Suppose that we randomly select 100 U. S. adults who are smokers and find that 25% of this sample started smoking before the age of 16. Is this much error surprising? Find the probability that a sample proportion will be an over- or underestimate of the parameter by more than 5%.