Hi Faith,
a. Two ways to do this: First way is by formula:
CI = (p^1 - p^2) +/- z* SE, SE= sqrt(p^1Q^1/n1 + p^2q^2/n2)
p^1 = Freshman Proportion who bought used = 95/109 = 0.8716
q^1= 1-p^1 = 0.1284
p^2 = Sophomore Proportion who bought used = 100/137 = 0.7299
q^2 = 1 - p^2 = 0.2701
SE= sqrt((0.8716*0.1284)/109) + (0.7299*0.2701)/137)) = 0.0497
z* = 2.326 (from z-table)
CI = (0.8716 - 0.7299) +/- [2.326 (0.0497)]
CI = (0.0261, 0.2573)
Second way is by TI-83 or 84 Plus calculator. Go to STAT-TESTS-2PropZInt--Enter 95 for x1, 100 for x2, 109 for n1, and 137 for n2. Interval will be slightly different due to rounding (0.0261, 0.2572).
b. I can see where this gave you trouble; all answers sound similar. Remember that most confidence interval interpretations begin with "We are x% confident that" and remember that we are dealing with difference between two population proportions here. Correct answer, given these, is:
We are 99% confident that, at this university, the proportion of first-year students who bought used textbooks is between 1.37% and 26.96% higher than the proportion of second-year students who bought used textbooks.
I hope this helps.