Confronted with a large data set one of the first wishes is to get a (graphical) overview that leads to insight in to the data and, hopefully, the system from which the data was obtained. Multivariate statistical analysis and multidimensional data analysis may help to fulfill this wish via dimension reduction and modeling of latent structures. Multivariate analysis studies the relationships between random variables (variates) of one or more sets on the basis of a sample of objects (patients, fields, plants). With n objects and, in total, m variates, the data are typically collected in an n x m matrix. The term multidimensional data analysis is often used when the emphasis is on the relationships between the objects (such as in multidimensional scaling), when the role of rows and columns of the data matrix is symmetric and also for methods that do not start with strict distributional assumptions.
Students will learn the key concepts and how to use the associated techniques for quantitative, ordinal and categorical data. Emphasis will be on interpretation and communication of results.
The course will consist of a two-hour lecture and a two-hour practical per week, for 14 weeks. In week five, students will be asked to analyze a practical data set, and hand in a report in week seven. There will be a written exam at the end of the course. Assessment of a student will be based on the case study report (1/3) and the written exam (2/3), with a minimum grade of 5 for the latter.
Johnson, R.A., and Wichem, D.W. 2007, Applied multivariate statistical analysis 6th edition. Pearson Higher Education, Paperback