2. Suppose we want to test whether Canadians and American viewers have the same reaction to a certain documentary film. A random sample of 6 Canadians and 7 American viewers were collected. Each are...

2. Suppose we want to test whether Canadians
and American viewers have the same reaction to a
certain documentary film. A random sample of 6
Canadians and 7 American viewers were
collected. Each are asked to watch and then rate
the film, from 1 (most negatively), to 10 (most
positively). The data are shown on the
accompanying table:
Canadians 2 10 1 4 6 1
Americans 4  7  7 5 5 3 8
a. Explain why we cannot test for the equality of means. That is, we cannot test Ho: Average rating among Canadians is equal to average rating among Americans.
b. Use Wilcoxon Rank Sum procedure to test whether Canadians and Americans tend to rate the film similarly. Include the test statistic. State your conclusion in the context of the problem. Use the 10% significance level.
c. Use Normal distribution to approximate the p-value. Would you recommend using it to carry out the test instead of the table?

Y) • Scale of measurement must be at least ordinal • No assumption needed. To test Ho: M1 = M2 • Scale of measurement must be at least interval • The two populations the samples are taken from must be of the same shape To test Ho: µi - H2 = do • Scale of measurement must be at least interval. The two population the samples are taken from must be of the same shape, and they both must by symmetric.


Extracted text: ank Sum Test Test Statistic: T1 = Sum of ranks from smaller sample If n 2 10 and n2 2 10, Tị is approx Normal with n, (N+1) n,n,(N+1) and 1 %D 2 12