EPSY 507

Multiple Regression

Semester: Spring, 1995
Professor: Patricia B. Elmore
Room:: Section 1, M W, 1:00-2:50 pm, Wham 317; Section 3, M W, 5:00-6:50 pm, Wham 317
Office: Educational Psychology & Special Education Department, Wham 223H
Office Phone: (618)453-6915
Office Hours: M W 4:00-5:00 pm, Tu 1:00-3:00 pm; and by appointment
E MAIL: PBELMORE@SIU.EDU
Textbook:: Pedhazur, E. J. (1982). Multiple Regression in Behavioral Research (2nd ed.). New York: Holt, Rinehart and Winston.
Reference:: SAS/STAT User's Guide - Volumes 1 & 2, Version 6, 4th Edition

The purpose of this course is to acquaint the student with the use of the general linear model as a data analytic tool. There will be two exams, a midterm and a final. Each exam will be worth 100 points. There will be a required research project for which the student will formulate a research problem, develop linear models consistent with the research questions, computer analyze the data using the models, and report the results of the analysis. This research project will be worth 50 points. There will be computer assignments worth a total of 100 points. The two exams, the computer assignments, and the one research project provide a possible total of 350 points. The grades that will correspond to point totals are as follows:

Heavy emphasis will be placed upon using the computer to analyze data. Numerous assignments will be made that will require the student to set up problems for analysis by the computer. The computerized results will be used for discussion in class, hence these analyses are also required assignments.

The course research project is a study designed and conducted on a topic in your area of specialization. The study should use one of the inferential statistical techniques discussed in the course. If the student uses a statistical technique covered in EPSY 506 (such as One-way ANOVA, t-test, or simple linear regression and correlation), the highest possible number of points assigned will be 40 points. To receive 50 points on the research project, a student must use a statistical technique covered in EPSY 507 and not covered in EPSY 506. A suggested outline for the project will be distributed and discussed in the course. The maximum number of double spaced typed pages is ten.

NOTE: STUDENTS WITH DISABILITIES REQUIRING ACCOMMODATIONS NEED TO INFORM THE INSTRUCTOR IMMEDIATELY.

TOPICS	CHAPTERS IN TEXTBOOK	DATE

I. Introduction to Linear Models	1	1/18
A. Simple Linear Regression	2	1/23
B. Introduction to CMS and SAS		1/25
C. Analysis of Variance vs. Correlational Methods		1/30
D. Multiple Regression Analysis - Two Independent Variables	3	2/1,6
E. General Method of Multiple Regression Analysis	4	2/8,13
1. General Linear Model and Least Squares Solutions
2. Full and Restricted Models
3. Full and Restricted Models
4. Error Sums of Squares (ESS)
5. Vector Algebra		2/15
F. Statistical Control: Partial and Semipartial Correlation	5	2/20,22,27
G. Stepwise Regression and Hierarchical Regression	6	3/1
Review for Midterm Exam		3/6
MIDTERM EXAM (100 points)	1-6	3/8

SPRING BREAK		3/13-17

II. Regression Analysis with Categorical and Continuous Predictor Variables
A. One-way ANOVA and A-priori Comparisons	9	3/20,22,27
B. Factorial Designs	10	3/29; 4/3,5,10,12
C. Trend Analysis	11	4/17,19
D. Analysis of Covariance	12,13	4/24,26
E. Repeatd Measures Designs	14	5/1
Review for Final Exam		5/3

RESEARCH PROJECT (50 points) - MUST BE TURNED IN MAY 3, 1995

FINAL EXAM (100 points)	1-6 & 9-14	5/8-12

NOTE: There will be a penalty for late computer assignments and research projects. Five points will be subtracted for each day the assignment or project is late.

Spring Conventions:: April 18-22 AERA Annual Meeting, San Francisco; April 27-30 ACA Annual Convention, Denver

Please submit any questions or comments to:

Patti Elmore

Updated 2/24/98 lv