the Very High group for CHD are the same as the Normal group. State the test you would use, and write the null and alternative hypothesis (using odds ratio). (:1. Let the pvalue from your test in part c. to be very small (19 m 0). Make a conclusion about the null and alternative in context of the problem. e. Now say each sampling unit had their sex recorded, Female or Male. Based on prior research, it is thought that sex is associated with CHD (females have lower rates of CHD) and also, that sex is associated with Cholesterol (males tend to have higher levels of cholesterol). Can sex be a potential confounder? In a sentence or two explain why or why not. f. It is found that the estimated odds ratio for CHD comparing Very High to Normal cholesterol is