Insurance companies are becoming concerned that increased use of cell phones is resulting in more car crashes, and they are considering implementing higher rates for drivers who use such phones. How large a sample of drivers should be selected so that a 98% confidence interval for the population proportion who use cell phones while driving has a margin of error of 0.5%? (Assume a prior study based on data from Accident Prevention magazine suggested that 30% of drivers talk on cell phones while driving)