UGA Bulletin Logo

Applied Correlation and Regression Methods in Education


Course Description

Nonexperimental and quasi-experimental research studies, including simple and multiple regression techniques, nonorthogonal analysis of variances, correlation techniques, and analysis of covariance.


Athena Title

App Corr and Regres Mth Educ


Equivalent Courses

Not open to students with credit in ERSH 8320E


Prerequisite

ERSH 8310


Semester Course Offered

Offered every year.


Grading System

A - F (Traditional)


Course Objectives

On successfully completing the course students should be capable of designing studies and analyzing data sets using regression techniques. Furthermore, students should feel comfortable reading journal articles which have used basic regression techniques in analyzing data sets. A word of caution however is needed. Students are not likely to fully understand all research studies using regression techniques. Regression analysis is a very powerful technique which can be appropriately used to analyze very complex problems. This course is designed to introduce students to the major concepts of regression analysis. For more thorough understanding of the analysis procedure further study will be necessary.


Topical Outline

Correlation and Simple Linear Regression Representing the relationship between two variables Identifying the best fitting straight line Estimation and hypothesis test Appropriateness of the model Prediction Summarizing relations with correlations Regression analysis with a qualitative independent variables Developing and interpreting a qualitative model Regression analysis with two or more quantitative explanatory variables Overview Meaning of the model Assumptions Multiple Correlation Coefficient Testing the overall model Partial F-test R_increase (part correlation squared) Partial Correlation Coefficient Testing for an interaction Identifying the "best" subset of independent variables Regression analysis with one continuous and one categorical independent variable Overview Identifying separate regression equations. Testing for the equality of the slopes Johnson-Neyman Technique Testing for the relationship between the continuous and dependent variables Testing for differences in the intercepts Identifying specific differences in treatments Regression Analysis with two categorical variables independent variables Factorial ANOVA Weighted and Unweighted marginal means Hierarchical vs Unique solutions


Syllabus