8.1.3 Exploration 8A: Regression Modeling Practice
StateSteins is a tourist trade vendor. The company manufactures hand-crafted beer steins with
state logos and images. They have facilities in each of the fifty states (and in Washington, DC).
The data file C08 Profit.xls [.rda] contains last year’s figures for profit, revenue, cost,
number of steins sold, and labor for each of the separate state facilities. Your job is to
investigate these data and determine which variable is the best predictor for the company’s
profits.
- Formulate and estimate linear regression models to predict profit as a function of each
of the four explanatory variables. Be sure that you have the routine construct the
important diagnostic graphs (Fitted v. Actual, Residuals v. Fitted) to help in your
analysis. (If you ae using StatPro/Excel, you will need to rename the worksheets with
these graphs in order to ensure that StatPro will not overwrite them each time you
compute a new regression model.)
- Interpret the slopes of each of the four regression models you created. Be sure to include
the units for each slope.
- Examine the diagnostic graphs to see what they tell you about the quality of your
model. You will learn more about how to interpret these graphs in the next section,
but for now, see what you can learn from them.