Linear Statistical Models is a two course sequence that covers linear statistical models. Regressionis taught in the Winter quarter and covers, primarily, topics in multiple regression. Analysis of Variance, taught Spring quarter, covers experimental design and analysis of variance.
The purpose of these courses is to provide solid and comprehensive training in quantitative methods. It is designed to prepare students to carry out and interpret research using a variety of quantitavie and statistical methods. It will cover key aspects of research design and statistical inference involving linear statistical models. The two quarters provide an integrated and unified approach to the application of linear statistical models in regression, analysis of variance, and experimental and quasi-experimental designs. This integrated approach will give students an understanding of how the analytic approaches are closely connected and will help them develop flexibility in applying quantitative methods correctly to a wide range of research problems. This sequence will also provide a strong foundation for further training in advanced statistical methods.
Satisfactory completion of Introduction to Research Design & Statistics or a passing score on the Linear Models Screening Exam.
Agresti, A. & Finlay, B. (1997). Statistical methods for the social sciences (3rd edition). New Jersey: Prentice Hall. ISBN 0-13-526526-6 (Ed230B)
Kirk, Roger E. (1998) Experimental Design: Procedures for the Behavioral Sciences, Third Edition. Monterey, California: Brooks/ColePublishing. ISBN 0-534-25092-0 (Ed230C)
Recommended readings for the aggressive graduate student:
Pedhazur, E.J. (1997). Multiple regression in behavioral research, Third edition. New York: Harcourt Brace College Publishers. ISBN 0-03-072831-2
Hamilton, L.C. (1992) Regression with graphics. Belmont, CA: Wadsworth. ISBN 0-534-15900-1
Chatterjee, S., Hadi, A., & Price, B. (2000) Ression analysis by example. New York: Wiley..] ISBN 0-471-31946-5
Weisberg, S. (1985) Applied linear regression. New York: Wiley. ISBN 0-471-87957-6
Regression with Stata Webbook (Ed230B)
Chen, X., Ender, P.B., Mitchell, M. & Wells, C. (2001) Regression with Stata Webbook Los Angeles: UCLA Academic Technology Services.
In each of the two quarters, Winter and Spring, there will be four assignments and at least two quizes. The assignments will consist of a mixture of data analysis projects, theoretical interpretations, and critiques of data analyses and published articles from the current research literature. Data analysis projects may be done using any of the major statistical packages, such as, Stata, SAS or SPSS. Students may work cooperatively in small groups (maximum of 3) on portfolio projects. The names of all the participants on a given project should appear on the project report.
Take home quizzes will be handed out several times each quarter so that the students and instructor will be able to monitor the students' progress to date. Students must work independently on their quizzes. There will be a final examination in the Ed230C course.
The instructors may also administer in-class tests should the instructor feel they are needed.
The final course grade for each quarter will be given based upon an evaluation of the assignments and performance on the quizzes and tests.
Internet Access and World Wide Web
Students will be required to have network access by one of the following means: 1) GSE&IS Computer Labs, 2) their own departmental computer labs with Internet access, or 3) their own personal computer at home with Ineternet access. In addition, students will need to use a World Wide Web browser. Recommended browsers include Internet Explorer or Netscape Navigator.
The course web pages will contain important course information, including assignments, datasets, examples, helpsheets, computer printouts, class discussion forums, and lecture notes. The URL is http://www.gseis.ucla.edu/courses/ed230bc1/230bc.html.
Proposed Course Schedule
Schedule of Readings
|Week||Topic||A & F|
|1-3||Simple Linear Regression (SLR)|
Overview and rationale, basic model, assumptions,
SLR residuals, nonlinearity, transformations.
Partial regression coefficients, partitioning
of variation, matrix approach, variable
selection, multicollinearity cross validation,
|Ch. 10, 11|
|7-8||Causal models and path analysis||Ch. 16|
|9-10||Analysis of covariance,|
aptitude treatment interaction
Completely randomized designs
|Ch. 5; 5.1-5.6, 5.9|
|2||Multiple comparisons||Ch. 4; 4.1-4.5, 4.7-4.8|
|3||Completely randomized factorial designs||Ch. 9; 9.1-9.6, 9.8, 9.10-9.11, 9.15
Ch. 10; 10.1-10.3
|4||Within group designs|
Randomized block designs
|Ch. 7; 7.1-7.5, 7.7, 7.10|
|5||Randomized block factorial designs||Ch. 10; 10.5-10.7|
|6||Split Plot Factorial designs||Ch. 12; 21.1-12.12, 12.15|
|Ch. 11; 11.1-11.3, 11.6, 11.9|
|8||Unbalanced designs||Ch. 9; 9.14|
|9||Analysis of covariance||Ch. 15; 15.1-15.6, 15.9, 15.12|
|10||Review for final|
Use of Teaching Assistants
Teaching Assistant (TAs) are assigned by the Department of Education to the Social Research Methods Division to assist students in our courses. This assistance covers explanations of concepts and procedures, help with assignments and computer runs, and review for examinations. TAs are assigned to particular courses and their office hours are listed on the TA office door. These are the times they are available to help you. In addition, you may use any other "on duty" TA to receive assistance when your class TA is "off duty" or tied up with other students. TAs, however, are not available to provide assistance when they are "off duty". At that time, they are just like you, working on courses and research projects. Out of courtesy, please do not ask for a "minute" of their time; go the TA office to discuss your work with an on duty TA. If you do ask for help from and "off duty" TA, he or she has been instructed to refer you to the TA office.
Linear Statistical Models Course
Phil Ender, 10Jan00