1) When we say we have confidence level of 95% what do you believe our "margin of error" is? 

2) If we move level of confidence from 95 to 99%, what will occur to our margin of error? What will occur to width of our confidence interval?

3) Would we have got better feel for amount of flour required if we'd used larger sample? Why (mathematically) would this occur?

4) President of American insurance Institute wishes to compare yearly costs of auto insurance offered by two leading companies. He chooses a sample of 15 families some with only single insured driver, others with many teenage drivers, and pays each family a stipend to contact two companies and ask for price quote. To make data comparable, certain features, like the deductible amount and limits of liability, are standardized. Sample information is reported below. At 0.10 significance level, can we conclude that there is difference in amounts quoted?