Econometric Methods for Business Research II

This is a course on multivariate statistical methods for econometricians and empirical researchers in other business disciplines. We begin with descriptive statistics (graphical and numerical summaries) for multivariate data, and review the basics of matrix algebra. We then develop some probability theory for random vectors and matrices (specifically the multivariate normal and related distributions), which gives us a foundation for inference about one, two, or many mean vectors (MANOVA). We then move on to exploratory and inferential multivariate methods including: discrimination and classification, canonical correlation, principal component analysis, factor analysis, clustering and multidimensional scaling. The emphasis throughout will be on the application of statistical methods to real data from marketing, management, accounting and finance.