Suppose that a highway patrol person wants to estimate the proportion of all drivers speed, and that she wants to be able to assert with probability of at least 0.95 that the error of her estimate will not exceed 0.07. How large a sample, n, will she require if she has no idea what the true proportion is.