# Psychology 777

PSYC 777: Multivariate Analysis of Behavioral Data

Prerequisites: Graduate standing and PSYC 610 or equivalent, or admission into Psy.D. program, or permission of instructor

Credit Hours: (3)

This course will provide a treatment of the most commonly used multivariate techniques for quantitative analysis of behavioral data. Students will learn the conceptual basis for these techniques, as well as receive instruction for conducting their own analyses using the SPSS software package.

Detailed Description of Course

The course will cover a variety of multivariate techniques for the analysis of behavioral data. An overarching theme of the course will be the place of each technique within the General Linear Model for conceptualizing relationships among variables. The order of topics covered in the course is presented below:

1.      Review of basic statistical techniques

a.      Descriptive statistics

b.      The logic of statistical inference.

c.       T-test and One-way Analysis of Variance (ANOVA)

d.      Correlational techniques

2.      Simple regression

a.      Least-squares solution for deriving a regression equation.

b.      Standard error of estimate.

c.       Analysis of Variance as a special case of regression.

3.      Multiple regression

a.      Least-squares solution for regression equation with multiple predictors.

b.      Unstandardized and standardized regression coefficients.

c.       Forward, Backward, and Stepwise algorithms for selecting predictors.

d.      Hierarchical regression to test the unique contributions of predictors.

e.      Partial and semi-partial correlation.

f.        Testing mediation effects using regression.

g.      Analysis of Covariance

4.      Testing Regression models containing categorical or continuous predictors

a.      Dummy, effect, and contrast coding to test categorical predictors.

b.      Effect coding to test moderation effects among predictor variables.

5.      Logistic Regression

a.      Regression with a categorical criterion variable.

b.      Comparison of Logistic Regression with multiple regressions.

c.       Testing interaction effects using Logistic regression.

6.      Factor Analysis

a.      Principle Components Analysis for extraction of factors.

b.      Interpretation of Eigenvalues and communities

c.       Orthogonal methods for factor rotation

d.      Non-orthogonal methods for factor rotation

e.      Exploratory vs. confirmatory factor analysis.

7.      Introduction to casual modeling

a.      Interpreting path diagrams.

b.      Conceptual basis of Path Analysis and Structural Equations Modeling

8.      Multivariate Analysis of Variance

a.      Assessing group differences with multiple dependent variables.

b.      Hotelling’s T-squared

c.       MANOVA

9.      Cluster Analysis

a.      Methods for assigning cases to clusters

b.      Conceptual limitations of Cluster Analysis

Detailed Description of Conduct of Course

Classes will be held in a computer lab where every student has access to the SPSS software package. The course will be conducted using two primary modes of instruction. The conceptual basis of each technique will be covered in a lecture-based format supplemented by immediate exposer to the use of that technique using SPSS. Whenever possible students will move from one step in the statistical output to another while the instructor explains the rationale behind each piece of information displayed. After the rationale for each technique has been presented students will be given a data set on which to apply the technique they have just learned about. In this way, students will receive immediate feedback about (a) the degree to which they understand the material conceptually and (b) their ability to use SPSS to apply the technique to real data.

Goals and Objectives of the Course

The primary goal of the course is to provide students with a conceptual understanding of the statistical techniques that are most commonly used to address research questions in Psychology. Students completing the course will be able to apply these techniques to their own data and to review critically the results sections of articles that use these techniques.

Assessment Measures

The grade for the course will be based on a number of assessment measures including in-class examinations, take-home examinations that test the ability to analyze data and write a results section based on these analyses, a semester-long paper assignment, and participation in class discussions of readings for the course.

Other Course Information

None

Review and Approval

December 2007