Case: The Construction Department at Croqpain
Assignment:
Using the data for this case, provided in the spreadsheet file entitled CROQ-PAIN.xls, do the following:
(a) Examine the operating earnings regression model output obtained from the 60 stores, as shown in Table 6.27. Try to improve the model by eliminating certain independent variables or by making any other changes that you think make good sense. You should strive for a model that is simple (i.e., has few independent variables), that does not violate any of the basic assumptions of multiple regression (e.g., multicollinearity, heteroscedasticity), but nevertheless has good predictive power.
(b) Michel thinks that a good way to validate the model obtained with data from the 60 stores is to see how a similar model, obtained from the 50 stores opened before 1994, would have performed in predicting the performance of the last ten stores opened. Step back one year, prior to the opening of the last ten restaurants. Amend the model you have developed using only the data from the first fifty stores. Using Croq'Pain's performance ratio target of 26%, which of the ten stores would you have opened in 1994?
(c) Croq'Pain's strategic planning group has developed a list of ten potential store locations for 1996 shown in Table 6.28. Which of these locations would you select, i.e., which locations meet or exceed the performance ratio? Use the most complete model (60 restaurants) for your analysis.
(d) Prepare a memorandum containing your recommendations as to whether, and how, regression models can be used in the process of location selection. Defend your recommendations with a discussion of the relative strengths and weaknesses of regression models in this setting.
Attachment:- croqpain.rar