My main problem is at part b, please provide details

Suppose we want to do a poll to see what is the percentage, p of voters that will

support president candidate A over candidate B. Assume that whether a person will vote for

candidate A over candidate B is a Bernoulli trial, and the chance of a person voting for either

candidate is equal.

a) How many people do you need to poll so that the percentage p is within 1% from its true

value with 95% confidence? [Hint: The percentage p is the mean of the random variables

which represent a vote for candidate A over candidate B]

b) If you poll 400 people, how many have to prefer candidate A to B so that the 90%

confidence interval is entirely above p=0.5?