Pilot Shop is a catalog business providing a wide variety of aviation products to pilots throughout the world. Maynard Shephard, the recently hired assistant controller, has been asked to develop a cost function to forecast shipping costs. The previous assistant controller had forecast shipping department costs each year by plotting cost data against direct labor-hours for the most recent 12 months and visually fitting a straight line through the points. The results were not satisfactory. After discussions with the shipping department personnel, Maynard decided that shipping costs could be more closely related to the number of orders filled. He based his conclusion on the fact that 10 months ago the shipping department added some automated equipment. Furthermore, he believes that using linear regression analysis will improve the forecasts of shipping costs. Cost data for the shipping department have been accumulated for the last 25 weeks. He ran two regression analyses of the data, one using direct labor-hours, and one using the number of cartons shipped. The information from the two linear regressions follows:
Regression 1 Regression 2
Equation Sc=804.3+ Sc=642.9+
15.68DL 3.92NR
R-squared .365 .729
Standard error of the 2.652 1.884
estimate
t-value 1.89 3.46
where:
SC = total shipping department costs
DL = total direct labor-hours
NR = number of cartons shipped

Required
1. Identify which cost function (regression 1 or regression 2) Pilot Shop should adopt for forecasting total shipping department costs and explain why.
2. If Pilot Shop projects that 600 orders will be filled the coming week, calculate the total shipping department costs using the regression you selected in requirement 1.
3. Explain two or three important limitations of the regression you selected in requirement 1, and identify one or two ways to address the limitations. Specifically include in your discussion the effect, if any, of the global nature of Pilot Shop’s business.