Suppose you intend to use the binomial distribution to test a hypothesis at the .01 level of significance with a two tailed test. What min. sample size would be required for you to possibly reject the null? Show calculations.

I have a horrible book that doesn't tell me anything. What would be the formula to use?