According to the Rule of Three, when we have a sample size n with xequals0 successes, we have 95% confidence that the true population proportion has an upper bound of StartFraction 3 Over n EndFraction . a. If n independent trials result in no successes, why can't we find confidence interval limits by using the methods described in this section? b. If 40 couples use a method of gender selection and each couple has a baby girl, what is the 95% upper bound for p, the proportion of all babies who are boys