Advanced Multivariate Statistical Methods
Instructor(s):
- Robert A. Henson, University of North Carolina at Greensboro
The purpose of this workshop is to discuss multivariate distributions and the role they play in modern methods for analyzing multivariate data. The course is designed to be an introduction to multivariate thinking, covering concepts that underlie many statistical models in widespread use today. The course will begin with a review of univariate linear models and an introduction to matrix algebra. Following that, the course will introduce the multivariate normal distribution and demonstrate its properties by covering classical multivariate methods such as principal components analysis, discriminant analysis, and factor analysis. To bring modern relevance to its use, the role of the multivariate normal distribution will then be described in methods such as structural equation modeling, multilevel modeling, and general linear mixed models. The course will conclude by discussing categorical multivariate distributions (such as those found in generalized linear mixed models and, more generally, finite mixture models) and classical multivariate techniques such as cluster analysis, multidimensional scaling, and correspondence analysis. A very strong background in statistics at least at the level of the ICPSR courses Regression Analysis II: Linear Models and Mathematics for Social Scientists II is necessary for this course. The level and breadth of coverage is roughly equivalent to that found in the following multivariate texts: Cooley and Lohnes, Multivariate Data Analysis; Tatzuoka, Multivariate Analysis; and Johnson and Wichern, Applied Multivariate Statistical Analysis
Fees: Consult the fee structure.
Tags: multivariate statistics, factor analysis, principal components analysis, cluster analysis, canonical correlation, multidimensional scaling, MANOVA, discriminant analysis,
Course Sections
Section 1 Location: ICPSR -- Ann Arbor, MI Date(s): June 23 - July 18 Time: 3:00 PM - 5:00 PM Instructor(s):
Syllabus: |