Due
Tuesday, March 6, 2000
1.
The price of advertising is different from one magazine to
another. Develop a multiple regression
model for forecasting ad cost using the projected audience (in thousands), the percent
of readership that is male, and the median household income of the readership
(see magazine.xls). State the multiple
regression model. Determine if the
model is significant. Determine of each
independent variable is significant in the presence of the others. Interpret the coefficients of the
model. Interpret the coefficient of
determination and the standard error of the estimate. Check the assumptions of the model. Use the model to forecast ad cost for a magazine with a projected
audience of 10,000, percent male equal to 20, and median household income of
50,000.
2.
A local carwash is trying to estimate the number of
customers that it will have on any given day of the week. The owner believes that demand is probably
determined by the temperature outside, whether or not rain is predicted, and if
it is a weekday or weekend (Saturday or Sunday). He has randomly selected 21 observations for you. Code the data using dummy variables as
needed. Do a multiple regression
analysis on this data. State the
multiple regression equation and interpret the coefficients. Test to see if each of the variables is
significant in the presence of the others using an alpha of .05. After giving it some thought, the owner has
decided to count Friday as part of the weekend. Change your coding and rerun the analysis. How have the results changed?