Quality Improvement Through Planned Experimentation 3/E

Home >
Browse >
Quality Improvement Through Planned Experimentation 3/E

Quality Improvement Through Planned Experimentation 3/E

Authors: Ronald Moen, Thomas Nolan and Lloyd Provost

Published: June 2012

eISBN: 9780071759670 0071759670 | ISBN: 9780071759663

Quality Improvement through Planned Experimentation

Contents

Foreword

Preface

Acknowledgments

Chapter 1: Improvement of Quality

1.1 Introduction

1.2 Building Knowledge and the Scientific Method

1.3 Defining Quality

Planned Experimentation

Prediction and Degree of Belief

1.4 Model for Improvement

The Three Questions

The PDSA Cycle

Step 1: Plan

Step 2: Do

Step 3: Study

Step 4: Act

1.5 Sequential Experimentation* Using the PDSA Cycle

1.6 Summary

References

Exercises

Chapter 2: Principles for Design and Analysis of Planned Experiments

2.1 Introduction and Definitions

2.2 T ypes of Planned Experiments

Analytic Studies and Prediction

2.3 P rinciples for Designing Analytic Studies

Objective of the Study

Sequential Approach

Partitioning of Variation

Degree of Belief

Simplicity of Execution

2.4 Tools for Experimentation

Experimental Pattern

Planned Grouping

Randomization

Replication

2.5 Form for Documentation of a Planned Experiment

2.6 Analysis of Data from Analytic Studies

Basic Principles for Analysis

2.7 Summary

References

Exercises

Chapter 3: Experiments with One Factor

3.1 General Approach to One-Factor Experiments

3.2 Using Run Charts for a One-Factor Design

Example 3.1

Example 3.2

3.3 Using Shewhart Charts for One-Factor Experiments

3.4 Paired-Comparison Experiments

Example 3.3

3.5 Randomized Block Designs

Example 3.4

Example 3.5

3.6 Incomplete Block Designs

Example 3.6

3.7 Summary

References

Exercises

Chapter 4: Experiments with More Than One Factor

4.1 Introduction to Factorial Designs

Example 4.1: A 22 Design: A Customer Survey

Example 4.2: 23 Study of a Dye Process

Estimating Effects Using the Design Matrix

Response Plots

Conclusions for Example 4.2

Example 4.3: A 2 4 Experiment to Maximize Email Click-Through Rates

Conclusions

4.2 Design of Factorial Experiments

Choosing the Number of Factors

Choosing the Levels for Each Factor

Considering Background Variables in the Design

Example 4.4: A Welding Process Experiment

Selecting the Amount of Replication

Randomizing the Order of the Tests

4.3 Advanced Topics in the Analysis of Factorial Experiments

Special Causes in the Run Chart

Example 4.5: Study of Automotive Emissions

Example 4.6: Design of a Throttle Return Mechanism

Special Causes Buried in the Data

Example 4.7: Pilot Line for a Tile Process

Missing Data

Nuisance Variables Dominate

Factors Interact with Conditions

Example 4.4 (Continued): A Welding Process

4.4 Summary

References

Exercises

Chapter 5: Reducing the Size of Experiments

5.1 Introduction to Fractional Factorial Designs

Example 5.1: The Email Experiment as a 24–1 Design

5.2 Fractional Factorial Designs—Moderate Current Knowledge

Example 5.2: A 2 5–1 design for a welding process

Run Charts

Dot Diagram

Conclusions for Example 5.2

5.3 Fractional Factorial Designs—Low Current Knowledge

Example 5.3: A 27–4 Design for Cell Culture Media

Run Charts

5.4 Using Blocking to Design a Sequence of Experiments

Example 5.4: A 23 Design in Two Blocks for the dye process

A 24 Design in Two Blocks of Size 8

A 27-3 Design in Two Blocks of Size 8

A 215-10 Design in Two Blocks of Size 16

Example 5.3 (Continued): Cell Culture Media Experiment

Conclusions for the Continuation of Example 5.3

5.5 Summary

Appendix: Development of Other Blocking Arrangements

References

Exercises

Chapter 6: Evaluating Sources of Variation

6.1 Applications of Nested Designs

Example 6.1 The Shewhart Control Chart as a Nested Design

Example 6.2 Nested Design to Study Measurement Variation

Example 6.3 Nested Design to Study Variation in Test Scores

6.2 Planning and Analyzing an Experiment with Nested Factors

Planning a Nested Experiment

Analyzing a Nested Experiment

Step 1: Run Chart or Shewhart Chart

Step 2: Dot-Frequency Diagram

Step 3: Analysis of the Dot-Frequency Diagram

6.3 More Complex Nested Designs

Example 6.4 Foundry Study—More Than Four Factors in a Nested Design

Example 6.5 Interlaboratory Study with Nested and Crossed Factors

6.4 Summary

Appendix 6A: Calculation of Variance Components

Example 6.4 Five Factor Foundry Study

Appendix 6B: Calculating and Combining Statistics (X, S, or R)

References

Exercises

Chapter 7: Sequential Experimentation—A Case Study

7.1 Improving a Milling Process—Getting Started

7.2 The First Improvement Cycle: Current Performance of the Mills

7.3 The Second PDSA Cycle: Sources of Variation

7.4 The Third PDSA Cycle: Evaluating Mill Cutter Vendors

7.5 The Fourth PDSA Cycle: Screening Process Variables

7.6 The Fifth PDSA Cycle: Evaluate Effect of Improvements on the Mill Process

7. 7 The Sixth PDSA Cycle: Evaluating Important Factors

7.8 The Seventh PDSA Cycle: Determining Optimum Levels

7.9 The Eighth PDSA Cycle: Confirmation of Improvements

7.10 Final Actions of the Mill Improvement Team

Exercises

Chapter 8: Using a Time Series Response Variable

8.1 Incorporating Experimental Patterns in a Time Series

8.2 Shewhart Charts

Interpretation of a Shewhart Chart

Types of Shewhart Charts

Example 8.1. Using a Shewhart Chart with a Sequential Planned Experiment

8.3 Designs for Sequential Experimentation Using Time Series Response Variables

Factorial Designs with Time Series Response Variables

Example 8.2: Factorial design for improving teaching of geometry

8.4 Summary

References

Exercises

Chapter 9: Experiments with Factors at More Than Two Levels

9.1 Factorial Designs with More Than Two Levels

Example 9.1: Microfinish Tolerances

9.2 Augmenting 2k Factorial Designs with Center Points

Example 9.2

9.3 Three-Level Designs for Quantitative Factors

Example 9.3

9.4 Experiments for Formulations or Mixtures

Using Factorial Designs for Mixture Experiments

Simplex Patterns When All Components Have No Constraints

Example 9.4: Improving the Density of Ceiling Tile

Mixture Experiments When the Components Are Constrained

Example 9.5: Fruit Punch Study

9.5 Experimental Designs for Complex Systems

Example 9.6: Visual Defects in a Manufacturing Process

9.6 Summary

References

Exercises

Chapter 10: Applications in Health Care

10.1 Introduction

10.2 Applications

Experimentation by an Individual

Example 10.1: Deciding on Cataract Surgery

Cycle 1: Regular Sunglasses

Cycle 2: Special Golf Sunglasses

Cycle 3: Repeat Cycle 2 Using the New Measurement Scales

Cycle 4: Prescription Sunglasses

Cycles 5–8: Maintaining the Gain

Cycle 9: Using a Colored Ball

Cycle 10: Colored Ball and Nonprescription Sunglasses

Lessons From This Example

Experimental Design Using Simulations

Simulation of a Health System

Experimentation on the Healthcare Delivery System

Example 10.2: Reducing “No Show” Rates at a Clinic

Conclusions

Example 10.3: Reducing Rehospitalizations

10.3 Summary

References

Exercises

Chapter 11: New Product Design

11.1 Introduction

11.2 Phase 0: Generate Ideas

Marketing

Example 11.1: Redesigning a Floor-Covering Product

Engineering

Operations

11.3 Phase 1: Develop Concepts and Define Product

Marketing

Example 11.1 (Continued): Conjoint Analysis on Potential Features for New Floor Covering

Engineering

Average

Standard Deviation

Average Versus Variation

Example 11.1(Continued): Setting of Control Factors for New Floor Covering

Example 11.2: Designing a New Service

Operations

11.4 Phase 2: Test

Marketing

Example 11.2 (Continued): Testing a New Service

Engineering

Example 11.1 (Continued): Follow-up Study on Interactions

Operations

Example 11.1 (Continued): The Field Test for New Floor Covering

11.5 Phase 3: Produce Product

11.6 Summary

Objective of Experiment

Response Variables

Factors Under Study

References

Exercises

Appendix A: Evaluating Measurement Systems

A.1 Introduction

A.2 Studying Measurement Processes

A.3 Data from Measurement Process

Measurement Discrimination and Rounding of Numbers

A.4 Monitoring and Improving a Measurement Process

Evaluating Precision of a Measurement Process

Precision for Classification Data

Precision for Rank Data

Precision for Continuous Data

Evaluating Bias of a Measurement Process

Effect of Precision and Bias of a Measurement Process

A.5 Using Planned Experimentation to Improve a Measurement Process

Case Study: Evaluating a Measurement Process

Develop Control Charts for the Measurement Process

Stability

Evaluation of Bias

Evaluation of Precision: Repeatability and Reproducibility

Planned Experiment for Measurement Process

Improvements to the Measurement Process

Updating Control Charts for Measurement Process

Monitoring the Measurement Process

A.6 Summary

Definitions of Terms Used with Measurement

References

Glossary

Index