(f) Construct a95%confidence interval for the true difference in test scores between the two kinds of training methods.   (g) Does this interval reaffirm your statistical decision from the hypothesis test? Explain.

Suppose you run a non-for-profit organization focused to keeping national parks free of litter. Crucialto getting donations to your non-for-profit is email advertising. One way to measure the success ofyour email advertising is to measure the amount of money coming from links to donate on your emails.Normally, your emails contain only words and numbers about litter in national parks. You believe thatincluding images of littered areas of national parks might increase the amount in donations that youget through emails. To test if this is true, you randomly assign 50 individuals to receive the new emailformat with images and 42 to receive your usual email format (this is called an A/B test). The newemail format had an average donation per recipient of$5.50 with a standard deviation of$2.75 while theold email format had an average donation per recipient of$3.50 with a standard deviation of$2.50. Doesthe average donation differ between the new and old email formats? Letμ1be the true mean donationsper recipient the for new email format andμ2be the true mean donations per recipient for the old emailformat. Test this hypothesis at theα= 0.01 level.

(f) Construct a95%confidence interval for the true difference in test scores between the two kinds of training methods.

(g) Does this interval reaffirm your statistical decision from the hypothesis test? Explain.