an insurance company is trying to estimate the average number of sick days that full-time food service workers use per year. A pilot study found the standard deviation to be 2.5 days. How large of a sample must be selected if the company wants to be 95% confident of getting an interval that contains the true mean with a maximum error of one day?