Total Quality Tools

and Statistical Thinking

J oseph Juran describes quality management as the “Quality Trilogy”: plan-

ning, control, and improvement. He says that most managers devote too
much attention to control, and too little to planning and improvement—
which may be the most important activities for meeting and exceeding cus-
tomer expectations and gaining competitive advantage. Quality practitioners
have adapted a variety of tools from other disciplines, such as statistics,
operations research, and creative problem solving to aid the planning and
improvement processes. These tools provide a means by which problems
can be viewed objectively, data can be used as a basis for fact-driven deci-
sions, and managers can deal with variation in a logical fashion. This chap-
ter describes and illustrates some of the most useful and popular tools and
describes how they support process-driven statistical thinking.
The tools we describe in this chapter can be broadly categorized into two
groups: tools for planning and tools for continuous improvement. Planning tools—
which include quality function deployment, concurrent engineering, and the
“new seven” management and planning tools—are designed to assist man-
agers in planning the quality effort and making efficient use of information.
They are often used by cross-functional teams in their quality planning
efforts. Tools for continuous improvement—which include the Deming
Cycle, tools for data analysis, poka-yoke (mistake-proofing), and benchmark-
ing—are means of improving the manufacturing and service systems. The
Deming Cycle and basic statistical tools are usually found in basic quality
training for all employees of an organization and are used by problem-solving
teams to attack specific quality problems. Benchmarking is used at all levels
88 Part I: Introduction to Total Quality

of the firm to better understand its processes relative to its competitors’ and
to make significant improvements in operations.
Since the focus of this book is management, organization, and strategy,
this chapter is intended to be only an elementary introduction. The list pre-
sented here is by no means exhaustive. The bibliography at the end of the
book provides supplementary reading on these and other tools for quality
improvement.
The objectives of this chapter are
• to describe how quality function deployment and concurrent engineer-
ing can improve the process of designing products and services to achieve
better customer satisfaction;
• to show how simple graphical tools can improve management planning;
• to describe and illustrate the Deming Cycle—a simple methodology for
continuous improvement;
• to illustrate the application of basic statistical tools, mistake-proofing
approaches, benchmarking, and reengineering for quality improvement;
• to discuss the importance of creativity and innovation for quality improve-
ment and the management environment that fosters these characteristics;
and
• to describe principles of statistical thinking as a basis for effective
management.

TOOLS FOR QUALITY PLANNING

Customers’ needs and expectations drive the planning process for products
and the systems by which they are produced. Marketing plays a key role in
identifying customer expectations. Once they are identified, managers must
translate them into specific product and service specifications that manufac-
turing and service delivery processes must meet. In some cases the product
or service that customers receive is quite different from what they expect. It
is management’s responsibility to minimize such gaps. Firms use several
tools and approaches to help them focus on their external and internal cus-
tomers. This section introduces three of these tools: quality function deploy-
ment, concurrent engineering, and the “new seven” management and plan-
ning tools.

Quality Function Deployment

Quality function deployment (QFD) is a methodology used to ensure that
customers’ requirements are met throughout the product design process and in
the design and operation of production systems. QFD is both a philosophy and
a set of planning and communication tools that focuses on customer require-
ments in coordinating the design, manufacturing, and marketing of goods. A
major benefit of QFD is improved communications and teamwork among all
constituencies in the production process—marketing and design, design and
manufacturing, and purchasing and suppliers. QFD allows companies to sim-
 Chapter 3: Total Quality Tools and Statistical Thinking 89

ulate the effects of new design ideas and concepts. This allows them to bring
new products into the market sooner and to gain competitive advantage.
QFD originated in 1972 at Mitsubishi’s Kobe shipyard site. Toyota began
to develop the concept shortly thereafter, and it has been used since 1977.
The results have been impressive: Between January 1977 and October 1979,
for example, Toyota realized a 20 percent reduction in start-up costs on the
launch of a new van. By 1982 start-up costs had fallen 38 percent from the
1977 baseline, and by 1984 they were reduced by 61 percent. In addition,
development time fell by one-third and quality improved.
In the United States, the 1992 Cadillac was planned and designed entirely
with QFD. The concept has been publicized and developed in the United
States by the American Supplier Institute, Inc., a nonprofit organization, and
by GOAL/QPC, a consulting firm in Massachusetts. Today QFD is success-
fully used by manufacturers of electronics, appliances, clothing, and construc-
tion equipment, and by firms such as General Motors, Ford, Mazda, Motorola,
Xerox, Kodak, IBM, Procter & Gamble, Hewlett-Packard, and AT&T.
The term quality function deployment represents the overall concept that
provides a means of translating customer requirements into the appropriate
technical requirements for each stage of product development and produc-
tion. The customers’ requirements—expressed in their own terms—are appro-
priately called the voice of the customer. These are the collection of customer
needs, including all satisfiers, delighters/exciters, and dissatisfiers—the
“whats” that customers want from a product.
For example, a consumer might ask that a dishwashing liquid be “long
lasting” and “clean effectively” or that an MP3 player have “good sound
quality.” Sometimes these requirements are referred to as customer attributes.
Under QFD, all operations of a company are driven by the voice of the cus-
tomer, rather than by top management edicts or design engineers’ opinions.
Technical features are the translation of the voice of the customer into
technical language. They are the “hows” that determine the means by which
customer attributes are met. For example, a dishwashing detergent loosens
grease and soil from dishes. The soil becomes trapped in the suds so dishes
can be removed from the water without picking up grease. Eventually the
suds become saturated with soil and break down. Thus, a technical feature
of a dishwashing liquid would be the weight of greasy soil that the suds gen-
erated by a fixed amount of dishwashing liquid can absorb before breaking
down. Another might be the size of the soap bubble (which, incidently, has
been found to be a key attribute of customers’ perception of cleaning effec-
tiveness!).
A set of matrices is used to relate the voice of the customer to technical
features and production planning and control requirements. The basic plan-
ning document is called the customer requirement planning matrix. Because
of its structure (Figure 3.1), it is often referred to as the House of Quality.
The House of Quality relates customer attributes to technical features to
ensure that any engineering decision has a basis in meeting a customer need.
Building the House of Quality requires six basic steps:
90 Part I: Introduction to Total Quality

FIGURE 3.1 THE HOUSE OF QUALITY

Interrelationships

Technical features

Voice of Relationships between customer Importance of Competitive

customer attributes and technical features customer evaluation
attributes

Priorities of technical features

1. Identify customer attributes.

2. Identify technical features.
3. Relate the customer attributes to the technical features.
4. Evaluate competing products.
5. Evaluate technical features and develop targets.
6. Determine which technical features to deploy in the production process.

The first step is identifying customer attributes. In applying QFD, it is impor-

tant to use the customer’s own words so as not to have customer needs mis-
interpreted by designers and engineers. Recall that not all customers are end-
users. For a manufacturer, customers might include government regulators,
wholesalers, and retailers. Thus, many classes of customer needs may exist.
The second step is listing the technical features that are necessary to
meet customer requirements. These technical features are design attributes
expressed in the language of the designer and engineer. They form the basis
for subsequent design, manufacturing, and service process activities. They
must be measurable, because the output will be controlled and compared to
objective targets.
The roof of the House of Quality shows the interrelationships between
any pair of technical features. Various symbols are used to denote these
 Chapter 3: Total Quality Tools and Statistical Thinking 91

relationships. A typical scheme is to use the symbol • to denote a very strong

relationship, ❍ for a strong relationship, and ∆ to denote a weak relation-
ship. These notations help determine the effects of changing one product
characteristic and enable planners to assess the tradeoffs between character-
istics. This process enables designers to focus on features collectively rather
than individually.
Next, a relationship matrix between the customer attributes and the
technical features is developed. Customer attributes are listed down the left
column, and technical features are written across the top. In the matrix itself,
a symbol is used to indicate the degree of relationship in a manner similar to
that used in the roof of the house. The purpose of the relationship matrix is
to show whether the final technical features adequately address the cus-
tomer attributes. This assessment may be based on expert experience, cus-
tomer responses, or controlled experiments.
Technical features can affect several customer attributes. The lack of a
strong relationship between a customer attribute and any of the technical
features suggests that the attributes are not being addressed and that the
final product will have difficulty meeting customer needs. Similarly, if a
technical feature does not affect any customer attribute, it may be redundant
or the designers may have missed an important customer attribute.
The next step is adding market evaluation and key selling points. This
step includes rating the importance of each customer attribute and evaluat-
ing existing products on each of the attributes. Customer importance ratings
represent the areas of greatest interest and highest expectations to the cus-
tomer. Competitive evaluation helps highlight the absolute strengths and
weaknesses of competing products. This step enables designers to seek oppor-
tunities for improvement. It also links QFD to a company’s strategic vision
and allows priorities to be set in the design process. For example, focusing
on an attribute that receives a low evaluation on all competitors’ products
can help to gain a competitive advantage. Such attributes become key sell-
ing points and help establish promotion strategies.
Next comes the evaluation of the technical features of competitive prod-
ucts and the development of targets. This is usually accomplished through in-
house testing and translated into measurable terms. These evaluations are
compared with the competitive evaluation of customer attributes to find
inconsistencies. If a competing product best satisfies a customer attribute, but
the evaluation of the related technical feature indicates otherwise, then either
the measures used are faulty or the product has an image difference (either
positive toward the competitor or negative toward the product) that affects
customer perceptions. Targets for each technical feature are set on the basis of
customer importance ratings and existing product strengths and weaknesses.
The final step in building the House of Quality is selecting technical fea-
tures to be deployed in the remainder of the process. This means identifying
the characteristics that have a strong relationship to customer needs, have
poor competitive performance, or are strong selling points. These charac-
teristics need to be deployed—or translated into the language of each
92 Part I: Introduction to Total Quality

function—in the design and production process, so that proper actions and
controls are taken to maintain the voice of the customer. Characteristics that
are not identified as critical do not need such rigorous attention.
A simple example of a House of Quality is shown in Figure 3.2 for the
hypothetical case of a quick-service franchise that wishes to improve its
hamburger. The voice of the customer consists of four attributes. The ham-
burger should
• be tasty,
• be healthy,
• be visually appealing, and
• provide good value.
The technical features that can be designed into the product are price, size,
calories, sodium content, and fat content. The symbols in the matrix show the
relationships between each customer attribute and technical feature. For
example, taste bears a strong relationship to sodium content, a moderate
relationship to fat content, and a weak relationship to caloric content. In the
roof of the house, price and size are seen to be strongly related (as size
increases, the price must increase). The competitive evaluation shows that
competitors are currently weak on nutrition and value; these can become key
selling points in a marketing plan if the franchise can capitalize on them.
Finally, at the bottom of the house, are targets for the technical features based

FIGURE 3.2 HOUSE OF QUALITY EXAMPLE

Importance

Competitive
Customer
Calories

Evaluation
Sodium
Price
Size

Fat

Us A B

Taste 4 3 4 5
Nutrition 4 3 2 3
Visual appeal 3 3 5 4
Good value 5 4 3 4
Our priority 5 4 4 4 5
Competitor A 2 5 3 2 4
Competitor B 3 4 4 3 3
Deployment
* * *
Legend: 1 = low, 5 = high

Very strong relationship

Strong relationship
Weak relationship
 Chapter 3: Total Quality Tools and Statistical Thinking 93

on an analysis of customer importance ratings and competitive ratings. The

features assigned asterisks will be deployed in subsequent design and pro-
duction activities.
The House of Quality provides marketing with an important tool to
understand customer needs and gives top management strategic direction.
However, it is only the first stage in the QFD process. The voice of the cus-
tomer must be carried throughout the production process. Three other houses
of quality are used to deploy the voice of the customer to component parts
characteristics, process planning, and production planning. These are

1. technical features deployment matrix, which translates technical features of

the final product into design requirements for critical components;
2. process plan and quality control charts, which translate component features
into critical process and product parameters and control points for each;
and
3. operating instructions, which identify operations to be performed by plant
personnel to assure that important process and product parameters are
achieved.

Most of the QFD activities represented by the first two houses of qual-
ity are performed by people in the product development and engineering
functions. At the next stage, the planning activities begin to involve super-
visors and production-line operators. This represents the transition from
planning to execution. If a product component parameter is critical and is
created or affected during the process, it becomes a control point. This tells
the company what to monitor and inspect and forms the basis for a quality
control plan for achieving those critical characteristics that are crucial to achiev-
ing customer satisfaction. The last house relates the control points to specific
requirements for quality assurance activity. This includes specifying control
methods, sample sizes, and so on, to achieve the necessary level of quality.

Concurrent Engineering
A topic closely related to QFD is concurrent engineering. This is the concept
that all major functions that contribute to getting a product to market have
continuing product-development involvement and responsibility from orig-
inal concept through sales.
The designer’s objective is to create a product that meets the desired func-
tional requirements. The manufacturing engineer’s objective is to produce
the designed product efficiently. The salesperson’s goal is to sell the product,
and that of finance personnel is to make a profit. Purchasing must ensure
that purchased parts meet quality requirements. Packaging and distribution
personnel must ensure that the product reaches the customer in good oper-
ating condition. Since all these functions have a stake in the product, they
must all work together.
Unfortunately, the product development process in many large firms is
carried out in a serial fashion with little cooperation among departments. In
94 Part I: Introduction to Total Quality

the early stages of development, design engineers dominate the process.

Later the prototype is transferred to manufacturing for production. Finally,
marketing and sales personnel are brought in.
This approach has several disadvantages. First, product development time
is long. Second, up to 90 percent of manufacturing costs may be committed
before manufacturing engineers have any input to the design. Third, the final
product may not be the best one for market conditions when it is introduced.
Ford Motor Company was the first U.S. automotive firm to move away
from this traditional approach when developing the Taurus/Sable.1 “Team
Taurus” took an approach in which representatives from all the various units—
planning, design, engineering, and manufacturing—worked together as a
group. Communication was dramatically improved, and many problems were
resolved much earlier in the process. For instance, manufacturing suggested
changes in design that resulted in higher productivity and better quality.
Comprehensive market surveys were conducted to determine customers’
wants and preferences. Ford even asked assembly-line workers for their advice
before the car was designed. Workers complained that they had trouble
installing doors because the body panels were made up of too many pieces.
Designers reduced the number of panels from eight to two. One worker sug-
gested that all bolts should have the same head size so that they would not
have to change wrenches constantly. The Taurus/Sable has been one of Ford’s
biggest success stories and has revolutionized the product development
approach. Chrysler and General Motors have moved toward similar styles of
product development. In fact, Chrysler’s expertise in product development has
been noted as one of the key reasons it was acquired by Daimler-Benz.
Multifunctional teams of 4 to 20 members typically comprise concurrent
engineering teams. The functions of such teams include the following:

1. Distinguishing the character of the product in order to determine appro-

priate design and production methods, and ensuring that the product
can be repaired easily.
2. Analyzing product functions so that all design decisions can be made
with full knowledge of how the item is supposed to work and so that all
team members understand it well enough to contribute.
3. Relating product function to production methods. Computer-aided design
tools allow a designer to simulate product performance by varying assump-
tions within a computer model.
4. Performing a design-for-manufacturability study to determine whether
the design can be made easier to produce without affecting performance.
5. Designing an assembly sequence that integrates quality control and ensures
that each part is designed so that its quality is compatible with the assem-
bly method.
6. Designing a factory system that fully involves workers in the production
strategy, operates on minimal inventory, and is integrated with suppliers’
methods and capabilities.
 Chapter 3: Total Quality Tools and Statistical Thinking 95

One example of an aggressive product-development effort based on con-

current engineering was pioneered by Cincinnati Milacron. Milacron code-
named its effort “Wolfpack” because wolves work in packs and have the
instinct to kill. Fifteen teams of managers drawn from different areas of the
company are responsible for developing globally competitive machines at
significantly lower costs. The Wolfpack concept has cut lead time for new
products from 3 years to 18 months and allowed Milacron to regain lost
market leadership in several areas.2

The New Seven Management and Planning Tools

Many of the problems in implementing Quality Function Deployment are
due to the way that American managers have become accustomed to plan-
ning and organizing businesses based on Frederick W. Taylor’s philosophy.3
The barriers to effective planning and quality improvement efforts have been

• strict departmentalization that has separated the planners (staff special-

ists, such as QA and industrial engineers) from the doers (line and func-
tional managers), thus limiting the abilities of both to make significant
improvements;
• relegating planning to a seat-of-the-pants approach, due to the percep-
tion that it is either too theoretical to be of practical use or too detailed to
be interesting or action-oriented;
• a lack of available tools to make planning available and timely for man-
agers to use.

The seven management and planning tools had their roots in post–World
War II operations research developments in the United States, but were com-
bined and refined by several Japanese companies over the past several
decades as part of their planning processes. They were popularized in the
United States by the consulting firm GOAL/QPC and have been used by a
number of firms since 1984 to improve their quality planning and improve-
ment efforts. They are new only to managers who have not previously seen
what powerful aids they can be in improvement processes.
These tools can be used to address problems typically faced by managers
who are called upon to structure unstructured ideas, make strategic plans,
and organize and control large, complex projects. They have helped to over-
come the barriers listed previously and have given managers tools appro-
priate to their specific needs for planning and implementing quality improve-
ment efforts. Due to space limitations, only a brief discussion of each tool
follows. (See books by Brossart, Brassard, and Mizuno4 for further details
and examples.)

Affinity Diagram/KJ Method

This is a technique for gathering and organizing a large number of ideas,
opinions, and facts relating to a broad problem or subject area. It enables
96 Part I: Introduction to Total Quality

problem solvers to sift through large volumes of information efficiently and

to identify natural patterns or groupings in the information. This method
was developed in the 1960s by Kawakita Jiro, a Japanese anthropologist.
“KJ” is a trademark registered by the Kawayoshida Research Center.
The technique requires that a group of six to eight people meet to con-
sider a broad issue, such as identifying the elements of poor quality cost for
their organization. Responses can be recorded on a flip chart or on small
cards that can be posted and moved around on a board. For example, in
determining the elements of quality cost, the group will probably list vari-
ous elements in a random fashion. Once many ideas have been generated,
they can be grouped according to their “affinity,” or relationship, to each
other. An example is shown in Figure 3.3. This technique helps managers
focus on the key issues and their elements rather than an unorganized col-
lection of information.
The Affinity Diagram/KJ Method is intended to be a creative, rather than
a logical process. It resembles brainstorming and “story-boarding,” a tech-
nique developed by Walt Disney to create cartoons and movies.

Interrelationship Digraphs
The purpose of an interrelationship digraph is to take a central idea and
map out logical or sequential links among related categories. It shows that
every idea can be logically linked with more than one idea at a time, and
allows for “lateral” rather than “linear” thinking. This technique often is

FIGURE 3.3 EXAMPLE OF AN AFFINITY DIAGRAM

Poor Quality Cost

Equipment Cost Failure Cost

Instruments Internal External

Complaint
Test and Inspection Scrap
Investigation

Computers for
Rework Recalls
Process Control

Downtime Returns
 Chapter 3: Total Quality Tools and Statistical Thinking 97

used after the affinity diagram has brought issues and problems into clearer
focus. Figure 3.4 shows an example of how failure costs are influenced by
other factors.
Like affinity diagrams, this technique also depends on getting together
a team of people who own the problem. Some of the same cards or flip-chart
lists developed in the affinity diagram can be duplicated and used in this
technique. New cards or lists of specific items must be added frequently as
the issue becomes more focused.

Tree Diagram
A tree diagram maps out the paths and tasks that need to be accomplished
to complete a specific project or to reach a specified goal. A planner uses this
technique to seek answers to such questions as “What sequence of tasks
needs to be completed to address the issue?” or “What are all of the factors
that contribute to the existence of the key problem?”
This technique brings the issues and problems disclosed by the affinity
diagram and the interrelationship digraph down to the operational planning
stage. A clear statement of the problem or process must be specified. From this
general statement, a team can be established to recommend steps required to
solve the problem or implement the plan. The “product” produced by this
group would be a tree diagram with activities and recommendations for tim-
ing the activities. Figure 3.5 shows an example of some of the key elements
in establishing a quality cost system.

Matrix Diagrams
These are spreadsheets that graphically display relationships between char-
acteristics, functions, and tasks in such a way as to provide logical connecting

FIGURE 3.4 EXAMPLE OF AN INTERRELATIONSHIP DIGRAPH

Rework Cost/Unit

Prevention Number Total Rework

Cost Reworked Cost

Failure
Cost
Defective Cost of Not
Units Reworking

Number Cost of Processing

Returned Returns
98 Part I: Introduction to Total Quality

FIGURE 3.5 EXAMPLE OF A TREE DIAGRAM

Control Over
Reporting Process
Responsible
Departments
Computerization of
Data Collection

Accounting System
Establishing Data
Quality Cost Collection Estimation
System
Special Systems

What to
Measure
Intangible
Costs
Coordination with
Marketing and
Customer Service

points between each item. The House of Quality is an example of one of the
many matrix diagrams now used for planning and quality improvement.
Matrix Data Analysis
This process takes data from matrix diagrams and seeks to arrange it quan-
titatively to display the strength of relationships among variables so that
they can be easily viewed and understood. Matrix data analysis is a rigor-
ous, statistically based “factor analysis” technique. GOAL/QPC personnel
feels that this method, although worthwhile for many applications, is too
quantitative to be used on a daily basis and they have developed an alterna-
tive tool called a prioritization matrix that is easier to understand and imple-
ment. This approach bears a lot of similarity to decision matrices that you
may have studied in a quantitative methods course. Interested readers should
consult Brassard’s book for further details.
Process Decision Program Chart (PDPC)
This is a method for mapping out every conceivable event and contingency
that can occur when moving from a problem statement to possible solutions.
It is used to plan for each possible chain of events that could occur when a
problem or goal is unfamiliar. A PDPC takes each branch of a tree diagram,
anticipates possible problems, and provides countermeasures that will pre-
vent the deviation from occurring or be in place if the deviation does occur.
Figure 3.6 shows one example.
 Chapter 3: Total Quality Tools and Statistical Thinking 99

FIGURE 3.6 EXAMPLE OF A PROCESS DECISION PROGRAM CHART

Goal: Install Quality Cost System

Establish Select
Expand
Installation Trial
System
Team Department

Obtain Identify
Steps: Cooperation Quality Cost
of Users Categories

Middle
What-Ifs?: Cost Not
Management
Available
Resistance

Possible Executive Modify Direct

Countermeasures: Order Accounting Observation
System
Focus on
Benefits

Arrow Diagrams
These have been used by construction planners for years in the form of CPM
and PERT project planning techniques. Arrow diagramming has also been
taught extensively in quantitative methods, operations management, and
other business and engineering courses in the United States for a number of
years. Unfortunately, its use has been confined to technical experts. By adding
it to the “quality toolbox,” it has become more widely available to general
managers and other nontechnical personnel.
Implementation of process improvements is an essential, but frequently
ignored, step. Process improvements often are not implemented because they
are too complex to work in practice or are not accepted by those who have the
responsibility to carry them out. These seven quality improvement tools assist
managers in implementing improvements through active involvement.

TOOLS FOR CONTINUOUS IMPROVEMENT

Many tools have been created or adapted from other disciplines (such as oper-
ations research and industrial engineering) to facilitate the process of con-
tinuous improvement. In this section we describe the most common ones
used in quality improvement applications.
100 Part I: Introduction to Total Quality

Tools for Data Collection and Analysis

Seven simple statistically based tools are used extensively to gather and ana-
lyze data. Like the seven management and planning tools, these tools—flow-
charts, check sheets, histograms, Pareto diagrams, cause-and-effect diagrams,
scatter diagrams, and control charts—are visual in nature and simple enough
for anyone to understand. Historically, these tools preceded the seven man-
agement and planning tools and often are called the “seven QC (quality con-
trol) tools.” The seven management and planning tools have been referred to
as the “new seven.”
Flowcharts
A flowchart is a picture of a process that shows the sequence of steps per-
formed. Figure 3.7 is an example. Flowcharts are best developed by the peo-
ple involved in the process—employees, supervisors, managers, and cus-
tomers. A facilitator often is used to provide objectivity, to ask the right
questions, and to resolve conflicts. The facilitator can guide the discussion
through questions such as “What happens next?,” “Who makes the decision
at this point?,” and “What operation is performed here?” Often the group does

FIGURE 3.7 EXAMPLE OF A FLOWCHART FOR TRAINING NEW PRINTING PRESS OPERATORS

Hire candidate

Training with operator checklist

Safety, quality, and procedures testing

No
Test passed?
Yes

Four-week evaluation

pass
Solo with lead operator support

fail
90-day evaluation Reevaluate
employee

pass
Press-certified
 Chapter 3: Total Quality Tools and Statistical Thinking 101

not agree on the answers to these questions, due to misconceptions about the
process or a lack of awareness of the “big picture.”
Flowcharts help the people involved in the process to understand it bet-
ter. For example, employees realize how they fit into a process—that is, who
their suppliers and customers are. By helping to develop a flowchart, workers
begin to feel a sense of ownership in the process and become more willing to
work on improving it. Using flowcharts to train employees on standard pro-
cedures leads to more consistent performance.
Once a flowchart is constructed, it can be used to identify quality prob-
lems as well as areas for improvement. Questions such as “How does this
operation affect the customer?,” “Can we improve or eliminate this opera-
tion?,” or “Should we control a critical quality characteristic at this point?”
help to identify such opportunities. Flowcharts help people to visualize sim-
ple but important changes that could be made in a process.
Check Sheets
These tools aid in data collection. When designing a process to collect data,
one must first ask basic questions such as:
• What question are we trying to answer?
• What type of data will we need to answer the question?
• Where can we find the data?
• Who can provide the data?
• How can we collect the data with minimum effort and minimum chance
of error?
Check sheets are data collection forms that facilitate the interpretation of
data. Quality-related data are of two general types—attribute and variable.
Attribute data are obtained by counting or from some type of visual inspec-
tion: the number of invoices that contain errors, the number of parts that con-
form to specifications, and the number of surface defects on an automobile
panel, for example. Variable data are collected by numerical measurement on
a continuous scale. Dimensional characteristics such as distance, weight, vol-
ume, and time are common examples. Figure 3.8 is an example of an attrib-
ute data check sheet, and Figure 3.9 shows a variable data check sheet.

FIGURE 3.8 EXAMPLE OF A CHECK SHEET FOR ATTRIBUTE DATA: AIRLINE COMPLAINTS

Type Week 1 Week 2 Week 3 Week 4

Lost baggage

Baggage delay

Missed connection

Poor cabin service

Ticketing error
102 Part I: Introduction to Total Quality

FIGURE 3.9 EXAMPLE OF A CHECK SHEET FOR VARIABLE DATA

Frequency
20
19
18
17
16
15
14
13 X
12 X
11 X X
10 X X X
9 X X X
8 X X X X
7 X X X X
6 X X X X
5 X X X X X
4 X X X X X X
3 X X X X X X X
2 X X X X X X X
1 X X X X X X X X X X
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Time to process loan request (days)

Histograms
Variation in a process always exists and generally displays a pattern that can
be captured in a histogram. A histogram is a graphical representation of the
variation in a set of data. It shows the frequency or number of observations
of a particular value or within a specified group.
Histograms provide clues about the characteristics of the population from
which a sample is taken. Using a histogram, the shape of the distribution can
be seen clearly, and inferences can be made about the population. Patterns can
be seen that would be difficult to see in an ordinary table of numbers.
The check sheet in Figure 3.9 was designed to provide the visual appeal of
a histogram as the data are tallied. It is easy to see how the output of the process
varies and what proportion of output falls outside of any specification limits.
Pareto Diagrams
Pareto analysis is a technique for prioritizing types or sources of problems.
Pareto analysis separates the “vital few” from the “trivial many” and pro-
vides help in selecting directions for improvement. It is often used to analyze
the attribute data collected in check sheets. In a Pareto distribution the char-
acteristics are ordered from largest frequency to smallest. For example, if the
data in Figure 3.8 is placed in order of decreasing frequency, the result is

1. Baggage delay
2. Poor cabin service
 Chapter 3: Total Quality Tools and Statistical Thinking 103

3. Missed connection
4. Lost baggage
5. Ticketing error

A Pareto diagram is a histogram of these data, as shown in Figure 3.10.

A cumulative frequency curve is usually drawn on the histogram, as shown.
Such pictures clearly show the relative magnitude of defects and can be used
to identify the most promising opportunities for improvement. They can
also show the results of improvement projects over time.

FIGURE 3.10 EXAMPLE OF A PARETO DIAGRAM

Frequency Percent
100

Cumulative Percent

23 50

14

4
2
Ba

Po

Mi

Lo

Tic
ss

st
o
gg

ke
rc

ed

ba

tin
ag

ab

gg

ge
co
ed

in

ag
nn

rro
ela

se

ec

r
rvi
y

tio
ce

n
104 Part I: Introduction to Total Quality

Cause-and-Effect Diagrams
The most useful tool for identifying the causes of problems is a cause-and-
effect diagram, also known as a fishbone or Ishikawa diagram, named after
the Japanese quality expert who popularized the concept. A cause-and-effect
diagram is simply a graphical representation of an outline that presents a
chain of causes and effects.
An example is shown in Figure 3.11. At the end of the horizontal line is
the problem to be addressed. Each branch pointing into the main stem rep-
resents a possible cause. Branches pointing to the causes are contributors to
these causes. The diagram is used to identity the most likely causes of a
problem so that further data collection and analysis can be carried out.
Cause-and-effect diagrams are usually constructed in a brainstorming set-
ting so that everyone can contribute their ideas. Usually small groups drawn
from operations or management work with an experienced facilitator. The
facilitator guides the discussion to focus attention on the problem and its
causes, on facts, not opinions. This method requires significant interaction
among group members. The facilitator must listen carefully to the partici-
pants and capture the important ideas.
Scatter Diagrams
Scatter diagrams illustrate relationships between variables, such as the per-
centage of an ingredient in an alloy and the hardness of the alloy, or the num-
ber of employee errors and overtime worked (Figure 3.12). Typically the
variables represent possible causes and effects obtained from cause-and-effect
diagrams.
A general trend of the points going up and to the right indicates that an
increase in one variable corresponds to an increase in the other. If the trend is
down and to the right, an increase in one variable corresponds to a decrease
in the other. If no trend can be seen, then it would appear that the variables

FIGURE 3.11 EXAMPLE OF A CAUSE-AND-EFFECT DIAGRAM

Client Time
Rush request
Poor
Unclear handwriting
directions Overload

Word
processing
errors

Inattention
No spell check

Did not
understand Training
directions
Typist
 Chapter 3: Total Quality Tools and Statistical Thinking 105

FIGURE 3.12 EXAMPLE OF A SCATTER DIAGRAM

Number
of errors

Volume of work

are not related. Of course, any correspondence does not necessarily imply
that a change in one variable causes a change in the other. Both may be the
result of something else. However, if there is reason to believe causation, the
scatter diagram may provide clues on how to improve the process.
Control Charts
These tools are the backbone of statistical process control (SPC), and were
first proposed by Walter Shewhart in 1924. Shewhart was the first to distin-
guish between common causes and special causes in process variation. He
developed the control chart to identify the effects of special causes. Much of
the Deming philosophy is based on the use of control charts to understand
variation.
A control chart displays the state of control of a process (Figure 3.13). Time
is measured on the horizontal axis, and the value of a variable on the verti-
cal axis. A central horizontal line usually corresponds to the average value of
the quality characteristic being measured.
Two other horizontal lines represent the upper and lower control limits,
chosen so that there is a high probability that sample values will fall within
these limits if the process is under control—that is, affected only by common
causes of variation. If points fall outside of the control limits or if unusual
patterns such as shifts up or down, trends up or down, cycles, and so forth
exist, special causes may be present.
As we noted in chapter 2, two fundamental mistakes that can be made
concerning variation are
1. treating special causes as common causes, and
2. treating common causes as special causes.
Control charts minimize the risk of making these two types of mistakes. As
a problem-solving tool, they allow workers to identify quality problems as
they occur and base their conclusions on hard facts.
The seven QC tools provide excellent communication vehicles both ver-
tically and horizontally across organizational boundaries (see box).
106 Part I: Introduction to Total Quality

FIGURE 3.13 EXAMPLE OF A CONTROL CHART

Percent shipped
within 24 hours

Upper
97% Control
Limit

93% Average

Lower
89% Control
Limit

Day
1 2 3 4 5 6 7 8 9 10 11 12 13

Shooting for Quality5

Timothy Clark observed that in basketball games, his son Andrew’s free-
throw percentage averaged between 45 and 50 percent. Andrew’s process
was simple: Go to the free throw line, bounce the ball four times, aim, and
shoot. To confirm these observations, Andrew shot five sets of 10 free
throws with an average of 42 percent, showing little variation among the five
sets. Timothy developed a cause-and-effect diagram (Figure 3.14) to identify
the principal causes. After analyzing the diagram and observing his son’s
process, he believed that the main causes were not standing in the same
place on the free-throw line every time and having an inconsistent focal point.
They developed a new process in which Andrew stood at the center of the
line and focused on the middle of the front part of the rim. The new process
resulted in a 36 percent improvement in practice (Figure 3.15). Toward the
end of the 1994 season, he improved his average to 69 percent in the last
three games.
During the 1995 season, Andrew averaged 60 percent. A control chart
(Figure 3.16) showed that the process was quite stable. In the summer of
1995, Andrew attended a basketball camp where he was advised to change
his shooting technique. This process reduced his shooting percentage during
the 1996 season to 50 percent. However, his father helped him to reinstall
his old process, and his percentage returned to its former level, also improv-
ing his confidence.
FIGURE 3.14 FREE-THROWING CAUSE-AND-EFFECT DIAGRAM

Materials People Measurement

Regulation Player Hit

backboard
and rim Nothing but net
Miss

Touched rim
Short Left
Basketball Coach
Right Long Low
free-throw
shooting
Shooting
Video percentage
Practice position
camera
Indoor
Technique

Games
Ritual
Outdoor Focus point

Environment Equipment Method

© 1997 American Society for Quality. Reprinted with permission.

FIGURE 3.15 FREE-THROWING SHOTS MADE BEFORE AND AFTER IMPLEMENTING THE
IMPROVEMENT (3/17/94–11/23/94)

10

9 Before = 42% After = 57%

7
Number of shots made

1 3 5 7 9 11 13 15 17 19 21 23 25

Practice session
© 1997 American Society for Quality. Reprinted with permission.

107
108 Part I: Introduction to Total Quality

FIGURE 3.16 DETERMINING WHETHER THE FREE-THROW PROCESS IS STABLE

(3/17/94–1/18/96)

50

45
Upper control limit
40

Number of shots made

35

30

25

20
Lower control limit
15

10

5
2 4 6 8 10 12 14 16 18 20
0
Practice session
© 1997 American Society for Quality. Reprinted with permission.

The Deming Cycle

Managers need systematic approaches to drive continuous improvement
programs. One approach that can be learned and applied by everyone in an
organization is the Deming Cycle. The Deming Cycle is a methodology for
improvement, based on the premise that improvement comes from the appli-
cation of knowledge.6 Knowledge of engineering, management, or operations
may make a process easier, more accurate, faster, less costly, safer, or better-
suited to customer needs. Three fundamental questions to consider are

1. What are we trying to accomplish?

2. What changes can we make that will result in improvement?
3. How will we know that a change is an improvement?

This methodology was originally called the Shewhart Cycle after Walter
Shewhart, its founder, but was renamed for Deming by the Japanese in 1950.
The Deming Cycle is composed of four stages: Plan, Do, Study, Act (Figure
3.17). Sometimes it is called the PDSA cycle.
The Plan stage consists of studying the current situation, gathering data,
and planning for improvement. In the Do stage, the plan is implemented on
a trial basis in a laboratory, pilot production process, or with a small group
of customers. The Study stage is designed to determine whether the trial
plan is working correctly and to see whether any further problems or oppor-
 Chapter 3: Total Quality Tools and Statistical Thinking 109

FIGURE 3.17 THE DEMING CYCLE

Plan Do Plan Do

Act Study Act Study

tunities can be found. The last stage, Act, is the implementation of the final
plan to ensure that the improvements will be standardized and practiced
continuously. This leads back to the Plan stage for further diagnosis and
improvement (see box).
As Figure 3.17 suggests, this cycle is never ending. That is, it is focused on
continuous improvement, so the improved standards serve as a springboard

Crying Out for the Deming Cycle7

Kevin Dooley and his wife applied the Deming Cycle and various quality tools
to help stop their infant daughter from crying whenever her diaper was
changed, which as any new parent or older sibling knows, can break your
heart or drive you crazy. Their first cycle involved creating an experiment to
determine the percentage of time crying while on the diaper changing table
(Plan); collecting data on 15 diaper changes and plotting them on a run chart
(Do); observing that the data appeared to be random (Study); and focusing on
the steps involved in the changing process (Act). The second cycle involved
developing a flowchart to document the steps in changing a diaper (Plan);
constructing the chart (Do); studying the process (which did not appear com-
plex or incorrect—Study); and deciding to seek other causes (Act). Cycles 3,
4, and 5 involved developing a cause-and-effect diagram, collecting data to
test the hypothesis that the type of outfit worn caused her to cry more (stud-
ied with a Pareto diagram), and looking for correlations between the time cry-
ing and the time since last changing. In cycle 6, the Dooleys collected data to
determine whether she cried less when being changed by her mother. His-
tograms confirmed a difference between the parents! Cycle 7 was to observe
what Kevin’s wife did differently (Plan); make a list of key differences (Do);
study the differences—his wife had captured the baby’s attention better—
and develop some attention-getting strategies (Act). The last cycle imple-
mented these, and data indeed confirmed an improvement! We’re sure the
Dooleys can’t wait to apply the Deming Cycle to their daughter’s driving . . .
110 Part I: Introduction to Total Quality

for further improvements. This distinguishes it from more traditional problem-

solving approaches and is one of the essential elements of the Deming
philosophy.

Poka-Yoke (Mistake-Proofing)
Poka-yoke is an approach for mistake-proofing processes using automatic
devices or methods to avoid simple human error. The poka-yoke concept was
developed and refined by the late Shigeo Shingo, a Japanese manufacturing
engineer who developed the Toyota production system. The idea is to avoid
repetitive tasks or actions that depend on vigilance or memory in order to free
workers’ time and minds to pursue more creative and value-adding activities.
Poka-yoke is focused on two aspects: prediction, or recognizing that a
defect is about to occur and providing a warning, and detection, or recog-
nizing that a defect has occurred and stopping the process. Many applica-
tions of poka-yoke are deceptively simple, yet creative. Usually, they are
inexpensive to implement. Many machines have limit switches connected to
warning lights that tell the operator when parts are positioned improperly
on the machine. In another example, a device on a drill counts the number
of holes drilled in a workpiece; a buzzer sounds if the workpiece is removed
before the correct number of holes has been drilled. As a final example, one
production step at Motorola involves putting alphabetic characters on a
keyboard, then checking to ensure each key is placed correctly. A group of
workers designed a clear template with the letters positioned slightly off
center. By holding the template over the keyboard, assemblers can quickly
spot mistakes.
Poka-yoke techniques are also applied to the design of consumer prod-
ucts to prevent inadvertent user errors or safety hazards. For example, a 3.5-
inch diskette is designed so that it cannot be inserted unless the disk is oriented
correctly (try it!). These disks are not perfectly square, and the bevelled right
corner of the disk allows a stop in the disk drive to be pushed away if it is
inserted correctly. Power lawn mowers now have a safety bar on the handle
that must be engaged in order to start the engine. Computer software such
as Microsoft Word will automatically check for any unsaved files before clos-
ing down. A proxy ballot for an investment fund will not fit into the return
envelope unless a small strip is detached. The strip asks the respondent to
check whether the ballot is signed and dated.
Richard B. Chase and Douglas M. Stewart suggest that the same concepts
can be applied to services.8 The major differences are that service mistake-
proofing must account for the customers’ activities as well as those of the
producer, and mistake-proof methods must be set up for interactions con-
ducted directly or by phone, mail, or other technologies, such as ATM. Chase
and Stewart classify service poka-yokes by the type of error they are designed
to prevent: server errors and customer errors. Server errors result from the
task, treatment, or tangibles of the service. Customer errors occur during prepa-
ration, the service encounter, or during resolution.
 Chapter 3: Total Quality Tools and Statistical Thinking 111

Task errors include doing work incorrectly, in the wrong order, or too
slowly, as well as doing work not requested. Some examples of poka-yoke
devices for task errors are computer prompts, color-coded cash register
keys, measuring tools such as McDonald’s french-fry scoop, and signaling
devices. Hospitals use trays for surgical instruments that have indentations
for each instrument, preventing the surgeon from leaving one of them in the
patient.
Treatment errors arise in the contact between the server and the customer,
such as lack of courteous behavior, and failure to acknowledge, listen, or
react appropriately to the customer. A bank encourages eye contact by requir-
ing tellers to record the customer’s eye color on a checklist as they start the
transaction. To promote friendliness at a fast-food restaurant, trainers pro-
vide the four specific cues for when to smile: when greeting the customer,
when taking the order, when telling about the dessert special, and when giv-
ing the customer change. They encourage employees to observe whether the
customer smiled back, a natural reinforcer for smiling.
Tangible errors are those in physical elements of the service, such as
unclean facilities, dirty uniforms, inappropriate temperature, and document
errors. Hotels wrap paper strips around towels to help the housekeeping staff
identify clean linen and show which ones should be replaced. Spell-checkers
in word processing software help eliminate misspellings (provided they are
used!).
Customer errors in preparation arise when customers do not bring nec-
essary materials to the encounter, do not understand their role in the service
transaction, or do not engage the correct service. Digital Equipment provides
a flowchart to specify how to place a service call. By guiding the customers
through three yes-or-no questions, the flowchart prompts them to have the
necessary information before calling.
Customer errors during an encounter can be due to inattention, misun-
derstanding, or simply a memory lapse, and include failure to remember
steps in the process or to follow instructions. Poka-yoke examples include
height bars at amusement rides that indicate rider size requirements, beep-
ers that signal customers to remove cards from ATM machines, and airplane
lavatory doors that must be locked to turn on the lights. Some cashiers at
restaurants fold back the top edge of credit card receipts, holding together
the restaurant’s copies while revealing the customer’s copy.
Customer errors at the resolution stage of a service encounter include
failure to signal service inadequacies, learn from experience, adjust expecta-
tions, and execute appropriate post-encounter actions. Hotels might enclose
a small gift certificate to encourage guests to provide feedback. Strategically
placed tray-return stands and trash receptacles remind customers to return
trays in fast-food facilities.
Mistake-proofing a service process requires identifying when and where
failures generally occur (see box). Once a failure is identified, the source
must be found. The final step is to prevent the mistake from occurring
through source inspection, self-inspection, or sequential checks.
112 Part I: Introduction to Total Quality

Mistake-Proofing the U.S. Election Process?9

The controversy surrounding the 2000 U.S. presidential election is well

known. A difference of roughly 500 votes out of 5.8 million cast in Florida is
smaller than the predictable number of errors for miscounted votes, miscast
votes, incorrectly rejected ballots, and other vote-casting and vote counting-
errors caused by systems and processes (people, equipment, methods,
materials, and environment). Several issues affect the quality of the results.

• The prescored punch cards commonly used for ballots require that voters
punch the cards in a way that meets machine specifications. The stylus
used to punch out the chad (the little piece that gets punched out of the
ballot card) could be inadequate in terms of shape or sharpness. The die
that prescores the cards during manufacturing wears down over time.
Cards may be too thick or thin. Cards exposed to too much humidity may
not be counted properly in machines.
• Process errors can occur in several places. The voter may not have actu-
ally cast a vote, or the voter may not have been able to vote for the can-
didate of choice because of confusing ballot design. The ballot may not
have been counted correctly. Manual recounts may not record the voter’s
intent correctly.
• Lack of uniform standards among voting jurisdictions make it difficult to
predict process error rates accurately.
Reducing such errors will require a systematic approach to problem solv-
ing built on quality principles and mistake-proofing ideas. While such thinking
has been successful in many organizations, it has not become a part of the pub-
lic policy dialog nor has it been institutionalized as part of any national debate.

CREATIVITY AND INNOVATION

Creativity is the ability to discover useful new relationships or ideas; innova-

tion refers to the practical implementation of such ideas.
Research studies have suggested that the achievement of business excel-
lence requires a change-oriented environment where creativity of employees
is nurtured, developed, and sustained.10
From the perspective of total quality, creativity and innovation are needed
to better respond to customer needs, particularly the “exciters/delighters”
that customers cannot articulate, and to develop the products and services
that will position an organization strategically ahead of its competitors. They
also are needed to support continuous improvement efforts, for example the
poka-yoke devices and methods discussed in this chapter. Finally, an envi-
ronment that fosters creativity and innovation can motivate employees more
than any extrinsic reward—“joy in work” as Deming used to say (see box).
Thus, creativity and innovation are instrumental in achieving the principles
of total quality discussed in the first chapter.
 Chapter 3: Total Quality Tools and Statistical Thinking 113

Creativity in the Heartland

In 1994, three companies received the Malcolm Baldrige National Quality

Award: AT&T Consumer Communications Services (CCS), GTE Directories Cor-
poration, and Wainwright Industries. AT&T CCS (the long-distance provider) and
GTE Directories (which publishes and sells advertising for telephone directo-
ries) are large, innovative firms, with sophisticated technologies and human
resource development activities that support empowered teams and foster
an atmosphere of creativity throughout their organizations. Both companies
also offer a wide variety of training and education courses, including courses
devoted specifically to creativity and innovation. It is not surprising that they
have achieved remarkable results in product and service quality, customer
satisfaction, and in various operational and financial measures.
Wainwright Industries, headquartered in rural St. Peters, Missouri, is con-
siderably different from CCS and GTE. Wainwright is a small, family-owned
business that manufactures stamped and machined parts for the automotive
and other industries. Since initiating continuous improvement processes in
1991, Wainwright has seen continual, and sometimes dramatic, improvements
in customer satisfaction, defect and scrap rates, work-related accidents, man-
ufacturing cycle times, and quality costs. At the same time, market share, pro-
ductivity gains, and profit margins have all increased.
Innovation is a way of life at Wainwright. Each associate averages more
than one implemented improvement per week! That’s over 50 each year, in an
industry that averages at most one suggestion per employee per year in the
United States. Wainwright does not have the comparable resources available to
large corporations like AT&T or GTE. The company spends a relatively high pro-
portion of its budget on training, some of which is outsourced, yet no formal cre-
ativity training is offered. What the company does have is a culture that exudes
a spirit of creativity and innovation. The plant has a folksy, Midwestern atmos-
phere. Everyone—up to the chairman of the board—wears a company uniform
with his or her first name stitched on it. The human resources function is called
“The People Zone.” The training director is known simply as “The Training Guy.”
And a stuffed duck is the company’s mascot and symbol of quality leadership.
Wainwright is an excellent example of how creativity, integrated within a
traditional American company, can lead to exceptional improvements in qual-
ity and business performance. More importantly, when visitors tour the plant,
they see clearly the spirit and enthusiasm exuded by Wainwright associates.
The associates are having fun! Improving quality—and work itself— should
be fun, and people have fun when they are creative.

In Japanese, the word creativity has a literal translation as dangerous

opportunity. In the Toyota production system, which has become the bench-
mark for world-class efficiency, a key concept is soikufu—creative thinking or
inventive ideas, which means capitalizing on worker suggestions. The chair-
man of Toyota once observed: “One of the features of Japanese workers is
114 Part I: Introduction to Total Quality

that they use their brains as well as their hands. Our workers provide 1.5
million suggestions a year, and 95 percent of them are put to practical use.
There is an almost tangible concern for improvement in the air at Toyota.”11
Creativity is often motivated by an individual’s or group’s need to invent
solutions from limited resources. The Japanese have shown remarkable cre-
ativity in developing solutions to manufacturing quality problems. This is
no wonder, given the limited natural resources in Japan and the Japanese
culture focused on eliminating waste and conserving every precious resource
available. The largest source of creativity in any organization is the frontline
employee. They gather a wealth of data and information about their work
every day. To tap into their knowledge, companies must make creativity a key
part of their culture and think of improvement as everybody’s job. This
requires companies to empower their employees to allow them to put their
ideas to work. We will address this further in chapter 8.
Creativity is the foundation of successful problem-solving teams in the
workplace. Quality circles and other forms of teams that address quality and
productivity problems often use a “scientific” approach to solving problems.
Often, the approach overshadows the need for creativity. Creative approaches
to problem solving rely on four key steps:

1. redefining and analyzing the perceived problem,

2. generating ideas,
3. evaluating ideas and selecting a workable solution, and
4. implementing the solution.

In redefining and analyzing a problem, information is collected and organized,

the data and underlying assumptions are analyzed, and the problem is reex-
amined from new perspectives. At this stage, the goal of the problem solver
is to collect facts and achieve a useful problem definition. The purpose of gen-
erating ideas is to develop novel solutions. After ideas have been generated,
they are evaluated and the best one is identified and selected. Finally, the
solution must be put to work. The Deming Cycle is just one specific example
of a creative problem-solving approach.
Innovation and creativity are important aspects of the Malcolm Baldrige
National Quality Award Criteria for Performance Excellence. Mechanisms
used to encourage innovation and creativity include

1. The nonprescriptive nature of the criteria, which encourages innovative

approaches and breakthrough thinking toward meeting the purpose of
the specific items in the criteria. This channels activities toward purpose,
not toward following procedures.
2. Customer-driven quality, which places major emphasis on the positive
side of quality, and stresses enhancement, new services, and customer
relationship management. Enhancing the positive side of quality relies
heavily on creativity, usually more so than steps to reduce errors and
defects, which tend to rely more on well-defined techniques.
 Chapter 3: Total Quality Tools and Statistical Thinking 115

3. Continuous improvement and cycles of learning, which are stressed as

integral parts of the activities of all work units. This encourages analysis
and problem solving everywhere within the company.
4. Strong emphasis on cycle time reduction in all company operations, which
encourages companies to analyze work processes, work organizations, and
the value-added contributions of all process steps. This fosters change,
innovation, and creative thinking in how work is organized and conducted.
5. Focus on future requirements of customers, which encourages compa-
nies to seek innovative and creative ways to serve needs.

Many examples of creativity and innovation are seen in firms that have
received the Baldrige Award. Among the many examples are
• The benchmarking process, pioneered by Xerox.
• Granite Xpress, an automatic loading system for rock, sand, aggregates,
and other construction materials, developed by Granite Rock. The sys-
tem is similar to an automatic teller machine and allows customers to
rapidly, accurately, and automatically order, load, and invoice materials
24 hours a day, seven days a week.
• Wainwright Industries’ practice of reenacting and videotaping work-
place accidents for study and prevention.

Creativity and Organizational Systems

Warren Bennis, a recognized expert on leadership, noted: “There are two ways
of being creative. One can sing and dance. Or one can create an environment
in which singers and dancers flourish.” While the tools discussed in this chap-
ter all help to promote creative thinking, they cannot be applied effectively in
a noncreative environment. Because management designs the organizational
systems, management is responsible for developing a climate conducive to
creativity and innovation. The organizational literature contains many differ-
ent recommendations for fostering creativity. Some of these are listed below:12
• Remove or reduce obstacles to creativity within an organization. These obsta-
cles include the various environmental blocks such as autocratic bosses,
distractions (constant meetings or phone calls), and lack of management
support. In addition, creative people should be relieved of routine duties
and administrative chores.
• Match jobs to individuals’ creative abilities. Some people work best alone;
others work better in groups. Some work well in 9-to-5 time frames; oth-
ers require flex time. Managers need to be tolerant of individual idiosyn-
crasies, nonconformity of dress codes, frequent coffee breaks, and so on.
• Tolerate failures and establish direction. Creative people need an atmosphere
that allows radical ideas without being harshly judged. Seemingly silly
ideas often turn into the best products. However, appropriate direction
must be given, and realistic goals and objectives must be set to maintain
a sense of urgency.
116 Part I: Introduction to Total Quality

• Improve motivation to increase productivity and solve problems creatively. Cre-

ative accomplishments should be recognized publicly, to peers, superi-
ors, and upper management. Such recognition increases both self-esteem
and motivation.
• Enhance the self-esteem and build the confidence of organization members.
Creative individuals are at their best when their minds are challenged—
not their security or ego. Job security, adequate wages, and job satisfac-
tion enhance an individual’s self-confidence and security.
• Improve communication so that ideas can be better shared. This is certainly
true within an organization. Creative individuals also have the need to
communicate with peers outside the organization; such activities should
be encouraged. Creative people need a sounding board for their ideas
and continuous feedback from their efforts.
• Place highly creative people in special jobs and provide training to take advantage
of their creativity. Establish career paths and financial rewards so as not to
disadvantage creative people who are not part of the line organization.
If you examine carefully the underlying principles of Deming’s 14 Points,
you see many similarities with attitudes and organizational structures that sup-
port creativity. This is particularly true in the elimination of fear, the removal of
barriers that inhibit joy in work, elimination of numbers driven and short-term
goal-driven management, a focus on continuous improvement, leadership, and
continual training and education. Fear, in particular, is a creativity killer. The
head of a corporate legal office commented that “when someone tries to use
fear to motivate me, I get all tied up. I get less creative and less willing to take
chances, and you’ve got to take risks to do good work.” A midlevel manager
observed, “Where there is a lot of fear of screwing up, people don’t change
behaviors or work systems. Creativity is inhibited. People work one day at a
time, rather than looking to the future.”13 Generally, the cultures of most organ-
izations do not support the conditions that lead to creative behavior. Cultural
change, education, and training are necessary to develop a creative climate.
Leadership, in particular, can have a significant influence on creativity.14
Researchers have noted that personality, cognitive style orientation, and level
of intrinsic motivation are core characteristics for employee creativity. Cogni-
tive style is the natural orientation or preferred means of problem solving. An
individual with an innovative cognitive style will seek and integrate diverse
information, redefine posed problems, and generate novel ideas. Those with
adaptive cognitive styles will tend to use data within a well-established
domain, accept problems as defined, and generate ideas consistent with
accepted convention. When people enjoy creativity-related tasks, their level of
creative output is high. Researchers have discovered that creativity does not
appear to be enhanced for employees working with supervisors having similar
cognitive styles. In particular, those with innovative cognitive styles receive lit-
tle benefit from working with supervisors who also exhibit creative tendencies,
probably because they already possess the skills, confidence, and values to be
creative. However, it appears that when employees work with supervisors who
 Chapter 3: Total Quality Tools and Statistical Thinking 117

possess a similar intrinsic motivation, creative performance is enhanced.

Researchers have also found that when employees with a low intrinsic moti-
vation for creativity are assigned to high intrinsic motivation supervisors, they
produced lower creative output, perhaps because such supervisors may unin-
tentionally intimidate or suppress creativity of these employees.
These findings have practical implications for selecting, assigning, and
training employees. Managers must consider individuals’ motivation to be
creative. Identifying and assigning employees with the appropriate motiva-
tional orientation for jobs involving creative tasks is likely to enhance inno-
vation. Placing a supervisor with a true appreciation for creative work among
employees with the motivation to create is also useful. The ability and will-
ingness of supervisors to create positive experiences conducive to creativity
may provide a powerful and effective means by which organizational cre-
ativity can be enhanced. We will discuss issues of leadership and motivation
further in later chapters.

STATISTICAL THINKING
Statistical thinking is at the heart of the Deming philosophy and is the basis
for good management. Statistical thinking is a philosophy of learning and
action based on the principles that

1. all work occurs in a system of interconnected processes,

2. variation exists in all processes, and
3. understanding and reducing variation are keys to success15.

Statistical thinking is more than simply applying statistical methods. Statistical

thinking focuses on understanding and reducing variation, not merely quanti-
fying it. Nevertheless, statistical methods are important to be a good statistical
thinker.
A sign that the president of the former Texas Instruments Defense Systems
and Electronics Group (now Raytheon Systems Company) had in his office
said: “Unless you change the process, why would you expect the results to
change?” By viewing work as a process, we can apply management-by-fact
and various quality tools to establish consistent, predictable processes, study
them, and improve them. By viewing processes as interconnected components
of a system, we avoid suboptimization—one of the key principles of Dem-
ing’s Profound Knowledge. When managers make decisions in isolation,
they often fail to see chains of events that might occur throughout the com-
pany because of their decisions. A typical example is designing a product
without consideration of the capability of processes to manufacture it or the
support systems required to service it in the field.
Recognizing and understanding variation is the essence of statistical think-
ing. We discussed principles of variation within the context of Deming’s Pro-
found Knowledge in chapter 2, particularly the differences between common
and special causes of variation. While variation exists everywhere, many
118 Part I: Introduction to Total Quality

business decisions do not often account for it. How often do managers make
decisions based on a single data point or two, seeing trends when they don’t
exist, or manipulating financial figures they cannot truly control (see box)?
The lack of broad and sustained use of statistical thinking in many organi-
zations is due to two reasons.16 First, statisticians historically have functioned
as problem solvers in manufacturing, research, and development, and thereby
focused on individual clients rather than on organizations. Second, statisticians
have focused primarily on technical aspects of statistics rather than emphasiz-
ing the focus on process variation that will lead to bottom-line results. Process
management—category 6 in the Baldrige criteria—includes process definition,
measurement, control, and improvement. Each of these are fundamental to sta-
tistical thinking. Understanding processes provides the context for determining
the effects of variation and the proper type of managerial action to be taken.

The VP’s Dilemma17

Brian Joiner, a noted quality management consultant, relates the following case:
Ed was a regional VP for a service company that had facilities around
the world. He was determined that the facilities in his region would get
the highest customer satisfaction ratings in the company. If he noticed
that a facility had a major drop in satisfaction ratings in one month or
had “below average” ratings for three months in a row, he would call
the manager and ask what had happened—and make it clear that next
month’s rating had better improve. And most of the time, it did!
As the average satisfaction score dropped from 65 to 60 between February
and March, Ed’s memo to his managers read:
Bad news! We dropped five points! We should all focus on improving
these scores right away! I realize that our usage rates have increased
faster than anticipated, so you’ve really got to hustle to give our cus-
tomers great service. I know you can do it!
As Joiner observed,
Do you look at data this way? This month versus last month? This
month versus the same month last year? Do you sometimes look at the
latest data point? The last two data points?
I couldn’t understand why people would only want to look at two
data points. Finally, it became clear to me. With any two data points, it’s
easy to compute a trend: “Things are down 2 percent this month from
last month. This month is 30 percent above the same month last year.”
Unfortunately, we learn nothing of importance by comparing two results
when they both come from a stable process . . . and most data of
importance to management are from stable processes.”
 Chapter 3: Total Quality Tools and Statistical Thinking 119

This variation is quantified through statistical analysis of process data, and

requires understanding the sources, magnitude, and nature of the variation.
Senior management needs to champion the use of statistical thinking by
defining the strategy and goals of the approach, clearly and consistently com-
municating the benefits and results, providing the necessary resources, coach-
ing others, and recognizing and rewarding the desired behavior. To help man-
agers work in this fashion, many organizations are creating core groups of
highly trained professionals (see box) who are skilled in statistical thinking
and can help others to use it effectively. This requires an environment con-
ducive to learning new behaviors and concepts.

Black Belts and Statistical Thinking at General Electric18

A “Black Belt” means one thing to karate students; at General Electric, Black
Belts (along with Green Belts and Master Black Belts) are employees who are
highly trained in quality improvement principles and techniques. They roam man-
ufacturing plants to improve quality as a part of a major initiative introduced by
CEO Jack Welch in 1996. This “Six Sigma” initiative, which was benchmarked
from Motorola, includes many TQ principles such as a better focus on cus-
tomers, data-driven decisions, improved design and manufacturing capabilities,
and individual rewards for process improvements. The effort aims to reduce
defect levels to only a few parts per million for strategic products and processes.
Accomplishing such a daunting task requires effective implementation of many
of the process management tools described in this chapter, as well as the prin-
ciples of TQ and management infrastructure discussed in chapter 1. As one
manager stated, “Because it’s data-driven, Six Sigma helps you to make better
decisions faster and ensures better results than the trial and error method.”
General Electric embarked on this initiative to achieve Six Sigma quality
levels by the year 2000, down from a starting level of about 35,000 defects
per million. All of GE’s businesses are engaged in Six Sigma process improve-
ments, ranging from making jet engine blades to executing credit transac-
tions and minimizing “dead air” between segments in broadcasting. The
Black Belts and their associates are working in teams to reduce variation and
defects by using a five-phase approach:
1. Define: Identify customers, their priorities, business objectives, and CTQs —
“critical-to-quality” characteristics that are most important to the customer.
2. Measure: Determine the frequency of defects, define performance stan-
dards, validate the measurement system, and establish product capabil-
ity. Measurements include output, process, and input (supplier) measures
to evaluate current performance.
3. Analyze: Understand when, where, and why defects occur by defining per-
formance objectives and sources of variation. This step includes process
continued on next page
120 Part I: Introduction to Total Quality

mapping, identifying potential root causes, discovering cause-effect rela-

tionships, and establishing operating tolerances.
4. Improve: Generate ideas and narrow the list of potential solutions, select
a solution, validate it, and develop an implementation strategy.
5. Control: Maintain improvements by validating the measurement system,
determining process capability, and implementing process control sys-
tems to monitor performance.
To address customer issues, a program known as Work-Out, in which an
individual or team devotes all their time and energy to solving one problem
and designing solutions with customers’ assistance, helps to minimize
bureaucracy in supporting the Six Sigma initiative.
From 1996 to 1998, GE has increased the number of Six Sigma projects
from 200 to 6,000. This effort has required massive training and new incen-
tives. A typical Black Belt training curriculum includes four weeks of training on
statistical thinking (tools such as quality function deployment), measurement
systems analysis, correlation, design of experiments, statistical process con-
trol, and mistake-proofing. Forty percent of managerial bonuses are tied to
progress in Six Sigma results. Most importantly, the effort is driven by top man-
agement leadership. CEO Welch has made it clear that if employees are not
enthusiastic about Six Sigma, then GE is not the right company for them. From
all these efforts, GE hopes to save $7 to $10 billion over a decade. In its annual
report for 2000, GE noted that “Six Sigma has turned the Company’s focus
from inside to outside, changed the way we think and train our future leaders
and moved us toward becoming a truly customer-focused organization.”

REVIEW AND DISCUSSION QUESTIONS

1. Why is good planning important to quality improvement? Why should

managers invest in the time to learn many of the tools presented in this
chapter?
2. Explain the benefits of the quality function deployment approach. How
does it help organizations to design better products and services?
3. Using whatever “market research” techniques you feel are appropriate,
define a set of customer attributes for (a) purchasing books at your col-
lege bookstore, (b) a hotel room used for business, and (c) a college reg-
istration process. How might QFD be used to improve these processes?
Define a set of “hows” and try to construct the relationship matrix for the
House of Quality for each of these examples.
4. Most organizations have well-defined mission statements that include a
set of goals for the firm and actions that the firm can take. How might
QFD be used to ensure that the actions are consistent with the goals? Find
some company’s mission statement to illustrate this.
 Chapter 3: Total Quality Tools and Statistical Thinking 121

5. What is concurrent engineering? What quality-related advantages does

it have?
6. What type of organizational culture would be required to make concur-
rent engineering successful?
7. Explain the purpose of the seven management and planning tools.
8. How might you use the seven management and planning tools in your
daily activities (schoolwork, fraternity or honor society operations, and
so on)? Provide specific examples.
9. Explain the purpose and uses of each of the “seven QC tools.”
10. Select a process that you do routinely (for example, washing your car,
preparing for an exam, preparing a meal) and draw a flowchart of it.
Explain how the flowchart helps you to understand and improve the
process.
11. A flowchart for a fast-food drive-through window is shown in Figure
3.18. Discuss the important quality characteristics inherent in this process
and suggest possible improvements.

FIGURE 3.18 FLOWCHART OF A FAST-FOOD DRIVE-THROUGH PROCESS

Customers pull up to sign

and decide what they
want to order

Bell rings inside

Employee says,
“May I help you?”

Employee repeats order Customer waits and

Customer gives order
and gives total drives through

Employees Customer and food Employee

process order reach window repeats total

Customer gives Employee gives Customer

money food and change pulls away
122 Part I: Introduction to Total Quality

12. Design a check sheet to help a high school student who is getting poor
grades on a math quiz determine the source of his or her difficulty.
13. Develop cause-and-effect diagrams for
a. a poor exam grade;
b. no job offers;
c. too many speeding tickets;
d. being late for work or school.
14. What is the Deming Cycle? How is it used to improve quality?
15. Many books in business describe some sort of problem-solving process.
Find two or three descriptions of systematic problem-solving processes.
How are they similar to or different from the Deming Cycle?
16. Choose some process in which you are involved. Devise a plan to use the
Deming Cycle to improve it.
17. How might a professor use the Deming Cycle to improve his or her
teaching performance?
18. Describe the purpose and role of benchmarking in business organiza-
tions. How much effort do you believe companies should spend in
benchmarking efforts?
19. Discuss how a college or university might apply benchmarking to
improving its operations. You might solicit views from academic admin-
istrators and from business people. (You might find some differences of
opinion!)
20. Explain how creativity is embodied in the various tools and approaches
described in this chapter.
21. In a recent book, Weird Ideas That Work (New York: The Free Press, 2002),
Robert I. Sutton suggests 111/2 practices for promoting, managing, and
sustaining innovation. These include
• hiring “slow learners” (of the organizational code);
• hiring people who make you uncomfortable, even those you dislike;
• encouraging people to ignore and defy superiors and peers;
• finding some happy people, and get them to fight;
• thinking of some ridiculous or impractical things to do, then plan-
ning to do them; and
• forgetting the past, especially your company’s successes.
Why do you think these practices work? If you have few ideas, you should
probably read the book!
22. What is statistical thinking? How might the traditional teaching of statis-
tics be improved by incorporating this notion? Draw your response from
your own experiences in learning statistics.
23. Identify several sources of errors in your personal life. Develop some
mistake-proofing approaches that might prevent them.
24. How might poka-yoke be applied to the U.S. election system based on
the information described in the example in this chapter? You might
wish to do some additional research on the subject or find out how your
local election process is performed.
25. At a university library, many activities take place. Some of these are
 Chapter 3: Total Quality Tools and Statistical Thinking 123

• processing request forms from patrons for copying of journal articles;

• reshelving books that readers have left on desks (books are picked
up, placed on carts, sorted, and reshelved); and
• locating missing books.
For these activities, identify potential problems that might arise and
what the library might do to “mistake-proof” its activities to provide
better service.

CASES

Scott’s Fitness Center

Figure 3.19 shows a partially completed House of Quality for a proposed fit-
ness center.

1. Examine the relationships in the roof of the House of Quality. Explain

why they make sense (or if you think they do not, explain why). How
would this assessment help in the design activity?
2. Complete the matrix in the body of the House of Quality. That is, examine
each pair of customer and technical requirements and determine whether
there is a very strong relationship, strong relationship, weak relationship,
or no relationship, and fill in the appropriate symbols in the matrix.
3. Suppose that the most important customer requirements identified through
surveys and focus groups are “Has programs I want,” “Family activities
available,” “Equipment available when desired,” “Easy to sign up for
programs,” and “Value for your money.” “Staff available when needed”
was ranked low, while the remaining were ranked moderate in impor-
tance. Based on this information, identify the most important technical
requirements that should be addressed in subsequent design activities.

Welz Business Machines19

Welz Business Machines sells and services a variety of copiers, computers,
and other office equipment. The company receives many calls daily for serv-
ice, sales, accounting, and other departments. All calls are handled centrally
by customer service representatives and routed to other individuals as appro-
priate. A number of customers had complained about long waits when calling
for service. A market research study found that customers became irritated if
the call was not answered within five rings. Scott Welz, the company presi-
dent, authorized the customer service department manager, Tim, to study this
problem and find a method to shorten the call-waiting time. Tim met with the
service representatives who answered the calls to attempt to determine the
reasons for long waiting times. The following conversation ensued:
Tim: This is a serious problem. How a customer phone inquiry is answered
is the first impression the customer receives from us. As you know, this
124 Part I: Introduction to Total Quality

FIGURE 3.19 HOUSE OF QUALITY FOR SCOTT’S FITNESS CENTER

Program offerings

Aml./types equip.
Very strong relationship

Maint. schedule

Internet access
Access control
Program times

Staff schedule
Facility hours

Fee structure
Fitness staff
Strong relationship

Facility size
Instructions
Maint. staff

Training

Lighting
Weak relationship

Has programs I want

Programs

Activities
and

Programs are convenient

Family activities available
Facilities

Clean locker rooms

Well-maintained equipment
Safe place to be
Atmosphere

Equipment available
when desired
Wide variety of equipment
Adequate parking
Friendly and courteous
Knowledgeable and
Staff

professional
Available when needed
Respond quickly to problems
Easy to sign up for programs
Other

Value for your money

company was founded on efficient and friendly service to all our customers.
It’s obvious why customers have to wait: You’re on the phone with another
customer. Can you think of any reasons that might keep you on the phone
for an unnecessarily long time?
Robin: I’ve noticed quite often that the person to whom I need to route the
call is not present. It takes time to transfer the call and to see whether it is
answered. If the person is not there, I end up apologizing and transferring
the call to another extension.
Tim: You’re right, Robin. Sales personnel often are out of the office on sales
calls, away on trips to preview new products, or away from their desks for a
variety of reasons. What else might cause this problem?
 Chapter 3: Total Quality Tools and Statistical Thinking 125

Ravi: I get irritated at customers who spend a great deal of time complain-
ing about a problem that I cannot do anything about except refer to someone
else. Of course, I listen and sympathize with them, but this eats up a lot of
time.
Lamarr: Some customers call so often, they think we’re long-lost friends and
strike up a personal conversation.
Tim: That’s not always a bad thing, you realize.
Lamarr: Sure, but it delays my answering other calls.
Nancy: It’s not always the customer’s fault. During lunch, we’re not all
available to answer the phone.
Ravi: Right after we open at 9 a.m., we get a rush of calls. I think that many
of the delays are caused by these peak periods.
Robin: I’ve noticed the same thing between 4 and 5 p.m.
Tim: I’ve had a few comments from department managers who received
calls that didn’t fall in their areas of responsibility and had to be transferred
again.
Mark: But that doesn’t cause delays at our end.
Nancy: That’s right, Mark, but I just realized that sometimes I simply don’t
understand what the customer’s problem really is. I spend a lot of time try-
ing to get him or her to explain it better. Often, I have to route it to someone
because other calls are waiting.
Ravi: Perhaps we need to have more knowledge of our products.
Tim: Well, I think we’ve covered most of the major reasons why many cus-
tomers have to wait. It seems to me that we have four major reasons: the
phones are short-staffed, the receiving party is not present, the customer
dominates the conversation, and you may not understand the customer’s
problem. Next we need to collect some information about these possible
causes. I will set up a data collection sheet that you can use to track some of
these things. Mark, would you help me on this?
Over the next two weeks the staff collected data on the frequency of rea-
sons why some callers had to wait. The results are summarized as follows:

Reason Total number

A. Operators short-staffed 172
B. Receiving party not present 73
C. Customer dominates conversation 19
D. Lack of operator understanding 61
E. Other reasons 10

Discussion Questions
1. From the conversation between Tim and his staff, draw a cause-and-
effect diagram.
126 Part I: Introduction to Total Quality

2. Perform a Pareto analysis of the data collected.

3. What actions might the company take to improve the situation?

The Quarterly Sales Report20

Suppose that Ron Hagler, the vice president of sales for Selit Corp., had just
gotten a report on the past five years of quarterly sales data for the regions
under his authority (see Figure 3.20). Not happy with the results, he got on
the phone to his secretary. “Marsha, tell the regional managers I need to
speak with them this afternoon. Everyone must attend.”
Marsha had been Hagler’s secretary for almost a decade. She knew by
the tone in his voice that he meant business, so she contacted the regional
managers about the impromptu meeting at 2 P.M. At 1:55 P.M., the regional
managers filed into the room. The only time they were called into a meeting
together was when Hagler was unhappy.
Hagler wasted no time. “I just received the quarterly sales report. North-
east sales were fantastic. Steve, you not only improved 17.6 percent in the
fourth quarter, but you also increased sales a whopping 20.6 percent over the
previous year. I don’t know how you do it!” Steve smiled. His philosophy to
end the year with a bang by getting customers to stockpile units paid off again.
Hagler had failed to notice that Steve’s first quarter sales were always sluggish.
Hagler continued: “Terry, Southwest sales were also superb. You showed
an 11.7 percent increase in the fourth quarter and an 11.8 percent increase
over the previous year.” Terry also smiled. She wasn’t sure how she did so
well, but she sure wasn’t going to change anything.
“Jan, Northwest sales were up 17.2 percent in the fourth quarter, but down
8.2 percent from the previous year,” said Hagler. “You need to find out what
you did previously to make your sales go through the roof. Even so, your per-
formance in the fourth quarter was good.” Jan tried to hide his puzzlement.
Although he had received a big order in November, it was the first big order he
had received in a long time. Overall, sales for the Northwest were declining.
Hagler was now ready to deal with the “problem” regions. “Leslie, North
Central sales were down 5.5 percent in the fourth quarter, but up 4.7 percent
from the previous year. I don’t understand how your sales vary so much. Do
you need more incentive?” Leslie looked down. She had been working very
hard the past five years and had acquired numerous new accounts. In fact,
she received a bonus for acquiring the most new business in 1998.
“Kim, Mid-Atlantic sales were down 3.2 percent in the fourth quarter
and down 2.6 percent from the previous year. I’m very disappointed in your
performance. You were once my best sales representative. I had high expec-
tations for you. Now, I can only hope that your first quarter results show
some sign of life.” Kim felt her face get red. She knew she sold more units in
2000 than in 1999. “What does Hagler know anyway,” she thought to herself.
“He’s just an empty suit.”
Hagler turned to Dave, who felt a surge of adrenaline. “Dave, South Cen-
tral sales were the worst of all! Sales were down 19.7 percent in the fourth
 Chapter 3: Total Quality Tools and Statistical Thinking 127

FIGURE 3.20 SALES BY REGION FOR SELIT CORP., 1998–2002

1998 Sales (in thousands)

Region First Quarter Second Quarter Third Quarter Fourth Quarter
Northeast $924 $928 $956 $1,222
Southwest $1,056 $1,048 $1,129 $1,073
Northwest $1,412 $1,280 $1,129 $1,181
North Central $431 $470 $439 $431
Mid-Atlantic $539 $558 $591 $556
South Central $397 $391 $414 $407

1999 Sales (in thousands)

Region First Quarter Second Quarter Third Quarter Fourth Quarter
Northeast $748 $962 $983 $1,024
Southwest $1,157 $1,146 $1,064 $1,213
Northwest $1,149 $1,248 $1,103 $1,021
North Central $471 $496 $506 $573
Mid-Atlantic $540 $590 $606 $643
South Central $415 $442 $384 $448

2000 Sales (in thousands)

Region First Quarter Second Quarter Third Quarter Fourth Quarter
Northeast $991 $978 $1,040 $1,295
Southwest $1,088 $4,322 $1,256 $1,132
Northwest $1,085 $1,125 $910 $999
North Central $403 $440 $371 $405
Mid-Atlantic $657 $602 $596 $640
South Central $441 $366 $470 $426

2001 Sales (in thousands)

Region First Quarter Second Quarter Third Quarter Fourth Quarter
Northeast $756 $1,008 $1,038 $952
Southwest $4,352 $1,353 $1,466 $1,196
Northwest $883 $851 $997 $878
North Central $466 $536 $551 $670
Mid-Atlantic $691 $723 $701 $802
South Central $445 $455 $363 $462

2002 Sales (in thousands)

Region First Quarter Second Quarter Third Quarter Fourth Quarter
Northeast $1,041 $1,020 $976 $1,148
Southwest $1,330 $1,003 $1,197 $1,337
Northwest $939 $834 $688 $806
North Central $588 $699 $743 $702
Mid-Atlantic $749 $762 $807 $781
South Central $420 $454 $447 $359
128 Part I: Introduction to Total Quality

quarter and down 22.3 percent from the previous year. How can you explain
this? Do you value your job? I want to see a dramatic improvement in this
quarter’s results or else!” Dave felt numb. This was a tough region, with a
lot of competition. Sure, accounts were lost over the years, but those lost
were always replaced with new ones. How could he be doing so badly?

Discussion Question
1. How can Hagler improve his approach by applying principles of statisti-
cal thinking? Use any analyses of the data that you feel are appropriate
to fully explain your thinking and help him.

The HMO Pharmacy Crisis21

John Dover just completed an intensive course, “Statistical Thinking for Con-
tinuous Improvement,” that was offered to all employees of a large health
maintenance organization (HMO). There was no time to celebrate, however,
because he was already under a lot of pressure. Dover worked as a phar-
macy assistant in the HMO’s pharmacy, and his manager, Juan de Pacotilla,
was about to be fired. Pacotilla’s dismissal appeared imminent because of
numerous complaints—and even a few lawsuits—over inaccurate prescrip-
tions. Pacotilla now was asking Dover for his assistance in trying to resolve
the problem.
“John, I really need your help,” said Pacotilla. “If I can’t show some
major improvement or at least a solid plan by next month, I’m history.”
“I’ll be glad to help,” replied Dover, “but what can I do? I’m just a phar-
macy assistant.”
“Your job title isn’t important. I think you’re just the person who can get
this done,” said Pacotilla. “I realize that I’ve been too far removed from day-
to-day operations in the pharmacy, but you work there every day. You’re in
a much better position to find out how to fix the problem. Just tell me what
to do, and I’ll do it.”
“But what about the statistical consultant you hired to analyze the data
on inaccurate prescriptions?” asked Dover.
“To be honest, I’m really disappointed with that guy. He has spent two
weeks trying to come up with a new modeling approach to predict weekly
inaccurate prescriptions. I tried to explain to him that I don’t want to predict
the mistakes, I want to eliminate them. I don’t think I got through, however,
because he said we need a month of additional data to verify the model
before he can apply a new method he just read about in a journal to identify
‘change points in the time series,’ whatever that means. But get this, he will
only identify the change points and send me a list. He says it’s my job to fig-
ure out what they mean and how to respond. I don’t know much about sta-
tistics. The only thing I remember from my course in college is that it was the
worst course I ever took. I’m becoming convinced that statistics really doesn’t
have much to offer in solving real problems. Since you’ve just gone through
the statistical thinking course, maybe you can see something I can’t. I realize
 Chapter 3: Total Quality Tools and Statistical Thinking 129

it’s a long shot, but I was hoping you could use this as the project you need
to officially complete the course.”
“I used to feel the same way about statistics, too,” replied Dover. “But the
statistical thinking course was interesting because it didn’t focus on crunch-
ing numbers. I have some ideas about how we can approach making improve-
ments in prescription accuracy. I think it would be a great project. But we
might not be able to solve this problem ourselves. As you know, there is a lot
of finger pointing going on. Pharmacists blame the doctors’ sloppy hand-
writing and incomplete instructions for the problem. Doctors blame the
pharmacy assistants, who do most of the computer entry of the prescrip-
tions, claiming that they are incompetent. Pharmacy assistants blame the
pharmacists for assuming too much about their knowledge of medical ter-
minology, brand names, known drug interactions, and so on.”
“It sounds like there’s no hope,” said Pacotilla.
“I wouldn’t say that at all,” replied Dover. “It’s just that there might be
no quick fix we can do by ourselves in the pharmacy. Let me explain what
I’m thinking about doing and how I would propose attacking the problem
using what I just learned in the statistical thinking course.”

Discussion Question
1. How do you think John should approach this problem, using what he has
just learned? Assume that he really did pick up a solid understanding of
the concepts and tools of statistical thinking in the course.

ENDNOTES

1. “How Ford Hit the Bull’s Eye with Taurus,” Business Week, June 30, 1986, pp. 69–70.
2. Mike Boyer, “Milacron Seeks Killer Instinct,” Cincinnati Enquirer, August 31, 1990, B6.
3. James L. Brossert, Quality Function Deployment: A Practitioner’s Approach, Milwaukee: ASQC
Quality Press/Marcel Dekker, Inc., 1991, Part 2.
4. Michael Brassard, The Memory Jogger Plus +, Meuthen, Mass.: GOAL/QPC, 1989; Brossert,
Quality Function Deployment; Shigeru, Mizuno, Management for Quality Improvement: The 7 New
QC Tools, Cambridge, Mass.: Productivity Press, 1988.
5. Timothy Clark and Andrew Clark, “Continuous Improvement on the Free Throw Line,” Qual-
ity Progress, October 1997, pp. 78–80.
6. Gerald Langley, Kevin Nolan, and Thomas Nolan, “The Foundation of Improvement,” Sixth
Annual International Deming User’s Group Conference, Cincinnati, Ohio, August, 1992.
7. Adapted from Kevin Dooley, “Use PDSA for Crying Out Loud,” Quality Progress, October
1997, pp. 60–63.
8. Excerpts reprinted from Richard B. Chase and Douglas M. Stewart, “Make Your Service Fail-
Safe,” Sloan Management Review, Vol. 35, No. 3, Spring 1994. Copyright 1994 by the Sloan Man-
agement Review Association. All rights reserved.
9. Howard R. Schussler, “Can Quality Concepts and Tools Fix the U.S. Election Process?” Qual-
ity Progress, April 2001, pp. 46–50.
10. See, for example, Jacob Eskildsen, Jens Dahlgaard, and Anders Norgaard, “The Impact of
Creativity and Learning on Business Excellence,” Total Quality Management, Vol. 10, Issue 4/5,
July 1999, S523–S530.
130 Part I: Introduction to Total Quality

11. Masaaki Imai, Kaizen: The Key to Japan’s Competitive Success, New York: McGraw-Hill, 1986,
p. 15.
12. Mark R. Edwards and J. Ruth Sproull, “Creativity: Productivity Gold Mine?” Journal of Cre-
ative Behavior, Vol.18, No. 3, 1984, pp. 175–184; and Michael K. Badawy, “How to Prevent Cre-
ativity Mismanagement,” Research Management, Vol. 29, No. 4, 1986, p. 28.
13. Kathleen D. Ryan and Daniel K. Oestreich, Driving Fear Out of the Workplace, San Francisco:
Jossey-Bass, Inc., 1991, pp. 63, 64.
14. Pamela Tierney, Steven Farmer, and George Graen, “An Examination of Leadership and
Employee Creativity: The Relevance of Traits and Relationships,” Personnel Psychology, Vol. 52,
No. 3, Autumn 1999, pp. 591–620.
15. Adapted from Galen Britz, Don Emerling, Lynne Hare, Roger Hoerl, and Janice Shade,
“How to Teach Others to Apply Statistical Thinking,” Quality Progress, June 1997, pp. 67–79.
16. Ronald D. Snee, “Getting Better Business Results: Using Statistical Thinking and Methods to
Shape the Bottom Line,” Quality Progress, June 1998, pp.102–106.
17. Adapted from Brian L. Joiner, Fourth Generation Management, New York: McGraw-Hill, 1994,
p. 129.
18. William M. Carley, “To Keep GE’s Profits Rising, Welch Pushes Quality-Control Plan,” The
Wall Street Journal, January 13, 1997, A1, A8; “Changing the Way Aircraft Engines Work,” GE
Aircraft Engines News, January 1998; Roger W. Hoerl, “Six Sigma and the Future of the Quality
Profession,” Quality Progress, June 1998, pp. 35–42; and presentation notes provided by Stefanie
A. Darlington, Leader—Total Product Quality, GE Aircraft Engines, Cincinnati, Ohio.
19. This problem was developed from a classic example published in “The Quest for Higher
Quality: The Deming Prize and Quality Control” by RICOH of America, Inc.
20. Adapted from Britz, et al. See Note 7.
21. Adapted from Britz, et al. See Note 7.