Statistical Methods for Quality Improvement, 2nd Edition

by Thomas P. Ryan (Acworth, GA )

• ISBN13: 9780471197751
• ISBN10: 0471197750
• Edition: 2nd
• Format: Hardcover
• Copyright: 2000-02-01
• Publisher: Wiley-Interscience

Summary

Very extensive and up-to-date references in each chapter, in addition to a bibliography of papers on a variety of control chart applications

Author Biography

THOMAS P. RYAN, PhD, is Director of Statistical Consulting in the Department of Statistics at Case Western Reserve University, and the author of Modern Regression Methods, available from Wiley.

Table of Contents

Preface	xxi
Preface to the First Edition	xxiii
PART I FUNDAMENTAL QUALITY IMPROVEMENT AND STATISTICAL CONCEPTS
Introduction	3	(8)
Quality and Productivity	4	(1)
Quality Costs (or Does It?)	4	(1)
The Need for Statistical Methods	5	(1)
Early Use of Statistical Methods for Improving Quality	5	(1)
Influential Quality Experts	6	(3)
Summary	9	(2)
References	9	(2)
Basic Tools for Improving Quality	11	(17)
Histogram	11	(4)
Pareto Charts	15	(3)
Scatter Plots	18	(3)
Variations of Scatter Plots	21	(1)
Control Chart	21	(2)
Check Sheet	23	(1)
Cause-and-Effect Diagram	23	(1)
Defect Concentration Diagram	24	(1)
The Seven New Tools	25	(2)
Affinity Diagram	25	(1)
Interrelationship Digraph	25	(1)
Tree Diagram	25	(1)
Prioritization Matrix	26	(1)
Matrix Diagram	26	(1)
Process Decision Program Chart	26	(1)
Activity Network Diagram	26	(1)
Summary	27	(1)
References	27	(1)
Basic Concepts in Statistics and Probability	28	(49)
Probability	28	(2)
Sample Versus Population	30	(1)
Location	31	(2)
Variation	33	(3)
Discrete Distributions	36	(12)
Binomial Distribution	37	(7)
Beta--Binomial Distribution	44	(1)
Poisson Distribution	44	(1)
Geometric Distribution	45	(1)
Negative Binomial Distribution	46	(1)
Hypergeometric Distribution	46	(2)
Continuous Distributions	48	(13)
Normal Distribution	48	(5)
t Distribution	53	(2)
Exponential Distribution	55	(1)
Lognormal Distribution	55	(2)
Weibull Distribution	57	(1)
Gamma Distribution	57	(1)
Chi-Square Distribution	58	(1)
Truncated Normal Distribution	58	(1)
Bivariate and Multivariate Normal Distributions	59	(1)
F Distribution	60	(1)
Beta Distribution	61	(1)
Uniform Distribution	61	(1)
Choice of Statistical Distribution	61	(1)
Statistical Inference	62	(8)
Central Limit Theorem	62	(1)
Point Estimation	63	(1)
Maximum Likelihood Estimation	63	(1)
Confidence Intervals	64	(1)
Tolerance Intervals	65	(1)
Hypothesis Tests	66	(1)
Probability Plots	66	(3)
Likelihood Ratio Tests	69	(1)
Bonferroni Intervals	69	(1)
Enumerative Studies versus Analytic Studies	70	(7)
References	71	(1)
Exercises	72	(5)
PART II CONTROL CHARTS AND PROCESS CAPABILITY
Control Charts for Measurements with Subgrouping (for One Variable)	77	(56)
Basic Control Chart Principles	77	(2)
Real-Time Control Charting Versus Analysis of Past Data	79	(2)
Control Charts: When to Use, Where to Use, How Many to Use	81	(1)
Benefits from the Use of Control Charts	81	(1)
Rational Subgroups	82	(1)
Basic Statistical Aspects of Control Charts	82	(1)
Illustrative Example	83	(16)
R Chart	86	(2)
R Chart with Probability Limits	88	(1)
s Chart	89	(1)
s Chart with Probability Limits	89	(4)
s2 Chart	93	(1)
X Chart	93	(3)
Recomputing Control Limits	96	(1)
Applying Control Limits to Future Production	97	(1)
Standards for Control Charts	97	(1)
Deleting Points	98	(1)
Target Values	98	(1)
Illustrative Example with Real Data	99	(1)
Determining the Time of a Parameter Change	100	(1)
Acceptance Sampling and Acceptance Control Chart	101	(5)
Acceptance Control Chart	102	(2)
Acceptance Chart with X Control Limits	104	(1)
Acceptance Charts Versus Target Values	105	(1)
Modified Limits	106	(1)
Difference Control Charts	107	(1)
Other Charts	108	(1)
Average Run Length	109	(2)
Weakness of the ARL Measure	110	(1)
Determining the Subgroup Size	111	(1)
Unequal Subgroup Sizes	112	(1)
Out-of-Control Action Plans	112	(1)
Assumptions for the Charts in This Chapter	113	(8)
Normality	114	(4)
Independence	118	(3)
Measurement Error	121	(1)
Summary	122	(11)
Appendix	123	(2)
References	125	(3)
Exercises	128	(5)
Control Charts for Measurements without Subgrouping (for One Variable)	133	(19)
Individual Observations Chart	133	(11)
Control Limits for the X Chart	134	(1)
X-Chart Assumptions	135	(1)
Illustrative Example: Random Data	136	(2)
Example with Particle Counts	138	(1)
Illustrative Example: Trended Data	139	(3)
Trended Real Data	142	(2)
Moving Average Chart	144	(1)
Controlling Variability with Individual Observations	145	(2)
Summary	147	(5)
Appendix	148	(1)
References	148	(1)
Exercises	149	(3)
Control Charts for Attributes	152	(34)
Charts for Nonconforming Units	153	(16)
np Chart	153	(2)
p Chart	155	(1)
Stage 1 and Stage 2 Use of p Charts and np Charts	155	(3)
Alternative Approaches	158	(1)
Arcsin Transformations	158	(5)
Q Chart for Binomial Data	163	(1)
Regression-Based Limits	163	(2)
ARL-Unbiased Charts	165	(1)
Variable Sample Size	166	(1)
Charts Based on the Geometric and Negative Binomial Distributions	167	(1)
Overdispersion	168	(1)
Charts for Nonconformities	169	(11)
c Chart	169	(2)
Transforming Poisson Data	171	(1)
Illustrative Example	171	(3)
Regression-Based Limits	174	(2)
u Chart	176	(1)
Regression-Based Limits	177	(1)
Overdispersion	177	(1)
D Chart	178	(2)
Probability-Type D-Chart Limits	180	(1)
Summary	180	(6)
References	181	(2)
Exercises	183	(3)
Process Capability	186	(30)
Data Acquisition for Capability Indices	186	(1)
Selection of Historical Data	187	(1)
Process Capability Indices	187	(5)
Cp	188	(1)
Cpm	188	(1)
Cpk	189	(2)
Cpmk	191	(1)
Other Capability Indices	191	(1)
Estimating the Parameters in Process Capability Indices	192	(3)
X Chart	193	(1)
X Chart	193	(1)
Case Study	193	(2)
Distributional Assumption for Capability Indices	195	(1)
Confidence Intervals for Process Capability Indices	195	(7)
Confidence Interval for Cp	196	(1)
Confidence Interval for Cpk	196	(2)
Confidence Interval for Cpm	198	(1)
Confidence Interval for Cpmk	198	(1)
Confidence Intervals Computed Using Data in Subgroups	198	(1)
Nonparametric Capability Indices and Confidence Limits	199	(1)
Robust Capability Indices	199	(1)
Capability Indices Based on Fitted Distributions	200	(1)
Data Transformation	201	(1)
Capability Indices Computed Using Resampling Methods	201	(1)
Asymmetric Bilateral Tolerances	202	(1)
Examples	202	(1)
Capability Indices That Are a Function of Percent Nonconforming	203	(5)
Examples	204	(4)
Modified k Index	208	(1)
Other Approaches	208	(1)
Process Capability Plots	209	(1)
Process Capability Indices Versus Process Performance Indices	209	(1)
Process Capability Indices with Autocorrelated Data	210	(1)
Summary	211	(5)
References	211	(3)
Exercises	214	(2)
Alternatives to Shewhart Charts	216	(37)
Introduction	216	(2)
Cumulative Sum Procedures: Principles and Historical Development	218	(17)
CUSUM Procedures Versus X Chart	218	(6)
Fast Initial Response CUSUM	224	(3)
Combined Shewhart--CUSUM Scheme	227	(2)
Computation of CUSUM ARLs	229	(1)
Robustness of CUSUM Procedures	230	(4)
CUSUM Procedures for Individual Observations	234	(1)
CUSUM Procedures for Controlling Process Variability	235	(2)
CUSUM Procedures for Nonconforming Units	237	(3)
CUSUM Procedures for Nonconformity Data	240	(4)
Exponentially Weighted Moving Average Charts	244	(5)
EWMA Chart for Subgroup Averages	245	(2)
EWMA Misconceptions	247	(1)
EWMA Chart for Individual Observations	247	(1)
Shewhart--EWMA Chart	248	(1)
FIR--EWMA	248	(1)
EWMA Chart with Variable Sampling Intervals	248	(1)
EWMA Chart for Grouped Data	248	(1)
EWMA Chart for Variances	248	(1)
EWMA for Attribute Data	248	(1)
Summary	249	(4)
References	249	(2)
Exercises	251	(2)
Multivariate Control Charts for Measurement Data	253	(36)
Hotelling's T2 Distribution	256	(1)
A T2 Control Chart	257	(11)
Identifying the Source of the Signal	258	(4)
Regression Adjustment	262	(1)
Recomputing the UCL	262	(1)
Characteristics of Control Charts Based on T2	263	(2)
Determination of a Change in the Correlation Structure	265	(1)
Illustrative Example	265	(3)
Multivariate Chart Versus Individual X Charts	268	(1)
Charts for Detecting Variability and Correlation Shifts	269	(3)
Application to Table 9.2 Data	270	(2)
Charts Constructed Using Individual Observations	272	(4)
Retrospective (Stage 1) Analysis	273	(1)
Stage 2 Analysis: Methods for Decomposing Q	274	(1)
Illustrative Example	275	(1)
When to Use Each Chart	276	(1)
Actual Alpha Levels for Multiple Points	276	(1)
Requisite Assumptions	277	(1)
Effects of Parameter Estimation on ARLs	277	(1)
Dimension Reduction Techniques	278	(1)
Multivariate CUSUM Charts	278	(1)
Multivariate EWMA Charts	279	(2)
Design of a MEWMA Chart	280	(1)
Searching for Assignable Causes	281	(1)
Applications of Multivariate Charts	281	(1)
Multivariate Process Capability Indices	282	(1)
Summary	282	(7)
Appendix	282	(1)
References	283	(3)
Exercises	286	(3)
Miscellaneous Control Chart Topics	289	(30)
Pre-Control	289	(3)
Short-Run SPC	292	(3)
Charts for Autocorrelated Data	295	(3)
Autocorrelated Attribute Data	298	(1)
Charts for Batch Processes	298	(1)
Charts for Multiple-Stream Processes	298	(1)
Nonparametric Control Charts	299	(1)
Bayesian Control Chart Methods	300	(1)
Control Charts for Variance Components	301	(1)
Neural Networks	301	(1)
Economic Design of Control Charts	301	(3)
Economic-Statistical Design	303	(1)
Charts with Variable Sample Size and/or Variable Sampling Interval	304	(1)
Users of Control Charts	304	(2)
Recent Control Chart Nonmanufacturing Applications	305	(1)
Health Care	305	(1)
Financial	306	(1)
Environmental	306	(1)
Clinical Laboratories	306	(1)
Analytical Laboratories	306	(1)
Civil Engineering	306	(1)
Education	306	(1)
Law Enforcement/Investigative Work	306	(1)
Lumber	306	(1)
Athletic Performance	306	(1)
Software for Control Charting	306	(13)
Bibliography	307	(8)
Exercises	315	(4)
PART III BEYOND CONTROL CHARTS: GRAPHICAL AND STATISTICAL METHODS
Other Graphical Methods	319	(18)
Stem-and-Leaf Display	319	(2)
Dot Diagrams	321	(1)
Digidot Plot	322	(1)
Boxplot	322	(3)
Normal Probability Plot	325	(3)
Plotting Three Variables	328	(1)
Displaying More Than Three Variables	328	(1)
Multi-Vari Chart	329	(1)
Plots to Aid in Transforming Data	330	(3)
Summary	333	(4)
References	333	(1)
Exercises	334	(3)
Linear Regression	337	(22)
Simple Linear Regression	337	(4)
Worth of the Prediction Equation	341	(2)
Assumptions	343	(1)
Checking Assumptions through Residual Plots	343	(1)
Confidence Intervals and Hypothesis Tests	344	(1)
Prediction Interval for Y	345	(1)
Regression Control Chart	346	(1)
Cause-Selecting Charts	347	(2)
Inverse Regression	349	(2)
Multiple Linear Regression	351	(1)
Issues in Multiple Regression	352	(3)
Variable Selection	352	(1)
Extrapolation	353	(1)
Multicollinear Data	353	(1)
Residual Plots	353	(1)
Regression Diagnostics	354	(1)
Transformations	354	(1)
Software for Regression	355	(1)
Summary	355	(4)
References	355	(2)
Exercises	357	(2)
Design of Experiments	359	(68)
A Simple Example of Experimental Design Principles	359	(2)
Principles of Experimental Design	361	(1)
Statistical Concepts in Experimental Design	362	(2)
t Tests	364	(4)
Exact t Test	365	(2)
Approximate t Test	367	(1)
Confidence Intervals for Differences	367	(1)
Analysis of Variance for One Factor	368	(9)
ANOVA for a Single Factor with More Than Two Levels	370	(4)
Multiple Comparison Procedures	374	(1)
Sample Size Determination	375	(1)
Additional Terms and Concepts in One-Factor ANOVA	376	(1)
Regression Analysis of Data from Designed Experiments	377	(5)
ANOVA for Two Factors	382	(8)
ANOVA with Two Factors: Factorial Designs	383	(1)
Effect Estimates	384	(2)
ANOVA Table for Unreplicated Two-Factor Design	386	(2)
Yates's Algorithm	388	(2)
The 23 Design	390	(5)
Assessment of Effects without a Residual Term	395	(2)
Residual Plot	397	(3)
Separate Analyses Using Design Units and Uncoded Units	400	(2)
Two-Level Designs with More Than Three Factors	402	(1)
Three-Level Factorial Designs	403	(1)
Mixed Factorials	404	(1)
Fractional Factorials	405	(8)
2k--1 Designs	405	(5)
2k--2 Designs	410	(2)
More Highly Fractionated Two-Level Designs	412	(1)
Fractions of Three-Level Factorials	412	(1)
Incomplete Mixed Factorials	412	(1)
Cautions	413	(1)
Other Topics in Experimental Design and Their Applications	413	(6)
Hard-to-Change Factors	413	(1)
Split-Lot Designs	413	(1)
Mixture Designs	413	(1)
Response Surface Designs	414	(1)
Designs for Measurement System Evaluation	415	(1)
Computer-Aided Design and Expert Systems	416	(1)
Sequential Experimentation	416	(1)
Supersaturated Designs and Analyses	417	(1)
Multiple Responses	417	(2)
Summary	419	(8)
References	419	(5)
Exercises	424	(3)
Contributions of Genichi Taguchi and Alternative Approaches	427	(47)
``Taguchi Methods''	427	(1)
Quality Engineering	428	(1)
Loss Functions	428	(3)
Distribution Not Centered at the Target	431	(1)
Loss Functions and Specification Limits	431	(1)
Asymmetric Loss Functions	432	(3)
Signal-to-Noise Ratios and Alternatives	435	(2)
Experimental Designs for Stage 1	437	(1)
Taguchi Methods of Design	438	(26)
Inner Arrays and Outer Arrays	439	(1)
Orthogonal Arrays as Fractional Factorials	439	(3)
Other Orthogonal Arrays Versus Fractional Factorials	442	(5)
Product Arrays Versus Combined Arrays	447	(6)
Application of Product Array	453	(10)
Cautions	463	(1)
Desirable Robust Designs and Analyses	463	(1)
Designs	463	(1)
Analyses	463	(1)
Determining Optimum Conditions	464	(4)
Summary	468	(6)
References	469	(2)
Exercises	471	(3)
Evolutionary Operation	474	(20)
EVOP Illustrations	475	(9)
Three Variables	484	(2)
Simplex EVOP	486	(3)
Other EVOP Procedures	489	(1)
Miscellaneous Uses of EVOP	489	(1)
Summary	490	(4)
Appendix	490	(1)
References	491	(1)
Exercises	492	(2)
Analysis of Means	494	(25)
ANOM for One-Way Classifications	495	(3)
ANOM for Attribute Data	498	(3)
Proportions	498	(2)
Count Data	500	(1)
ANOM When Standards Are Given	501	(1)
Nonconforming Units	501	(1)
Nonconformities	501	(1)
Measurement Data	502	(1)
ANOM for Factorial Designs	502	(5)
Assumptions	505	(1)
Alternative Way of Displaying Interaction Effects	506	(1)
ANOM When at Least One Factor Has More Than Two Levels	507	(8)
Main Effects	507	(4)
Interaction Effects	511	(4)
Use of ANOM with Other Designs	515	(1)
Nonparametric ANOM	515	(1)
Summary	516	(3)
Appendix	516	(1)
References	516	(1)
Exercises	517	(2)
Using Combinations of Quality Improvement Tools	519	(11)
Control Charts and Design of Experiments	520	(1)
Control Charts and Calibration Experiments	520	(1)
Six Sigma Programs	520	(7)
Components of a Six Sigma Program	524	(1)
Six Sigma Applications and Programs	524	(1)
Six Sigma Concept for Customer Satisfaction	525	(1)
Six Sigma Training	525	(1)
Related Programs/Other Companies	526	(1)
SEMATECH's Qual Plan	526	(1)
AlliedSignal's Operational Excellence Program	526	(1)
Statistical Process Control and Engineering Process Control	527	(3)
References	528	(2)
Answers to Selected Exercises	530	(2)
Appendix: Statistical Tables	532	(13)
Author Index	545	(6)
Subject Index	551