Bookstore Physical stores face increasing competition from online rivals. To compete, a campus bookstore collected data on sales of newly released trade books displayed on shelving near the store entrance. These data give weekly sales per week of trade books over the last two years, the number of titles on display, the amount of shelf space (in feet), and the total sales in the store.
(a) What patterns are apparent in the timeplot of Trade Sales?
(b) The store manager built a regression model with Trade Sales as the response, using Week, Total Sales, Shelf Space, and Titles as explanatory variables. What would you recommend to improve this model? (Hint: Some claim that the typical amount of space per book is more important than the quantity of books.)
(c) If the manager uses this model to evaluate a new display for trade books in the next few weeks, would you expect the model to give a fair evaluation of the new display? Why? (Hint: Look at the timeplot of the residuals.)