Course Syllabus

Psychology 480.14 --- Multivariate Topics --- Spring, 2000

Tu,Th 9:35 am -- 10:50 am --- DeGarmo 18

Instructor

Office Hours

Dr. Matthew Hesson-McInnis
415 DeGarmo Hall
309.438.7266 (Voice)
309.438.5789 (Fax)
mshesso@ilstu.edu
http://www.ilstu.edu/~mshesso/
Wednesday, 9:00 am to 10:00 am
Wednesday, 3:00pm to 4:00 pm
or by appointment
and subject to change

Texts

Bollen, K. A. (1989). Structural Equations with Latent Variables. New York: Wiley.

Approximate Course Schedule

Week of

Topic

1/18 Matrix Algebra Review
1/25 Multiple Regression and the GLM
2/1 Exploratory Factor Analysis
2/8 Bollen, Chapters 1--3
2/15 Bollen, Chapter 4
2/22 Bollen, Chapter 5
2/29 Bollen, Chapter 6
3/7 Bollen, Chapter 7
3/14 Spring Break
3/21 Bollen, Chapter 8
3/28 Bollen, Chapter 9
4/4 Fit Statistics (Student Presentations)
4/11 Fit Statistics (Student Presentations)
4/18 Model Identification \& Causation Revisited
4/25 Student Literature Presentations
5/2 Student Literature Presentations
5/9 Oral Final Exam

Policies and Procedures

Course Components

Homework --- 20%

Students are expected to complete weekly homework assignments on time. The assignments are to be word processed when appropriate. Some assignments, however, would be difficult to word process and may be hand written (such as matrix algebra). Each student may turn in any one assignment one week late without asking for permission beforehand. Word processed assignments should be spell checked and proof read for grammar; hand written assignments should neat, clear, and very legible.

Course Project --- 20%

Each student will complete a course project, typically involving the analysis of a data set using one of the techniques studied during the course. Students should start looking for a good data set immediately. If you do not have a suitable data set already, you may want to talk with your advisor, other faculty, or the instructor. One could also tackle a project geared to address a more methodological issue. The course project does not need to be a small dissertation; extensive literature reviews are not necessary. Students are expected to discuss their projects with the instructor frequently.

Presentations --- 20%

Students are expected to make several presentations during the course of the semester.

Class Participation --- 20%

As a seminar class, students are expected to participate very actively each day. This participation can take the form of making comments, asking questions, or responding to another students' question or comment. Students will need to keep a record of their participation, in the form of a journal, to be turned in at the end of the semester. It is especially important that students participate during other students' presentations.

Students are expected to attend every class session, although it is to be expected that conferences and other professional activities often take us away from the classroom. Students do not need to ask for permission to be absent from class when they are not scheduled to present, but excessive absences will have a negative effect on the class participation grade.

Students are also expected to read the appropriate chapters in the text, as well as the journal articles supplied by all students (to be placed on reserve later in the semester) so as to be prepared to discuss the material and participate fully in the seminar. Especially with the textbook, students do not need to read every word laboriously and repeatedly until every concept is crystal clear. If we could do that, none of us would need this class. You should read the chapters closely enough, however, to formulate coherent questions or at least to know where you "lost it."

Final Oral Exam --- 20%

Students will make a brief (5 minute) presentation about their course project and answer questions from the instructor and other students as an oral final exam.

Grading Policies

All assignments will be graded on a check-plus, check, or check-minus system. Assignments that are well written, coherent, complete, and accurate will be given a check-plus. Assignments that are basically acceptable but with minor errors or omissions will be given a check, and assignments that have major errors, have substantial omissions, or are poorly prepared will be given a check-minus.

Final grades will be assigned using the "camel" distribution. Graduate courses typically exhibit a bimodal distribution in which the students form two separate clumps: "A" students in the higher clump; "B" students in the lower clump, and the occasional "C" student as an lower-tail outlier.

Computers, Software, Labs

The psychology maintains a small lab with marginally adequate (i.e., old) PowerMac computers. Although the computers are problematic in many regards, some of them do have EQS installed (and a full version that allows you to save models, print, etc...). Multivariate Software, Inc. has made a demo-version of EQS available for various platforms. Students may want to dowload this demo from the web page http://www.mvsoft.com/index.html, but you should be aware that the demo version does not let you print output. You can order your very own copy of EQS from this website if you've got the bucks, but the demo version can let you debug your EQS commands and procedures at home before going to the lab to run the program. Do be aware that the department can only run one coy of EQS at a time regardless of the number of computers on which the software is installed. You will also find that running EQS is one of the most efficient ways of making your computer crash.

At some point this semester, the University will be installing several copies of AMOS in select labs. AMOS is more limited than EQS (or LISREL) but is easier to use for simple models. As an alternative to bridge the gap between EQS and a future installation of AMOS (WIntel only), I will be providing an SPSS Matrix Syntax program for estimating covariance structure modeling parameters. Although this code will be fairly limited

Academic Dishonesty

Academic dishonesty, in any form, will not be tolerated. Copying homework, presenting someone else's work without citation as one's own, or other forms of plagiarism, cheating, or academic dishonesty will result in a grade of ``F'' and referral to the student judicial office.

Last modified Tuesday, January 18, 2000 9:33 AM