This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Stats for Strategy Fall 2008 HOMEWORK 10 (covers Topic 11 and Topic 12, Part 1. Due Monday, Dec. 1) DIRECTIONS: • Part of this assignment requires reading in the supplementary textbook Chapter 17: Logistic Regression. Find this chapter as a PDF file on the main course webpage. • Some of the exercises listed below include special instructions which modify or clarify textbook instructions. • Data sets for some exercises are available on the Data Sets link from the course website. (Remember to open data files as worksheets in MINITAB .) Homework Part 1: Topic 11 (Logistic Regression) A. Mail-Order Sales • Open the PDF file for the supplementary textbook Chapter 17: Logistic Regression on the main course webpage. Read the description of mail-order sales on the first page. Answer these questions: (a) There are 5 predictor variables and a response variable described for this logistic regression. Let’s call them x 1 , x 2 , x 3 , x 4 , x 5 , and y . Write an English definition for all 6 variables, including the meaning of “0” and “1” for binary variables. (b) Which of the variables are binary ? Which are quantitative (i.e., numerical )? B. College Student Drinking • The textbook introduces logistic regression with the same data on drinking as used in Example 2 of the Topic 11 notes. Quickly glance over the notes you wrote for Example 2 on pages 8-11 of the Topic 11 notes. • Now open page 17-2 of the textbook Chapter 17 PDF file. Read pages 17-2 and 17-3 at a leisurely pace. Soak up what the textbook says and occasionally glance at the Topic 11 notes to confirm that the book’s answers are the same as ours. • Next read Model for logistic regression at the bottom of page 17-4 and all of page 17-5. Then read Fitting and interpreting the logistic regression model on pages 17-6 and 17-7. Pay particular attention to the way the book calculates b and b 1 at the top of 17-7. Are these the same estimates (except for rounding) that MINITAB logistic regression produces in the output shown on page 9 of the Topic 11 notes? • Lastly, read Section 17.1 Summary at the bottom of page 17-4 and top of page 17-5. • Open MINITAB and type the drinking data into the worksheet: C1 C2 C3 C4 Gender Binge Total Men 1 1630 7180 Women 1684 9916 1 • Reproduce the MINITAB output shown on page 9 of the Topic 11 notes. (Recall that in class we made some corrections to the Minitab instructions shown in the notes.) Here are the correct MINITAB steps: Stat > Regression > Binary Logistic Regression > (Click radio button “Response in event/trial format”) > (Enter Binge for Number of events, Total for Number of trials) > (Select all predictor variables for the Model : Gender ) > (Select all binary predictor variable as Factors : Gender ) > OK Notice that there’s some extra output at the bottom which you can ignore....
View Full Document
- Spring '11
- Regression Analysis