1.

A nationwide retailer wants to test whether new product shelf facings are effective in
increasing sales volume. New shelf facings for the soft drink Country Time are tested at a
random sample of 15 stores throughout the country. Data on total sales of Country Time for
each store, for the week before and the week after the new facings are installed, are given
below:

Store 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Before 57 61 12 38 12 69 5 39 88 9 92 26 14 70 22
After 60 54 20 35 21 70 1 65 79 10 90 32 19 77 29

Using the 0.05 level of significance, do you believe that the new shelf facings increase sales
of Country Time?

2. Travel & Leisure conducted a survey of affordable hotels in various European countries. The
following list shows the prices (in U.S. dollars) for one night of a double hotel room at
comparable paired hotels in France and Spain.

France 258 289 228 200 190 350 310 212 195 175 20 190
0
Spain 214 250 190 185 114 285 378 230 160 120 22 105
0

Conduct a test for equality of average hotel room prices in these two countries against a
two-tailed alternative. Which country has less expensive hotels? Back your answer using
statistical inference, including the p-value.

3. New corporate strategies take years to develop. Two methods for facilitating the
development of new strategies by executive strategy meetings are to be compared. One
method is to hold a two-day retreat in a posh hotel; the other is to hold a series of informal
luncheon meetings on company premises. The following are the results of two independent
random samples of firms following one of these two methods. The data are the number of
months, for each company, that elapsed from the time an idea was first suggested until the
time it was implemented.

Hotel 17 11 14 25 9 18 36 19 22 24 16 31 23
On-site 6 12 13 16 4 8 14 18 10 5 7 12 10

Test for a difference between means, using 5% level of significance.

4. The following data are independent random samples of sales of the Nissan Pulsar model
made in a joint venture of Nissan and Alfa Romeo. The data represent sales at dealerships
before and after the announcement that the Pulsar model will no longer be made in Italy.
Sales numbers are monthly.

Before 329 234 423 328 400 399 326 452 541 680 456 220
After 212 630 276 112 872 788 345 544 110 129 776 -

Do you believe that the variance of the number of cars sold per month before the
announcement is equal to the variance of the number of cars sold per month after the
announcement?