A retailer of motorcycle accessories wants to develop a model to make sales forecasts based on the dataset in the table below. The first column is sales for each geographic region. Units are millions of euros. We have data for 11 regions; there are more regions for which we have no data. The second column is number of retail outlets in each region. The third column is the number of motorcycles registered in each region. Units are millions of motorcycles.

a. First, please analyze the correlation matrix.

b. Can you develop an effective predictive equation from the results of multiple regression on these data?

c. In region 12 the number of outlets is 2,500 and the number of motorcycles is 20.2 million. Can sales be forecast for region 12?

d. Discuss the accuracy of the forecast for region 12 (item c above).

e. How can this regression equation be improved? Sales 52.3 26 20.2 16 30 46.2 35 3.5 33.1 25.2 38.2 Outlets 2011 2850 650 480 1694 2302 2214 125 1840 1233 1699 Motorcycles 24.6 22.1 7.9 12.5 9 11.5 20.5 4.1 8.9 6.1 9.5