Researchers frequently collect measurements on several variables simultaneously. Multivariate data analysis is focused on the analysis of these simultaneous measurements. It generalizes the ideas of univariate data analysis to create analyses that are more powerful both in a statistical as well as a practical sense. This power comes with the added costs of multivariate notation and computing effort. Since statistical software can readily handle the complex statistical calculations that are necessary, the goal for this course is provide students with the supporting knowledge to interpret these results, select appropriate techniques, and evaluate the strengths and weaknesses of these approaches. The course covers the following topics: multivariate normal distribution, general linear model, multivariate regression, MANOVA, principal components and factor analysis. Additional topics such as multinomial choice models, cluster analysis, and multidimensional scaling may be covered if time permits. This course is specifically designed for graduate students who intend to do empirical research.
Lecture: 100min/wk and Recitation: 50min/wk