Six Sigma is a proven, powerful methodology for significant quality improvement. Thousands have been trained in it, learning fundamental statistical tools and becoming Black Belts and Master Black Belts. Yet little has been written for those desiring to go beyond their basic training to enhance their problem solving techniques and project improvement efficiency.
Beyond Six Sigma Statistics helps fill that gap by providing enhancements to the traditional training, demonstrating useful advanced techniques on all the Six Sigma statistical tools and revealing important insights that will empower Black Belts to be more effective.
Particularly targeted for engineers, Beyond Six Sigma Statistics is packed with ideas that involve the practical application of statistics, yet does not demand a strong statistical background to understand. As such, any operational Black Belt will find many useful ideas for process improvement, and Master Black Belts will discover new and better ways to approach old problems.
Writing from 30 years of practical industrial statistical consulting, including 10 years as a Master Black Belt, Dockendorf reveals many additional techniques that could enhance a Black Belt's data analysis. These include little known methods, new statistical tools, and revelations about the weaknesses of some of the standard data analyses approaches.
For example, he demonstrates that most correlation studies give erroneous results, but a correctly run correlation study can actually provide spec relief. And he shows why nearly all MSA and regression results are substantially incorrect, while providing methods to improve those results.
New statistical tools include a novel and nearly perfect capability metric for Six Sigma, one that rewards both centering the mean and achieving minimum variation. Also, he demonstrates how data from batch processes can be handled in a statistical meaningful way, including performing appropriate hypothesis tests. And he provides valid methods for determining an appropriate sample size and sampling frequency for SPC.
Additionally, the book contains commentary on many of the standard statistical techniques- such as what can truly be stated after a hypothesis test has been run or what one should really look for in a Pareto Analysis. Finally there are chapters on the role of critical thinking and discussions on the appropriateness of many of the advanced statistical tools, such as reliability, binary logistic regression, time series analysis and Monte Carlo simulation.
In all, this book will guide practitioners to new and better ways to pull more information from the data they have obtained, leading to better conclusions and an improved bottom line.
Lyle Dockendorf has been an internal statistical consultant at IBM and Seagate Technology for over thirty years. Since 1998 he has been intimately involved with Seagate's Six Sigma initiative, becoming a Master Black Belt with the principal responsibilities of training, course development and statistical consulting. Besides Black Belt training, he has co-developed classes in Advanced Statistics, Advanced Measurement Systems Analysis, Reliability, and Green, Orange and White Belt training.