MA207-B Spring 2008 Regression Modelling Project - Problem 5
A large discount retail chain wants to evaluate the effectiveness of a telivision advertising campaign. The method used will be to compare the difference in same store sales for two months, June and July, that are normally quite similar.
The ad campaign will be run during the first week of July, and the predictor variable will be the number of minutes of prime time advertising that was broadcast in the area of each store.
At the end of July, the net change in sales in thousands of dollars is computed for each of 85 stores, and paired with the
number of minutes of prime time advertising that was broadcast in the 85 corresponding areas.
Use the data in your comma delimited file to construct a regression model.
You should answer the questions on the assignment page that apply to all problems.
For this specific problem, answer the following questions:
- How would you interpret the slope of the regression line in this problem?
- How would you interpret the intercept of the regression line in this problem?
- Were there any outliers or questionable data points? If so, how did you handle them?
- Did you use a log transform in this model? If so, explain why you chose to use a transform.