Assume that 0.50 of all births are boys. Suppose we look at a sample of 20 births in a city each

day for 200 days.

(a) Suppose we set up a 99% confidence interval each day for the proportion of all births that

are boys, based on the sample proportion of boys born that day. About how many of the

200 confidence intervals should contain the true proportion 0.50?

(i) 0 (ii) 1 (iii) 2 (iv) 10 (v) 20 (vi) 50 (vii) 99 (viii) 100 (ix) 190 (x) 198 (xi) 200

(b) Suppose that each day we test the true null hypothesis that the population proportion is

0.50 against the two-sided alternative, based on the sample proportion of boys born that

day. About how many of these 200 tests should reject at the = .01 level?

(i) 0 (ii) 1 (iii) 2 (iv) 10 (v) 20 (vi) 50 (vii) 99 (viii) 100 (ix) 190 (x) 198 (xi) 200