1. In a study for hangover recovery, ten students were given equal amounts of alcohol and then

given either treatment A or B. Alcohol in the blood was measured one hour later. The data

is below (each sample is ordered for your convenience). Treatment A: 0.76 0.79 0.90 0.92 1.02

Treatment B: 0.70 0.98 1.00 1.10 1.25 Use an apprOpriate test to determine whether Treatment A tends to have less variability in

alcohol levels. State your hypotheses, p-value, and conclusion for or = 0.05. . The data in “bus.dat” are samples of waiting times from three bus routes (R, V, and S). The data can be read into R with the following command: (a)

(b) (C) (d) x = read.table(file.choose(), header=T, sep=" ") Find the ANOVA F statistic and p—value for testing equal mean waiting times (Ho :

m; = m; = ‘as). What assumptions are necessary for this p—value to be valid? Apply the permutatiOn F—test to the data. You may use the code from “5 k sample

permit”, which ﬁnds the p—value by randOm permutations. What assumptions are

necessary for this p-value to be valid? Provide histograms of the data for each of the three bus routes. Comment on the

assumptions required for each test. Are they appropriate here? Which p—value do you

prefer, and what conclusion would you make with 05 = .05? State the hypotheses for the Kruskal—Wallis (KW) test. Find the KW test statistic and conclusion, using the approximate chi-square distribution

with a = .05. Compare with the conclusion in the problem above.