1. (a) The sign test and the Wilcoxon signed-rank test can both be used to test
hypotheses about the location of a population.
i. When is it more appropriate to use the sign test over the Wilcoxon signed-
rank test? Refer to the shape of the population distribution in your an-
swer.
ii. What (if any) limitations does the sign test have?
iii. Let $x_1$, $x_2$,..., $x_n$ be a sample taken at random from a population. Describe
the steps in computing the Wilcoxon signed-rank test statistic for these
data.
(15 marks)
(b) Ten cows were used in a study to examine methane emissions under two dietary
regimes. Five cows were randomly assigned to each regime and the response
($CH_4$ flux in grams per day) was recorded. The dataset is given in the following
table.
Regime 1 210 220 254 261 280
Regime 2 275 310 330 475 520
A Kolmogorov-Smirnov test was used to compare the two regimes. The test
was implemented using R software, with output:
Two-sample Kolmogorov-Smirnov test
data: r1 and r2
D = 0.8, p-value = 0.07937
alternative hypothesis: two-sided
i. State the null and alternative hypothesis.
ii. Show how the test statistic was computed.
iii. State the conclusion, using $\alpha$ = 0.05.
(18 marks)