Multivariate Statistical Methods in Quality Management

The word multivariate not only means many variables, but also means that these variables might be correlated. Statistical method refers to a range of techniques and procedures for analyzing data, interpreting data, displaying data, and making decisions based on data. Therefore, multivariate statistical methods are collections of methods and procedures that analyze, interpret, display, and make decisions based on multivariate data. The foundation of multivariate statistical methods was gradually developed, starting from the beginning of the twentieth century. Multivariate statistical methods are playing important roles in the areas of psychology, sociology, and many others.
Statistical methods are playing important roles in quality engineering practice. Statistical process control and sampling inspection methods are among the first applied statistical methods developed for quality assurance practice. Statistically designed experiments (design of experiments) and regression analysis are among the most important work horses in Six Sigma movement and their applications contributed a great deal to improving the quality and profitability of many prominent multinational corporations in the world.
However, almost all the popular statistical methods that are used in quality engineering practice are univariate statistical methods. Very few industrial engineering departments in the United States offer multivariate statistical methods courses. Currently, the certification requirements of certified quality engineers and certified Six Sigma black belts do not include multivariate statistical methods in the body of knowledge.
On the other hand, with the development of sensor technology and computer information technology, gigabytes upon gigabytes of data are...