The dataset Blood1 contains data from a random sample of 500

The dataset Blood1 contains data from a random sample of 500 adults.

The dataset contains 3 variables, 2 of which we will consider.

SystolicBP: Systolic blood pressure (mm of Hg) Smoke: 1=smoker or 0=non-smoker

We will assume that is a representative sample of the population of all of Canada and that blood pressure is normally distributed.

Let us create a new random variable D which is the difference in these values. If you took another sample (with the same number of smokers and nonsmokers) you probably would get a different value for D.

D = Sample average BP of smokers - Sample average BP of nonsmokers
Since we are assuming that both sample averages are normal, D is the difference between

two normal variables, and is thus also normal.

(i) Find the variance of D

(j) If smoking has no effect on blood pressure, what is the expected value of D?

(k) Based on your above calculations. If smoking has no effect on blood pressure, what is the

probability that you got a random sample the showed such a big effect just by chance?

(l) Do you think smoking has a significant effect on blood pressure?

.

Solution