Medical Multivariate Statistics and Statistical Software

Author: Sun Shanggong
Publisher:
Publishing Date: 2000-10-01
Features: The purpose of writing this book is: (1) Practicality and accessibility. Statistics has a strong dual nature: it is both concrete and abstract, practical and theoretical, easy to learn and difficult to learn. For example, it is very easy for us to understand and calculate the "arithmetic mean," but to have a good understanding of such fundamental terms as "theoretical mean" and "conditional mean" is much more difficult, let alone the computational challenges. However, most of these difficulties come from mathematics. This book does not introduce the derivation of mathematical formulas and does not intend to include overly complex mathematical formulas. Instead, it emphasizes the meaning of statistical models, basic concepts, and the understanding of statistical terminology, with particular focus on interpreting computational results. (2) It is written in conjunction with internationally recognized statistical software SAS and SPSS. The terminology, symbols, and formulas introduced in the book are as consistent as possible with those in SAS and SPSS. However, the book primarily focuses on SAS statistical software, with some content centered on SPSS. (3) It incorporates as many internationally mature new methods as possible. For example, "latent variable analysis" has developed rapidly since the 1980s and is highly popular among practitioners. There are already over 50 books on latent variables published abroad, but few formal publications in China. The "uniform experimental design" (a new multivariate statistical experimental method) created by Chinese mathematician Fang Kaitai and internationally popular, as well as certain statistical methods created by the author, will also be included in the book. (4) This book is not merely an introduction to statistical methods but also aims to share the author's decades of experience and insights from medical statistics work. The Encyclop?dia Britannica defines statistics as "the art of collecting and analyzing data." The author strongly agrees with the statement that "statistics is an art." In the author's experience, statistical analysis is a highly creative discipline. Those unfamiliar with statistical analysis often believe that after collecting data, they can simply select a method in a computer program and obtain results within seconds, thinking the analysis is complete. This is a very amateurish view or approach. The author's experience is that to truly complete a specific statistical analysis task, it is unlikely to be done well in just a few days or months. Statistical analysis, especially multivariate statistical analysis, is highly flexible. We should identify the most suitable analytical method for the data from among the possible multiple methods and results.