Visual SIX SIGMA. Making Data Analysis Lean. Ian Cox Marie A. Gaudard Philip J. Ramsey

Transcription

1 Visual SIX SIGMA Making Data Analysis Lean Ian Cox Marie A. Gaudard Philip J. Ramsey Mia L. Stephens Leo T. Wright

2

3 Visual Six Sigma Making Data Analysis Lean IAN COX MARIE A. GAUDARD PHILIP J. RAMSEY MIA L. STEPHENS LEO T. WRIGHT John Wiley & Sons, Inc.

4 Copyright C 2010 by SAS Institute, Inc. All rights reserved. Published by John Wiley & Sons, Inc., Hoboken, New Jersey. Published simultaneously in Canada. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) , fax (978) , or on the web at Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) , fax (201) , or online at Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) , outside the United States at (317) or fax (317) Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. For more information about Wiley products, visit our web site at Library of Congress Cataloging-in-Publication Data: Cox, Ian, 1956 Visualsixsigma:makingdataanalysislean/IanCox...[et al.]. p. cm. (Wiley & SAS business series) Includes bibliographical references and index. ISBN (cloth) 1. Six sigma (Quality control standard) 2. Decision support systems. 3. Decision making Statistical methods. 4. Organizational effectiveness. I. Title. HD C dc ISBN Printed in the United States of America

5 Wiley & SAS Business Series The Wiley & SAS Business Series presents books that help senior-level managers with their critical management decisions. Titles in the Wiley and SAS Business Series include: Business Intelligence Competency Centers: A Team Approach to Maximizing Competitive Advantage by Gloria J. Miller, Dagmar Brautigam, and Stefanie V. Gerlach Case Studies in Performance Management: A Guide from the Experts by Tony C. Adkins CIO Best Practices: Enabling Strategic Value with Information Technology by Joe Stenzel Credit Risk Scorecards: Developing and Implementing Intelligent Credit Scoring by Naeem Siddiqi Customer Data Integration: Reaching a Single Version of the Truth by Jill Dyché and Evan Levy Enterprise Risk Management: A Methodology for Achieving Strategic Objectives by Gregory Monahan Fair Lending Compliance: Intelligence and Implications for Credit Risk Management by Clark R. Abrahams and Mingyuan Zhang Information Revolution: Using the Information Evolution Model to Grow Your Business by Jim Davis, Gloria J. Miller, and Allan Russell Marketing Automation: Practical Steps to More Effective Direct Marketing by Jeff LeSueur Performance Management: Finding the Missing Pieces (to Close the Intelligence Gap) by Gary Cokins Performance Management: Integrating Strategy Execution, Methodologies, Risk, and Analytics by Gary Cokins Credit Risk Assessment: The New Lending System for Borrowers, Lenders, and Investors by Clark R. Abrahams and Mingyuan Zhang Business Intelligence Success Factors: Tools for Aligning Your Business in the Global Economy by Olivia Parr Rud The Data Asset: How Smart Companies Govern Their Data for Business Success by Tony Fisher Demand-Driven Forecasting: A Structured Approach to Forecasting by Charles Chase The New Know: Innovation Powered by Analytics by Thornton May For more information on any of the above titles, please visit iii

6

7 Contents Preface Acknowledgments ix xi PART I BACKGROUND CHAPTER 1 Introduction 3 What Is Visual Six Sigma? 3 Moving beyond Traditional Six Sigma 4 Making Data Analysis Lean 4 Requirements of the Reader 5 CHAPTER 2 Six Sigma and Visual Six Sigma 7 Background: Models, Data, and Variation 7 Six Sigma 10 Variation and Statistics 13 Making Detective Work Easier through Dynamic Visualization 14 Visual Six Sigma: Strategies, Process, Roadmap, and Guidelines 16 Conclusion 21 Notes 21 CHAPTER 3 A First Look at JMP R 23 The Anatomy of JMP 23 Visual Displays and Analyses Featured in the Case Studies 39 Scripts 44 Personalizing JMP 47 Visual Six Sigma Data Analysis Process and Roadmap 47 Techniques Illustrated in the Case Studies 50 Conclusion 50 Notes 50 v

8 vi Contents PART II CASE STUDIES CHAPTER 4 Reducing Hospital Late Charge Incidents 57 Framing the Problem 58 Collecting Data 59 Uncovering Relationships 62 Uncovering the Hot Xs 90 Identifying Projects 103 Conclusion 103 CHAPTER 5 Transforming Pricing Management in a Chemical Supplier 105 Setting the Scene 106 Framing the Problem: Understanding the Current State Pricing Process 107 Collecting Baseline Data 112 Uncovering Relationships 121 Modeling Relationships 147 Revising Knowledge 152 Utilizing Knowledge: Sustaining the Benefits 159 Conclusion 162 CHAPTER 6 Improving the Quality of Anodized Parts 165 Setting the Scene 166 Framing the Problem 167 Collecting Data 169 Uncovering Relationships 183 Locating the Team on the VSS Roadmap 196 Modeling Relationships 197 Revising Knowledge 210 Utilizing Knowledge 229 Conclusion 231 Note 232 CHAPTER 7 Informing Pharmaceutical Sales and Marketing 233 Setting the Scene 235 Collecting the Data 235 Validating and Scoping the Data 237 Investigating Promotional Activity 263 A Deeper Understanding of Regional Differences 282 Summary 291 Conclusion 292

9 Contents vii Additional Details 292 Note 301 CHAPTER 8 Improving a Polymer Manufacturing Process 303 Setting the Scene 305 Framing the Problem 307 Reviewing Historical Data 314 Measurement System Analysis 320 Uncovering Relationships 334 Modeling Relationships 345 Revising Knowledge 366 Utilizing Knowledge 378 Conclusion 388 Note 389 CHAPTER 9 Classification of Cells 391 Setting the Scene 393 Framing the Problem and Collecting the Data: The Wisconsin Breast Cancer Diagnostic Data Set 394 Uncovering Relationships 395 Constructing the Training, Validation, and Test Sets 417 Modeling Relationships: Logistic Model 443 Modeling Relationships: Recursive Partitioning 460 Modeling Relationships: Neural Net Models 467 Comparison of Classification Models 480 Conclusion 483 Notes 484 Index 485

10

11 Preface The purpose of this book is to show how, using the principles of Visual Six Sigma, you can exploit data to make better decisions more quickly and easily than you would otherwise. We emphasize that your company does not need to have a Six Sigma initiative for this book to be useful. Clearly there are many data-driven decisions that, by necessity or by design, fall outside the scope of a Six Sigma effort, and in such cases we believe that Visual Six Sigma is ideal. We seek to show that Visual Six Sigma can be used by a lone associate, as well as a team, to address data-driven questions, with or without the support of a formal initiative like Six Sigma. To this end, we present six case studies that show Visual Six Sigma in action. These case studies address complex problems and opportunities faced by individuals and teams in a variety of application areas. Each case study was addressed using the Visual Six Sigma Roadmap, described in Chapters 2 and 3. As these case studies illustrate, Visual Six Sigma is about exploration and discovery, which means that it is not, and never could be, an entirely prescriptive framework. As well as using the case studies to convey the Visual Six Sigma Roadmap, we also want to use them to illustrate Visual Six Sigma techniques that you can reuse in your own setting. To meet this goal, sometimes we have deliberately compromised the lean nature of the Visual Six Sigma Roadmap in order to take the opportunity to show you extra techniques that may not be strictly necessary to reach the conclusion or business decision. Striking the balance this way means that you will see a wider repertoire of techniques from which to synthesize an approach to Visual Six Sigma that works for you. Because of its visual emphasis, Visual Six Sigma opens the doors for nonstatisticians to take active roles in data-driven decision making, empowering them to leverage their contextual knowledge to pose relevant questions, get good answers, and make sound decisions. You may find yourself working on a Six Sigma improvement project, a design project, a data mining inquiry, or a scientific study all of which require decision making based on data. After working through this book, we hope that you will be able to make data-driven decisions in your specific situation quickly, easily, and with great assurance. How This Book Is Organized This book is organized in two parts. Part I contains an introductory chapter that presents the three Visual Six Sigma strategies, a chapter on Visual Six Sigma, and ix

12 x Preface a chapter introducing JMP statistical software (from SAS Institute), which will be used throughout the case studies. Case studies are presented in Part II. These case studies follow challenging real-world projects from start to finish. Through these case studies, you will gain insight into how the three Visual Six Sigma strategies combine to expedite project execution in the real world. Each case study is given its own chapter, which can be read independently from the rest. A concise summary of the storyline opens each case study. Although these case studies are real, we use fictitious names for the companies and individuals to preserve confidentiality. Within each case study, visualization methods and other statistical techniques are applied at various stages in the data analysis process in order to better understand what the data are telling us. For those not familiar with JMP, each case study also contains the relevant how-to steps so that you may follow along and see Visual Six Sigma in action. The data sets used in the case studies are available at visualsixsigma. Here you can also find the exhibits shown in the case studies, allowing you to see screen captures in color. Additional Visual Six Sigma resource materials will be made available on the Web site, as appropriate. We have used different fonts to help identify the names of data tables, of columns in data tables, and commands. Data table names are shown in bold Times New Roman, while the names of columns (which are variable names) are shown in Helvetica. The names of commands are shown in bold Helvetica. A Word about the Software The ideas behind Visual Six Sigma are quite general, but active learning in our view the only kind of learning that works requires that you step through the case studies and examples in this book to try things out for yourself. For more information about JMP and to download a trial version of the software, visit JMP is available on Windows, Mac, and Linux platforms. The step-by-step instructions in this book assume that you are working in Windows. Mac and Linux users should refer to the JMP documentation for details on differences. This book is based on JMP version

13 Acknowledgments Central to this book are the case studies that allow us to show Visual Six Sigma in some specific, real-world settings. We would like to thank all those who had a hand in shaping the case studies, giving specific mention to Joe Ficalora (Global Services for Sigma Breakthrough Technologies Inc., SBTI), who contributed the Reducing Hospital Late Charge Incidents case study in Chapter 4, Andrew Ruddick (Catalyst Consulting) for the Transforming Pricing Management in a Chemical Supplier case study in Chapter 5, and Andy Liddle (Catalyst Consulting) for the Improving a Polymer Manufacturing Process case study in Chapter 8. Their contributions were valuable in making the book more relevant and useful to prospective readers and are indicative of a consulting approach that is both pragmatic and enlightened. The book was substantially improved by suggestions from the following individuals: Mark Bailey, Monica Beals, Duane Hayes, Bradley Jones, Ann Lehman, Paul Marovich, Tonya Mauldin, Malcolm Moore, Heath Rushing, Laura Ryan, and Scott Wise. We greatly appreciate their time, interest, valuable feedback, and insights. This project was greatly facilitated by Stacey Hamilton and Stephenie Joyner, of SAS Publishing. Their support, encouragement, and attention to detail at every step of this adventure were invaluable. Finally, we would like to thank John Leary, Dave Richardson, and Jon Weisz for their support and encouragement in the writing of this book. A special thankyou goes to John Sall and the JMP Development Team for their work on a visionary product that makes Visual Six Sigma possible. xi

14

15 Background PART I