DSC 410/510, 8:00-9:50am Tuesday/Thursday, 312 Lillis
Instructor: Iain Pardoe, 474 Lillis (346-3250), e-mail: ipardoe at
lcbmail.uoregon.edu
In order to make informed decisions, business decision-makers need to identify and quantify relationships between multiple information sources that are usually interrelated. Inflation, for instance, is related to taxes, interest rates, the money supply, oil prices, the business cycle, foreign wars, and a good deal more. Buyers' reactions to an advertisement may be related to the price of the item, competitors, warranty terms, previous experiences with the product, conversations with neighbors, and season of the year. A complete system of complicated relationships can exist in any single problem, with decisions affected by many variables that all interact with one another.
In striving to make some sense out of our environment, to seek some order to the chaos of inter-relationships, there are many approaches, including intuition, logical reasoning, and making simplifying assumptions. This course instead relies on the accepted framework of modern multivariate analysis techniques as it applies to business decisions. In particular, we consider conjoint analysis, cluster analysis, discriminant analysis, logistic regression, multidimensional scaling, correspondence analysis, and discrete choice analysis; all can be very powerful means of achieving parsimonious descriptions, explanations, and predictions of reality.
| Tuesday | 4-5pm | Also by appointment if you cannot make any of these times |
| Wednesday | 12-2pm | |
| Thursday | 4-5pm |
"Multivariate Data Analysis" (1998, 5th edition) by Hair, Anderson, Tatham, Black. The book is available from the University Bookstore on 13th and Kincaid for about $114, or you could try getting the much cheaper black and white paperback "international edition" from the "used & new" link on amazon.com.
The tentative course outline is as follows:
| Week | Date | Class | Topics |
|---|---|---|---|
| 1 | U 9/28 | 1. | Introduction to course and examining your data |
| H 9/30 | 2. | ||
| 2 | U 10/5 | 3. | Conjoint analysis |
| H 10/7 | 4. | ||
| 3 | U 10/12 | 5. | H 10/14 | 6. |
| 4 | U 10/19 | 7. | Cluster analysis |
| H 10/21 | 8. | ||
| 5 | U 10/26 | 9. | |
| H 10/28 | 10. | Discriminant analysis and logistic regression | |
| 6 | U 11/2 | 11. | |
| H 11/4 | 12. | ||
| 7 | U 11/9 | 13. | |
| H 11/11 | 14. | Multidimensional scaling and correspondence analysis | |
| 8 | U 11/16 | 15. | |
| H 11/18 | 16. | ||
| 9 | U 11/23 | 17. | |
| H 11/25 | - | Thanksgiving Holiday - no class | |
| 10 | U 11/30 | 18. | Discrete choice analysis |
| H 12/2 | 19. | ||
| 11 | U 12/7 | - | Final exam at 8:00 am |
Computing is an integral part of this course. We'll use "Microsoft Excel" as well as "SAS" statistical software. No prior knowledge of SAS is expected, and the software is installed on the computers in the LCB Business Technology Center. Information on the use of the Business Technology Center is available at http://lcb.uoregon.edu/btc/. You can also obtain a copy of SAS for use on your own computer from the University Computer Center - see http://sas.uoregon.edu/.
You will receive instruction on software use during the lectures. Information on the use of SAS is also available from the software help itself. All data-sets discussed in the text and to be used in the course are available at the course web-site (see below). There will also be handouts on the software at this web-site.
You are expected to attend all lectures, to prepare for examples that we'll do in class, to keep up with assigned reading of the text, to self-test your understanding of the material, to complete one (undergraduates) or two (graduates) projects, and (optionally) to take a final exam.
Class time will be spent discussing various multivariate statistical techniques (see "Course Outline" above), with emphasis on demonstrating how to use the techniques with real data. Some of the real data that we'll use will come from you, so you'll be asked to prepare for some lectures by obtaining data that we'll then analyze in class. Somewhere in the region of 30 to 35 pages of reading from the text-book will be assigned per week. This may be supplemented with other sources of material.
Graded homework is a required part of this course, and will be entirely e-mail based. Also, questions to self-test your understanding of the material covered will be assigned, and suggested solutions will be provided to allow you to check your answers. Preparing for class (by, for example, obtaining some data) and class attendance will form part of your grade. Finally, there will be one or two projects based on analyzing data that you obtain. (Undergraduates are only required to do one project, while graduate students need to do two.) You can do the projects by yourself or in a group of up to four. For those doing project-work in groups, each member of the group contributing to the project gets the same score; if you don't contribute to your group's project you get zero.
There is no mid-term exam, but there is an optional final exam for those wanting an "A" from the class (see "Grading" below). The final exam is scheduled for 8:00-10:00am Tuesday, December 7; this will be comprehensive and so cover everything from the term.
The class will be graded on the A-F scale using the following guidelines:
| Graduate students: | Undergraduates: |
|---|---|
| Homework, class preparation, attendance: 200 | Homework, class preparation, attendance: 300 |
| Project 1: 250 | Project: 400 |
| Project 2: 250 | |
| Comprehensive final exam: 300 | Comprehensive final exam: 300 |
To get a "B" you'll have to score about 670 points, while a "B+" will require about 800 points, an "A-" will require about 870 points, and an "A" will require about 930 points. Anything less than 500 will get a "C" or worse. Note that an unsatisfactory project (or failing to do a required project) will score minus 100 points so that it is not possible for graduate students to get a "B" with just one satisfactory project, or for undergraduates to get a "C" without a satisfactory project.
The web-site for this course is at http://lcb1.uoregon.edu/ipardoe/teaching/dsc510/. There you will find course announcements, assignments, copies of handouts, and data used in class.
The University of Oregon is an equal opportunity, affirmative action institution committed to cultural diversity and compliance with the Americans with Disabilities Act. If you have a documented disability and anticipate needing accommodations in this course, please make arrangements to meet with the instructor soon. Please request that the Counselor for Students with Disabilities send a letter verifying your disability. This syllabus will be made available in alternative formats upon request.
The views and opinions expressed in this page are strictly those of the page author. The contents of this page have not been reviewed or approved by the University of Oregon.
