Columbia University

Department of Statistics

Statistics W4315: Linear Regression Models

Fall 2004

Prerequisite: probability and statistics at the level of W4150 or (both W4105 and W4107) taken concurrently; linear algebra; calculus.

Course information: Theory and practice of regression analysis: simple and multiple regression including testing, estimation and confidence procedures; modeling; regression diagnostics and plots; model selection; and non-parametric regression. Software implementation of the computations is also emphasized.

· Lectures:

Room: 301 Pupin Laboratories

Time: Mon and Wed 5:45-7:00 pm

· Textbooks:

Primary text:

o “Applied Linear Regression Models, third edition”, 4th ed., Kutner et al., Irwin. (Available at Labyrinth Bookstore on 112^th between Broadway and Amsterdam.)

Recommended SAS text:

o (Optional) “The little SAS book: a primer”, 3^rd ed., Delwiche and Slaughter, SAS publishing. (Available at Labyrinth Bookstore on 112^th between Broadway and Amsterdam.)

o (Optional) “Learning SAS in the Computer Lab”, 2^nd ed., Rebecca J. Elliott, Duxbury Press.

· Class website: http://courseworks.columbia.edu (Registered students: log on using your UNI and click on “W4315 Fall 2004”; auditing students: please talk to me after class so that I can list you as unofficial students to grant your access to this website.)

Instructor: Tian Zheng

Office: Room 1007, SSW Building

(1255 Amsterdam Avenue, between 121^st and 122^nd street)

Telephone: 212-851-2134

Email: tzheng@stat.columbia.edu

Office hours: Mon 1:30 – 3:00 pm, Room 1007, SSW Building

Teaching assistants:

Yu Zheng

Office: Room 1020 SSW Bldg

Email: zheng@stat.columbia.edu

Office hours: TBA

Grading: homework 40%, midterm 20%, and final 40%

Homework: problem sets will be assigned weekly (posted electronically on courseworks) and will be collected at the beginning of lecture on the due day. Each set of homework problems will consist of both paper-based practice and computer implementation of the theory and calculation covered in class or as reading assignments.

Exams: midterm (In class, Wednesday, October 20, 2004), final (TBA, during Dec 16-23, 2004). Do not make travel plan on exam dates.

Lecture schedule (as 8/27/04): most update schedule will be posted and modified on courseworks.

Lecture	Date	Reading	Assignments
1	Wednesday, September 08, 2004	Syllabus, Ch 1.1-1.5	W4315 pre-test
2	Monday, September 13, 2004	Ch 1.6-1.8 SAS Tutorial
3	Wednesday, September 15, 2004	Ch 2. 1-2.2	Pre-test due; Problem set 1
4	Monday, September 20, 2004	Ch 2.3-2.6
5	Wednesday, September 22, 2004	Ch 2.7-2.11 SAS	PS 1 due; Problem set 2
6	Monday, September 27, 2004	Ch 3.1-3.4
7	Wednesday, September 29, 2004	Ch 3.5-3.8	PS 2 due; Problem set 3
8	Monday, October 04, 2004	Ch 3.9-3.11
9	Wednesday, October 06, 2004	SAS implementation of Ch 3	PS 3 due; Problem set 4
10	Monday, October 11, 2004	Ch 4.1-4.7
11	Wednesday, October 13, 2004	Ch 5.7-5.13	PS 4 due; Problem set 5
12	Monday, October 18, 2004	Ch 6.1-6.7
13	Wednesday, October 20, 2004	Midterm
14	Monday, October 25, 2004	Ch 6.8-6.9, SAS, Ch 7.1-7.3
15	Wednesday, October 27, 2004	Ch 7.4-7.6	PS 5 due; Problem set 6
16	Monday, November 01, 2004	Academic Holiday
17	Wednesday, November 03, 2004	Chapter 8	PS 6 due; Problem set 7
18	Monday, November 08, 2004	Chapter 8, SAS
19	Wednesday, November 10, 2004	Chapter 9	PS 7 due; Problem set 8
20	Monday, November 15, 2004	Chapter 9, SAS
21	Wednesday, November 17, 2004	Chapter 10	PS 8 due; Problem set 9
22	Monday, November 22, 2004	Chapter 10, SAS
23	Wednesday, November 24, 2004	Chapter 11	PS 9 due; Problem set 10
24	Monday, November 29, 2004	Chapter 11, SAS
25	Wednesday, December 01, 2004	Chapter 13	PS 10 due; Problem set 11
26	Monday, December 06, 2004	Chapter 13, SAS
27	Wednesday, December 08, 2004	Chapter 14 (selected)	PS 11 due;
28	Monday, December 13, 2004	Review
	Final Date	Final

Class policy:

· Late homework is NOT accepted under any circumstances. One lowest homework grade can be dropped from final grade calculation.

· You may discuss on homework assignments or work together on some demanding problems. However, you must prepare and write-up your own solution. Any evidence of cheating will be reported.

· Prepare your solution to homework assignment in a neat and readable fashion. Include steps of proof or calculation; attach interpretation and explanation to computer outputs. No homework should exceed 10 letter-size pages (including computer outputs, thus be selective on what to turn in).

· No make-up or alternative exams will be given. Don’t make travel plan on exam dates.

· Emails related to homework problems will not be answered individually, instead, correspondences will be posted on the discussion board on the coursework website. Please check there before you send in emails to see whether your question has been answered.

· It is the responsibility of the students to come to lectures, finish homework before due dates, teach self the course material BEFORE the exams.

Explanation on the usage of SAS in homework problems: Most of the time, all you need to do is to modify the codes on the in-class handout for new datasets. This should not take much time each week. SAS programming will not be tested in exams. However, you should be able to read, understand and interpret SAS outputs.

If you really want to learn SAS:

1. A good website for beginner and intermediate user of SAS

http://www.ats.ucla.edu/stat/sas/

2. Both the instructor and primary teaching assistant will provide SAS help during their office hours.

3. It is important to know how to use SAS help.

4. It is recommended to work in groups.

5. Do not simply copy-paste others’ codes and produce something you don’t understand. It is CHEATING.

6. If the SAS in the computer lad that you go to is not working, please report immediately to AcIS.