QATAR UNIVERSITY

COLLEGE OF ARTS AND SCIENCES

DEPARTMENT of MATH, STAT& PHYSICS

STATISTICS PROGRAM

STAT 459: Multivariate Analysis

Spring 2010

Course Information

Course Title: Multivariate Analysis

Course Number: STAT 459

Credit Hours: 3 Credit Hours

Course Status: Program Compulsory Course

Time & Location: Lecture, 10-11, (Sun., Tue. Thur.) C201

Lab, 11-12, (Sun.) A211

Required Text: Applied Multivariate Statistical Analysis. Richard A. Johnson and Dean W. Wichern, 6th edition, Prentice Hall.

.

Faculty Information

Name: Dr. Ayman Baklizi

Academic Title: Associate Professor

Office Location: C213, Arts and Sciences Building

Telephone Number: 2185

Email Address:

Office Hours:

Course Description

The course discusses the analysis of multivariate data. Multivariate distributions and inference about means are considered. Techniques like principal components, factor, cluster and discriminant analyses were studied with examples. The use of computer packages is emphasized in this course. Real life data are often used to illustrate the power and applicability of multivariate methods.

Course Objectives

The course aims at:

1-  Enabling the students to organize multivariate data into an array and calculate its mean vector, covariance matrix, and generalized variance.

2-  Introducing the students to Multivariate Normal (mn) distribution and distribution of sample mean and covariance from an mn distribution and the associated inferences.

3-  Helping students acquire knowledge of principal components analysis (PCA) and Factor Analysis (FA).

4-  Introducing the elements of discriminant anlaysis and canonical correlation.

5-  Familiarizing the students with the concepts of cluster analysis and multidimensional scaling.

Learning Outcomes

By the end of this course, students will be able to:

1-  Calculate statistical quantities like the multivariate mean, covariance matrix and the generalized variance.

2-  Test hypotheses about the parameters of the multivariate normal distribution.

3-  Compare several multivariate normal mean.

4-  Perform principal component analyses and factor analyses.

5-  Apply the techniques of canonical correlation and discriminant analysis.

6-  Apply the techniques of cluster analysis and multidimensional scaling.

Content Distribution

Lectures and Lab schedule

Week / Week Start / Topics
1 / Feb. 21 / Multivariate Data: Summary Statistics
2 / Feb. 28 / Multivariate Data: Plots
3 / Mar. 7 / The Multivariate Normal Distribution
4 / Mar. 14 / Inference about multivariate means
5 / Mar. 21 / Multivariate analysis of variance
6 / Mar. 28 / Examples and Discussion
7 / Apr. 4 / First Exam: 6-4-2010
8 / Apr. 11 / Holiday
9 / Apr. 18 / Principal Components Analysis
10 / Apr. 25 / Factor Analysis
11 / May 2 / Examples
12 / May 9 / Second Exam: 11-5-2010
13 / May 16 / Canonical Correlation
14 / May 23 / Cluster Analysis
15 / May 30 / Examples

Delivery Methods

1.  Course documents, lectures

2.  Interactive teaching

3.  Lab activities

Learning Resources & Media

·  Class meetings: with expected participation and discussion.

·  Assignments and quizzes.

·  Minitab and R: to understand and interpret solutions of real life problems.

Assessment Policy and Tools

This course will be assessed by the active participation of the students during lectures, assignments and exams:

Assessment

/

Points

Exams

/

First Exam: 20 points

Second Exam: 20 points

Final = 40 points

Assignments

/

20 points

Exams

First Exam: 6-4-2010

Second Exam: 11-5-2010

Final Exam: As Scheduled

Grades for the course will be assigned as follows: 90-100= A, 85-89.9= B+, 80-84.9= B, 75-79.9= C+, 70-74.9=C, 64-69.9=D+, 60-64.9=D, 59.9-0=F.

Learning Activities and Tasks

1.  Lab practical work

2.  In-class group work

3.  Take-home assignments

Course Regulations

Student Responsibilities and Attendance Policies and Procedures

Class attendance is compulsory. In accordance with University regulations, a student’s absence cannot exceed 25% of the total number (entire semester) of class meetings. If your absence rate exceeds 25%, including both excused and unexcused absences, you will NOT be allowed to take the final examination and will receive an ‘F barred’ grade for the course.

Students are expected to be punctual (every 3 late class arrivals will be counted as 1 class absence) in class attendance and to conduct themselves in an adult and professional manner.

Homework assignments and library assignment should be worked independently. Exchanging ideas are permitted orally but don't require any kind of copying.

Homework assignment should be submitted in organized way and any late assignments may be assessed and corrected but the grade will be zero.

Plagiarism (Academic Dishonesty)

All students are expected to turn in work that is their own. Any attempt to pass off another's work as your own will constitute an "F" in the entire course.

Using part of, or the entire work, prepared by another or turning in a homework assignment prepared by another student or party are examples of plagiarism.

You may discuss assignments and projects with each other, but you should do the work yourself. In the case of group projects, you will be expected to do your share of the work. If you use someone else's words or ideas, you must cite your sources.

Plagiarism is considered a serious academic offence and can result in your work losing marks or being failed. QU expects its students to adopt and abide by the highest standards of conduct in their interaction with their professors, peers, and the wider University community. As such, a student is expected not to engage in behaviours that compromise his/her own integrity as well as that of QU. You may discuss assignments and projects with each other, but you should do the work yourself. In the case of group projects, you will be expected to do your share of the work. If you use someone else's words or ideas, you must cite your sources.

Plagiarism includes the following examples and it applies to all student assignments or submitted work:

·  Use of the work, ideas, images or words of someone else without his/her permission.

·  Use of someone else's wording, name, phrase, sentence, paragraph or essay without using quotation marks.

·  Misrepresentation of the sources that were used.

·  For further information see: http://www.plagiarism.org/

The instructor has the right to fail the coursework or deduct marks where plagiarism is detected

Classroom Discipline

The use of mobile telephones inside the classroom is NOT allowed.

Any student disciplinary issues, which may arise, will be referred to the head of the Department.

Additional Sources

Reference Books

1- Multivariate Statistical Methods: A Primer

Bryan Manly, 2nd edition, 1994, Chapman & Hall/CRC.

2-Multivariate Statistical Methods and It’s Applications.

A. C. Rencher, 1998, John Wiley and sons, Inc.

Online Resourses

www.statsci.org

7

Appendices

Matrix of Objectives and Outcomes

OBJECTIVES / LEARNING OUTCOMES / Assessment Tools
Enabling the students to organize multivariate data into an array and calculate its mean vector, covariance matrix, and generalized variance. / Calculate statistical quantities like the multivariate mean, covariance matrix and the generalized variance. / Exams
Homworks
Introducing the students to Multivariate Normal (mn) distribution and distribution of sample mean and covariance from an mn distribution and the associated inferences. / Test hypotheses about the parameters of the multivariate normal distribution.
Compare several multivariate normal mean. / Exams
Homeworks
Helping students acquire knowledge of principal components analysis (PCA) and Factor Analysis (FA). / Perform principal component analyses and factor analyses. / Exams
Homeworks
Introducing the elements of discriminant anlaysis and canonical correlation. / Apply the techniques of canonical correlation and discriminant analysis. / Exams
Homeworks
Familiarizing the students with the concepts of cluster analysis and multidimensional scaling. / Apply the techniques of cluster analysis and multidimensional scaling. / Exams
Homeworks

Assignments Rubric

Total Score (20)

CATEGORY / 4 / 3 / 2 / 1
Organization / Information is very organized with well-constructed paragraphs and subheadings. / Information is organized with well-constructed paragraphs. / Information is organized, but paragraphs are not well-constructed. / The information appears to be disorganized. 8)
Amount of Information / All topics are addressed and all questions answered with at least 2 sentences about each. / All topics are addressed and most questions answered with at least 2 sentences about each. / All topics are addressed, and most questions answered with 1 sentence about each. / One or more topics were not addressed.
Mechanics / No grammatical, spelling or punctuation errors. / Almost no grammatical, spelling or punctuation errors / A few grammatical spelling or punctuation errors. / Many grammatical, spelling, or punctuation errors.
Diagrams & Illustrations / Diagrams and illustrations are neat, accurate and add to the reader's understanding of the topic. / Diagrams and illustrations are accurate and add to the reader's understanding of the topic. / Diagrams and illustrations are neat and accurate and sometimes add to the reader's understanding of the topic. / Diagrams and illustrations are not accurate OR do not add to the reader's understanding of the topic.

7