Fall Semester 2004

Chapter 8: Regression and Correlation.

1. What can you calculate using the Excel functions INTERCEPT and SLOPE? How? Give an example.

2. The workbook BCancer.xls contains data from a 1965 study analyzing the relationship between mean annual temperature and the mortality rate for a certain type of breast cancer in women. The subjects came from 16 different regiones in Great Britain, Norway, and Sweden. You've been asked to determine whether there is evidence of a linear relationship between the mean annual temperature in the region and the mortality index. Is the mortality index different for women who live in regions with different temperatures? Explain your answer.

3. Quote from the book: "One point that cannot be emphasized too strongly is that a significant regression is not proof that these assumptions haven't been violated. To verify that your data does not violate these assumptions is to go through a series of tests, called diagnostics." What assumptions does the quote refer to? Do you need to run any diagnostics with respect to your answer at the previous question? Why and how would you do that (if you have to).

4. What's Spearman's rank correlation coefficient and why (or when) might we be better off using it instead of Pearson's?

5. Consider the Calc.xls workbook. This file contains data collected to see how performance in a freshman calculus class is related to various predictors. Analyze the workbook and answer the following question: can you say that taking calculus in high school causes a better grade in college? Is there any relationship between taking calculus in high school and getting good grades in college? What is it? Explain your answers.

6. For which variables in the workbook from the previous question is it reasonable to say that correlation and linear regression are appropriate when predicting whether calculus has been taken in high school or not?