| Problem Set |
Date Assigned |
Chapter |
Problems |
|---|---|---|---|
| 21 | 12/1 | BPS 28 | Read (again from Chapter 28 on the CD-ROM) pp. 42-50. |
| ∗5 | Carry out the exercises on this sheet. | ||
| 12/3 | BPS 28 | Read pp. 54-61 | |
| BPS 28 | Do Exercises [31-40]. | ||
| ∗6 | Problem 2 from the first part of this problem set (i.e., from the list of problems assigned on Dec. 1) contained the results of multiple regression for a bear's weight as predicted by 7 variables: neck, age, length, sex, chest, headlen, and headwid. Using the StatCrunch output supplied in that problem, construct 90% confidence intervals for the true parameters (multipliers) associated with neck and headlen. One issue is whether or not the number 0 (zero) lies inside these intervals. How can you know in advance of constructing the CI whether it is or not? | ||
| ∗7 | Suppose you did multiple regression on a dataset with n = 25 observations (rows) using p = 4 predictors. If the resulting F-statistic is 4.803, what does Table D (pp. 688-691) tell about the resulting P-value? What do you conclude about the model? | ||
| n | manditory problem |
| ∗n | problem not found in text (but still manditory) |
| (n) | helper problem |
| [n] | ungraded problem |
| {n} | optional problem |