STA 301F97 Assignment 10: Quiz in tutorial Friday Nov 21

For this assignment and this assignment only, it's okay to write anything you want on your printout.

You will use the restaurant data again. We need to start by making sure we have the same data (missing values and everything). I did some descriptive stats on the variables we will be using; look at a part of my list file and verify that your means and frequencies are the same as mine.

Now suppose that an advertising agency is interested in finding new business doing commercials for restaurants. What they want is to predict amount of money spent on advertising (NOT advertising as a percent of sales). So create a new variable representing advertising in thousands of dollars. Did you do it right? Here's how you can look at just the first 10 cases to see. Suppose your new variable is called dweezel9.

list cases = 10 / variables = sales ads dweezel9.

Now do analyses to answer these questions.

1. Do these data provide evidence that if we control for estimated market value of business, advertising expenditure depends on the type of restaurant (Fast food vs supper club vs other)?

2. Do these data provide evidence that if we control for market value and type of restaurant, type of ownership (sole proprietorship vs partnership vs corporation) predicts advertising expenditure?

3. One advertising executive suspects that restaurants ready to spend money on capital improvements might also be willng to spend money on advertising. Is there evidence that if we control for market value, type of restaurant and type of ownership, new capital invested will help predict advertising expenditure?

4. On a fresh sheet of paper, make a big table with three rows and three columns. These will represent the 9 combinations of type of restaurant and type of ownership. In each cell, put mean advertising expenditure corrected for market value of the business (that is, Y-hat with market value of the business held constant at its mean level. For this question, you should bring the sheet with the table to the quiz. In addition, bring at least one other sheet of paper showing your calculations.

5. In these data, does the slope of the linear relationship between market value of the business and advertising expenditure depend significantly on the type of restaurant?

6. Using the output for the question above (not running 3 separate regressions), give equations for the regression line for (a) Fast food restaurants, (b) Supper clubs, and (c) Other restaurants. In your equations, let X represent market value of the business. All the b coefficients should have numerical values based on the printout.

Again, for this assignment and this assignment only, it's okay to write anything you want on your printout.