No label defined (Q4942159)

The second edition of Ryan's book is significantly enlarged compared with the first edition published in 1988. The book is divided into three parts. The first part makes available the methods, the second part treats almost exclusively control charts, and finally the third part is devoted to various topics as for example linear regression, experimental design and the Taguchi approach. Each chapter is completed by a summary, references and exercises. Most of the methods introduced are illustrated by numerical examples. The first part (73 pages) begins with a short introduction to the concept of quality, followed by a list of basic (graphical) tools for representing complex situations in a simpler way in order to get insight and improve the situation. This part is concluded with the third chapter being devoted to stochastic models (distributions) and statistical methods (estimation and testing). The second part (239 pages) starts with a detailed treatment of control charts for measurements (\(s\)-chart, \(R\)-chart, \(s^2\)-chart, \(\overline x\)-chart, acceptance control chart and difference control chart) based on samples (subgroups). Additionally, the topics ``normality'', ``independence'' and ``measurement error'' are discussed. Next, control charts for individual observations are introduced followed by control charts for attributes (\(np\)-chart, \(p\)-chart, \(Q\)-chart, \(c\)-chart, \(u\)-chart, \(d\)-chart). The next chapter describes process capability indices \((C_p,C_{pk}, C_{pm},C_{pmk})\) and their use in industry, followed by an introduction to CUSUM-charts and EWMA-charts. So far only the case of one-dimensional quality characteristics is considered. Chapter 9 is devoted to multivariate control charts. After discussing the \(T^2\)-control chart for controlling the process mean and charts for controlling process variability, multivariate CUSUM and EWMA charts are briefly considered. A variety of topics, among them pre-control, short run production, correlated process variables, multiple stream processes and the economic design of control charts are mentioned in Chapter 10, which completes the second part. The third part (211 pages) starts with some more advanced graphical tools, e.g., stem-and-leaf, boxplot, normal probability paper, multi-vari chart, and turns to linear regression and experimental design. However, it should be noted that the book contains only rudimentary treatments of these topics which cannot replace more detailed studies. Chapter 14 is devoted to the Taguchi approach, which is critically analyzed and compared with alternative approaches. It follows the so-called evolutionary operation (EVOP) which is based on experimental designs with moderate levels which can be run during the normal production and which aims at gradually improving continuous processes. Next, the analysis of means (ANOM) which in some sense can be looked upon as an alternative to the analysis of variances is introduced and its application in various fields is illustrated by examples. The heading of the final chapter reads: Using Combinations of Quality Improvement Tools, and deals with various programs developed and implemented by industry, among them Motorola's well-known six sigma program.