(c) what sample size would be required in each population if you wanted to be 95% confident that the error in estimating the difference in mean road octane number is less than 1?