A few years ago I was asked to help a company estimate its shipping costs. The company is in the business of providing authentic rebuilt parts for a classic car. They get orders from all around the world for both small and large parts. They needed a method to estimate the shipping costs for the ordered parts instead of telling customers “TBD”, which was becoming less and less acceptable given the wide variation depending on the size of the part and the shipping destination. The file Shipping_Cost.xlsx contains a random sampling of some recent orders (and includes only orders for the 48 contiguous states). The data included are the dollar amount of the merchandise to be shipped (Merchandise Total), the weight of the order (Shipping Weight), and the actual cost of shipping incurred (Shipping Charge).
For this question:
a. Run the regression model with only the weight as the explanatory variable.
b. Run the regression model with only the merchandise total as the explanatory model.
Upload one pdf document that includes the output for both analyses. Is one model better than the other? Explain your answer fully – include an analysis of the standard deviation of the prediction errors, etc.