The Quality Approach - Usaf
The Quality Approach - Usaf
1 AUGUST 1996
Command Policy
Supersedes AFH 90-502, Sept. 1994 and AFH 90-503, Sept. 1994
OPR: AFQI/OS (Lt Col Phillip Dove)
Certified by: AFQI/CC (Col Henry Fiumara)
Pages: 186/Distribution: F
Forward...
Over 1,000,000 copies of the previous editions of The
Quality Approach and Process Improvement Guide are in
print. With that many books in print, why are we
publishing a third edition? Simply put, in a dynamic world
its not enough to merely keep pace with our adversaries,
we must constantly seek better ways of accomplishing our
mission. To be effective, we have to be out front, seeking a
better, faster and cheaper way of doing business. The third
edition of this book is designed to help you find better ways
to accomplish your mission.
The biggest change youll notice in the two books is that
they are now one book. Thats right, based on input from
numerous customers, weve combined AFH 90-502, The
Quality Approach and AFH 90-503, The Process
Improvement Guide. The new handbook is titled AFH 90502, The Quality Approach. You also told us you wanted
to see more how to information on benchmarking, its in
here. Some of you wanted to see information on the
Action Workout process, its in the new edition also.
Additionally, weve updated the information on the Air
Force Core Values and modified the QAF Model by placing
the core values at the center of the model, where they
belong.
The third edition team charter also included a direction to
make this book useful to everyone from the lowest ranking
airman on the flightline to the wing commander. Hopefully,
we succeeded. As with the previous editions, team
members used the continuous improvement process from
identifying the improvement opportunity to planning for
future editions to match the product to your needs and
expectations.
Contents
Chapter 1.
Quality Air Force
Chapter 2.
Leadership
11
Quality Leadership
11
11
12
Principles
12
Operating Style
13
QAF Roles
14
Chapter 3.
Quality Focus
19
19
20
Chapter 4.
Quality in Daily Operations
27
Definition
27
Metrics
28
Assessment Process
30
Chapter 5.
Improvement Process
31
Shewhart Cycle
31
32
34
35
Action Workout
41
Benchmarking
46
Process Reengineering
52
Chapter 6.
Quality Air Force Assessments
Quality Air Force Criteria
57
57
65
131
159
167
171
Index
175
Feedback Form
1
Quality Air Force
What is Quality Air Force?
Quality Air Force (QAF) is a leadership commitment and operating
style that inspires trust, teamwork and continuous improvement
everywhere in the Air Force . The QAF principles and techniques
provide the tools needed to make those improvements. These
concepts have a proven track record in the Air Force and in worldclass organizations around the globe .
Quality is not a static description, but a dynamic process
for an attitude of continuous improvement within the
constraints of available resources. Quality does not stand
alone, but is measured by its contribution to the Air Force
mission.
Dr. Sheila E. Widnall
Secretary of the Air Force
Ex
cel
len
ce
Integrity
hip
ers
ad
Le
Quality
focus
ce
rvi
Se
Le
ade
rsh
ip
QAF
Improvement
process
Quality in
daily operations
Leadership
10
2
Leadership
Quality Leadership
Quality is not a new way of conducting business or a new
management tool. Military leaders of many eras have recognized
the value of quality leadership.
Never tell people how to do things. Tell them what to do
and they will surprise you with their ingenuity.
Gen. George S. Patton, Jr.
United States Army
A mission describes the task you face, the forces you bring to the
fight, what you hope to achieve and the medium in which you
operate. It is ongoing, enduring and symbolizes goals that will
stand for long periods of time.
Chapter 2/Leadership
11
12
Chapter 2/Leadership
13
QAF Roles
Senior Leader Roles
Senior leaders set the pace of an organization by identifying the
values through strategic planning. In addition to planning for the
future, senior leaders talk with customers, focus on work to
complement the Air Force vision and measure significant objectives
necessary to accomplish the goals. They know the benefit of
creating a short-term plan (12-24 months) as well as a long-term
plan (over three years). These plans help keep the organization on
track. Turnover in leadership shouldnt change these plans,
however its realistic to expect the inevitable mid-course
corrections.
Senior leaders can better plan for the future by listening to feedback
from the workforce and addressing individual and group needs.
Feedback through strong, positive communication helps leaders and
workers better understand the organizations processes.
Therefore, senior leaders must create a cooperative environment by
implementing policies and procedures that enable and energize the
work force.
14
Chapter 2/Leadership
15
Chapter 2/Leadership
17
18
3
Quality Focus
Phases of Quality Focus
A quality organization must start with a sound systematic plan.
Planning is a leadership-driven cooperative effort between the
leadership team and the organization. Establishing a strong quality
focus requires substantial time and effort from the leadership team
as they formulate, deploy, implement and review their vision,
mission and plans.
Formulation Phase ... consists of defining the mission, vision,
values and developing goals and objectives from a macro or
system perspective. Envision an organization that fulfills its
mission, exceeds customer needs and expectations that may
constantly be changing. The ability to meet those demands
depends on actively listening and communicating with customers
regularly to help tailor performance.
Start setting customer-driven goals and objectives by answering the
following questions. Who are your key customers? What major
product or service characteristics (timeliness, accuracy, reliability,
etc.) are they most concerned about? Can you and your customers
agree on indicators to how well youre doing in those areas? Once
you reach agreement, monitor those indicators and use the
information to identify improvement. Finally, develop supporting
action, human resource and budget plans to move you toward the
goals and vision.
Deployment Phase ... builds a support structure to integrate the
mission, vision, values and goals throughout the organization. Its
important that communication be two-way to ensure understanding
and buy-in. A technique known as catchball is very useful to
achieve wide scale deployment. Catchball is a give-and-take
Chapter 3/Quality Focus
19
4. Envision the
Future
2. Values
Assessment
1. Planning
to Plan
11. Annual
Review
6. Gap Analysis
Deploy Down
7. Develop
Strategic Goals
and Objectives
8. Develop
Functional Plans
10. Periodic
Review
5. Assess Current
Capabilities
9. Implement
Plans
21
22
23
24
25
26
4
Quality in Daily Operations
Definition
Quality in daily operations (QDO) implies repeatedly doing the best
at a job according to internal and external evaluation criteria. You
should be able to determine whether you are doing a good job
regardless of the frequency or duration of your tasks. In order to
determine this, you must know how your job is part of a process
that supports the larger Air Force mission. To do a good job in
daily operations look at the following components: customer and
suppliers interaction, metrics and an assessment process.
Customer/Supplier Model
This model is a simple illustration that demonstrates the flow of
inputs and outputs through a process.
Customer/Supplier Model
Needs
Suppliers
Needs
Process/Job
Input
Customers
Output
Understanding/Agreement/Feedback
Using the model, ask the following questions about the products
and services you provide:
27
28
Examples of Metrics
Most successful organizations are able to continually produce
quality products and services very quickly. They have processes in
place that allow workers to meet high standards in their products
and services. What metrics might organizations use to evaluate
themselves to ensure success? How is customer satisfaction
translated into metrics that can guide and inspire their individual
workers? The following are a few examples of metrics that might
be worthwhile in some organizations:
1. An outcome/customer satisfaction metric ... such as the
percentage of customers who respond excellent service
provided on random surveys over a pre-determined period of
time. The goal would be to keep the percentage high. This
survey could also help identify specific areas for improvement.
29
Assessment Process
Metrics measure processes over time. An assessment method is
needed to measure the success of the QAF system. To help track
your progress, the QAF roadmap includes assessment and awards
programs. You can use the QAF Criteria to measure your
progress. There are many formal assessment processes that the AF
uses to evaluate organizations. These assessments include IG
inspections, operational readiness inspections, unit self assessments,
award and recognition programs such as the Secretary of the Air
Force Unit Quality Award and the Chief of Staff Team Excellence
Award (see Chapters 1 and 6 for more detailed explanations).
These assessment processes incorporate metrics and may be of
assistance in developing metrics in your unit.
30
5
Improvement Process
Once you have defined your metrics and begun to achieve
maximum performance in your daily operations you will find
opportunities to improve your processes. Four approaches to
exploiting these opportunities are the continuous improvement
process (CIP), action workout, benchmarking, and reengineering.
The concept of process improvement is based on the Shewhart
cycle.
Shewhart Cycle
The Shewhart cycle, also referred to as the Ishikawa Circle or
Deming Wheel, is a systematic approach to achieving continuous
improvement. The approach involves repetition and is often
represented graphically as a circle. The circle has four quadrants:
plan, do, study and act (PDSA).
Act
Plan
Study Do
31
Plan Study the process flow and any existing data. Formulate
possible improvements and experiments or decide on methods to
gather data. When working with a new process, be willing to
concede additional time to this quadrant to identify and think about
the processes.
Do Implement the improvement effort youve planned using a
small-scale test. Working from too large a base of change can
make it difficult to assess the effects of the change. Train the
people responsible for implementation. They need to know what
the goal of this change is, and how they impact the implementation
process.
Study Measure the results of the improvement effort. Analyze
the data collected. Study the results to see if the process was
improved. This quadrant gives you the chance to see if you
measured the right things and can also give you helpful clues about
variables.
Act If the result was a clear improvement, the team should agree
to make the change permanent, standardizing and documenting all
actions. If the results werent successful, identify the cause then go
back to the plan quadrant and start over again with the newly
acquired knowledge.
Teams in the Work Place
One of the most effective ways to use the PDSA cycle is in a team
setting. Teams are a critical part of the QAF culture. They
improve existing processes, solve problems or develop new plans or
procedures. Team members benefit from a wide range of ideas and
an increased knowledge of the system. Four of the more common
types of teams are tiger teams, process action teams, developmental
teams and natural working groups. The teams can include workers
from different organizations, but everyone must share a given
process.
32
33
34
Im p r o v e
process
capabilities
7. Plan for future
6. Standardize
solution
2. Evaluate process
Act
Plan
Study
Do
3. A n a l y z e
5. Study results
4. Take action
35
36
37
Checkpoints:
Perform cause-and-effect analysis of the problem
Analyze potential root causes
Select the root cause that has the greatest probable
impact
Verify the root causes (use data)
Step 4: Take Action
Plan and implement actions that correct root causes. The team can
propose improvements by using an action plan matrix to identify
specific methods to use in attacking root causes. The action plan
should address what, who, how and when plus identify the resources
needed. The methods should be feasible, effective and cost
beneficial.
Checkpoints:
Develop and evaluate possible actions
Ensure the actions are cost-beneficial
Develop an action plan
Test actions (if possible) before fully implementing them
Get the cooperation and approval needed
Implement the action plan
Step 5: Study Results
Confirm that the actions taken achieved their target results. Its
important to understand why the target was or wasnt met. If the actions
were not effective, additional actions may have to be implemented.
38
Checkpoints:
Confirm the indicator was the same one used to identify
the process
Determine if the action results met or exceeded the
target
Discuss why the target was or wasnt met
If the target wasnt met, confirm additional actions
Step 6: Standardize Solution
Maintain the improved level of performance. Integrate the teams
improvement efforts into the organization; make improvements a
regular part of daily operations. A control system can help by
outlining the process, tasks involved, improvement efforts and
targets.
Checkpoints:
Publish revised methods and procedures
Conduct training on new processes
Create periodic process review points
Consider areas for replication
Step 7: Plan for Future
Plan what to do with remaining problems and evaluate the teams
effectiveness. The improvement process allows the team the
opportunity to review the work accomplished, address remaining
issues and evaluate effectiveness. Additionally, the team can review
lessons learned in problem-solving, interpersonal communications
and group dynamics.
39
Checkpoints:
Analyze and evaluate any remaining issues
Plan any future actions necessary
Evaluate the teams problem-solving skills and their
effectiveness
QAF Tools
There are a number of tools that can be effectively used in each of
the 7-steps of the CIP. Although some of the tools work better in
one area than another, it is up to each team to choose based on its
comfort level with each tool. A skilled facilitator can help
determine the tools a team should use.
The Continuous Improvement Storyboard
Storyboards visually and logically present the appropriate steps in a
plan or task. As an alternative to a lengthy narrative,
storyboarding captures and tracks basic ideas, plans, activities
and process actions. This technique is applicable to a wide range of
process improvement or planning activities and requires minimal
resources, materials and equipment.
Storyboards are usually visually appealing in their layout and
design; they stimulate the readers imagination. (Creating a
storyboard can be a team building exercise in itself!) Another useful
consequence of storyboards is the immediate feedback received.
Ideas, suggestions and constructive criticism from coworkers not
directly involved on the team can be posted using sticky-back notes.
Additionally, they aid in documenting the teams work and are an
excellent briefing tool.
40
Team info
Good
Schedule
12
10
8
6
4
2
2. Evaluate process
C
Event
Tally
A
B
C
D
E
F
3
1 2 3 4 5 6 7 8 9 10111213
3. Analyze
Total
5
2
8
7
15
3
4. Take action
6. Standardize
solution
5. Study results
Good
After
Schedule
100%
Act
75%
Plan
Study Do
50%
25%
14
12
10
UCL
0%
AVG
4
2
LCL
1 2 3 4
Lessons
Learned:
+/-
6 7 8 9 10 11 12 13 14 15
Action Workout
An Action Workout (AWO) is a rapid, concentrated, high-energy,
team effort to make dramatic productivity improvements in any
organization.
The AWO is driven by an urgent need to improve a process. Units
can identify an AWO candidate through their Unit Self Assessment
process or functional managers for systemic problems.
41
Waste
Work
Waste
Work
Work
Work
Work
Work
Working Harder
Waste
Waste
Work
Waste
Identify Waste
Eliminate Waste
Replace with Value-added Work
Work
Work
Waste
Working Smarter
Work Content
The diagram above depicts two distinct options to take in
approaching daily work. Most work usually consists of productive
activity and significant amounts of waste. However, when faced
with additional work, we often take the path depicted by the top
42
arrow and cram the extra work into what is an already very busy
and wasteful day. Consequently, this approach leads to longer
days, additional shifts, requests for more personnel, etc., because
systemic waste is never addressed. The path shown by the lower
arrow is the ideal path to take. Process owners and stakeholders
must determine what is necessary to eliminate waste in order to
work smarter, not harder. This approach creates space for
increased work capacity, multi-skilled workers and recovery time.
There are three classifications of work content: (1) value-added, (2)
non-value-added, and (3) queue or wait time.
Value-added activities are those tasks the customer is willing
to pay for that affects the form, fit, or function of the product.
Compliance tasks typically fall into this category. Usually less
than 10% of any process is truly value-added. Teams should not
look for short-cuts but for more efficient ways (reduce nonvalue-added steps) to get the job done while maintaining safety and
conformance standards. Units should first attempt to creatively reengineer a process and remove non-value-added rather than buy
newer and faster equipment.
Reduce Waste
Valud
Added
Typical Processes
Non-value added
5%
95%
World-class Process
NVA
50%
VA
50%
43
44
Production Waste
Over production producing more than needed or faster than
needed
Correction/rework repair of product or service not meeting
customer expectations
Motion movement that does not directly add value to the
product or service
Processing work or effort that adds no value to the
user/customer
Inventory materials exceeding what is required to work
productively via standard operations
Transportation Waiting people waiting for work or work
waiting in batches for people
Administrative Waste
Unnecessary approval
Batch processing
Multiple/manual systems
Multiple hand-offs
Waiting times
Not knowing who the customer is or their requirements
Paper vice electronic means of transferring information
The administrative process itself unnecessary movement of
product or information
AWO is systematic and results oriented. It begins with identifying
the targets of opportunity the processes that contain non-valueadded work and narrowing the focus down to a manageable level.
This allows you to get results real-time. Once customer and
mission requirements are confirmed document the actual work.
Chapter 5/Improvement Process
45
Next, tear into the work content, identifying value-added and nonvalue-added work along with queue and wait time. Finally,
systematically eliminate the non-value-added work and queue time
through a reiterative effort of brainstorming and trying of new
ideas.
Benchmarking
Another process available for continuous improvement is
benchmarking. Benchmarking is an organized way for Air Force
units to become more innovative an essential ingredient to infuse
excellence in all we do.
The defense of our country requires the aggressive pursuit of bestin-class service in all operational and support areas. Process
improvement must be accelerated wherever necessary to assure
operational efficiency and effectiveness at the benchmark level.
Executive Order 12862, Setting Customer Standards, directs
benchmarking customer service performance against the best in
business. Best in Business shall mean the highest quality of
service delivered to customers by private organizations providing a
comparable analogous service.
Benchmarking directly impacts the QAF Criteria, the framework for
improving overall unit performance. Benchmarking is addressed
throughout the criteria as an essential component to develop
appropriate goals and attaining superior processes and, most
importantly, results.
Definition
David T. Kearns, former CEO of Xerox, provided one of the
earliest definitions of benchmarking ... the continuous process of
measuring products, services, and practices against the toughest
competitors or those companies recognized as industry leaders.
46
47
Methodology
The benchmarking process involves four phases: planning the
benchmarking effort, collecting information, analyzing the data and
adapting best practices. An adaptation of the Shewhart Cycle.
48
Adapt
Plan
Analyze Collect
49
50
51
Br
Im eakt
pr hro
ov
em ugh
en
t
Performance
!
lass
C
t-in
Bes
CIP
Time
52
Radical
Change?
Plan
Analyze Collect
Reengineer!
Reengineering Methodology
Reengineering may involve many steps and sub-steps depending on
the level of complexity of the reengineering effort. The phases of
this methodology, like benchmarking, can also be linked to the
Shewhart Cycle of Plan-Do-Study-Act through the phases of
planning, redesigning, analyzing and implementation.
53
Implement
Plan
Analyze
Redesign
54
55
56
6
Quality Air Force
Assessments
In the past, military inspections focused on individual components
of an organization looking for conformance to a predetermined
standard. However, Quality Air Force assessments focus on the
entire systems level of performance . Both a Quality Air Force
Assessment (QAFA) and Unit Self Assessment (USA) allow units
to evaluate their success in applying the principles of Quality Air
Force. The QAFA an external assessment and the USA an
internal assessment identify organizational strengths and
opportunities for improvement. They help focus improvement
efforts on areas having the most impact on the units ability to meet
mission requirements. Assessments ultimately provide leaders a
nonprescriptive feedback report describing the organization and
its current performance levels. These assessments are based on the
Quality Air Force Criteria.
Quality Air Force Criteria
The QAF Criteria, based on the Malcolm Baldrige National Quality
Award (MBNQA), are grouped into seven broad categories. The
categories provide the framework to tie together the multitude of
seemingly independent tasks units perform to accomplish the
mission and assess their relative effectiveness. The four elements
of the QAF Criteria framework are as follows:
57
Drivers
Planning
Process
5.0 Process
Management
7.0 Customer Focus
and Satisfaction
1.0 Leadership
Goals
6.0 Performance
Results
2.0 Information and
Analysis
58
This Criteria Systems Model, derived from Figure 1.1 in AFI 90501, depicts the connection and integration of the categories that
comprise the whole system. The driver of the system is leadership.
The organizations senior leaders focus efforts on meeting current
and future mission and customer requirements. Inputs to the
planning process come from data on the current and future mission
and customer requirements, as well as organizational performance
measures. The measures of progress include: data from the
customers perspective; operational performance measures; human
resource and development measures; and financial performance
measures. All of these inputs are tracked in the information and
analysis function. The goals of the system are to continuously
improve the organizations operational performance, mission
accomplishment and customer satisfaction.
The QAF Criteria offer a well-designed framework to help raise
performance standards and expectations and help in planning,
training and assessment. The following are specific ways applying
the criteria can benefit an organization:
Help best execute the mission by becoming more
effective, efficient and focused
Emphasize how different units act in concert to produce
a units end product mission
Facilitate communications based upon a shared
knowledge of quality and mission requirements
Improve overall operational performance within the
organization
Enable a thorough, accurate assessment of an
organizations systems and how they affect customer
and supplier relationships and mission accomplishment
59
60
61
62
63
results (6.0) and external results for customers (7.4 and 7.5).
Depending on the results improvement processes can be used to
ensure the products and services meet mission requirements.
Quality Air Force
The language of Quality is heard at all levels of the Air Force.
There are many different methods for implementation. This
handbook is just one reference to use to sustain the quality journey
towards process improvement and mission enhancement. These
concepts, models and measurement tools have proven to be of
value in the pursuit of improving operations and quality of life.
As we look to the future, a quality culture will continue
to be important to our Air Force. We will employ its
tools and its techniques. We will help our service become
leaner, more flexible, more agile. We will operate in a
decentralized manner to keep up with the pace of events.
People at all levels will have to know their business.
Youll have to be ready to seize the initiative to exploit
opportunities as they arise. And youll need to be bold, to
think outside the box, to seek out fresh and innovative
ways to capture the promise of air and space.
General Ronald R. Fogleman
USAF Chief of Staff
64
A
Process Improvement Guide
Tools for Generating Ideas
Brainstorming, asking Why?, and mental imaging are all tools to
help generate ideas. These tools work well for individuals or
groups. Flexibility is critical. Be ready to change techniques to
suit the groups mood and purpose. Heres a quick look at tools
for generating ideas:
Brainstorming
Generates multiple ideas about a problem
or topic
Works well in groups of all sizes
Five Whys
Discovers the root cause of a problem
Shows how causes of a problem might be
related
Mental Imaging
Helps visualize key relationships and
obstacles
Creates a detailed picture of an ideal
situation
Brainstorming
This tool, which doesnt involve analysis, builds a shopping list of
ideas about a specific problem or topic in a short time. Heres
what to do:
65
66
Five Whys
The Basics: Ask Why?. Thats the key to finding the root
cause (or causes) of a problem. This technique can also help spur
understanding of how different causes might be related. Another
advantage of using this technique is focusing on the process instead
of personalities. Heres what to do:
The Process: Describe the problem in specific terms. Everyone
should have a chance to contribute opinions. Be specific in the
choice of words to help keep everyone focused.
Ask Why? Why did the problem occur? The question might
have to be asked more than once. Just keep asking Why? until
the root causes have been identified. (Youve identified the root
cause when asking Why? doesnt yield any more useful
information.)
An Example: Office workers wanted to discover why they missed
their Initial Operating Capability (IOC) date. So they asked
Why? and heres what they learned:
We missed the IOC.
Why?
Our contract delivery date slipped.
Why?
There were numerous engineering changes.
Why?
The contractor didnt understand our initial requirements.
Why?
We took only one week to prepare it.
67
68
69
Pairwise Ranking
Prioritizes items in a short list
Reaches decisions by consensus
Force Field Analysis
Identifies the significant forces of the process
Helps identify improvement opportunities
Effective Decision-Making
Meet in a suitable place. Schedule the decision-making meeting in
a room quiet and free from interruptions (telephones, co-workers,
etc.). Have plenty of flipcharts, markers and other supplies.
Combine items where possible. Look at the list. One method
might be to place each idea on a post-it-note then put them on a
flipchart or board so everyone can touch and move them around.
Are any items similar? Give group members a minute or two to
review the items. Then ask them for their opinions. Work as a
group to combine similar ideas where it is feasible.
Number each item on the list. This makes it easy for group
members to refer to item, and saves time as well.
Base decisions on data. Try to keep intuitive decisions to a
minimum. Data-driven decisions can speed your process.
Understand the politics of decision-making. When people make
decisions as a group, the decision-making process usually fits one
of seven styles ranging from no decision to consensus.
The following is a guide to consequences when groups make
decisions. The types of decision-making are listed with the
percentage of group involvement and a profile of the decision style.
70
Multivoting
The Basics: Prioritize the items on a list by multivoting. This
simple, fast technique works best for large groups and long lists.
The benefit to prioritize the list without creating a win-lose
situation for group members. This tool also helps you separate the
vital few from the trivial many on a large list. Heres what to
do:
The Process: Empower. First, count the number of items in the
list. Divide that number in half; thats the number of votes each
team member receives. For example, there would be ten votes for
a 20-item list.
Vote. Each member votes for the items they believe have high
priority. Then compile the votes.
71
Select the top items. Identify the top four, five or six items. If
there is still a problem identifying the top few, then drop the items
with the fewest votes and re-vote. Once the top items have been
identified discuss and prioritize these items.
An Example: Members of a system program office (SPO) often
participated in meetings held across the nation. Unfortunately, the
meetings were often unproductive. The division chiefs, hoping to
improve this situation, scheduled a brainstorming session. Heres
the list they generated:
1. No agenda
2. No clear objectives
4. Extraneous topics
5. Unproductive
13. No parking
Since there were 14 items, each team member had seven votes they
could use to reduce the list to a manageable size. Heres how they
voted:
72
1. No agenda
|||||
|||
2. No clear objectives
||
||
|||||
4. Extraneous topics
||
5. Unproductive
13. No parking
73
P e rson
3
29
#2
13
#5
#7
23
#4
32
#1
10
#6
25
#3
P ro b l e m
T o tal
P rio rity
Lowest
priority
H ighest
priority
Pairwise Ranking
The Basics: If there is a small list of items, try pairwise ranking.
This structured method ranks a small list of items in priority order.
It can help make decisions in a way thats consensus-oriented.
Heres what to do:
The Process: Construct a pairwise matrix. Each box in the
matrix represents the intersection (or pairing) of two items. Lets
say the list has five items. A pairwise matrix would look like this
(the top box represents idea 1 paired with idea 2):
1
2
3
4
2
3
4
74
1
2
3
2
3
1 and 2 compared;
2 is better...
5
1 and 3 compared;
1 is better...
5
5
5
5
4 and 5 compared;
5 is better.
Alternative
Count
Rank
1
2
2
3
3
1
4
0
5
4
Alternative 5
appears four times
in the matrix
Rank all items. Rank the alternatives by the total number of times
they appear. Do two ideas appear an equal number of times?
Break the tie; look at the box in which those two ideas are
compared.
1
2
3
Alternative
Count 2
3
1
Rank 3rd 2nd 4th
4
5
0
4
5th 1st
Alternative 5
ranks first overall
75
1. Fort Huachuca
4. Nellis AFB
2. Edwards AFB
5. Eglin AFB
3. Kirtland AFB
6. Hanscom AFB
1
2
3
2
1
2
3
Site
1
Count
2
Rank 3rd
2
1
6th
3
1
5th
4
5
1st
5
4
2nd
6
2
4th
76
Inhibiting forces
H abit
Need for nicotine
Need to have fingers occupied
Need to have something
i n m o u th
G ain w eight w hen I try to quit
After completing the force field analysis, the quality advisor worked
with the sergeant to prioritize the forces. That helped the sergeant
decide which forces to strengthen and which to weaken.
77
Affinity diagram
Shows relationship between factors in a problem
Identifies areas where you most need
improvement
Cause-and-effect diagram
Determines causes of a particular effect
Identifies areas where more information is needed
Thematic content analysis
Summarizes raw data into useful categories
Shows patterns in raw data
Pareto chart
Identifies and separates major and minor
problems
Prioritizes problems
Flowchart
The Basics: Want to see what a process looks like from start to
finish? A flowchart is a graphical representation of all major steps
of a process. To understand the complete process, identify critical
stages within a process, and locate problem areas. Flowcharts
also show relationships between different steps in the process.
Heres what to do:
The Process: Identify the process. First, define start and finish
points for the process being examined.
Describe the current process. From the starting point, chart the
entire process. Work slowly and include every step along the way,
right through to the finish. Use standard flowchart symbols to
improve the clarity of the flowchart, but theyre not essential.
78
79
Meaning
Examples
Start/stop
Decision point
Approve/disapprove
Accept/decline
Yes/no
Pass/fail
Activity
Document
Connector
(to another
page or part
of the diagram)
A
B
After identifying the basic flow of the process, put the flowchart on
a grid. Grids categorize information about geography,
organizational structure, time or other categories of information.
The goal is to see the process in different and useful ways. Heres
an example:
80
Days
from
arrival
Configuration
control
division
Functional
divisions
Contractor
submits report
for review
1
Contracting
division
Log report
in
Assign
control
number
3
Review and
comment
(CC division)
Review and
comment
(technical)
7
Review and
comment
(legal)
18
Is data
approved
?
22
25
Contractor
receives
approval
letter
35
Yes
Sign
approval
letter
No
Develop
clarification
request
Sign
disapproval
letter
Contractor
receives
disapproval
letter
81
Copy machine
stops
Check
paper
feed area
Close
access
panel
Press
print
No
Open
access
panel
Check
toner
area
Close
access
panel
Press
print
No
Open
access
panel
Check
collator
area
Close
access
panel
Press
print
No
Call
for
repairs
Open
access
panel
Yes
Yes
Yes
Does it
w o rk?
Does it
w o rk?
Does it
w o rk?
Copy machine
operable
82
Copy machine
stops
Open access
panel
Check paper
feed area
Check toner
area
Check collator
area
Close access
panel
Press
print
Does it
work?
No
Call for
repairs
Yes
Copy machine
operable
Affinity Diagram
The Basics: Exploring all aspects of an issue is easy to do when
using an affinity diagram. This diagram takes verbal information,
and organizes it into a visual pattern. Start with specific ideas, and
work toward more broad categories. (This is the opposite of a
cause-and-effect diagram, which starts with broad causes and
works toward specifics.) Affinity diagrams can also help identify
key areas needing improvement. First, identify the problem and
generate ideas. Heres what to do:
83
The Process: Cluster the ideas into related groups. Use stickyback notes or cards anything easy to sort and move. Which ideas
are similar? Which ideas seem connected to other ideas? Questions
like these can help you group the ideas.
Create affinity cards. For each group, create a card that has a
short statement describing the entire group of ideas.
Cluster related cards. Put the group of related cards under the
groups affinity card. Group the affinity cards into broader
groups? Keep creating groups until the definition of group grows
too broad to have comfortable meaning.
Create an affinity diagram. Lay out all the ideas and affinity cards
on a single piece of paper; use a blackboard or table. Put the
affinity cards at the top of each group, and draw an outline of the
group. A hierarchical structure will develop that can offer valuable
insight into the problem.
An Example: A publications team hoped to reduce typographical
errors. Heres the list (generated during a brain-storming session)
of factors affecting the error rate:
84
Computers
No feedback
Proofreading skill
Printers
Noise
Short deadlines
Lighting
Typewriters
Chair height
Comfort
Desk height
Time of day
Spelling
Interruptions
Handwriting
Grammar
Slang
Technical jargon
Draft copy
Punctuation
Distribution
Font
Final copy
Editing skill
Computer skill
Typing skill
No measurement
E rg o n o m ic s
N o ise
L ig h tin g
D e s k h e ig h t
C h a ir h e ig h t
C o m f o rt
O rig in a l d o c u m e n t
A u thor skill
H a n d w ritin g
G ra m m a r
P u n c tu a tio n
S p e llin g
R e q u ire m e n ts
D ra f t c o p y
F inal copy
D istrib u tio n
Font
T e c h n ic a l jarg o n , s l a n g
E q u ip m e n t
C o m p u te r s
P rin te rs
T y p e w rite rs
T ra in in g
T y p in g s k ill
E d itin g s k ill
C o m p u te r s k i l l
P ro o freading skill
N o d e f in itio n
o f q u a lity
N o measurement
N o feedback
85
Manpower, personnel,
staffing, etc.
Materials, policies,
regulations, etc.
The
problem,
objective,
goal, etc.
Methods, procedures,
specifications, etc.
Machine, plant,
equipment, etc.
The
problem,
objective
goal, etc.
86
Methods
Materials
Copy paper
Pencil
hardness
Original
settings
Poor storage
Drying
time
Sharpness
Bad paper
handling
Paper quality
Writing
pressure
Humidity
Paper
alignment
Hands
dirty
Bad originals
Poor maintenance
Operating
hours
Poor photocopy
quality
Bad settings
Not level
Roll condition
Speed
Not
cleaning-up
Contamination
Environment
People
Lamp dirty
Machine
87
88
# responses
Pareto Chart
The Basics: Pareto charts are bar charts based on the Pareto
Principle: 20 percent of the problems have 80 percent of the
impact. Those 20 percent are the vital few. Separating the
problems or issues in this way helps you focus on the improvement
process. Why? A Pareto chart allows you to arrange data
according to priority or importance. This takes the guesswork out
of the process. Heres what to do:
The Process: Identify the possible problems. Use brain-storming,
mental imaging or ask Why? to generate ideas. List all the
possible problems in a particular process.
Use existing reports or collect new data on the process. Group
existing data by consistent units of measure. That means dollars,
percentages, pounds, etc.
89
100
20
75
15
50
10
25
Culmulative
Percentage
Frrequency
of
occurrences
25
0
Doublespaced
90
Garbled
Lost
text
Extra
blank
space
Other
80
60
40
20
0
D E
Defect
Costs ($)
60
40
20
0
A M
Defect
91
1600
1200
800
400
A. Assembly 5.
0
Assy. 5 Assy. 2
Percent
80
60
40
Assy. 1
Assy. 9
20
0
Elect
Assy.
Solder
M. parts
Percent
30
20
A. The K2 relay.
10
0
K2
Percent
80
60
40
20
0
92
U101 Q101 Q7
Q5 Other
C o n tractors
F iscal quarters
93
Histogram
Displays the underlying distribution of process
Illustrates the total variability of the process
Scatter diagram
Shows relationship between variables
Identifies possible causes of problems
Run chart
Shows changes in a process over time
Helps recognize abnormal behavior in
a process
Boxplot
Shows center point and variation of a set of scores
Can be displayed repeatedly to show trend
Control chart
Gives detailed look at trends and variation
Shows changes in a process over time
Includes process driven control limits
Process capability ratio
Links control charts to customer requirements
Relates process variability to tolerance
requirements
Checksheet
The Basics: An organized method of collecting data is a
checksheet. This simple form helps convert the raw data into
readily useful information. An excellent advantage to checksheets
is that they help translate opinion into fact. Heres what to do:
94
M ay 6
M ay 7
17
17
M ay 8
O ld office sym b o l
Total defects
18
Total
defects
11
13
21
7
52
A
A
A2
A3
B
B
C
C
C2 C 3
Scratch
Chipped paint
Tar
Dent
Total defects
Tally
Total
7
11
7
0
25
A - Scratch
B - Chipped paint
C - Tar
D - Dent
Note: A 3 = 3 scratches in close proximity, etc...
95
Days
Upper
Upper
Upper
Ctr
Ctr
Ctr
Low
Low
Low
1-30
31-60
61-90
91120
Fails
18
Most failures (18 of 24) occurred in boxes along the center of the
equipment racks, and most of these failures were type C
(shutdown during test). A facility inspection revealed a heating
duct ran directly behind these boxes. The resulting high
temperatures caused the equipment to overheat and fail.
Histogram
The Basics: To show the central tendency and variability of a data
set use a graph called a histogram sometimes referred to as a
frequency distribution. A histogram can help you determine the
underlying distribution of a process. Histograms also help
understand the total variability of a process. When using
histograms, each data point appears in only one interval. The
number of intervals can influence the pattern the data will take.
Dont expect histograms to be a perfect bell curve; expect
variations. Heres what to do:
96
Use this
many data
number of
points:
intervals:
< 50
5-7
51-99
6-10
100-249
7-12
>250
10-20
Determine the range. Study the data set. Subtract the smallest
value from the largest. This value is the range of the data set.
Determine the interval width. Divide the range by the number of
intervals. Round answers up to a convenient value. For example, if
the range of the data is 17 and nine intervals are used, then interval
width is 1.88. Round this interval to 2.0. Its a good idea to carry
the intervals to one decimal place more than the data collected.
97
point plus the interval width. For example, if the smallest data
point is 10 and the interval width is two, then the starting point for
the second interval is 12. Label intervals along the horizontal axis.
Plot the data. Count the number of data points that fall within each
interval; plot this frequency on the histogram. Remember: each
data point can appear in just one interval. For example, if the first
interval begins with 10.0 and the second with 12.0, then all data
points that are equal to or greater than 10.0 and still less than 12.0
are counted in the first interval.
An Example: During aerobic testing, evaluators weighed 80 Air
Force officers. Heres a histogram and a table showing the
distribution of the data. Which format is most useful?
16
14
Frequency
12
10
8
6
4
2
0
110
120
130
140
150
160
170
180
190
200
Weight (lbs)
98
210
W eights of 80 officers
208
155
159
153
115
159
150
127
145
190
130
165
147
189
139
175
180
180
171
181
156
109
206
130
150
200
128
187
150
163
149
189
139
165
141
180
173
179
166
172
156
208
155
159
201
150
185
150
163
149
190
137
165
145
188
180
171
169
185
178
128
158
129
201
159
127
159
161
191
144
165
147
189
139
166
169
170
180
169
175
Scatter Diagram
The Basics: Recognize the relationship between two variables
with a scatter diagram. These diagrams are graphs that reveal
possible relationships and also help identify possible causes of
problems. An important note: while this method shows a
relationship exists, it wont show that one variable causes another.
Further analysis using other statistical techniques will quantify the
strength of a relationship between two variables. Remember, that
when a relationship exists between two variables, theyre
correlated. (Both positive and negative correlations can be useful
for continuous process improvement.) Heres what to do:
The Process: Collect the data in pairs. Find two different
variables (X and Y) that appear to have a relationship. Each point
on the scatter diagram is an (X,Y) pair of values. There will be
many (X,Y) datapoints on one scatter diagram.
99
X
No
correlation:
(Y doesnt
appear to be
related to X.)
X
Positive
correlation:
(Increase in X
may be related
increase
to in Y.)
Negative
correlation:
(Increase in X
may be related
decrease
to in Y.)
Number of Errors
10
8
6
4
2
0
30
40
50
60
70
Run Chart
The Basics: To show changes in a process measurement over
time, use a run chart. A run chart may also help recognize
abnormal behavior in a process. Heres what to do:
The Process: Construct the chart. Label the vertical axis with the
key measurement of the process you want to measure. Label the
horizontal axis with units of time.
Plot the data. After collecting the data, plot each data point on the
chart.
Interpret the chart. One signal that shows the process has
significantly changed: six steadily increasing or decreasing points in
a row. Another possible signal: nine points in a row that are on the
same side of the average.
An Example: Heres an example of a run chart tracking the
number of people who travel each week. Its important to
recognize the variability inherent in any process. In this process,
the variability is the number of people traveling (four to nine
people). Take a look:
10
Number TDY
9
8
7
6
5
4
3
1
10
WEEK
101
Boxplot
The Basics: A boxplot, sometimes called a box and whisker plot,
is a graph that offers a detailed picture of the center point and
variability of a distribution of data. Boxplots can also help identify
the changes in a process measured over a period of time or show
differences in similar processes (instructor ratings for different
instructors covering the same material). Heres what to do.
The Process: Understand the terms. 100 percent of the data
points fall below the highest data point. 75 percent of the data
points fall below the upper quartile, with 25 percent above it. The
number data points are divided equally at the median 50 percent
below and 50 percent above it. At the lower quartile, 25 percent of
the data points below, with 75 percent of the data points above it.
And finally, 100 percent of the data points are above the lowest
data point.
Collect the data. Collect the data by subset weeks, days,
questions, etc.
Put the data in sequential order. For each subset, put the data in a
row or column. Start with the lowest data point and end with the
highest. For example, this subset has nine data points:
2, 4, 6, 6, 7, 7, 7, 8, 10
Determine the median. The median is the mid-point of the
sequentially ordered data set. In other words, its the point where
half the data points are above and half are below.
If there are an odd number of data points the median is the middle
data point and the same number of data points are on either side of
the median. For example, the underlined 7 is the median for this
dataset with nine datapoints:
2, 4, 6, 6, 7, 7, 7, 8, 10
102
Does the subset have an even number of data points? Then the
median is the average of the two middle data points. For example
with 10 datapoints in which the median is the average of 6 and 7:
1, 2, 4, 6, 6, 7, 7, 7, 8, 10
Determine the quartiles. Quartiles are the points at which there is
25% of the data. If there are an odd number of data points in a
subset, the median is in both the upper and lower quartiles. Heres
an example the median and the upper and lower quartiles are
underlined:
2, 4, 6, 6, 7, 7, 7, 8, 10
If there are an even number of data points in a subset the quartiles
are the midpoints of the upper and lower halves. Heres an
example the quartiles are underlined, and the letter M
represents the median (M is 6.5, which is not a datapoint and is
acceptable):
1, 2, 4, 6, 6, M, 7, 7, 7, 8, 10
Determine the mean. To obtain the mean (X), or average, add all
numbers in the subset and divide that sum by the number of data
points in the subset. For example, heres the mean for the nine data
points in this subset:
2, 4, 6, 6, 7, 7, 7, 8, 10
2 + 4 + 6 + 6 + 7 + 7 + 7 + 8 + 10 = 57
X = 57 (sum of the scores) 9 (number of scores) = 6.33
Plot the data. Draw the axes for the entire data set. For the first
subset, put a dot at the highest data point, the lowest data point, the
quartiles and the median. Draw a box in the space between the two
quartiles. Then draw a line at the median. Draw lines from the
quartiles to the highest and lowest data points. Finally, draw a dot
at the mean. Heres an example:
103
Upper quartile
Mean
Median
4
3
Lower quartile
1
Subset 1
Subset 2
Subset 3
Subset 4
Subset 5
6
5
4
3
2
1
Week 1
Week 2
Week 3
Week 4
Week 5
104
Days
16
14
12
10
Q. Do we have an eight-day
average turnaround?
A. Maybe.
8
6
4
2
1
Week
Strongly
disagree
Strongly
agree
2. W o rk is completed in a timely
fashion.
105
Responses
1.
6 6 6 5 6 6 6 6 5 6 5 6 6 5 6 6 6 5 6 5 6 6
2.
3.
4 3 4 4 4 3 4 5 6 5 4 3 4 3 5 6 6 5 4 5 5 4
5 6 5 4 3 4 5 5 5 4 4 4 3 2 3 3 3 6 4 4 4 3
4.
5 6 5 5 6 6 5 5 5 4 3 5 6 5 5 5 5 4 4 5 5 5
1.
2.
3.
5 5 5 5 5
3 3 3 3 4
2 3 3 3 3
5 6 6 6 6 6
4 4 4 4 4 4
3 3 4 4 4 4
M 6 6 6 6 6
M 4 4 5 5 5
M 4 4 4 4 5
6 6 6 6 6 6
5 5 5 6 6 6
5 5 5 5 6 6
4.
3 4 4 4 5
5 5 5 5 5 5
5 5 6 6 6 6
5 5 5 5 5
106
Lowest
data
point
Lower
quartile
Median
Mean
(average)
Upper
quartile
Highest
data
point
1.
6+6/2=6
5.8
2.
4+4/2=4
4.4
3.
4+4/2=4
4.4
4.
5+5/2=5
5.0
6
5
M
M
3
2
1
M = Median
= Mean (average)
Courtesy
Timeliness
Q1
Q2
Accuracy
Q3
Overall
Goal
Q4
Control Chart
The Basics: Control charts show how a process varies over time
so variability can be recognized, understood and controlled. They
also help identify special causes of variation and changes in
performance. Control charts can also help avoid fixing a process
which is not broken.
Control charts are similar to run charts, but control charts have
statistically calculated upper and lower control limits. A process is
in statistical control when the process measurements vary randomly
within the control limits. That means the variation is consistent and
Appendix A/Process Improvement Guide
107
108
Still not sure what kind of control chart to use? Take a look at this:
Need to construct
a control chart
Variables
n = 1
Sample
size
XmR chart
Variables data
or
attributes data
A ttributes
n > 1
Defective
Defect
Defects or
defective
x-R chart
Yes
Constant
sample
size
np-chart
No
p-chart
Yes
No
Constant
sample
size
u-chart
c-chart
109
Control Chart
Type
Strengths
Concerns
varible data
XmR
X-
Simple to construct
and desirable if
other factors limit
the sample size.
Increased Type I
and Type II error
if used for attribute
data.
X-R
Difficult and
costly
Costly
c and u
counted data
Little value from
a prevention
perspective.
Good management
overview tool.
110
One or more
points outside the
control limits.
Seven or more
consecutive points on
one side of the centerline.
Fourteen points
alternating up
and down.
Fifteen points in a
row within the center
third of the control
region.
111
X + 3/d2(m R )
Center Line
X - 3/d2(m R )
mR chart
Upper Control Limit(R) UCL mR =
D4(m R )
Center Line
(m R )
CL mR =
D3(m R )
Plot the X values and draw the computed control limits and your
X chart is complete. Do the same for the mR chart.
An Example: This table shows individual customer service times
(X) and moving range values for a process observed 15 times.
Note that each mR is the absolute difference (no negative numbers)
between the current X value and the previous X value (e.g., l
7.47-7.89 l = 0.42).
112
7.47
7.89
0.42
7.34
0.55
5.92
1.42
5.81
0.11
7.03
1.22
9.45
2.42
7.08
2.37
6.03
1.05
10
5.27
0.76
11
6.77
1.50
12
7.53
0.76
13
6.04
1.49
14
6.88
0.84
15
7.53
0.65
= 6.94
Moving Range
m R =1.11
The X and mR charts created from the data above are on the
following page:
113
X chart
Service Time
9.89
UCL
6.94
3.98
LCL
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
mR Chart
3.63
1.11
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Observation
X R charts
114
X=
x1 + x 2 +...+ xn x
=
n
n
R = x max x min, by subgroup
X=
X
k
Take the sum of all the subgroup averages divided by the number of
subgroups.
Determine the average value of the range ( R ):
R=
R
k
Take the sum of all the subgroup ranges divided by the number of
subgroups.
Calculate the control limits and centerline using these formulas
and constants:
X chart
Upper Control Limit( X ) UCL x =
X + A2 R
Center Line ( X )
CL x =
X - A2 R
115
R chart
Upper Control Limit(R) UCL R =
Center Line
CL R =
R
D3 R
And for an R chart...
Use this
value for
n
A
2
3
4
5
6
7
8
9
10
D4 R
1.88
1.02
0.73
0.58
0.48
0.42
0.37
0.34
0.31
Use this
value for
D4
D3
0
0
0
0
0
0.08
0.14
0.18
0.22
3.27
2.57
2.28
2.11
2.00
1.92
1.86
1.82
1.78
Plot the charts. Always use these charts in tandem. The X chart
shows sample-to-sample changes in a process, and the R chart
shows the variability within each sample. Recalculate the control
limits if there is a significant change in the process average or
variability one of the signals mentioned earlier.
Customers complained the program office took too much time to
process documents submitted for approval. Thats why the
program manager decided to analyze the offices approval process.
He studied ten weekswork and noted the turnaround times for five
documents in each week. Take a look:
116
Doc. 1
1
2
3
4
5
6
7
8
9
10
36 days
31
43
41
34
59
31
40
25
37
Doc. 2
Doc. 3
33 days
50
41
40
26
33
41
40
47
48
Doc. 4
43 days
33
46
36
33
47
52
38
50
46
Doc. 5
51 days
54
26
56
42
51
38
65
61
61
33 days
35
37
29
28
65
40
51
56
49
39.2 days
40.6
38.6
40.4
32.6
51.0
40.4
46.8
47.8
48.2
18 days
23
20
27
16
32
21
27
36
24
x = 42.56
R = 24.4
Here are the variables and constants used to prepare the chart:
X = (39.2 + ... + 48.2)/10 = 42.56
UCLR =
D4 R = (2.11)(24.4) = 51.48
CLR =
R = 24.4
LCLR =
D3 R = (0)(24.4) = 0
Where
k = 10 (number of weeks)
n = 5 (number of documents/week)
117
UCL
56.71
42.56
LCL
28.41
Week
1
UCL
10
R chart
51.48
24.4
LCL
118
Use charts for defects when determining the items quality by the
number of defects in the item, or by counting the number of
occurrences of some event per unit of time. These charts involve
the degree to which an item is defective (for example, the number
of bad solder joints on a circuit board).
Decide which chart to use. This chart can help. (It is also
acceptable to use XmR charts to analyze attribute data.)
When charting:
sample size:
Defectives: A unit with
one or more defects.
Example: A memo with
five words misspelled.
Defects: An individual
failure to meet a single
requirement.
Constant
Example: A misspelled
word in a memo.
119
p=
UCL = p + 3
LCL = p 3
p(1 p )
= upper cont rol limit (varies by subgroup)
n
p (1 p)
= lower control limit (varies by subgroup)
n
plot p for each subg roup
Note: Use an average subgroup size to obtain a single set of control limits if the largest
subgroup size is less than twice the average subgroup size and the smallest is more than
half the average subgroup size.
u-chart
u
n
u
n
plot u for each subgroup
LCL = u 3
120
c-chart
9 10
Number of
specifications
changed
15
c=
56
= 5. 6 (changes per week)
10
UCL = 5. 6 + 3 5. 6 = 12. 7
LCL = 5. 6 3 5. 6 = 1. 5 => 0
n = 50 active contracts
121
The c-chart showed that week six contains a special cause for
variation. The special cause was a design review held the previous
week. Ten of the 15 changes in week six came from this design
review. Pleased with the charts results, the program manager
adopted this control chart as a permanent management tool.
15
UCL = 12.7
10
c = 5.6
5
c
(the number of
specifications
changed)
LCL = 0.0
1
5
6
Week
10
122
obtain :
R = overall av erage of s ubgroup ra nges
d
( c e n terline on R c h a r t )
= a constant based on the subgro up size
n = s u b g r o u p s ize
10
d2
1.128
1.693
2.059
2.326
2.534
2.704
2.847
2.970
3.078
R
= estimate of process standard deviation
d2
Cp =
USL LSL
6s
where
123
USL
LSL
6o
Cp < 1
This is a capable process. The process variability is equal to the
specification limits. If the process remains centered, specifications
will be met about 99 percent of the time.
LSL
USL
^
6o
Cp = 1
124
USL
^
6o
Cp > 1
Compute actual process performance. After determining the
potential capability of the process, measure the actual performance
of that process. The actual performance of a process is measured
using the performance index. Compute this:
Cp =
Cpl =
USL x
3
x LSL
3
Cpk = the minimu m of {C pu,Cpl }
where x is the overall aver age of subgroup averages
(centerlin e on x chart)
Appendix A/Process Improvement Guide
125
LSL
Defectives
X
3 ^0
3 ^0
C p = 1, C
pk
<1
LSL
X
3 ^0
3 ^0
Cp = 1
C pk = 1
126
The manager used the data from that chart for this process
capability ratio. He used the data to relate current process
performance to the 30- to 60-day turnaround expected by
contractors.
R = 24. 2; d 2 = 2. 326; and x = 42.56
(data from the x R chart)
24. 4
R
=
= 10. 49
d 2 2. 326
With a USL of 60 days and an LSL of 30 days:
USL LSL 60 30
=
= 0. 48
Cp =
6(10. 49)
6
=
Cpu =
Cpl =
USL x
3
x LSL
60 42. 56
= 0.55
3(10. 49)
42.56 30
= 0. 40
3(10. 49)
3
Cpk = minimum {Cpu,Cpl} = 0.40
127
USL
Early documents
represent inefficient
use of resources.
Late documents
produce customer
dissatisfaction.
30
60
The manager showed another graph and said, The answer is not to
speed up the entire process to get few documents taking longer
than 60 days. This would shift the process to the left, leaving us
with even more inefficient use of resources and wed still have
too much variability. Heres the graph illustrating the managers
point:
LSL
Even more
documents
early
128
USL
Fewer late
documents
USL
LSL
Few late
documents
Few early
documents
30
45
60
129
B
QAF Glossary
Action plan -- Specific method or process to achieve the results
called for by one or more objectives.
Action Workout -- An Action Workout is a rapid, concentrated,
high-energy, team effort to make dramatic productivity
improvements in any organization by reducing cycle time.
Affinity diagram -- A management tool that assists with general
planning. It makes disparate language information understandable
by placing it on cards and grouping the cards together in a creative
manner. Header cards are used to summarize each group.
Air Force Mission --To defend the United States through control
and exploitation of air and space.
Air Force Vision --Air Force people building the worlds most
respected air and space force...global power and reach for America.
Alignment --The process of improving a system so that all
elements contribute to the aim.
Appendix B/Glossary
131
132
Appendix B/Glossary
133
Appendix B/Glossary
135
136
Appendix B/Glossary
137
138
139
140
141
142
Appendix B/Glossary
143
144
Just-in-time (JIT) -- A concept where an item is delivered, justin-time, where and when it is needed.
Just-in-time inventory -- The minimum inventory required to
meet production schedules.
Appendix B/Glossary
145
146
Appendix B/Glossary
147
148
Appendix B/Glossary
149
150
Appendix B/Glossary
151
152
Appendix B/Glossary
153
154
Appendix B/Glossary
155
156
Appendix B/Glossary
157
C
Case Study
A Quality Fable About Improvement
Once upon a time at a fighter wing in the American southwest, the
duty day was sunup to sundown. Nobody knew why, but it had
been that way for as long as anyone could remember. The long
hours were frustrating for families, friends, flightline workers and
aircrews. There had to be a better way. But while there were
plenty of complaints, no one really knew what to do about the long
duty days. Operations blamed maintenance, maintenance blamed
the schedulers, and the families blamed everybody.
CIP step 1 Enter Capt F.W. Taylor. Capt Taylor had
worked at the base quality office, and noticed several processes in
his F15E squadron that needed improvement. He knew any
process worth improving must meet two criteria. First, the process
had to be consistent with Air Combat Command goals. Secondly,
Capt Taylor needed a process that was significant to his units
mission, but not overly ambitious.
So Capt Taylor went to work. With help from Capt Deming, who
worked in the quality office, and with inputs from the operations
group commander, they decided to focus improvement efforts on
the length of the crews duty day. Some folks werent sure the
duty day needed attention. After all, the flyers knew maintenance
could solve that problem by speeding up their turn times. On the
contrary, said maintenance; they were convinced the schedulers
needed to be more precise. Capt Taylor had to ask himself, Would
improving this process contribute towards the Air Combat
Command mission or even the squadrons mission?
159
160
Procedures
Policies
Range scheduling
Ordinance checks
Maintenance
Night flying
Refueling
Maintenance crews
Fuel crews
Pilots
People
Bombing range
Plant
161
Job type
M
Frequency
Receive instruction
from faster group
G ive instruction
to slow g r o u p
T im e t o
fin i s h a j o b
25
162
30
35
40
45
50
55
163
UCL
Avg
LCL
164
UCL
Avg
165
months. Eventually the flying window was cut by hours. That led
to a corresponding drop of four hours in the duty day. Before the
team worked on the process, the flying window exceeded eight
hours 40 percent of the time. After the improvements, the flying
window rarely exceeded eight hours. The maintenance crews were
happy with the shorter hours and increased efficiency, the aircrews
were happy with the shorter hours and increased training and that
made the families happier, too. Air Combat Command enjoyed
increased combat capability, the units enjoyed better relations and
everyone enjoyed an increased quality of life. This story has a
happy ending.
166
D
Benchmarking Code of Conduct
Preamble To guide benchmarking encounters and enhance the
professionalism and effectiveness of benchmarking, the
International Benchmarking Clearinghouse, a service of the
American Productivity & Quality Center, and the Strategic Planning
Institute Council on Benchmarking have adopted this common
Code of Conduct. We encourage all organizations to abide by this
Code of Conduct. Adherence to these principles will contribute to
efficient, effective, and ethical benchmarking.
Individuals agree for themselves and their organization to abide by
the following principles for benchmarking with other organizations.
1. Principle of Legality.
167
Refrain from the acquisition of trade secrets from any means that
could be interpreted as improper, including the breach or
inducement of a breach of any duty to maintain secrecy. Do not
disclose or use any trade secret that may have been obtained
through improper means or that was disclosed by another in
violation of a duty to maintain its secrecy or limit its use. Do
not, as a consultant or client, extend one benchmarking effort's
findings to another organization without first obtaining
permission from the parties of the first effort
2. Principle of Exchange
3. Principle of Confidentiality
168
4. Principle of Use
7. Principle of Preparation
169
8. Principle of Completion
Understand
170
E
Suggested Reading
1. Benchmarking: The Search for Best Practices that Lead to
Superior Performance by Robert Camp (ASQC Press, 1989)
2. The Deming Management Method by Mary Walton (The
Putnam Publishing Group, 1986)
3. Employee Driven Quality: Releasing the Creative Spirit of
Your Organization Through Suggestion Systems by Robin E.
McDermott, Raymond J. Mikulak, and Michael R. Beauregard
(Quality Resources, 1993)
4. The Empowered Manager: Positive Political Skills at Work
by Peter Block (Jossey-Bass Inc., 1987)
5. Excellence in Government by David K. Carr and Ian D.
Littman (Coopers and Lybrand, 1991)
6. The Fifth Discipline by Peter M. Senge (Doubleday
Publishing, 1990)
7. Five Pillars of TQM by Bill Creech (The Penguin Group,
1994)
171
172
173
174
Index
Action Workout - 31, 41, 131
Affinity Diagram - 77, 78, 83, 84, 85, 131
Assessments - 30, 57
Awards
Chief of Staff Team Excellance Award - 6, 30
Secretary of the AF Award - 6, 30
Benchmarking - 24, 46, 48, 52, 133, 167
Boxplot - 94, 102, 133
Brainstorming - 34, 65, 71, 73, 86, 133
Cause and effect diagram - 85, 134
Checksheet - 94, 134
Continuous improvement process - 35, 51 55, 135
Control chart - 107-111, 118-122, 136
Customer/supplier model - 27, 138
Empowerment - 5, 13, 139
Environmental scan - 22
Index
175
176
Index
177
Somewhat
Dissatsified
2
Satisfied
Very
Satisfied
Satisfied
12345
12345
12345
12345
12345
-------------------------------------------------------------------------------