54. For a random sample of n subjects, explain why it is about 95% likely that the sample proportion has an error of ne more than 1/n as an estimate of the population proportion (Hint: To show this "In Rule." find two standard enors when 7 = .5. and explain how this compares to two standard erions at other values of x.) Using this result, show that a=1/B is a safe sample size for estimating a proportion to within B units with 95% confidence