Homework 6

due in class, Tuesday February 24

You must SHOW ALL WORK for full credit. Include only the most relevant parts of output from R (or whatever package you are using). Do not include a transcript of your entire session, just cut and paste (electronically or literally) the necessary parts. Please staple your assignment if it is more than one page.

Required problems:

1) 13.6 (p. 497) Use a computer package (preferably R) for part (b).
2) 13.17 (p. 503)
3) 13.25 (p. 509) part (b) only
4) 13.40 (p. 520)
5) 13.52 part (b) only (p. 525)
6) Show that if the slope (b1) is known and only the intercept (b0) is to be fitted, the least squares fit for b0 is b0-hat = Y-bar - b1*X-bar. (To do this, you need to differentiate the sum of squared errors (residuals) with respect to b0, set it equal to 0, and solve for b0-hat.)

Optional problems:

7) 13.60 (p. 535)
8) 13.68 parts (a)-(c), (e)-(h), (j), and (k) (p. 539)

If you have the 9th edition, the corresponding problems are:
1) 13.6 (p. 493), 2) 13.17 (498), 3) 13.25 (503), 4) 13.40 (514), 5) 13.54 part (b) (519), 6) (above), 7) 13.62 (531), 8) 13.70 (532-3)