Competitive Innovation and Improvement
Statistical Design and Control
Prices & shipping based on shipping country
Competitive Innovation and Improvement: Statistical Design and Control explains how to combine two widely known statistical methods—statistical design and statistical control—in a manner that can solve any business, government, or research problem quickly with sustained results. Because the problem-solving strategy employed is pure scientific method, it makes integration into any existing problem-solving or research method quite simple.
The material in the book is presented in a manner that anyone can read and immediately put to use, including executives, managers, statisticians, scientists, engineers, researchers, and all of their supervisors and employees. Organizations can apply the concepts discussed with existing staff to release latent energy rather than adding to their workload. Optional footnotes provide the opportunity for more advanced technical insight.
Supplying readers with an understanding of orthogonal design, the book illustrates key ideas through large-scale case studies. The book’s 12 case studies examine the coupling of statistical design with economic control across a range of industries and problem types.
The book suggests the real world, rather than mathematics alone, to reveal how things work and how to make them work better. Innovation and improvement by design is explained, which will help readers open up left-brain analytics to more right-brain creativity.
Although mathematics (as advanced as needed to solve the problem) is used throughout the text, it is translated into simple arithmetic without any mathematical notation. The book limits references to a few essential texts and papers that readers can refer to as they become more experienced in statistical design and control.
Table of Contents
Simplicity of Statistical Design and Control
Making a Start
How Does It Work?
Care Management Case: Improving Health for Thousands of People
Care Management Statistical Design
Managing the Test
What Might the Results Mean?
Findings Are Often Surprising
Significance of the Results
Innovation Uses More Right Brain than Left
Retailing Case: New Product Sales
Preparing for the Test
Retail Furniture Statistical Design and Its Management
Exploratory Analysis and Inference
What Might the Results Mean?
Ironing Out Some Possible Wrinkles
Predicting and Delivering the Improvement
Retailing Designed Innovation Case: Conclusion
Using Statistical Control
Practical Use of Statistical Control
Digression into Causality
Concluding Scientific Work in the Care Management Case
False Alarm Rate Is Neither Known Nor Useful in Statistical Control
Statistical Control Terminology
Statistics Breaks Down in Unstable Processes
Economic Loss without Statistical Control
Cost Explosion Story Unexploded
Tests for Statistical Control
Statistical Control Integrated with Statistical Design
Managing Statistical Control Schemes
Mechanics of Statistical Control
Where Did Statistical Control Originate?
Measurement Error and Control
All Measurement Systems Are Inherently Flawed
Clinical Care Case: Initial Measurement Study and Long-Term Controls
Establishing a Measurement Control Scheme
Advantages of Large Statistical Design
Full Factorial Designs
Fractional Factorial Designs
Managing the Test
What Might the Initial Results Mean?
Solving the Puzzle
Analysis of All Pair Interactions
Measurement Problem Found and Fixed after the Test
Using Sales Change as the Test’s Measurement
Calculating Precision and Sample Size Before the Test
Diagnosing Unusually High or Low Results in a Statistical Design Row
Guidance on Fractional Factorial Designs
Care Management Case: More Analytical Insight
Geometric versus Nongeometric Designs
Aliasing Scheme for the Care Management Design
Augmenting Multifactorials to Also Estimate Pair Interactions
Uniqueness and Stumbling Around
Where Did Statistical Design Originate?
Statistical Design and Control: A Dozen Large-Scale Case Studies
Selection of Cases
Solving Complex Problems Simply
Simultaneous Design Idea
Science Education Case
Simultaneous Statistical Designs for Science Classes
Pair Interactions across Designs and an Easier Analysis
Rules for Simultaneous Designs
General Multichannel Optimization Case
Simultaneous Design Procedure
Scientific Method, Randomization, and Improvement Strategies
Simplicity of the Scientific Method
Scientific Method with Statistical Design and Control
Proof Isn’t in the Pudding
What Science Lies beneath Implementation Being the Hardest Part?
Common Improvement Strategies
Randomized Control Trials (RCT)
Statistical Design and Control Are for Real Problems with Everyone Contributing
Managing Improvement and Innovation
Speed without Net Resources
How to Manage Specific Improvements/Innovations
Statistical Design and Control Summary
Appendix: Answers to Exercises
Kieron Dey studied mathematics and statistics at Reading University, England and management at Rensselaer Polytechnic Institute, New York. He was on the experimental staff at Hirst Research Center, London, England (an early specialized center for applied scientific research), and apprenticed with Joan Keen, a pioneer in industrial statistics. He later joined IIT Research Institute, another contract research organization, serving in several roles including scientific advisor. He has held technical leadership positions in corporations up to $2 billion in size, now with Nobigroup Inc. He has vast experience with corporate and government leaders. Dey is a Fellow of the Royal Statistical Society.
—Dr. Randy Brown. Director of Health Research, Mathematica Policy Research, Inc.
Work that is way ahead of others… interesting and energetic writing. A lot of hands-on as well as technical wisdom – a very rare combination. The material is new, challenging, and important.
—Dr. Brian L. Joiner. Minitab Co-Inventor, former Professor of Statistics, University of Wisconsin, Madison
A wonderful book with unique insights into an area of enormous potential. ... not duplicated anywhere to the best of my knowledge. Writing style makes it easy to follow each topic.
—Dr. Steve Grady, Consulting Econometrician
The concept is simple (it makes you wonder why others haven’t tried it). The large scale is unique ... not discussed in current literature.
—Tim Baer, Principal Statistician, Roche Diagnostics