P654 Applied Multivariate Statistical Analysis

Instructor: Jerome R. Busemeyer, Professor of Psychology

Office: Room 328 Psychology

Phone: 855-4882

email: jbusemey@indiana.edu

Notes will be available on World Wide Web from address below:
http://mypage.iu.edu/~jbusemey/home.html

Text:

Stevens, J. (1996) Applied multivariate statistics for the social sciences. Erlbaum.

Additional References

Course Content:

Introduction to the multivariate general linear model, principle component

analysis, factor analysis, latent structural equation modeling, categorical data

analysis, Bayesian Classification, and discriminant function analysis.

Applications selected from a wide range of areas including measurement theory,

causal modeling, signal processing, longitudinal data analysis, classification

theory, and repeated measures designs.

Evaluation: Grades will be based on bi-weekly homework assignments that involve statistical analyses using Matlab, SAS, and SPSS and written reports.

Click Here to Get an Account on STEEL

Requirements: P553 and P554 or equivalent training in basic statistical theory, regression, and analysis of variance. Programming will be covered in labs so that familiarity with statistical packages is not required.

Schedule (subject to announced changes):

1. Matrix Algebra

a) Interpretations of vectors and matrices

b) distance between vectors

c) Special matrices (identity, ones, square, symmetric)

d) addition, multiplication, transpose, inversion of matrices

e) eigenvalues and eigenvectors of a matrix

2. Random Vectors

a) Expectation Operation

b) Variance - Covariance Matrices

c) Variance of Linearly Transformed Vectors

d) sample statistics

e) Multivariate Normal Distribution

f) Multivariate Central Limit Theorem

3. Multivariate General Linear Model

a) multivariate regression

b) multivariate ANOVA

c) repeated measure analysis

4. Analysis of Covariance Structures

a) Principle Components Analysis

b) Exploratory Factor Analysis

c) Latent Structural Equation Models

d) Longitudinal Analysis

5. Categorical Dependent Variables

a) Categorical data analysis

b) Bayesian Classification

c) Discriminant function Analysis