P

Putting Best Practices to Work www.qualityprogress.com | February 2017

QUALITY PROGRESS | FEBRUARY 2017

Game Theory
For Effective
Resource
Management
p. 30

QUALITY PROGRESS

Just My Type
LEAN AND SIX SIGMA

Organizing employees
by process vs. function
grows efficiency p. 16

Plus:
Bolstering your chances
VOLUME 50/NUMBER 2

for Six Sigma success p. 24

Understanding the revised

automotive standard p. 48
 Certification
Training Materials

PRIMERS
Our Primers are the most widely used texts for certification training.
They can be taken into the exam. QCI offers 16 different Primers.

SOLUTION TEXTS
Detailed solutions to all questions in the corresponding Primer.

CD-ROMS
Interactive software to assist students preparing for ASQ exams.

QUALITY COUNCIL OF INDIANA

Online Orders: www.qualitycouncil.com
Phone Orders: 800-660-4215
Fax Orders: 812-533-4216
Mail Orders: QCI Order Department,
602 W. Paris Ave., W. Terre Haute, IN 47885-1124
 Statistics can be
powerful.

Learn the statistics you need to make

data-driven decisions with Quality Trainer,
Minitab’s e-learning course. Understand
how to analyze your data and realize the
strength to improve your business.

e-learning for statistics

w w w. q u a l i t y t r a i n e r. c o m
 Contents
Putting Best Practices to Work | February 2017 | www.qualityprogress.com

FEATURES
16 SERVICE QUALITY
Come Together
The pod concept and lean principles streamlined the customer
experience for one insurance organization.

by Ryan Marcus and Michelle Harris

24 CASE STUDY
Subject to Review
A Virginia hospital pinpointed the factors that improved the odds of
success for its Six Sigma projects.

by Todd Creasy
16
30 PROJECT MANAGEMENT
Good Shepherds
The tragedy of the commons theory on resource sharing will give
organizations added insight into managing staff and projects more
efficiently and effectively.

by Kathryn D. McIver ONLY @

www.qualityprogress.com

36 CASE STUDY
Small Changes, Big Results
• Added Insight
To detect defects in electronic products, design a highly accelerated
stress screen and understand what adjustments to make to build a Additional figures and tables
better product. to illustrate the sources of
variation in statistical analyses,
by Daniel K. Sarosky the topic of this month’s Back
to Basics (“Uncovering the Truth,”
p. 64).

30
• Volviendo a los
Fundamentos
Back to Basics translated
into Spanish.

• Questions Answered
Check out QP’s archive of the
Expert Answers department.

• Daily Fix
Read Quality News Today for
the latest quality-related headlines.
 DEPARTMENTS

10
6

8
LogOn
• The natural force of PDCA.

Expert Answers
• Sample size confusion.

Keeping Current
QP
Mail
Quality Progress/ASQ
600 N. Plankinton Ave.
QUALITY PROGRESS

• Augmented reality advances. Milwaukee, WI 53203

• Auditing standard under review. Telephone Fax
800-248-1946 414-272-1734
414-272-8575
14 Mr. Pareto Head
Email
58 QP Toolbox Follow protocol of first initial and full last

10
name followed by @asq.org (for example,
vellifson@asq.org).
60 QP Reviews
Article Submissions
Quality Progress is a peer-reviewed publica-
tion with 85% of its feature articles written
by quality professionals. For information
about submitting an article, call Valerie

COLUMNS
Ellifson at 800-248-1946 x2139, or email
manuscripts@asq.org.

Author Guidelines
5 Up Front 44 Statistics Roundtable To learn more about the manuscript review
Keep it simple. Using sensitivity analyses to design process, helpful hints before submitting a
better studies. manuscript and QP’s 2017 editorial planner,
click on “Author Guidelines” at www.

48
qualityprogress.com under “Tools and
48 Standards Outlook Resources.“
Getting a handle on changes to the
automotive standard ISO TS 16949. Photocopying Authorization
Authorization to photocopy items for
internal or personal use or the internal or
personal use of specific clients is granted by
63 One Good Idea Quality Progress provided the fee of $1 per
Measuring effective-quality copy is paid to ASQ or the Copyright Clear-
to improve processes. ance Center, 222 Rosewood Drive, Danvers,
MA 01923, 978-750-8400. Copying for other
purposes requires the express permission
64 BONUS of Quality Progress. For permission, write
Back to Basics Quality Progress, PO Box 3005, Milwaukee,
Addressing variation in statistical WI 53201-3005, call 414-272-8575 x7406, fax
analyses. 414-272-1734 or email reprints@asq.org.

Photocopies, Reprints
And Microform
Article photocopies are available from ASQ
at 800-248-1946. To purchase bulk reprints
(more than 100), contact Barbara Mitrovic at
ASQ, 800-248-1946. For microform, contact
SPECIAL SECTION ProQuest Information and Learning, 300 N.
Zeeb Road, Ann Arbor, MI 48106, 800-521-
ASQ’S 2017 SIX SIGMA 0600 x2888, international 734-761-4700,
RESOURCE GUIDE p. 54 www.il.proquest.com.

Membership and Subscriptions

For 70 years, ASQ has been the worldwide
provider of information and learning oppor-
NEXT MONTH tunities related to quality. In addition, ASQ
membership offers information, networking,
- LIFE AFTER DISRUPTION - TENSION AND COMPRESSION certification and educational opportunities
Workflow transformation and the relationship Bringing process improvements to help quality professionals obtain practical
between technology and access. and innovation initiatives to life. solutions to the many problems they face
each day. Subscriptions to Quality Progress
are one of the many benefits of ASQ mem-
bership. To join, call 800-248-1946.
ASQ’s Vision: By making quality a global priority, an organizational imperative and List Rentals
a personal ethic, the American Society for Quality becomes the community for everyone Orders for ASQ’s member and nonmember
who seeks quality technology, concepts or tools to improve themselves and their world. buyer lists can be purchased by contacting
Michael Costantino at the Infogroup/Edith
Roman List Management Co., 402-836-6626
Quality Progress (ISSN 0033-524X) is published monthly by the American Society for Quality, 600 N. Plankinton Ave., Milwaukee, WI 53203.
Editorial and advertising offices: 414-272-8575. Periodicals postage paid at Milwaukee, WI, and at additional mailing offices. Institutional
or fax 845-620-1885.
subscriptions are held in the name of a company, corporation, government agency or library. Requests for back issues must be prepaid and
are based on availability: ASQ members $17 per copy; nonmembers $25 per copy. Canadian GST #128717618, Canadian Publications Mail
Agreement #40030175. Canada Post: Return undeliverables to 2835 Kew Drive, Windsor, ON N8T 3B7. Prices are subject to change without
prior notification. © 2017 by ASQ. No claim for missing issues will be accepted after three months following the month of publication of the
issue for domestic addresses and six months for Canadian and international addresses.
Postmaster: Please send address changes to the American Society for Quality, PO Box 3005, Milwaukee, WI 53201-3005. Printed in USA.

4 QP • www.qualityprogress.com
UPFRONT
QP QUALITY PROGRESS

Keep It Simple DIRECTOR OF KNOWLEDGE

Streamlining for clarity and efficiency PRODUCTS
Seiche Sanders

DECLUTTER. DESTRESS. REMOVE complication. It’s something to strive for, but ASSOCIATE EDITOR
Mark Edmund
often easier said than done. Many of us are overscheduled, overburdened and burning the
candle at both ends most of the time. And because it’s so commonplace nowadays, there ASSISTANT EDITOR
Tyler Gaskill
is a whole movement that has sprung up to help people rid their lives of physical things,
activities and even people that create “static.” To commit to being less “busy.” MANUSCRIPT COORDINATOR
Valerie Ellifson
Essentially, that is the goal of lean and Six Sigma initiatives as they apply to manufac-
COPY EDITOR
turing, service, healthcare or other industries. This month, we highlight some of those Susan E. Daniels
methods and tools to help simplify processes and streamline productivity, removing the
ART DIRECTOR
noise from the everyday. Mary Uttech
In “Come Together,” p. 16, learn about how the pod concept and lean principles help GRAPHIC DESIGNER
organize employees by process rather than function, allowing organizations (service, Sandy Wyss
in the example provided) to focus more wholly on the customer. The configuration can PRODUCTION
provide more visibility to processes end-to-end, and provide staff with direct customer Cathy Milquet

feedback. The article explains how to implement pods and their potential pitfalls, then ADVERTISING PRODUCTION
Barbara Mitrovic
provides a brief case study explaining how pods worked for one service organization.
Healthcare settings are popular venues for implementing Six Sigma for improvement DIGITAL PRODUCTION SPECIALISTS
Julie Stroik
projects. “Subject to Review,” p. 24, provides a case study of a 200-bed Virginia hospital Julie Wagner
that conducted several Six Sigma projects over a four-year period. Often, projects are MEDIA SALES
considered either successes or failures, and everyone moves on. This hospital instead Naylor LLC
Lou Brandow
focused on rigorous analysis to gauge the relative successes or failures of its projects, Krys D’Antonio
helping it to determine what was needed to justify future projects. Nicholas Manis
Norbert Musial
“Good Shepherds,” p. 30, explains the tragedy of the commons theory and how it Erin Pande
can be applied to resource management in scheduling staff and projects. Again, if done EDITORIAL OFFICES
effectively, workers receive clearer direction in assignments, allowing them to com- Phone: 414-272-8575
Fax: 414-272-1734
plete tasks more effectively without becoming overworked. More is not always better.
ADVERTISING OFFICES
In this month’s Standards Outlook, you’ll find part two of the series on the automotive Phone: 866-277-5666
standard, IATF 16949:2016. Author R. Dan Reid outlines what quality professionals must
ASQ ADMINISTRATION
know about the changes, and how they should be engaging their organizations’ leadership. CEO
William J. Troy

If you’re going to lose a good team member, it sure lessens the sting if it’s to a good
cause. This month, we say goodbye to Assistant Editor Tyler Gaskill, who leaves us for
a position in the healthcare industry where he’ll get to build and manage his own quality
program. I’m sure all he’s learned during his time with ASQ and QP will give him a great
jump-start. Best of luck, Tyler! QP
Senior Leadership
Andrew Baines
Kalleen Bruch
Dick Palmersheim
Shontra Powell
To promote discussion of issues in the field of quality and
ensure coverage of all responsible points of view, Quality
Progress publishes articles representing conflicting and minor-
ity views. Opinions expressed are those of the authors and not
necessarily of ASQ or Quality Progress. Use of the ASQ logo in
advertisements does not necessarily constitute endorsement of
that particular product or service by ASQ.

Seiche Sanders
Editor

February 2017 • QP 5
 LOGON
Seen&Heard
Quality commitment
“Building Assurance” (January 2017, pp.
37-43) was a good example of apply-
ing quality management principles to a
construction project up front and ahead
of schedule to deliver improved quality
outputs and cost savings on a project.
It’s vital for
organizations to
do it right the
first time with
pride in their
commitment and
workmanship.
Martin Searson
Dublin, Ireland
Everyday quality
The law of gravity example in “Back to
Basics: The Gravity of PDCA” (January
2017, p. 64) is so appropriate. Everyone
applies the plan-do-check-act cycle in
their daily lives, regardless of whether
they know its name, scholastic theory
or steps.
After reading this article, it’s possible
to easily convince people to use quality
management’s many tools and approach-
es in their everyday tasks without label-
ing it appropriately.
Catalina Tutulea
Targoviste, Romania
Resisting ‘Resistance’
With regard to “Expert Answers: Com-
bating Quality Resistance” (January
2017, p. 8), I cannot agree wholly with
the author’s claim that quality manage-
6 QP • www.qualityprogress.com
Tune In
ment (QM) must only report noncon-
formities and should not intervene
in shipment decisions. Ideally, this
might work. But departments make
mistakes on their decisions. In some
circumstances, QM may need more
strength and even have to report to
higher levels of the organization to
stop a bad decision. It all depends on ASQ TV’s latest episode offers
the situation. some highlights from QP’s 2016
I’m speaking from my experience salary survey report, including the
in the food industry. I know it sounds average salary for
silly, but after complaints come in, U.S. quality profes-
upper management many times sionals and how
blames QM for not doing anything. holding certifica-
Of course, it also is silly to rely too tions can influence
much on inspection and rejection. QM salaries.
should be more proactive in helping Visit http://videos.asq.org
(and if possible ensuring) departments to access the full video library.
design and execute good processes—
beginning with raw materials. That is
a more preventive approach.
Francis Watson StayConnected
Ica, Peru Find the latest news, quips and targeted
content from QP staff.
Playing games
Director of Knowledge Products
In response to “Gaming the System” Seiche Sanders:
@ASQ_Seiche
(December 2016, pp. 44-48):
From a quality standpoint, the
great game of business could be Associate Editor Mark Edmund:
@ASQ_Mark
played on cost of quality alone. For
example, your employees might
editor@asq.org
receive a profit sharing bonus. The
quality costs could have a bottom-
line impact on their bonuses that www.facebook.com/
groups/43461176682
they can see and feel empowered to
affect.
Edward Oshaben www.linkedin.com/groups/quality-
progress-magazine-asq-1878386
Euclid, OH
 QP ONLINE
EXTRAS@
www.qualityprogress.com
QP QUALITY PROGRESS

PAST CHAIR
Looking for Career Advice or a New Job? Patricia La Londe (retired – CareFusion)
Subscribe to ASQ’s career-focused monthly enewsletter, Career Connection, to receive
articles, videos and other resources to help grow your professional quality career CHAIR
and improve your workplace. Career Connection also delivers the latest quality job Eric Hayler, BMW Manufacturing
postings and helps your organization find top talent. Subscribe for free at http://asq.org/
newsletters.
CHAIR-ELECT
Elmer Corbin, IBM

Experts Are Here to Help TREASURER

Feel stressed about a tough quality question? Relax—QP’s experts can help. Submit Francisco “Paco” Lopez, Metalsa
your question to editor@asq.org or at http://tinyurl.com/qpexpertanswers, and a subject
matter expert will provide a solution. SECTION AFFAIRS COUNCIL
Sylvester (Bud) Newton, Jr., Alcoa

TECHNICAL COMMUNITIES COUNCIL

QUICK POLL RESULTS Heather Crawford, Apollo Endosurgery
Each month at www.qualityprogress.com, visitors can take an informal survey. Here are
SECRETARY
the results from last month‘s Quick Poll: William J. Troy, ASQ

What fundamental quality tool or concept must all quality professionals master? DIRECTORS
Donald Brecken, Ferris State University
Jim Creiman, Northrop Grumman Corp.
6.1% Root cause analysis. Ha C. Dao, Emerson Climate Technologies Inc.
71.5% Value stream mapping. David B. Levy, Boyce Technologies Inc.
10.7%
Control chart. Austin S. Lin, Google
11.5% Fishbone diagram. Raul Molteni, Molteni Consulting
Luis G. Morales, Daimler Trucks North America
Mark Moyer, CAMLS
Daniella A. Picciotti, Veridiam
Visit www.qualityprogress.com for the latest question: Barrie Simpson, Genentech Access Solutions
Joann Sternke, Pewaukee School District
What’s the most critical factor in overcoming resistance to Six Sigma programs? Sunil Thawani, Quality Indeed Consulting FZE
• Involving front-line employees. John Vandenbemden, Q-Met-Tech
Allen Wong, Abbott
• Accurately communicating the program’s objectives.
• Top management support. QP EDITORIAL REVIEW BOARD
• Demonstrating success early. Randy Brull, chair

Administrative Committee
Brady Boggs, Randy Brull, Jane Campanizzi,
Larry Haugh, Jim Jaquess, Gary MacLean,

QualityNewsTODAY
R. Dan Reid, Richard Stump

Technical Reviewers
Naveen Agarwal, Ashraf Ali, Suresh Anaganti,
Recent headlines from ASQ’s global news service M. Onur Artan, Andy Barnett, Matthew Barsalou,
David Bonyuet, David Burger, Brooks Carder, Bernie
Medical Device Makers Urged to Fight Hacking Threats
Carpenter, L.N. Prabhu Chandrasekaran, Ken Co-
Preventing cyberattacks on medical devices requires more than great designs. In today’s gan, Linda Cubalchini-Travis, Ahmad Elshennawy,
connected world, device makers and hospitals must continually look for problems in Mark Gavoor, Kunita Gear, Daniel Gold,
existing devices, and communicate early and clearly when potential risks emerge, ac- T. Gourishankar, Roberto Guzman, Ellen Hardy,
Lynne Hare, Victoria Jones, Trevor Jordan, Ray
cording to federal regulators. (http://tinyurl.com/qnt-hacking)
Klotz, Tom Kubiak, William LaFollette, Scott Laman,
Patients Get Bar Codes to Avoid Hospital Blunders Pradip Mehta, N.S. Narahari, Arind Parthasarathy,
Larry Picciano, Gene Placzkowski, Tony Polito,
Bar codes are being used in routine hospital operations for the first time to help reduce Peter Pylipow, Imran Ahmad Rana, John Rich-
human error. The technology is being tested by regulators in England to “revolutionize ards, James Rooney, Ayman Sakr, Mahboubeh
patient safety.” Bar codes enable parts used in a major surgery to be easily traced if faults Samghabadi, Brian Scullin, Abhijit Sengupta, Ami-
are found years later. (http://tinyurl.com/qnt-bar-codes) tava Sengupta, Mohit Sharma, A.V. Srinivas, Adrian
Tan, Joe Tunner, Balan Vaithiyanathan, Manu Vora,
Keith Wagoner, Jack Westfall, Doron Zilbershtein

WANT THE LATEST QUALITY-RELATED NEWS AND ANALYSIS? The QNT Weekly
enewsletter, available exclusively to ASQ members, delivers it every Friday.
Subscribe now at http://asq.org/newsletters.

February 2017 • QP 7
 EXPERTANSWE
Sample size confusion
Q: Our customer expects specific levels
of inspection, such as pulling 157 bottles
for visual testing, but they also want 20
pulled for dimensional testing. Can the
20 additional bottles be a subset of the
original testing sample?
When calculating the lot, should I pull
the samples before or after my calcula-
tions? Are the samples included in the
produced quantity? For example, if a
customer orders 10,000 bottles and the
level-two inspection pulls 200 bottles, this
drops the total shipped to the customer to
9,800. If 10,200 bottles are produced, the
inspection level increases to 315 bottles
that must be pulled for testing. What is
the correct sample size and production
number?
A: You asked some great, relevant
questions. For your first question about
whether you can combine visual and
dimensional checks in a single sample, the
simple answer is yes. But there is a catch.
The sampling scheme should be defined
prior to collection. The subset of dimen-
8 QP • www.qualityprogress.com
sional samples must not change based on
the result of the visual inspection.
Sometimes, visual and dimensional
characteristics are correlated. For
example, process upsets in injection
molding can simultaneously cause visual
(cosmetic) defects and dimensional con-
cerns. This is why care should be taken to
ensure that an inspector does not remove
cosmetic defects before selecting the
subset for the dimensional samples.
As a general rule, I advise my clients to
collect a single sample. In many regulated
industries, one critical defect is suffi-
cient to reject a batch. Suppose you are
conducting the visual inspection, and you
discover a critical defect. You must reject
the batch even though the purpose of the
inspection was to check for cosmetic de-
fects. If you collect separate samples, you
are actually increasing the total number
of units that must be inspected. This will
increase the probability of rejecting the
batch.
Some people recommend using ran-
dom sampling to select samples. Random
samples must be collected with respect
to the entire batch, and the best way to
ensure samples are random is to use a
random-number generator and perform
collections based on the production
sequence.
This may be difficult to do in practice
because the times of collection will be
unpredictable. It is much easier to collect
stratified samples such as one-third of the
total samples required at each of three
intervals—the beginning, middle and end
of a production run. Studies show that
stratified sampling is more appropriate if
the process quality drifts over time.1
With regard to whether samples
should be included in the production to-
tal: In most cases, visual and dimensional
checks are nondestructive. So if there
is no contractual requirement to retain
the samples, there is no reason why you
can’t ship the acceptable samples to the
customer. This means you should include
the samples in the production total.
Some facilities conduct large produc-
tion runs and sample once to approve the
entire lot. Later, as orders come in, they
break the lot into smaller sublots for ship-
ping. In this case, I recommend counting
the entire production run when determin-
ing the sample size.
Based on the sample sizes you men-
tioned in your question, it sounds like you
are using American National Standards
Institute (ANSI) and ASQ quality control
standards Z1.4 and Z1.9.2 These include
index tables that increase the sample size
as a function of the production total. If
you comply with these standards, there
will be cases in which they direct you
to use nonzero acceptance criteria. This
is inappropriate for regulated industries
such as pharmaceuticals—in which
a single critical defect is sufficient to
reject the batch. For this reason, all of
the double sampling plans and multiple
RS
sampling plans in the ANSI standard are
unacceptable for critical defects. This is
because all of these plans have nonzero
acceptance standards.
I advise my clients to use only “ac-
cept = zero” (A = 0) sampling plans. This
greatly simplifies the process of selecting
a sampling plan. The process can be sum-
marized as follows:
1. Determine the acceptance quality limit
(AQL) per the quality manual or policy
directive.
2. Enter the ANSI table in the column
corresponding to the AQL.
3. Go down the column until you reach
the A = 0, R = 1 plan.
4. Look to the left to obtain the sample
size.
This approach to attribute sampling
does not use the ANSI sampling plan
“code letters.” Because there is only one
sampling plan with A = 0 for any given
AQL, the sample size is independent of
the lot size. Keep in mind that the confi-
dence associated with a sampling plan
derives from the size of the sample, not
the size of the lot. It, therefore, doesn’t
sample—so there is no need to index the
sample size in proportion to the lot size.
If your organization is not bound by
regulatory or contractual constraints, you
can consider using nonzero acceptance
standards. Table 2 shows that the per-
formance of the sampling plan improves
somewhat as the sample size increases.
Like Table 1, this table shows the prob-
ability of passing the sampling plan if the
population quality is equal to the AQL.
Table 2 shows that the sampling plan’s
performance improves somewhat as
the sample size increases. For example,

Hypergeometric matter whether the lot size is 1,000 or

1 million.
increasing the sample size from n = 32 to
n = 125 increases the probability of pass-
probabilities of It is a common misconception that a ing by about 3%. Users should consider
acceptance / TABLE 1 sample size should grow as a lot size in- whether a four-fold increase in sampling
creases. This is false. Use the A = 0, R = 1 costs is justified by a small increase in
AQL = 0.4 Lot size Probability
plan for critical defects, and use this plan the probability of passing. I recommend
plan of
acceptance regardless of the lot’s size because, quite using the smallest sample size possible,
32 A = 0, R = 1 1,000,000 0.8796 simply, it is the only ANSI-based sampling which will be a plan with A = 0.

32 A = 0, R = 1 100,000 0.8796 plan that is appropriate if there is a possi- Some organizations use the ANSI
bility of detecting a critical defect during attribute-sampling tables for attribute
32 A = 0, R = 1 10,000 0.8795
the inspection. and dimensional checks. While they are
32 A = 0, R = 1 5,000 0.8793
The ANSI attribute-sampling tables are free to use any statistically valid sampling
32 A = 0, R = 1 2,500 0.8789
based on binomial probability, but the plan, the attribute sampling plans are
32 A = 0, R = 1 1,000 0.8778
binomial approximation is not accurate inefficient for inspection by variables,
32 A = 0, R = 1 500 0.8760 if a lot size is small or the sample size including dimensional checks. These
AQL = acceptance quality limit is large relative to the lot size. Table 1, organizations may be able to reduce the
therefore, shows the exact hypergeomet- sample size by at least 50% by switching

Performance ric probabilities of acceptance.

The hypergeometric probabilities
to the ANSI/ASQ Z1.9 standard for inspec-
tion by variables.
improves as were calculated using Minitab statistical Andrew Barnett
sample size software, and Table 1 shows the prob- Director of quality systems

increases / TABLE 2 ability of passing the sampling plan if the NSF Health Sciences Pharma Biotech
population quality is equal to the AQL. Houston
AQL = 0.4 Acceptance Probability For this table, the sampling plan is based
sample criteria of on AQL = 0.4. It is important to note that REFERENCE AND NOTE
size acceptance 1. Paula Lagares Barreiro and Justo Puerto Albandoz,
similar performance would be obtained “Population and Sample: Sampling Techniques,” Manage-
n = 32 A = 0, R = 1 0.8796 ment Mathematics for European Schools, http://tinyurl.
for other AQL values. com/sampling-teq-qp.
n = 125 A = 1, R = 2 0.9101 2. For more information on sampling, read the American
Table 1 shows the probability of ac-
National Standards Institute (ANSI) and ASQ standard,
n = 200 A = 2, R = 3 0.9529 ceptance is nearly constant regardless of ANSI/ASQ Z1.4 Sampling Procedures and Tables for
n = 315 A = 3, R = 4 0.9611 the lot size. As stated earlier, the power
Inspection by Attributes (2008) and ANSI/ASQ Z1.9–2003
(R2013): Sampling Procedures and Tables for Inspection
n = 500 A = 5, R = 6 0.9837 of sampling is due to the size of the by Variables for Percent Nonconforming (2003).

February 2017 • QP 9
 KEEPINGCURRE
TECHNOLOGY
Facing New Realities
T
Augmented reality could reinvent manufacturing processes, employee training
The technology behind last summer’s Pokémon Go craze is get-
ting serious. Augmented reality (AR)—overlaying computer-gen-
erated graphics onto reality—is transitioning from a kids’ game
to military and industrial applications, such as helping Lockheed
Martin engineers assemble F-35 aircraft or allowing factory-floor
workers to see 3-D blueprints
overlaid as holograms on ma-
chines they’re repairing.
As this new technology
becomes more affordable and
mainstream, it has the potential
to create factory-floor efficiencies
and employee training programs
that were previously considered
the stuff of science fiction.
From fiction to reality
According to Forrester Research
Inc., by 2025 nearly 15 million U.S.
workers will use smart glasses, such as Google Glass and Micro-
soft’s HoloLens, which is an increase of 400,000 users from 2016.
Forrester also estimates large organizations will spend $3.6 billion
on smart glasses in 2025, up from $6 million in 2016.1
AR’s ability to provide visual representations of data or sche-
matics on the factory floor could be a game changer. A mainte-
nance worker who is examining a pressure valve, for example,
could be wearing lenses that display the valve’s temperature and
pressure as text that hovers above the part, said Mike Campbell,
executive vice president of industrial software firm PTC’s Vuforia
augmented reality segment.2
This concept just scratches the surface of AR’s potential
applications, Campbell said. The lenses could overlay graphics
that also show how a part’s pieces fit together, instructions for
disassembling it and how it could connect to different machines.3
Several organizations already are using AR in this way.
AGCO Corp., a Minnesota agricultural equipment manufacturer,
has equipped some of its employees with Google Glass, which
shows diagrams and instructions to help them perform quality
checks on tractors. This method has made their quality checks
20% faster.4
Boeing is trying a pilot program in which Google Glass helps
10 QP • www.qualityprogress.com
mechanics wire airplanes’ complex electrical systems. Instead of
trying to follow paper or laptop instructions, mechanics’ Google
Glass headsets show them wire placement diagrams and instruc-
tions. So far, the pilot program has reduced production time 25%
and errors have been reduced to zero.5
AR’s benefits also extend
to data analysis. Holograms
are useful for visualizing data
because they engage the spatial
awareness part of the brain that
allows humans to understand
complex concepts more quickly
and promotes greater reten-
tion, according to Brian Mullins,
co-founder and CEO of Daqri,
a maker of augmented reality
smart helmets.6
“In cases where you have
to react with very complex
amounts of information, it can tell a much better story than just a
bar graph,” Mullins said.7
IoT meets AR
Incorporating the Internet of Things (IoT) with AR could add yet
another level of possibilities to how manufacturers could use the
technology. Many factories are using IoT machines and devices to
create a connected environment that sends large amounts of data,
such as power consumption, operating status or dates the ma-
chines or devices were last serviced. Campbell said using IoT alone,
however, lacks context and that AR can contextualize IoT data.8
“Getting that data is not as hard these days as making sense
out of it for the people that need it to make decisions,” said Peggy
Gulick, director of business-process improvement at AGCO Corp.9
For example, an employee wearing AR gear could view any
IoT-enabled machine and receive visual cues about whether
it’s experiencing any problems and where an issue is occurring
without even opening the machine for repairs. Experts believe
this would not only save organizations time and labor, but service
technicians could use AR to determine whether a faulty part is
salvageable or should be replaced.10
“One of the big things that IoT is changing is that people will
E NT
not need to interact only with other people, but also with machine
and devices,” said Luigi De Bernardini, CEO of manufacturing software
provider Autoware. “The more intelligent machines are, the more users
will need to interact with them, and the more they will expect to do it
in a natural way.”11
Training advantages
Bernardini said he sees AR as a solution to a major concern most
manufactures face: the aging workforce. As experienced employees
leave the workforce, new ones will struggle to gain their predecessors’
levels of knowledge.
A more experienced engineer, for example, equipped with an AR
headset with a camera, could record his or her maintenance or trou-
bleshooting approach on a piece of equipment, and a newer engineer
could later use the video for training.12
“Augmented reality could be a very effective way in delivering the
knowledge in the right place at the right time,” Bernardini said. “Seeing
something is much more effective than reading about it or listening to it.”13
AR has allowed Lockheed Martin to skip years of training that were
previously required for employees building its F-35 aircraft. Using AR
glasses equipped with cameras and sensors, engineers can see render-
ings overlaid onto their working environments that show them how
to assemble components. After implementing this method, engineers
increased their accuracy 96% while working 30% faster.14
In healthcare, AR could be used in operating theaters to improve
surgeon training—typically an expensive and difficult process—and re-
duce mistakes in complicated procedures. Last month, Touch Surgery,
a developer of surgical training applications, unveiled an AR application
that allows surgeons wearing AR headgear to practice surgeries on
digital patients in immersive, virtual reality environments.15
The application also can be used in live procedures to create an
overlay on patients to help guide surgeons. Some experts, however,
worry this type of technology could leave patients questioning sur-
geons’ competency if they’re being walked through surgeries by their
glasses.16
Looking ahead
Some industry experts said realizing AR’s full potential has a long way
to go. Most of the technology that will make future AR experiences
impactful has yet to be created, and it also will require connected
services and applications that have yet to be conceived.
Many companies are still tackling how wearable AR devices can
more effectively capture human interactions without a keyboard,
mouse or touchscreen, and how their cameras can capture subtle
hand and finger movements used to manipulate digital images or data.
(continues on p. 12)
Q
Who’s Who in
NAME: Lucita Kahn.
RESIDENCE: North Bergen, NJ.
EDUCATION: Master’s degrees in man-
agement and project management, both from the Florida
Institute of Technology in Melbourne.
INTRODUCTION TO QUALITY: Kahn learned about quality
during her first job as a quality improvement inspector and
total quality management instructor in the U.S. Air Force
Reserve.
PREVIOUS JOBS: Retired as a master
sergeant in the U.S. Air Force Reserve
after 25 years. Retired as a product
quality manager at the U.S. Army Arma-
ment Research, Development and Engi-
neering Center after nearly 31 years.
ASQ ACTIVITIES: Involved in ASQ’s
North Jersey Section as the past chair, performance awards
and recognition chair, spring quality conference chair and an
ASQ fellow mentor. Serves as ASQ North Jersey Section rep-
resentative on the North East Quality Council (NEQC) and is
a past NEQC secretary. Active member of technical commu-
nity council and belongs to five divisions: quality manage-
ment, audit, lean enterprise, innovation, and electronics and
communication. Also an electronics and communication
division (ECD) past secretary, treasurer and chair elect. Now
serves as ECD chair.
OTHER ACTIVITIES: Volunteering and helping people with
their public speaking skills through Toastmasters Interna-
tional.
RECENT HONORS: Elected to ASQ’s 2016 class of fellows.
Received section chair appreciation award from the ASQ
North Jersey section in 2016. Received the 2016 NEQC R.
Shaw Goldthwait Award.
PUBLISHED WORKS: Several articles in U.S. Army and U.S.
Air Force newsletters.
PERSONAL: Married; six children.
FAVORITE WAYS TO RELAX: Reading inspirational books
and keeping up with quality news.
QUALITY QUOTE: Quality is everyone’s responsibility.
February 2017 • QP 11
 KEEPINGCURRENT
SHORTRUNS
THE BALDRIGE PERFORMANCE Excellence Program re-
leased the 2017-2018 Baldrige Excellence Framework (Busi-
ness/Nonprofit, Healthcare and Education) booklets. All three
versions include the Baldrige criteria, core values and concepts,
and guidelines for evaluating an organization’s processes and
results. Baldrige Excellence Framework printed booklets costs
$30. For more information, visit http://asq.org/2017baldrige.
THE AUTOMOTIVE INDUSTRY Action Group’s southern au-
tomotive quality summit will be held March 9 in Huntsville, AL.
The one-day event will be the official roll-out of the new Inter-
national Automotive Task Force 16949 standard to the supply
base in the southeast United States. For updates on this event,
visit http://bit.ly/SouthernAutoQuality2017 (case sensitive).
A NEW RISK AND regulatory services innovation center has
been established by PricewaterhouseCoopers and Carnegie
Mellon University to provide training to organizations and
perform research on risk and compliance-related issues. The
center also will address issues related to audit innovation, data
analytics, information privacy and security. For more informa-
tion, visit http://tinyurl.com/new-innovate-center.
MANUFACTURING
SURVEY: COMPANIES STAYING
WITH THEIR SUPPLIERS
While 83% of manufacturers say they have been adversely
affected in the past by their suppliers’ inabilities to meet
their needs, only one-third anticipate a shortage of parts or
services this year, according to a recent ASQ survey.
ASQ’s 2017 Manufacturing Outlook Survey also shows
that 66% of manufacturers expecting a problem with sup-
pliers are working closely with providers to resolve issues,
while 35% are working with their suppliers’ competition.
Some manufacturers are stockpiling parts, while others are
expanding operations to create the necessary parts them-
selves.
“Supply chain plays a critical role in manufacturing, and
companies simply can’t risk being without the material they
need to be successful,” said Pat La Londe, past chair of ASQ’s
board of directors. “Companies need to carefully consider
multiple options when faced with a shortage of materials or
with suppliers that can’t meet their needs.”
More than 1,125 manufacturing professionals from around
the world responded to ASQ’s 2017 Manufacturing Out-
look Survey, which was conducted online in November and
December. Respondents to the survey represented several
industries such as aerospace, automotive, food and medical
devices.
For more information from the survey, visit http://tinyurl.
com/asq-outlook-survey.

Advances in augmented reality technology (continued from p. 11)

Ensuring long-lasting battery life for AR
devices is another hurdle tech companies
are struggling to clear.17
Still, many experts agree that it’s a mat-
ter of when—not if—AR will significantly
change industries and people’s lives.
“I believe that every walk of life and
every human activity, which today is medi-
ated by computers, will be transformed
by AR,” said Microsoft CEO Satya Nadella.
“This is not happening tomorrow. It’s not
happening next year. But that is the con-
tinuum we are on.”18
Dan Arczynksi of Index AR Solutions,
a developer of AR applications, said AR
isn’t a technology that will just provide
incremental, 2 to 5% percent savings.
“It’s a game-changing application with
measurable savings ranging from 25 to
more than 90%,” he said. “[This is because]
when you augment people with AR, you’re
not buying equipment: You’re freeing up
cash flow because you’re speeding up
processes.”19 —compiled by Tyler Gaskill,
assistant editor
REFERENCES
1. Sara Castellanos, “Augmented Reality, Hologram-Like
Images Enter the Workplace,” Wall Street Journal, Dec. 12,
2016, http://tinyurl.com/wsj-qp-augmented.
2. Matt Weinberger, “Augmented Reality Is Going From
‘Pokémon Go’ to the Factory Floor,” Business Insider, Oct.
25, 2016, http://tinyurl.com/bi-qp-augmented.
3. Ibid.
4. Castellanos, “Augmented Reality, Hologram-Like Images
Enter the Workplace,” see reference 1.
5. Ibid.
6. Ibid.
7. Ibid.
8. Weinberger, “Augmented Reality Is Going From ‘Pokémon
Go’ to the Factory Floor,” see reference 2.
9. Castellanos, “Augmented Reality, Hologram-Like Images
Enter the Workplace,” see reference 1.
10. K.R. Sanjiv, “How Augmented Reality Can Revolutionize
Manufacturing,” IndustryWeek, Sept. 29, 2016, http://
tinyurl.com/iw-qp-reality.
11. Luigi De Bernardini, “Augmented Reality Is Key for Smart
Manufacturing,” AutomationWorld, June 20, 2016, http://
tinyurl.com/qp-luigi.
12. Jeff Kerns, “Will Virtual Reality Replace the Desktop?”
Machine Design, Jan. 10, 2017, http://tinyurl.com/qp-kerns.
13. Ibid.
14. Sanjiv, “How Augmented Reality Can Revolutionize Manu-
facturing,” see reference 10.
15. Ted Ranosa, “Doctors Can Now Practice Surgical Pro-
cedures Through Augmented Reality,” Tech Times, Jan. 7,
2017, http://tinyurl.com/ar-surgery.
16. Ibid.
17. Tom Mainelli, “The Challenge and Opportunity of Aug-
mented Reality,” Recode, May 8, 2016, http://tinyurl.com/
qp-recode.
18. Gerard Baker, “Microsoft CEO Envisions a Whole New
Reality,” Wall Street Journal, Oct. 30, 2016, http://tinyurl.
com/wsj-qp-new-reality.
19. David Greenfield, “Augmented Reality for Industry,”
AutomationWorld, June 8, 2016, http://tinyurl.com/qp-
automation-kc.

12 QP • www.qualityprogress.com
 ASQ
DATE IN QUALITY HISTORY
QP looks back on a person or event that made a difference in the
history of quality.
Feb. 23, 1947
The International Organization for Standardization (ISO), the
world’s leading developer of international industry standards, was
founded in Geneva on this date.
ISO was born from the union of two organizations—the Inter-
national Federation of the National Standardizing Associations,
established in 1926, and the United Nations Standards Coordinating
Committee, established in 1944.
In October 1946, delegates from 25 countries, meeting at the
Institute of Civil Engineers in London, decided to create a new inter-
national organization. Their objective was to “facilitate the interna-
tional coordination and unification of industrial standards.” The new
organization, ISO, officially began operations four months later.
Today, ISO is a network of the national standards institutes of
157 countries, one member per country, with a central secretariat
in Geneva that coordinates the system. Each country is represented
by its own national body. The American National Standards Institute
represents the United States.
SOURCE
ISO, www.iso.org/iso/about.htm.
ASQ CONFERENCE
LEAN SIX SIGMA EVENT
TO BE HELD IN PHOENIX
ASQ’s Lean and Six Sigma Conference will be held Feb. 27-28 in
Phoenix.
The theme of this year’s conference is “Leading the Culture of
Excellence Through Lean and Six Sigma.”
Focus areas at the two-day event include: lean and Six Sigma fun-
damentals; lessons learned—implementation of lean and Six Sigma;
doing more with less; and tips and tricks—sus-
taining results.
Jacob Stoller is one of the keynote speakers
who will appear at the event.
Stoller is an author, speaker and consultant
who specializes in communication between
experts and outsiders in areas such as lean
management, IT, accounting and engineering.
He is also the author of The Lean CEO
STOLLER (McGraw-Hill Education, 2015).
To learn more about the conference or to register, visit http://asq.
org/conferences/six-sigma.
ASQ FINALIZES 2017 BOARD
ASQ has named its 2017 board of directors. They are:
• Eric Hayler, chair, BMW Manufacturing.
• Patricia La Londe, past chair, CareFusion (retired).
• Elmer Corbin, chair-elect, IBM.
• Francisco “Paco” Lopez, treasurer, Metalsa.
• Sylvester “Bud” Newton Jr., section affairs council chair,
Arconic.
• Heather Crawford, technical communities council chair,
Apollo Endosurgery.
• Donald Brecken, Ferris State University.
• Jim Creiman, Northrop Grumman Corp.
• Ha C. Dao, Emerson Climate Technologies Inc.
• David Levy, Boyce Technologies.
• Austin S. Lin, Google.
• Raul Molteni, Molteni Consulting.
• Luis G. Morales, Daimler Trucks North America.
• Mark Moyer, Center for Advanced Medical Learning and
Simulation.
• Daniella A. Picciotti, Veridiam.
• Barrie Simpson, Genentech Access Solutions.
• Joann Sternke, Pewaukee (WI) School District.
• Sunil Thawani, Quality Indeed Consulting.
• John Vandenbemden, Q-Met-Tech.
• Allen Wong, Abbott Laboratories.
WORDTOTHEWISE
To educate newcomers and refresh practitioners and
professionals, QP occasionally features a quality term and
definition each month:
Seiban
The name of a Japanese management practice taken from
the words sei, which means manufacturing, and ban, which
means number. A seiban number is assigned to all parts,
materials and purchase orders associated with a particular
customer’s job, project or anything else. This enables a man-
ufacturer to track everything related to a particular product,
project or customer, and facilitates setting aside inventory
for specific projects or priorities. That makes it an effective
practice for project and build-to-order manufacturing.
SOURCE
“Quality Glossary,” Quality Progress, June 2007, p. 56.
February 2017 • QP 13
 KEEPINGCURRENT
HEALTHCARE
ANALYSIS AIMS AT WAYS
TO REDUCE VARIATION IN
QUALITY MEASUREMENTS
The National Quality Forum (NQF) says its new report
provides guidance on reducing unnecessary variation
in healthcare quality measurements so measurements
can be more uniform and have greater impact on im-
proving care.
The recent analysis identified 1,367 quality measure-
ments in use across 48 different state and regional
programs in the United States.
Of these measurements, 509 were distinct. The
remaining 800-plus measurements overlapped or had
similar focus, with one or more variations in the specifi-
cations.
“Quality measurements are essential building blocks in
large-scale public and private-payer efforts to reform the
nation’s healthcare system. But slightly different versions
of the same measure contribute to waste through report-
ing burden for providers and make performance com-
parisons more difficult,” said Helen Darling, NQF’s interim
president and CEO.
A 16-person expert panel developed a decision-logic
framework to reduce measurement variation, with a
focus on how best to improve the ability to compare and
interpret measurements, while driving transparency and
innovation and reducing burden.
For more from the study, visit http://tinyurl.com/
nqf-measure-report.
Mr. Pareto Head BY MIKE CROSSEN
14 QP • www.qualityprogress.com
18 MEMBERS INCLUDED IN ASQ
2017 CLASS OF FELLOWS
The ASQ Board of Directors has named 18 fellows who join nearly
675 current fellows. The 2017 class includes:
• Eric Alden, Xerox Corp., Canandaigua, NY.
• David Dewayne Butler, TI Automotive, Auburn Hills, MI.
• Rajeev Chadha, National Research Council of Canada,
Edmonton, Alberta, Canada
• Julie Furst-Bowe, Chippewa Valley Technical College,
Eau Claire, WI.
• Vinay Goyal, Abbott Laboratories, Fullerton, CA.
• Michael Hamada, Los Alamos National Lab, Los Alamos, NM.
• Bart Hamilton, consultant, Akron, OH.
• Michael Kent Hart, consultant, Kanata, Ontario, Canada.
• James Miller, Roche, Muncie, IN.
• Mark Neal, Global Quality Systems, Prosper, TX.
• Owen Ramsay, Lloyd’s Register Quality Assurance, Laurelton, NY.
• Abbas Saghaei, Azad University, Tehran, Iran.
• Donald Singer, GlaxoSmithKline Pharmaceuticals, Phoenixville,
PA.
• Nicholas Skovran, Quality Horizons LLC, Pittsburgh.
• Jennifer Joy Stepniowski, Pro QC International, Tampa, FL.
• Sandra Storli, Zimmer Biomet, Gurnee, IL.
• William Taraszewski, Meda Pharmaceuticals, Decatur, IL.
• Rajesh Kumar Tyagi, Telfer School of Management, University of
Ottawa, Kirkland, Quebec, Canada.
STANDARDS
ISO AUDITING STANDARD UNDERGOING REVISIONS
The International Organization for Standardization’s (ISO) popular
standard for auditing management systems is under revision
and has just reached the first voting stage.
ISO 19011—Guidelines for auditing management systems is
currently being revised to reflect the growing number of man-
agement system standards (MSS) and the recent revisions of
some of the most widely used ones, such as ISO 9001 for quality
and ISO 14001 for the environment. It has just reached commit-
tee draft stage, meaning those countries involved in its revision
have an opportunity to make comments on the draft.
“As organizations see the benefit and need for management
systems, there has been an increase in the number of sector-
specific standards to respond to the mandate,” said Denise
Robitaille, chair of ISO project committee 302, the ISO project
committee responsible for the revision. “There are now MSSs
that cover areas such as health and medical, environment,
services, information technology and more. In addition, the two
most popular MSSs—ISO 9001 and ISO 14001—have recently
been updated, so the auditing of these systems needs to reflect
the variety and number of standards being developed.”
For more information and updates, visit the ISO website at
http://tinyurl.com/iso-audit-stand.

QP’s 2017 editorial calendar has been

released, and now we’re looking for authors to
write on these topics:

Write HERE June September

Write NOW Managing Change Big Data

July October
Engagement and Standards and Auditing
Culture
November
August Customer Experience
A Stronger Supply
Chain December
Salary Survey

If you intend to write for a specific issue, submit

your article at least three to six months before the
issue date to allow enough time for the article to
be reviewed, revised and scheduled.

But if you miss that deadline, don’t worry. Send

the article anyway. All accepted articles will be
published.

Visit www.qualityprogress.com and click on

“Author Guidelines” at the bottom of the page
for more details.

February 2017 • QP 15
Come
Together Using pods to
streamline the
customer experience

by Ryan Marcus and

Michelle Harris
 SERVICE QUALITY

MANY SERVICE ORGANIZATIONS struggle

to become more customer-centric due to highly fragmented work-
flows that get in the way of delivering seamless customer service.
Creating cross-functional teams and organizing them as pods is an
effective strategy to strengthen customer focus, improve productiv-
ity and increase flexibility.
Most organizations that strive to deliver exceptional customer
service are confronted with issues such as convoluted processes,
inflexible staffing models and unacceptable turnaround times. If
work is organized by function rather than process, customer needs
can easily fall through the cracks.

In 50 Words
Or Less
• The pod concept
organizes employees
in a single location by
process rather than
function.
• By relying on lean
principles, pods help
employees see a pro-
cess end-to-end, receive
direct feedback from
the customer and gain
autonomy.
• This can increase
employee engagement,
eliminate handoffs and
improve customer satis-
faction.

February 2017 • QP 17
 The pod concept aims to organize employees by
process instead of function, allowing for efficient and
effective coordination across process steps. The pod
design is based on Henry Ford’s assembly line, which
minimized wait time by moving the product along a
line to be worked on by technicians at each step. This
same benefit can be achieved in a service organization
by having team members—who represent each func-
tion and understand the work of other functions—or-
ganized into one pod. In essence, a customer is the
product that’s moving down the assembly line with
minimal wait times.
In addition to increasing customer focus, pods
also increase flexibility and scalability. As demand
increases, decreases or changes, the amount and fo-
cus of pods can be easily adjusted. They also increase
employee engagement and empowerment because the
pods “own” the customer.
Lean principles are at the core of the pod concept.
Lean seeks to understand value from the viewpoint of
the customer and eliminate activities that don’t add
value—waste. Pods apply lean principles such as:
• Creating physical co-location, which minimizes
communication delays.
• Eliminating waste, such as handoffs.
• Minimizing customer wait time.
Pods can help achieve drastic improvements in per-
formance. Table 1 shows how adopting a pod structure
improved the customer intake process for a healthcare
services business. Table 1’s data show substantial and
immediate improvements. In our experience, pods
have the potential to reduce cycle time by 50% and
costs by 20%.
Adopting a pod structure also can increase em-
Customer intake
improvements / TABLE 1
Performance Description Pre-pod Post-pod
indicators Q1 Q2
Activation Percentage of referrals received
78% 85%
rate that were activated.
Turnaround
Number of days from therapy
time to 13 5.1
creation date to scheduled date.
schedule
Percentage of referrals that are
Denial rate 16%  8%
denied.
Unknown Percentage of referrals without
34% 17%
severities severity captured.
18 QP • www.qualityprogress.com
ployee engagement. For example, a case manager at
an insurance organization we worked with said, “Pods
definitely helped me build better relationships with
other functions in the organization because I now have
more face-to-face contact. Better relationships and
knowing your co-workers make for a more productive
and rewarding work day.”
Implementing pods
There is a 10-step approach for implementing pods
(see Figure 1), and you should start by considering its
scope.
1. Scope the project—Selecting the right process
to tackle is imperative to success. You should consider
selection criteria such as:
• Clear benefits to the customer.
• Process stability.
• Probability of success.
• Cost to implement.
• Team availability.
Starting small can reduce risks and increase the
likelihood of success. Start with the customer perspec-
tive and select an end-to-end process with clear out-
puts.
2. Manage the change from the start—Weigh
the tradeoffs of implementing a radical change versus
one that’s gradual. Consider the amount of resistance
you will encounter from those who are affected. Cre-
ate a communication plan. Messaging should be clear,
concise, fact-based and sent through several channels.
Train frontline managers on how to best handle ques-
tions and concerns from employees.
3. Develop a project plan—Create a plan with
high-level activities needed and the appropriate tim-
ings for completion. Start with the end state in mind
and plan backward. Account for setbacks: Working in
pods is often a radical departure from the status quo
and will likely result in an initial decrease in perfor-
mance levels as the members of the pod learn how to
be effective in the new operating model.
4. Redesign the process—Map how work current-
ly gets done and how it will be completed in the new
process. Creating a current-state map is the founda-
tion for developing the future-state process. The future
state should be based on the principle of continuous
flow—high levels of efficiency and low levels of wait
time. New tools to help pod members complete their
work also are usually needed. Take a look at what sys-

Steps for implementing pods
1. Scope the project
2. Manage the change from the start
3. Develop a project plan
4. Redesign the process
5. Develop metrics and establish a baseline
tems, equipment and templates are in use and update
them as necessary.
5. Develop metrics and establish a baseline—
Determine how you will measure success. Ask whether
you have a reliable baseline from which to improve. If
you don’t, put together a data collection plan to create
one. When selecting measurements of success, ensure
the process-level metrics align with your organization-
wide measurements of success. Do not move forward
with a pilot until you have the right metrics and mea-
surement system in place. Senior leadership will want
to see results from the initiative, and you must be pre-
pared to show them.
6. Assess staffing needs—A pod will be cross-
functional, so you will be drawing on resources from
several areas. Create a staffing model to allow for fact-
based decisions about how many employees it will
take to complete the work. Develop different staffing
scenarios using the staffing model, and let the key de-
cision makers come to an agreement on what makes
sense.
7. Redesign jobs—As you decide on the staffing
needs, consider how roles could be consolidated or
expanded. What are the competencies (knowledge,
skills and abilities) needed for each role? Will the
current staff be able to perform the redesigned jobs?
What responsibilities may need to shift? Leverage the
SERVICE QUALITY
/ FIGURE 1
6. Assess staffing needs
7. Redesign jobs
8. Address training needs
9. Prepare the pod workspace 10. Pilot
expertise of the HR department as you redesign jobs to
ensure there is correct salary grading, and don’t forget
how managers will be affected. Because pods require
less coordination, the number of managers and their
roles often change dramatically.
8. Address training needs—As you map the new
process, develop new tools and redesign jobs, think
about the competencies required at each process step
and develop a training plan that addresses what con-
tent is required, who is to be trained and how the par-
ticipants will be trained—such as virtually or face to
face. Pod members must have a detailed understand-
ing of how the process works from end to end.
9. Prepare the workspace—Prepare for how a
pod structure will physically change your office. Con-
sider how many members each pod will have and what
the layout will look like. Determine pods’ locations,
and mobilize resources to support the move and setup.
Don’t assume everything will be ready on day No. 1.
Give pod members time to get settled and ensure their
phones and computers work. The last thing you want
is a pod in place without the necessary tools to service
customers.
10. Pilot—The purpose of the pilot is to test the
process and design assumptions prior to launch Have
the project team available to help answer any ques-
tions. Use a whiteboard to visualize the major steps in
February 2017 • QP 19
 New customer-centric process / FIGURE 2

Case manager Case manager

Enrollment strategy
• Relationships • Deliver output to
• Requirements the client
Regional work team • Teach client how to
• Enrollment
coordinator do business with us

Centralized services
Sales coordinator

• External
Sold case

communication Contracts and Billing

• Manage handoffs system setup coordination
Contract
Work
• Coordinate call and Inputs Outputs certificates bill
distribution
external timelines lead
Sold case Enrollment EDI
package processing coordinator

EDI = electronic data interchange

Broker    Communication to broker and client  Client

the pod process, and collect data on performance and
the status of customer deliverables to identify break-
downs and opportunities for improvement. Refine the
design before scaling up.
What to avoid
When adopting the pod concept, there are many com-
mon pitfalls that must be avoided:
Underestimating the amount of organizational
change—Any process redesign inevitably creates a
need to change. Understand the amount of change re-
quired and its implications. Determine what effect co-
locating employees will have on reporting lines.
Delaying communication—Waiting to communi-
cate about pods until all pieces of the project are in
place might sound like a good idea, but that can create
anxiety and fear.
The more knowledge and understanding people
have, the more rapidly your organization can adopt
the pod process. Use various opportunities to explore
the reservations, concerns and complaints about the
change. Set up a display pod for prospective team
members to see and experience the changes ahead.
Overplanning—Project plans are important.
Spend time and effort creating a plan while keeping in
mind that it is only a plan, so remain flexible enough to
adjust course as needed.
Skipping current-state analysis—A common
mistake is to jump right into designing the future state.
Spend time mapping and understanding the current
state and pain points that must be addressed in the fu-
ture state.
Measuring everything because you can—There
are many things you can measure, such as turnaround
time, rework rates, production rates, queue size, staff-
ing levels or service levels. Measuring everything can
lead to confusion, however, so decide early on how
to best track progress and invest in building a robust
baseline.
Expecting immediate results—After the first
week, you will likely find turnaround times worsen
and there is some failure to complete customer orders.
Be prepared to deal with underperformance and set
realistic expectations with regard to a timeline for im-
proved performance. Remember, any process change
will inevitably decrease productivity in the short run.
Relying too much on the staffing model—The
tendency is to take the initial model at face value, but
remember it’s a model and only as good as the accu-
racy of the data. The real test is to set up a pilot pod
and start testing assumptions.
Underestimating manager’s resistance—Man-
agers are often protective of their turf. There is going
to be inevitable resistance to giving up resources for
the pods. Anticipate this and manage it effectively.
Underinvesting in team building—Teamwork is

20 QP • www.qualityprogress.com
 SERVICE QUALITY

a key element of a successful pod.

Create an environment in which
Projected future headcount / TABLE 2
this is accepted and encouraged, es- Future roles
pecially if it hasn’t been in the past.
Work
Investing in teambuilding early Sales Case distribution
helps to avoid the gravitational pull Current headcount coordinator manager lead Remaining
of the functional areas. Sales coordinator 73 66  7 - (0)
Ignoring the lessons learned Account coordinator 21 21 - - -
from the pilot phase—You should SWAT coordinators 8  8 - - -
note any pain points and changes Regional enrollment manager 7 -  7 - -
that must be made. It is important Enrollment case manager 11 - 11 - -
to aggregate the learnings from
Enrollment manager 9 -  9 - -
the pilot and make the necessary
Regional service specialist 38 9  4  4 21
changes before scaling up.
Implementation specialist—Northeast 8 -  3  5 -
Implementation specialist—National 5 -  2  3 -
Real-life example
Let’s look at a real-life application Implementation specialist—Southeast 7 -  3  4 -

of the 10-step implementation ap- Implementation specialist—West 6 -  2  4 -

proach and how it delivered effi- Implementation specialist—Great Lakes 7 -  3  4 -
ciencies for an insurance organiza- Implementation specialist—Other 13 -  5  8 -
tion striving to service its customers Totals 213 104 56 32 21
more effectively. SWAT = specialized work action team
Establishing a goal—Histori-
cally, the organization’s customer
service was below industry aver-
Percentage of future work / TABLE 3
age, and this was characterized by Future roles
the presence of incorrect customer Sales Case Work
information, long lead times for Current roles coordinator manager distribution lead
case implementation and convo- Sales coordinator 90% 10% 0%
luted processes. The organization, Account coordinator 100% 0% 0%
therefore, set a strategic target to SWAT coordinators 100% 0% 0%
improve customer experience and Regional enrollment manager 0% 100% 0%
decrease its rework rate of 42% to
Enrollment case manager 0% 100% 0%
match a 25% increase in sales for
Enrollment manager 0% 100% 0%
employee-paid insurance products.
Regional service specialist 25% 10% 10%
To reach this goal, the organiza-
Implementation specialist—Northeast 0% 40% 60%
tion focused on reducing customer
touchpoints, steering employees to Implementation specialist—National 0% 40% 60%

the right work, and driving service Implementation specialist—Southeast 0% 40% 60%
levels and turnaround times to in- Implementation specialist—West 0% 40% 60%
dustry standards. Implementation specialist—Great Lakes 0% 40% 60%
Pod kickoff—Leadership creat- Implementation specialist—Other 0% 40% 60%
ed an initiative framework divided SWAT = specialized work action team
into three tracks that would reduce
customer touch points, shared services and functional customer and organized the customer delivery by pro-
initiatives. The team focused on reducing customer cess rather than function.
touch points. To do this, it supported and scoped an To understand the intricacies of the process and
operational process redesign that created value for the create a foundation for how it would continue in the

February 2017 • QP 21
 SERVICE QUALITY
Managers had the opportunity to discuss the staffing
model that caused the pitfall of ‘relying too
much on the staffing model’ to surface.
future state, the initiative began with a facilitated,
cross-functional team event at the operations site. Af-
ter the meeting, a project plan was created and design
sessions were conducted with cross-functional teams
to align new roles, processes, handoffs and logistics
for the pods. Figure 2 (p. 20) illustrates the vision for
a new process that co-located key functions to work
together in one seamless, customer-centric process.
An all too common pitfall—The team agreed on
the pods’ foundational components and structures at
the design sessions. Confidence in the new process and
new roles, such as a case manager and sales coordina-
tors for work distribution, immediately led to a com-
mon pitfall: underestimating managers’ resistance.
With the new process, 213 staff members, ranging
from implementation specialists to enrollment manag-
ers, transitioned to roles that required new skill sets.
In turn, managers realized they would have different
teams (see Tables 2 and 3, p. 21). Their initial reactions
were reluctance and resistance.
Escaping the pitfall—The team immediately
called the organization’s training department to de-
sign a plan that detailed how and when pod members
would be trained and what they’d be trained on. A staff-
ing model was developed to provide a forward-look-
ing view of the staffing allocation in the pod process,
which was based on the assumption that the organiza-
tion’s current talent pool could perform the proposed
roles and responsibilities.
During the design session, managers had the op-
portunity to discuss the staffing model that caused the
pitfall of “relying too much on the staffing model” to
surface. We alleviated the tension after recommending
we test the model’s assumptions using a pilot.
Leadership and the design team agreed to a mul-
tiregional pilot of the pods. To identify improvement
opportunities, the pods were implemented in key re-
gional sales and home offices to collect data and test
the concept. The insights gathered from the pilot gen-
erated process enhancements that ensured success for
22 QP • www.qualityprogress.com
an organizationwide rollout.
Success and scalability—The pilot pods’ success
led to a decrease in the rate of rework from 42 to 33%
and led to a full pod rollout across all regional sales
offices. The takeaways from the pilot and the scalabil-
ity of the pod process were leveraged to ensure each
office became a center of excellence for a new and im-
proved customer experience.
Centers of excellence
Six steps that helped ensure future success of the cen-
ters of excellence in our real-life example included:
1. Providing start-up support for the new process.
2. Obtaining the right resources and placing the right
people in the right roles.
3. Appointing a pod leader with cross-functional
knowledge.
4. Managing change proactively.
5. Tracking the right metrics to measure success.
6. Rotating pod roles to facilitate cross-training and job
enrichment.
Implementing pods can help increase customer sat-
isfaction, reduce cycle times, achieve higher levels of
quality, and increase flexibility and scalability. More-
over, adopting pods results in higher levels of employ-
ee engagement because employees see the end-to-end
process, receive direct feedback from the customer
and gain autonomy. QP
RYAN MARCUS is a consultant at Valeocon Manage-
ment Consulting in New York City. He earned a
bachelor’s degree in economics from Wheaton College
in Norton, MA.
MICHELLE HARRIS is assistant vice president, cus-
tomer delivery, at Lincoln Financial Group in Omaha,
NE. She earned a bachelor’s degree in business ad-
ministration from the University of Nebraska in Lincoln.
 CHECK OUT THE NEW
BOOKS AVAILABLE FROM
ASQ QUALITY PRESS!
THIS MONTH’S PICK!
The Certified Supplier
Quality Professional
Handbook
The importance of a safe, reliable,
cost-effective, high-quality, and
integral supply chain cannot be
overstated. ASQ’s new Certified
Supplier Quality Professional
(CSQP) certification provides
valuable credentials to quality
professionals in the growing field
of supplier quality engineering.
Item: H1520
Learn more about these books through the
Quality Press bookstore at asq.org/quality-press.
The Certified Six Sigma Black Belt
Handbook, Third Edition
The best Six Sigma Black Belt handbook has
been fully revised, updated, and expanded!
Additions include more exercises, new tables
and figures, new content in the DMAIC parts
of the book, and more.
Item: H1494
We Move Our Own Cheese!:
A Business Fable About
Championing Change
This book is about creating change. The
cheese is a metaphor for what we have in life
and what we believe we want more of. In a
business context, it represents the business we
are in and what it gives us.
Item: H1523

In 50 Words
Or Less
• An emphasis on Factors that lead to
success or failure
“systemness”—a well-
functioning whole rather
than well-functioning
parts—can be instrumen-
tal in controlling costs in
healthcare settings.
in healthcare
• A Virginia hospital’s
handling of several Six
Sigma projects illustrates
projects
how improvement teams
need proper training to
ensure right tools are
used to improve the
probability of project suc-
cess outcomes. by Todd Creasy
 CASE STUDY

HEALTHCARE DELIVERY
has changed dramatically over the years: from simple
house calls to health maintenance organizations, point
of service plans, preferred-provider organizations and Affordable
Care Act exchanges. Not surprisingly, the common tread in the
most recent delivery systems is cost containment.
Controlling costs, in fact, has become the priority for health-
care executives. “The challenge for industry executives is to
continue to control spending even in the face of countervailing
winds such as expensive new innovations, improved consumer
confidence, and an aging society that requires more medical care
and services,” according to the 2015 PricewaterhouseCoopers
(PWC) Health Research Institute report.1

February 2017 • QP 25
 Authors of the report also suggest the need for
“systemness.” When referring to the many services
offered by a typical hospital, this new buzzword em-
phasizes the need for a well-functioning whole rather
than well-functioning disparate parts. Care teams
should strive to achieve more by working together,
eliminating redundancies and reinforcing common
goals through clinical and administrative standard-
ization, which should lead to lower operating costs,
PWC suggests. Healthcare organizations and hos-
pitals are no strangers to the need to standardize,
economize and seek to do more with less.2-4 For sev-
eral years, healthcare projects have been initiated to
reduce costs, increase quality of care and improve
patient, clinical staff and employee satisfaction.
Despite these attempts to improve the effective-
ness and efficiency of healthcare delivery systems,
however, many projects fail to deliver on the oppor-
tunity. Several reports support how project opportu-
nities often fall short of expectations:
• A McKinsey Quarterly report in 1994 examined
100 organizations that had engaged in significant
projects and how the organizations’ executives
believed too much had been invested with too few
bottom-line results.5
• A 1994 report cited in ComputerWorld said nearly
85% of top executives were dissatisfied with the
results of process re-engineering.6
• A Minitab survey of 200 Six Sigma professionals
looked to pinpoint the main reasons projects fail.
The results were narrowed into 41 failure themes,
which were then reduced to the top eight rea-
sons.7
• Multiple studies report that up to 70% of projects
fail. For total quality management projects, the
failure rate ranges from 60 to 80%.8
Looking to reduce waste, standardize practices,
improve quality of care, and advance clinical or staff
productivity, many hospital CEOs have turned to
process improvement or re-engineering to achieve
operational goals. Many of these efforts have paid
dividends, but others have not.
Considering the results of the PWC report and
its call for systemness, this article seeks to identify
two dynamics while showing why some healthcare
projects fail and others succeed. It is hoped that les-
sons learned can help reduce failure percentages and
boosting healthcare project success rates.
Case and method
A study was conducted at a 200-bed rural Virginia
hospital that had engaged in Six Sigma projects
for about four years. The hospital’s investment in-
cluded lean Six Sigma (LSS) White Belt training for
all senior managerial staff, Green Belt (GB) training
for about 25 personnel and Black Belt training for
one senior manager.
During the four-year span, Six Sigma projects
were classified as either:
• An unqualified success—a project such as im-
proving a med-cart exchange between nursing
and pharmacy, the time to get obstetrics supplies,

Survey for successful and moderately successful

projects / TABLE 1
Options when explaining successful or moderately successful project outcomes
  1. Size of the project (sized appropriately—not too large).
  2. Senior management support and prioritization.
  3. Process owner involvement and commitment.
  4. Project leader (provided good leadership).
  5. Process complexity minimal (perhaps had low-hanging fruit).
  6. Project goals (properly defined and clear).
  7. Spirit of collaboration between department heads or within department.
  8. Good cohesion among project team.
  9. Measuring system was good (able to collect and report valid data).
10. Good control system for process.

26 QP • www.qualityprogress.com

Healthcare organizations and hospitals
are no strangers to the need to
standardize, economize and seek
to do more with less.
patient meal tray preparation or preadmission test-
ing, met the project charter’s stated goal.
• A moderate success—a project, such as improv-
ing patient central scheduling for outpatient sur-
gery, emergency room (ER) wait time, 5S in supply
closets, dirty linen processing or patient snack or-
dering, had some success but could have achieved
more.
• Unsuccessful or disappointment—a project such
as 5S on a particular floor, operating room sched-
uling, lab testing quality and turnaround time, or
collaboration between first and second-shift house-
keeping, failed to meet the charter goal.
The research method involved conducting half-day,
group interviews with several GBs with different ex-
perience levels and healthcare positions. Queries were
made about hospital projects the GBs knew about,
either through personal involvement (as a leader or
team member) or second-hand information. Projects
with outcomes that could be classified into one of the
success categories were selected by senior manage-
ment, department heads, and LSS practitioners and
leaders.
During the meeting and for each project under re-
view, the particular project was named and leaders
chosen. The goals and the department on which the
project would be focused also were named. To ensure
valid responses by meeting attendees, each participant
was given an anonymous survey at the beginning of
each project’s dialogue.
Table 1 shows the survey instrument for successful
and moderately successful projects. Table 2 shows a
similar survey instrument for projects considered dis-
appointments or failures.
Using these surveys before each project’s discus-
sion, the meeting attendees were asked to identify the
top three reasons that would explain the improvement
projects’ success, moderate success or failure. After
CASE STUDY
votes were collected, conversations offered additional
insight into each project’s success level.
Link to leadership
Depending on the type of project considered (suc-
cessful or moderately successful), there were 10 pos-
sible survey responses for each project. When consid-
ering the projects categorized as a failure, there were
11 possible survey responses. In Table 3 (p. 28), the
top reasons (reflecting a scoring tie) for each project
outcome category are listed in descending order of
importance.
In this table, project success and failure anchor op-
posite ends of the continuum with moderate success in
the middle. Initially focusing on the anchor points and
examining the reasons for success and failure, both
seem to be strongly related to leadership.
When considering the successful projects, project
leadership, project team cohesion—which can be di-
Survey for unsuccessful or failed
projects / TABLE 2
Options when explaining failed or unsuccessful project outcomes
  1. Size of the project (too large, not manageable).
  2. Low senior management support and prioritization.
  3. Little to no involvement of process owner.
  4. Project leader (Green Belt) skills and interest lacking.
  5. Overly complex process.
  6. Lack of clearly defined project goals.
  7. Little or no collaboration between departments or within department.
  8. Continually changing project goals.
  9. Low team cohesion.
10. Poor measuring system.
11. Little or no control mechanism to keep process performing well.
February 2017 • QP 27
 rectly influenced by the project or departmental lead-
er—and process owner involvement all exhibit some
form of leadership. A 2002 survey of 126 organizations
showed top management sponsorship and commit-
ment were the most important contributors to any suc-
cessful change initiative. 9
Within a healthcare environment, process owners
are usually clinicians: for example, nurse managers,
lab supervisors, physical therapy department leaders,
registered dietitians, doctors and surgeons, and ER
directors. Without the involvement and support of
these employees, success at the clinical level will be
difficult.
In terms of leadership’s importance related to
project failure, similar results were found. Project
management, process owner involvement and col-
laboration among departments—which is directly in-
fluenced by that department’s manager—each relate
directly to leadership. The absence of leadership can
have a profound negative effect on projects, accord-
ing to recent research:
• When examining the top eight reasons projects fail,
the Minitab survey of 200 Six Sigma professionals
noted the top failure mode, not surprisingly, was the
lack of management support.10
• “Little to no support from the leadership … imple-
menting the initiative” is one of the three leading
reasons for quality initiative failures, wrote authors
Jamison Kovach and Jerry Mairani.11
• The lack of executive leadership, support and par-
ticipation is a big influence when projects fail, wrote
author Martin Smith.12
Based on this sample’s collective opinion and in
support of previous research, the balance between
project success and failure depends heavily on the
department leader and the project manager (in this
case, a GB). As such, it is imperative the process
owner—most likely a clinician—and the Six Sigma
project leader be held responsible for the improve-
ment initiative.
Interestingly, for moderately successful projects—
those that could have attained more but didn’t—notice
that no forms of leadership were selected to justify
that ranking. This section is classified as “project dy-
namics,” which could mean that even with lesser forms
of leadership, if a project team has grounded itself se-
curely with several fundamental tenets, there is a good
change of some project success:
• Properly defined goals was one of those funda-
mentals and seems to support Smith’s research that
noted “the most important contributors to re-engi-
neering success were a clear mission statement and
understanding of the business re-engineering defini-
tion.”13
• A good measuring system was the second rea-
son given for a moderately successful project. This
factor is supported by research that noted an ab-
sence of data was the No. 4 reason for a project to
fail.14
• The size of the project was ranked high to ensure
moderate project success. Selecting a manageable-
sized project was supported by Louis Johnson, who
noted large project size as being the No. 5 reason
projects fail,15 and Smith, who wrote that projects
“kept small and manageable” had a greater chance
of success.16

Top 4 reasons for project outcomes / TABLE 3

Rank Successful projects Moderately successful projects Failed projects

Effective project leader
1
(Green Belt)
2 Cohesion among
project team
3 Measuring system good
(able to collect valid data)
4 Process owner
involvement/commitment
(Project leadership)
Properly defined and clear project
goals
Measuring system good
(able to collect valid data)
Project sized appropriately—not
too large
Process complexity was low (had
low-hanging fruit)
(Project dynamics)
Little to no involvement by
process owner
Little or no collaboration
between departments or within
department
Overly complex process
Ineffective project leader
(Green Belt)
(Project leadership)

28 QP • www.qualityprogress.com

If there is no data capturing system,
Six Sigma is not an option.
Finally, to achieve some level of success, the re-
spondents indicated low project complexity and get-
ting some early wins (via low-hanging fruit activities)
were necessary. Most Six Sigma practitioners and
coaches advise that early wins can sustain and fuel the
process improvement initiative. It is possible that with
enhanced leadership at departmental or project lev-
els, these moderately successful projects could have
moved to the successful column.
What have we learned?
Leadership at the project level is critical. It is impera-
tive that your LSS belts are well trained and capable.
Cohesion among project team members may depend
on the ability of the project leader.
Leadership at the department level—or process-
owner level—is even more critical. Without their buy-in
and full cooperation, even the best-trained LSS practi-
tioner will struggle. Additionally, if cross-departmental
collaboration is essential for project success, its pos-
sibility will be dampened by disinterested department
managers.
Perhaps most importantly, the basic tenets of a well-
defined, adequately sized and manageable project that
offers less complexity and some early-win opportuni-
ties with proper data collection mechanisms may af-
ford some level of success—even with a nonparticipa-
tory department leader.
“Lean is inductive; Six Sigma is deductive,” wrote
Neil Nilakantasrinivasan and Arun Nair.17 Considering
Six Sigma’s data requirements, a poor measuring sys-
tem in terms of frequency and validity is a significant
handicap to a LSS practitioner. If there is no data cap-
turing system, Six Sigma is not an option.
The right training
With the dynamic nature of healthcare in the 21st cen-
tury and the need to improve patient outcomes while at
the same time reducing costs and improving the finan-
cial strength of our healthcare institutions, the need
for process improvement in healthcare is not going to
diminish in the near future. Healthcare practitioners,
CASE STUDY
clinicians, consultants and various stakeholders engag-
ing in continuous improvement must pay careful atten-
tion to their improvement teams and those tasked with
working with those teams.
Improvement teams should get proper training to
ensure the necessary tools are offered to improve the
probability of project success outcomes.
Based on the sample data presented in this article,
these factors should improve healthcare patient and
process initiatives, resulting in fewer project failures
and more project success, thus improving the likeli-
hood of producing healthcare systemness. QP
REFERENCES
1. PricewaterhouseCoopers Health Research Institute, “Medical Cost Trend:
Behind the Numbers 2015,” June 2014, http://tinyurl.com/pwc-report-
behind-numbers.
2. Todd Creasy and Sarah Ramey, “Don’t Lose Patients,” Quality Progress,
February 2013, pp. 42-47.
3. Nina Braswell, Jill Koroseta and Amy Grogan, “Lessons in Labeling,” Quality
Progress, February 2013, pp. 29-35.
4. Jim and Mary Beth Buckman, “Improving on Excellence,” Quality Progress,
July 2012, pp. 34-41.
5. Martin Smith, “Business Process Design: Correlates of Success and Failure,”
Quality Management Journal, 2003, Vol. 10, No. 2, pp. 38-49.
6. Ibid.
7. Louis A. Johnson, “Falling Short: Survey Ranks Top Eight Reasons Why Six
Sigma Projects Fail,” Six Sigma Forum Magazine, August 2009, Vol. 8, No. 3.
pp. 19-22.
8. Jamison V. Kovach and Jerry Mairani, “Exploring Quality Initiatives’ Success
and Failure,” Journal for Quality and Participation, 2012, Vol. 45, No. 10. pp.
24-28.
9. Smith, “Business Process Design: Correlates of Success and Failure,” see
reference 5, p. 41.
10. Johnson, “Falling Short: Survey Ranks Top Eight Reasons Why Six Sigma
Projects Fail,” see reference 7, p. 22.
11. Kovach, “Exploring Quality Initiatives’ Success and Failure,” see reference
8, p. 24.
12. Smith, “Business Process Design: Correlates of Success and Failure,” see
reference 5, p. 41.
13. Ibid.
14. Johnson, “Falling Short: Survey Ranks Top Eight Reasons Why Six Sigma
Projects Fail,” see reference 7, p. 22.
15. Ibid.
16. Smith, “Business Process Design: Correlates of Success and Failure,” see
reference 5.
17. Neil Nilakantasrinivasan and Arun Nair, “DMAIC Failure Modes: Don’t
Waste DMAIC’s Usefulness on the Wrong Problems,” Six Sigma Forum
Magazine, November 2005, Vol. 4, No. 3, pp. 30-34.
TODD CREASY is an associate professor at Western
Carolina University in Cullowhee, NC, and principal at
Process Serum in Nashville, TN. He holds a doctorate
in business management from Case Western Reserve
University in Cleveland. An ASQ member, Creasy is an
ASQ-certified Six Sigma Black Belt.
February 2017 • QP 29
Good
Shepherds

by Kathryn D. McIver
 PROJECT MANAGEMENT

Applying tragedy of the

commons theory for more
effective resource management

GAME THEORY IS the use of mathematical In 50 Words

models to show conflict and collaboration with regard to Or Less
• Organizations can apply
strategic decision making. Within game theory, the tragedy tragedy of the commons
(TOTC) theory—which
of the commons (TOTC) is a scenario that can be directly focuses on how public
resources are overused
applied to resource management and project-based work. to the detriment of
society—to more effec-
TOTC is considered the foremost argument for natural tively manage staff and
resource management and has been the starting point for projects.
• TOTC can create a pull
countless discussions about the effect of development on system that prevents
employees from becom-
natural resources.1 TOTC occurs if a common resource ing overworked, giving
them assignments
(in the case of project-based work, an organizational re- when they’re available
source) is exploited beyond the socially effective level.2 to complete them.

February 2017 • QP 31
 For this article, the only resource being discussed pasture for grazing their sheep. For the good of the
is employees who are working normal workweeks. community, the farmers have a set number of sheep
Resource management—such as the availability of they are allowed to graze in this pasture. This set
team members—is a crucial factor a lean practitioner number of sheep is the output of a formula the town
or project manager must consider when reviewing a elders used to determine exactly how many sheep
current-state process or planning an upcoming proj- the town can put in that pasture and still allow the
ect. The core principle of resource management is to pasture to remain usable year after year.
maximize outputs while minimizing inputs, which are A farmer decides he wants to add one more sheep
complementary goals to reducing waste in process im- because he thinks “just one” won’t hurt and, after all, he
provement efforts. needs the extra wool for his oldest son’s college fund.
Similar to an improvement project in which each Of course, the problem with this mindset is that every
process element contributes to outcomes, an organi- farmer gets the same idea, and they all add one more
zation’s resources contribute to organizational perfor- sheep. After every farmer adds a sheep, the pasture is
mance and must be assessed as assets and resources quickly overgrazed and cannot support sheep, which
that can be exploited. The element that’s missing from harms the village’s income.
this view of resource management, however, is the ap- This example shows the overexploitation of a re-
plication of lean manufacturing’s pull principles. Pull source—the pasture. The rationale behind the TOTC
is a cornerstone of process improvement and stream- is simple: If a resource is commonly owned, it will be
lined operations in which a need is filled in a just-in- overused. A more tangible and relevant example of the
time manner, rather than at the pace of an upstream TOTC is internet usage.
factor. If you think of employees as resources, pull When internet providers’ pricing structures switched
means they receive tasks when they’re available to from consumption-driven or time-based pricing to a
complete them, rather than having the work pushed flat rate, internet use spiked and caused more users to
or assigned independent of their current workload or download at all times, which slowed everyone’s down-
capacity. load speeds.
The TOTC is most frequently used to model or ex-
plain the effects of consumption and resources on a Applying TOTC in organizations
global scale. It’s considered highly applicable to the Applying this analogy to an organization and project-
study of sustainability and environmental resource level work requires viewing the organizational re-
management. source as a pasture, departments as the farmers, an
organization’s product as the wool and projects as
TOTC and counting sheep sheep that draw on the organization’s resources. These
The name “tragedy of the commons” goes back to resources can include budgets, equipment, human cap-
medieval times when communities had common land ital or any other resource that’s used in the course of
for the villagers to raise their livestock. Sheep farm- producing a deliverable.
ers, therefore, provide a classic example of TOTC: TOTC depends on the number of consumers or proj-
Imagine a village in which all farmers use a common ects (sheep), as well as available resources (pasture)
and production requirements (mak-
ing the wool). Dependencies on
Organizational project roadmap the number of projects linked to a
example / TABLE 1 resource can be seen at a high level
in portfolio management but may
not be clear at the project level. The
Resource Project Timeline Level of effort
number of projects that draw from
Bob Standard work March 2014 to July 2014 10 hours/weekly a specific resource is equivalent to
Bob Streamlining April 2014 to July 2014 10 hours/weekly the number of sheep in that pasture.
Tim Standard work March 2014 to July 2014 5 hours/weekly The overuse of a specific resource
Anne Standard work March 2014 to July 2014 5 hours/weekly is the scenario that must be consid-

32 QP • www.qualityprogress.com

Waste is one of the most difficult
things for organizations to identify
and isolate.
ered when identifying TOTC in an organization.
The medieval solution for TOTC was to divide and
parcel land, making farmers responsible for their ac-
tions. The decision to decentralize an organization,
however, requires input and insight from several dif-
ferent sources, and medieval economics probably
shouldn’t be the basis for that decision.
Solving TOTC from the organization’s pasture (resources).
The generally accepted solution for TOTC is to
regulate access to a resource or—as feudal lords
did—privatize it. Unfortunately, neither of these are
good solutions for TOTC in the workplace. Employ-
ees work 40 hours a week and tend to use a first-
come, first-served approach in doing work for their
organizations. Privatization could cause a vertical
silo mentality that most quality professionals are
against.
A good starting point for minimizing TOTC in the
workplace is to have a central repository of resource
requests, or in the case of project-level resources, a
reference list for which resources are being tapped
for each project.
Duplication or excessive use indicates an exploita-
tion of the resource, which can lead to failing work
performance in other areas, biased project feedback
(getting input from only one source) or team member
burnout. To supplement this, an organization must
have a listing of available resources to draw on and
balance the lean project workload.
Systemically, organizations that have TOTC sce-
narios are more likely to experience delays in project
completion (competing resources) or changeover of
necessary resources (either through competing pri-
orities or employee turnover). Neither of these symp-
toms benefits an organization’s long-term goals, and
they will slowly erode the success of projects.
Another key to minimizing TOTC with organiza-
tional resources is having an organizational project
roadmap that outlines resources, their level of effort
PROJECT MANAGEMENT
(time spent) on each project and their cumulative ef-
forts for any point in time (see Table 1).
Table 1 shows Bob is the overexploited resource,
and giving him additional project work would cause a
TOTC—if he’s not already there—and would be detri-
mental to the outcomes of all of the projects to which
he is assigned. The roadmap provides a high-level
snapshot of how many sheep (projects) are drawing
As with all game theory scenarios, the most nega-
tive outcome affects the parties or departments
that choose the more selfish or individually minded
course of action rather than the systemic course of
action. In this case, the most selfish outcome would
be for one of the projects to require more of Bob’s
time, which would force him to decrease his effort on
other projects.
To integrate pull and use a project roadmap, em-
ployees would volunteer for projects based on their
personal availability. This creates a pull system for
team management and diversifies project teams,
which will help disseminate this culture throughout
an organization.
Diversifying team members throughout organi-
zational projects also could decrease the hidden
waste of intellect. Waste is one of the most diffi-
cult things for organizations to identify and isolate.
Intellect waste typically is measured and seen on
employee engagement surveys. One of the battle
cries against wasted intellect is that by respecting
employees and their feedback, there is no waste.
Something missing from this, however, is the devel-
opment of staff talent.
Decreasing intellect waste also is challenging be-
cause culture often factors into why or how waste
began. Having employees pull assignments and
opening projects to a wider group of resources is a
passive way of resolving waste of intellect and en-
couraging independent staff development. Combat-
ing the waste of intellect isn’t just a matter of using
February 2017 • QP 33
 PROJECT MANAGEMENT
While TOTC is a thought experiment
that helps explain how people use
common resources, it has a practical
application in business for evaluating
resource management.
employees’ latent talents, however. It also helps de-
velop staff to be more effective resources for an orga-
nization’s projects.
Case study
A real-life case study of TOTC occurred during an or-
ganization’s annual hoshin kanri3 strategic planning
session. After the annual, top-level priorities were
identified and the second-level initiatives were named,
the matter of assigning owners became the issue at the
forefront. Rather than playing a game of hot potato—in
which everything is a great idea as long as someone
else runs it—leaders were anxious to have influence
over every project. Many top-level leaders signed up
to sponsor and champion three or even five initiatives,
including projects below those high-level initiatives.
After projects gained momentum, a key barrier—in
addition to the expected resource constraints—was
executive sponsors’ lack of availability. This caused
delays and unanticipated roadblocks that decreased
projects’ impacts and affected their timelines. Even-
tually, 75% of that year’s strategic projects were on
hold or discontinued entirely due to resources’ limited
availability starting at the executive-sponsorship level.
This organization learned TOTC is real. Instead of
having razor-sharp focus on resource availability, ex-
ecutive leadership was overexploited and unable to
sponsor the projects with the passion needed to be ef-
fective.
SHARE YOUR THOUGHTS
Have comments or opinions about this article? Let’s hear it! Post your
remarks on this article’s webpage at www.qualityprogress.com, and it
could appear in the next Seen&Heard.
34 QP • www.qualityprogress.com
Practical application
While TOTC is a thought experiment that helps explain
how people use common resources, it has a practical
application in business for evaluating resource man-
agement. The crucial takeaway from these scenarios
is that overusing a single resource serves short-term
rather than long-term goals.
Consider TOTC a cautionary tale that illustrates a
need to create a framework for a pull system for proj-
ect team members that can help prevent overexploita-
tion. The theory can be applied to any resource and
certainly is something all managers and leaders should
be aware of when evaluating finance or production ca-
pability. Decreasing resource exploitation can result in
better quality work from project team members and a
greater project success rates. QP
REFERENCES AND NOTE
1. Garrett Hardin, “Tragedy of the Commons,” Econlib.org, http://tinyurl.com/
totc-qp.
2. Ibid.
3. Hoshin kanri involves a selection of goals, projects to achieve the goals,
designation of people and resources for project completion and establish-
ment of project metrics. For more information, visit ASQ’s quality glossary
at asq.org/glossary/h.html.
BIBLIOGRAPHY
Brogan, Jesse W., “Improving Lean Six Sigma Process With Lean Six Sigma,”
isixsigma.com, http://tinyurl.com/brogan-qp.
Davis, Matthew M., “Right, Privilege—Or Tragedy of the Commons,” Robert
Wood Johnson Foundation, April 13, 2013, http://tinyurl.com/davis-qp.
Ilkılıç, Rahmi, “Network of Common Property Resources,” Economic Theory,
Vol. 47, No. 1, 2011, pp. 105-134.
Miller, James, Game Theory at Work, McGraw-Hill Education, 2003.
KATHRYN D. MCIVER is the senior director of clinical
operations for Professional Case Management in
Denver. She holds a bachelor’s degree in science
and technology from the University of Denver. McIver
is a certified lean Six Sigma Master Black Belt from
Villanova University and past chair of the ASQ Denver
Section.
2017 LEAN AND SIX SIGMA CONFERENCE
February 27 – 28, 2017 | Phoenix, AZ

All Signs Point to Phoenix!

Excellence is waiting for you at ASQ’s 2017 Lean and Six Sigma Conference in
sunny Phoenix, AZ!

Join hundreds of other global quality and process improvement professionals like
you for two exciting days as we focus on:

Leading the Culture of Excellence Through Lean and Six Sigma

Learn how to help create and support an organization-wide culture in which
excellence is an everyday practice, and demonstrate how quality is more than just
a role, department, initiative, or project. Hot topics you don’t want to miss include:

• Lean and Six Sigma Fundamentals

• Lessons Learned: Implementation of Lean and Six Sigma
• Doing More With Less
• Tips and Tricks: Sustaining Results
• Masters Series

Visit asq.org/lssc-qp to learn more and register now!

SMALL CHANGES
Big Results
Design and tune
HASS effectively to
detect manufacturing
defects in electronics
by Daniel K. Sarosky
 CASE STUDY

THE HIGHLY ACCELERATED stress screen (HASS) is a stan-

dard industry process used to precipitate and detect a product’s early-life defects. HASS
is not designed, however, to capture all defects. In fact, a well-constructed HASS profile
has a limited probability, varying by failure mode in its ability to
capture product defects.1 Increased stresses caused when trying In 50 Words
to precipitate harder-to-reach failure modes can lead to damage. Or Less
• The highly accelerated
There are five steps to take to effectively develop and tune stress screen (HASS) can
help assess a product’s
a HASS profile. This approach can be applied in other product reliability and reduce or
sectors, although the stresses and HASS modifications may need eliminate product fail-
ures.
customized approaches depending on the base technology. • HASS involves creating a
product profile and put-
To determine whether HASS tuning should be applied, an as- ting varying stresses on
sessment between field data and production data must be per- the product to determine
whether to make adjust-
formed. The results will show the efficiency of the current HASS ments.
• There are five steps to
profile. If upward trends in field failures attributed to manufac- effectively develop and
turing defects and production test yields are decreasing, HASS tune a product’s HASS
profile.
tuning should be considered.2

February 2017 • QP 37
 The design and development of HASS profiles are pri- ent air temperature of the environmental chamber up
marily based on the results of a highly accelerated life in 10º C increments. Dwell times at each temperature
test (HALT). There are several necessary steps that lead level can vary depending on the product, but a mini-
up to and include HASS tuning: mum of 10 minutes is recommended following ther-
mal stabilization (see Figure 1).
1. HALT 3. Rapid thermal transitions—The minimum thermal
To understand the various aspects of HASS tuning, it is cycle range is within 5º C of the lower and upper ther-
pertinent to review the basic construction of a typical mal operating limits discovered during the cold and
HASS profile, starting with a HALT.3 hot-step stress tests. A minimum of three cycles is
A HALT is a limit test, not a pass-fail test. The main recommended with minimum dwell times of 10 min-
objective of a HALT is to determine the product’s envi- utes at each thermal limit (see Figure 2).
ronmental (temperature and vibration) limits.4 Also, it’s 4. Vibration-step stress—Start at 5 root mean square
used to measure a product’s robustness. acceleration [g(RMS)] (lower vibration level—that
There are five tests to perform during a HALT that are is, 1 g(RMS) for delicate products and increment in 5
required to develop a HASS profile. The temperature and g(RMS) steps. The dwell time should be a minimum of
vibration levels and dwell times are suggested starting five minutes and can vary depending on the product’s
points and can be tailored to the product. The five tests mass. The ambient air in the chamber should be at
are: room temperature. Continue stepping the vibration to
1. Cold-step stress—Begin at 20º C and step the am- the point at which the product becomes out of specifi-
bient air temperature of the environmental chamber cation or reaches its destruction limit (see Figure 3).
down in 10º C decrements. Dwell times at each tem- 5. Combined environment—A minimum of five cycles
perature level can vary depending on the product, but is required unless a failure occurs prior to completing
a minimum of 10 minutes is recommended following the five cycles. The thermal cycle should be within 5º
thermal stabilization (see Figure 1). C of upper and lower thermal limits. The starting vi-
2. Hot-step stress—Begin at 30º C and step the ambi- bration level is determined by dividing the maximum

Temperature step stress test / FIGURE 1

200
Temperature
180

160

140

120

100

80
Temperature

60

40
Cold-step stress Hot-step stress
20

0
8:00:00
8:10:00
8:20:00
8:30:00
8:40:00
8:50:00
9:00:00
9:10:00
9:20:00
9:30:00
9:40:00
9:50:00
10:00:00
10:10:00
10:20:00
10:30:00
10:40:00
10:50:00
11:00:00
11:10:00
11:20:00
11:30:00
11:40:00
11:50:00
12:00:00
12:10:00
12:20:00
12:30:00
12:40:00
12:50:00
13:00:00
13:10:00
13:20:00
13:30:00

-20

-40

-60

-80

-100
Time

38 QP • www.qualityprogress.com
 CASE STUDY

vibration level reached during the vibration-step

stress by five (see Figure 4, p. 40). The vibration is
Rapid thermal transitions test / FIGURE 2
140
increased by the same number for each subsequent Temperature set
120
thermal cycle—that is, if the maximum vibration
100
during vibration-step stress was 50 g(RMS), begin
80
at 10 g(RMS).
60

Temperature
2. HASS 40

The results from the HALT are used to develop the 20

HASS profile. An expected starting point for minimum 0

12:25:00
12:27:30
12:30:00
12:32:30
12:35:00
12:37:30
12:40:00
12:42:30
12:45:00
12:47:30
12:50:00
12:52:30
12:55:00
12:57:30
13:00:00
13:02:30
13:05:00
13:07:30
12:10:00
13:12:30
13:15:00
and maximum temperatures would be 80% of the mini- -20
mum and maximum operating temperatures deter- -40
mined during the cold and hot-step stress tests.5 These -60
temperatures are the levels at which the unit under test -80
was fully functional during the HALT. Time
The maximum vibration level is typically 50% of the
upper destruct limit determined during a vibration step
Vibration-step stress test / FIGURE 3
stress test. Figure 5 (p. 41) illustrates a general HASS
60 60
profile shape. The square wave is the temperature, and Temperature
Vibration
the trapezoidal wave is the vibration. 50 50
The HASS profile in Figure 5 shows two HASS cy-
cles. For screening products, there is no set number of 40 40
Temperature

Vibration
cycles or industry standard to use.6 It’s dependent on
the type of product being manufactured and the vol- 30 30

ume of product being produced.

20 20
If production throughput is critical, which is the
case for most organizations, a minimum of two cycles
10 10
is suggested. If the goal is to capture more latent de-
fects, three to five cycles are recommended. The ul- 0 0
timate goal is to try to capture all the defects in the
13:30:10
13:35:10
13:40:10
13:45:10
13:50:10
13:55:10
14:00:10
14:05:10
14:10:10
14:15:10
14:20:10
14:25:10
14:30:10
14:35:10
14:40:10
14:45:10
14:50:10
14:55:10
15:00:10
15:05:10
15:10:10
first cycle of the HASS profile. If this is achieved, the
remaining cycles can be removed, thus reducing test Time
time and increasing throughput.
The HASS profile has two parts or stages (see Fig-
ure 6, p. 42). The first is the precipitation stage, which relevant to manufacturing defects.
attempts to bring out any latent manufacturing or com- Each failure would be seeded one at a time and be
ponent defects in a short period of time. The second is screened under the proposed HASS profile to deter-
the detection stage, which identifies the defects. Dur- mine the profile’s ability to detect the seed. If the num-
ing this stage, the product is functionally tested. Higher ber of seeded failures that are detected is low (fewer
stress levels are used in the precipitation stage than in than five out of 10), the profile will need to be modified
the detection stage. and testing repeated.7

3. Seeded failure testing 4. Safety of screen

To determine whether the proposed HASS profile is
effective in detecting manufacturing defects, proof of
screen, or seeded failure testing, is required. To help
determine what failures should be seeded, you should
look at their organization’s top five to 10 field failures
After a HASS profile is developed and proof of screen
is completed, you must ensure the profile is not too
stressful and won’t remove too much usefulness from
the product. Safety of screen (SOS) is a test in which a
product is repeatedly subjected to the proposed HASS

February 2017 • QP 39
 Combined environment stress test / FIGURE 4
140 60
Temperature
Vibration
120

50
100

80

40
60
Temperature

Vibration
40
30
20

0
15:10:10
15:20:10
15:30:10
15:40:10
15:50:10
16:00:10
16:10:10
16:20:10
16:30:10
16:40:10
16:50:10
17:00:10
17:10:10
17:20:10
17:30:10
17:40:10
17:50:10
18:00:10
18:10:10
18:20:10
18:30:10
18:40:10
18:50:10
19:00:10
19:10:10
19:20:10
19:30:10
19:40:10
19:50:10
20:00:10
-20

-20

-40
10

-60

-80 0
Time

profile for a predetermined number of cycles to ensure in stress levels are strategically applied to the base HASS
the profile is not too stressful. profile to increase detection and precipitation probabil-
A good starting point for the number of cycles is 40. If ity for a given failure mode.9 The stress level changes de-
time allows, a product can be exposed to as many cycles pend on the type of defects escaping into the field.
as it takes before the first failure occurs. Determining
the amount of useful life that is removed from a prod- Tuning example
uct after being exposed to the HASS profile is a simple, An electrical assembly with an inadequate or insufficient
straightforward calculation and expressed in a percent- solder joint on a resistor fails after only several hours of
age as:8 product operation out in the field (see Figure 7, p. 42). The
Useful life = (# of cycles in HASS profile / # cycles is product originally passed a HASS and all other produc-
SOS) * 100. tion testing without issues before being shipped to the
For example: customer. Examination of the failed board clearly shows
Useful life = (2 / 40) *100 = 5%. a manufacturing defect (missing solder), which is exactly
If the production HASS profile has two cycles and the type of anomaly that HASS is designed to capture.
there were 40 cycles performed during SOS, the amount Addressing this workmanship defect, a stress level
of useful life removed for a product therefore is no more test is performed on the HASS profile to initially consider
than 5%. modifying its vibration. When changing stress levels, it is
recommended that modifying the amplitude should be
5. HASS tuning considered before increasing the dwell time.10
Despite implementing a HALT/HASS program and fol- If just changing the amplitude doesn’t detect the de-
lowing all the standard guidelines, organizations still will fect, consider adjusting the dwell time. Changes in the
experience initial or early failures that escape into the vibration level should be made incrementally—for exam-
field, albeit a notable reduction in defects versus a prod- ple, 5 g(RMS) steps. After modifying the vibration level,
uct program lacking HASS. the defect (the poor solder joint) must be seeded in a
In HASS tuning and timing, slight changes or tweaks product and screened using the modified HASS profile.

40 QP • www.qualityprogress.com
 CASE STUDY

Standard highly accelerated stress

screen profile / FIGURE 5
120
Temperature setpoint
Vibration setpoint
100

80

60
Temperature/Vibration

40

20

0
17:47:51
17:49:31
17:51:11
17:52:51
17:54:31
17:56:11
17:57:51
17:59:31
18:01:11
18:02:51
18:04:31
18:06:11
18:07:51
18:09:31
18:11:11
18:12:51
18:14:31
18:16:11
18:17:51
18:19:31
18:21:11
18:22:51
18:24:31
18:26:11
18:27:51
18:29:31
18:31:11
18:32:51
18:34:31
18:36:11
18:37:51
18:39:31
18:41:11
18:42:51
18:44:31
18:46:11
18:47:51
-20

-40

-60
1 cycle 1 cycle

-80
Time

If the profile detects the defect, the profile was success- at an accelerated ramp rate (for example, 60º C/minutes)
fully tuned. If not, the stress level must be changed and during the HASS profile operation, it would create a low-
the process repeated (see Figure 7). cycle fatigue condition in the highly stressed area.
Increasing the vibration level may not be the solution When modifying the thermal limits, it’s important not
for detecting poor solder joints. This can be attributed to to exceed the upper and lower temperature operating
an insufficient mass of the leads and secondary attach- limits identified during HALT because the product po-
ment methods (for example, spring-force lead forming). tentially can be stressed to destructive-limit levels. Modi-
Low-level vibration excitation—commonly referred to as fying the temperature to near or at the upper operating
tickle vibration—has been proven to detect defects char- limit identified during HALT, for example, may weaken
acterized with intermittent connections (for example, electrical components and cause them to fail or wear out
mating direct-current power connectors).11 prematurely. Again, after the thermal changes are made,
Low-level random vibration is defined between 1 and seeded failure testing would need to be performed.
5 g(RMS). The unit g(RMS) is the square root of the area If changing only the vibration stress level or only the
under the acceleration spectral density curve in the fre- temperature stress level does not improve the profile’s
quency domain. The tickle vibration excitations should detectability, the temperature and vibration stresses may
be implemented in the detection stage of the HASS pro- need to be changed for the profile to detect the defects.
file, but also can be implemented in the precipitation After the tuned HASS profile has demonstrated it can
stage.12 The theory behind tickle vibration is that low vi- repeatedly detect the defect, SOS must be revisited—as
bration levels will ferret out certain defects, which would originally for the HASS profile construction—to ensure
otherwise go undetected at higher vibration levels. the improved profile does not remove useful life from the
If tuning the vibration stress levels does not improve product. SOS is the process of exposing the product to
defect detection, modifying the thermal stresses may ap- repeated HASS cycles.
ply. In the poor solder joint example, another possible
approach to improving detection would be to increase Improving reliability
the upper and lower-temperature limits of the HASS pro- HASS tuning is engineering craftsmanship. There are
file. This increases the temperature delta which, if used no specific industry guidelines documented on how to

February 2017 • QP 41
 CASE STUDY

Stages of highly accelerated stress HASS tuning

screen profile / FIGURE 6 flowchart / FIGURE 7
Temperature and vibration HASS profile (base)
125.0
Precipitation stage Detection stage Temp. setpoint
No
Temp. product
100.0
Temp. air
Defects
escaping to
75.0 Vib. setpoint field?
Vib. product
50.0 Yes

HASS tuning
25.0
X axis - grid lines
Y axis - grid lines Safety of screen
0.0

-25.0 Print Improved HASS profile

14 10:00 20:00 30:00 40:00 50:00 1:00:00 1:15:00

Temp. = temperature Vib. = vibration HASS = highly accelerated stress screen

tune HASS profiles. Success in HASS tuning will come
from experience and strategic test trials. It is the cor-
rect incremental iteration to address ineffective HASS
profiles. The tuning of HASS profiles is based on the ar-
chitecture and complexity of the electronic products.
Slight changes to the stress levels can increase the de-
tectability of a profile.
The product field performance is an excellent gauge
and a telltale sign that HASS profiles are either effec-
tive or need improvement. An acceptable failure per-
centage may vary based on the organization’s product
volume. A desired annualized failure rate for complex
products (greater than 1,000 electrical components) is
less than 2%.
HASS tuning does have its shortcomings, however,
such as introducing unwanted damage. If the profile’s
stress is increased too much, the added stress will
cause damage or even latent defects that will cause the
product to fail prematurely in the field, leading to cus-
tomer dissatisfaction.
All changes will need to be validated via proof of
screen, or seeded failure testing, and SOS. If the prod-
performed to ensure the changes to do not damage the
product or remove useful life (more than 3 to 5%) from
the product.
Improving the reliability of electronic products does
not end at the design phase. Environmental screening
plays an important role in keeping the product’s quality
and reliability at desirable levels. After being correctly
implemented, HALT/HASS programs will increase
product reliability. QP
REFERENCES
1. Gregg K. Hobbs, Accelerated Reliability Engineering: HALT and HASS, John
Wiley, 2000.
2. Harry W. McLean, HALT, HASS & HASA Explained: Accelerated Reliability
Techniques, ASQ Quality Press, 2000.
3. David Rahe, HALT Standard Document Revision 5.0, Professional Testing,
1999.
4. Wikipedia, “Random Vibration,” http://en.wikipedia.org/wiki/random_
vibration.
5. Ibid.
6. McLean, HALT, HASS & HASA Explained: Accelerated Reliability Techniques,
see reference 2.
7. Ibid.
8. Rahe, HALT Standard Document Revision 5.0, see reference 3.
9. Hobbs, Accelerated Reliability Engineering: HALT and HASS, see reference 1.
10. McLean, HALT, HASS & HASA Explained: Accelerated Reliability Tech-
niques, see reference 2.
11. Hobbs, Accelerated Reliability Engineering: HALT and HASS, see reference 1.
12. Ibid.

uct fails during these tests, the proposed changes must

be reevaluated.
DANIEL K. SAROSKY is a principal reliability engineer
Small changes, such as HASS tuning, can make big at MKS Instruments in Rochester, NY. He holds a
improvements in product quality and reliability and, bachelor’s degree in electrical engineering technol-
ogy from Rochester Institute of Technology in New
in turn, will keep your customers satisfied. Again, any York. An ASQ senior member, Sarosky is also an
ASQ-certified reliability engineer and software
changes (temperature, vibration, dwell times and ramp quality engineer.
rates) made to a HASS profile will need to have SOS

42 QP • www.qualityprogress.com
 8
Online, self-paced
training courses.

Accelerate your quality

tools know-how with
PDSA plus QToolsTM suite.

Watch a demo video

and find pricing at
asq.org/qualitytools.

SET YOURSELF APART

Give yourself a strong, competitive advantage with the PDSA plus QTools™
suite. Register for Plan-Do-Study-Act (PDSA) and any of the seven quality tools
for instant, on-demand access to these essential resources on your PC, laptop, or
tablet for an entire year. Each online course will help you gain proficiency and
confidence in applying quality tools to create solutions and drive improvements.
MEMBER DISCOUNTS AVAILABLE!

TRAINING CERTIFICATION CONFERENCES MEMBERSHIP PUBLICATIONS

 STATISTICS ROUNDTABLE
Is Heterogeneity Your Friend?
Using sensitivity analyses to design improved studies
IDENTIFYING, ELIMINATING or
controlling heterogeneity is a fundamental
principle in many statistical techniques.
Heterogeneity and its opposite, homoge-
neity, refer to how consistent or stable a
particular data set or variable relationship
are. Having statistical heterogeneity is
not a good or bad thing in and of itself for
the analysis; however, it’s useful to know
to design, choose and interpret statisti-
cal analyses. Indeed, the comparison of
heterogeneity often will be the outcome of
interest, especially in quality fields.
There are pros and cons to having
heterogeneity in a sample or analysis.
Statisticians are trained to analyze vari-
ability (analysis of variance), decompose
variability (variance components), control
variability (modeling confounders and
covariates) and to use other methods. In
general, controlling for or lowering hetero-
geneity can remove potential confounders
or noise to increase sensitivity for the
measured outcomes.
High heterogeneity, in contrast, is often
more realistic for modeling the messy
real world and may give better results or
identify subpopulations. In clinical trials of
new therapeutics, for example, homoge-
neous patient groups are sought to clearly
identify efficacy. In market research and
polling, however, heterogenous samples are
requested to get a full picture.
Traditionally, statisticians will aim for
homogeneity of controlled experiments
Summary statistics for
6 data sets / TABLE 1
X1 X4 Y1
Mean 9.00 9.00 7.50
SD 3.317 3.317 2.032
SD = standard deviation
44 QP • www.qualityprogress.com
BY I. ELAINE ALLEN AND
JULIA E. SEAMAN
because it allows for straightforward testing
and conclusions. While reducing variability
and aiming for homogeneity in testing is
important, recognizing and examining the
diversity should play a key role in any analy-
ses. The reputation of heterogeneity has
been increasing as evidenced in the meta-
analysis field’s phrase, “Yes, heterogeneity is
your friend,” commonly used when examin-
ing and performing meta-analyses. Includ-
ing and even inviting variability in analyses
ultimately can create stronger and more
representative models and conclusions.
Ignoring heterogeneity:
Anscombe’s quartet
Basic Anscombe data analysis:
Anscombe introduces six data sets with
two showing identical X summary statistics
and four showing similar Y summary statis-
tics. In examining these summary statistics,
we assume we understand quite a bit about
the distribution of the data set.
Table 1 gives the summary of four small
data sets with similar means and standard
deviations. The addition of the median
(Table 2) provides more information show-
ing some differences in the distributions
and differences from the mean values indi-
cating skewness in the data. Only after the
raw data are examined do the differences
become completely apparent, and when
plotted against Y1 – Y4 do the differences
in relationships between X and Y become
completely transparent (Figure 1).
Median for the
6 data sets / TABLE 2
Y2 Y3 Y4 X1
7.50 7.50 7.50 Median 9.00
2.032 2.030 2.031
Anscombe regression analysis:
Anscombe created this data set to illustrate
linear regression similarities with disparate
data sets. His famous set of linear regression
equations illustrates the perils of ignoring
the variability in the data. Creating almost
identical regression equations, Anscombe
shows that highly heterogeneous data can
deceive the researcher. The four regressions
shown in Table 3 are all almost identical to y
= 0.5x + 3 with a r-squared fit near 0.66.
Based on the equations alone, you
might assume the data came from the
same source on four different times and
represents a consistent output. Can you
reasonably assume, however, that the data
that generated these models are similar?
In fact, the data are quite heterogeneous as
shown in Figure 1.
Equation one is a typical scatter plot for
a linear regression, and equation two indi-
cates clear nonlinearity. Most interesting
are equations three and four because they
show data with clear outliers. Assuming
these are part of a clinical study, investigat-
ing these points can identify patients with
important characteristics, or simply inac-
curately entered data. An analysis strategy
to eliminate these outlier data points, and
analysis of equation two using a quadratic
function will give different results than the
initial four equations.
While Anscombe’s quartet1 is usually
used to represent the limitations of linear
regressions, the variability between and
X4 Y1 Y2 Y3 Y4
8.00 7.58 8.14 7.11 7.04
within the data sets are a good study for
heterogeneity. If equations one to three
one suggested thresh-
old). The FEM assumes
4 linear regression
measured a different variable for a given that all the studies have equations with
x1, this demonstrates heterogeneity in the a “true” effect size that Anscombe’s data / TABLE 3
relationships of the variables. Within each is identical and the only
Equation 1: Y1 = 3.00 + 0.500 X1
equation, there are examples of heteroge- variation in a study’s
Predictor Coef SE coef T P
neity. Equations one and two are examples results is sampling error.
Constant 3.000 1.125 2.67 0.026
of heterogeneity in the residuals to the The REM assumes each
X1 0.5001 0.1179 4.24 0.002
regression, known as heteroscedasticity; study provides informa-
S = 1.23660 R-sq = 66.7%
the pattern in the heteroscedasticity of tion about different effect
Equation 2: Y2 = 3.00 + 0.500 X1
equation two also would indicate the poor sizes and controls for this
Predictor Coef SE coef T P
fit for linear regression. between-study variability Constant 3.001 1.125 2.67 0.026
For equations three and four, the outlier in its summary values. X1 0.5000 0.1180 4.24 0.002
points may be interesting or mundane. It When reporting results S = 1.23721 R-sq = 66.6%
could simply (and often) be an entry mistake as an REM, researchers Equation 3: Y3 = 3.15 + 0.486 X1
or calculation error. If it is a trustworthy data can assume they have Predictor Coef SE coef T P
point, however, it is worthwhile to investi- controlled for heteroge- Constant 3.152 1.136 2.77 0.022
gate. These points may hint at underlying neity between studies X1 0.4861 0.1191 4.08 0.003
subpopulations that have a distinctly differ- but have not explained S = 1.24948 R-sq = 64.9%
Equation 4: Y4 = 3.07 + 0.496 X2
ent response (heterogeneity in the sample or why the studies vary. It is
Predictor Coef SE coef T P
independent variables), or the measured out- even possible that when
Constant 3.071 1.082 2.84 0.019
come is distinctly different at that measure- switching from a FEM to
X2 0.4963 0.1134 4.38 0.002
ment (heterogeneity in the function, output a REM, the results will
S = 1.18948 R-sq = 68%
or dependent variable), or both. switch conclusions from
Coef = coefficient SE = standard error
While it is often frustrating to find these favoring one treatment to P = p-value T = test statistic
issues in an analysis, heterogeneity can favoring another.
lead to new questions and conclusions In a meta-analysis comparing the For this meta-analysis, several covariates
that create better models and discoveries. persistence of asthma over and under age were available: year of the study, location of
For multivariate linear models, diagnostic 12, seven studies were identified and a the study, percentage of females in the study,
statistics such as Cook’s D, leverage and risk difference (RD) (> age 12 - <= age 12) lost to follow-up study and year of asthma
residual analyses can point to issues with was synthesized over these studies. The onset. Because of the small number of stud-
the relationship between independent and results (see Figure 2) show that with the ies, each was examined separately. Initially,
dependent variables.2 FEM, the overall effect was significant (p < a jackknife procedure omitting individual
0.001) and was negative (significantly lower studies was performed, and the omission
Controlling heterogeneity: persistence over age 12). For the REM, the of one study (the only study from Iceland)
Using REMs in meta-analyses overall effect was inconclusive and not reduced the heterogeneity to 51%.
The majority of meta-analyses are conduct- significant (p = 0.513). Highly significant After further examination, this study
ed using either fixed effects models (FEM) heterogeneity was seen with the variation (with the largest positive effect size) had an
or random effects models (REM). The deci- in the RD attributable to heterogeneity = imbalance of females and had many more
sion to switch from FEM to REM is often 76% (Cochran’s Q statistic p < 0.0001). The patients lost to follow-up than the other six
based solely on the values of Cochran’s REM was chosen as the appropriate model studies. The final decision of the authors
Q-statistic or the I statistic.
2
and presented by one of the authors. was to include the REM as well as an
Cochran’s Q statistic represents the total Selecting the appropriate model for re- extensive sensitivity analysis to identify the
variance between the studies, while I2 is a porting, however, should not be the final step potential causes of heterogeneity.
measure of how much heterogeneity there in the process. Identifying possible causes of
is between the studies.3 Researchers will heterogeneity and analyzing the results—in- Improving future studies
use the REM if Cochran’s Q is significant or cluding important covariates in the model— These examples illustrate the problem
if the I2 statistic is large (greater than 50% is should be part of the analysis plan. that simply examining results of statistical

February 2017 • QP 45
 analyses—even those as simple as sum-
Plots of Anscombe data / FIGURE 1
mary statistics—can hide heterogeneity
in the data set. Several rules of thumb for Equation 1 Equation 2
Scatter plot of Y1 vs. X1 Scatter plot of Y2 vs. X1
statistical analyses can be suggested by 10
11
these examples: 10 9

1. Always examine your raw data. If the 9 8

8 7
data set is large, take a random sample

Y1

Y2
7 6
and examine the distribution. 6 5

2. When fitting a model, ensure that the 5 4

3
relationships between the independent 4
5.0 7.5 10.0 12.5 15.0 5.0 7.5 10.0 12.5 15.0
and dependent variables are linear (for X1 X1
4/29/2007 3:37:23 pm 4/29/2007 3:37:23 pm
general linear models) and don’t include
outliers that are influencing the result- Equation 3 Equation 4
ing regressions. Examine residuals and Scatter plot of Y3 vs. X1 Scatter plot of Y4 vs. X2
13 13
regression diagnostic measures. 12 12
11 11
3. Identify the possible causes of hetero-
10 10
geneity. Analyze your data with and 9 9
Y3

without outliers, and attempt to identify
why the data point is an outlier. Include
Y4
8 8
7 7
6 6

covariates and confounders in your 5 5

5.0 7.5 10.0 12.5 15.0 8 10 12 14 16 18 20
analyses to explain the heterogeneity in X1 X2
4/29/2007 3:37:23 pm 4/29/2007 3:37:23 pm
the model.
Use the results of these sensitivity
analyses to design improved studies in the and Douglas G. Altman, “Measuring Inconsistency in Meta- JULIA E. SEAMAN is a research scien-
Analyses,” British Medical Journal, Sept. 6, 2003, https://www. tist at the University of California, San
future. QP ncbi.nlm.nih.gov/pmc/articles/PMC192859. Francisco, focused on pharmaceutical
chemistry. She also is a statistical
I. ELAINE ALLEN is professor of biosta- consultant for the Babson Survey
REFERENCES tistics at the University of California, Research Group. She earned her doc-
1. Wikipedia, “Anscombe’s quartet,” https://en.wikipedia.org/ San Francisco and emeritus professor torate in pharmaceutical chemistry
wiki/Anscombe’s_quartet. of statistics at Babson College in and pharmacogenomics from the
2. “Research Methods II: Multivariate Analysis,” Journal of Wellesley, MA. She also is director of University of California, San Francisco.
Tropical Pediatrics, Oxford Journals, chapter five, “Regres- the Babson Survey Research Group
sion Diagnostics,” www.oxfordjournals.org/our_journals/ at Babson College. She earned a
tropej/online/ma_chap5.pdf. doctorate in statistics from Cornell
3. Julian P.T. Higgins, Simon G. Thompson, Jonathan J. Deeks University in Ithaca, NY. Allen is a member of ASQ.

Fixed and random effects meta-analysis

on persistence of asthma / FIGURE 2
A: Fixed effects model B: Random effects model
Study Percentage Study Percentage
ID RD (95% CI) weight ID RD (95% CI) weight

Ballardini (2012) 0.03 (0.02, 0.05) 26.18 Ballardini (2012) 0.03 (0.02, 0.05) 15.36
Burr (2013) -0.01 (-0.04, 0.02) 4.32 Burr (2013) -0.01 (-0.04, 0.02) 14.21
Finnbogadóttir (2012) 0.10 (0.06, 0.15) 1.83 Finnbogadóttir (2012) 0.10 (0.06, 0.15) 12.31
Gough (2010) -0.02 (-0.05, 0.02) 3.41 Gough (2010) -0.02 (-0.05, 0.02) 13.65
Nissen (2013) 0.03 (-0.00, 0.06) 2.87 Nissen (2013) 0.03 (-0.00, 0.06) 13.95
Williams (1998) -0.05 (-0.05, -0.04) 50.36 Williams (1998) -0.05 (-0.05, -0.04) 15.51
Ziyab (2010) 0.01 (-0.01, 0.03) 11.03 Ziyab (2010) 0.01 (-0.01, 0.03) 15.01
Overall -0.01 (-0.02, -0.01) 100.00 Overall 0.01 (-0.02, 0.05) 100.00
(I-squared = 96.3%, p = 0.000) (I-squared = 96.3%, p = 0.000)

-0.15 0 0.15 -0.15 0 0.15

Note: Weights are from random effects analysis
CI = confidence interval
RD = risk difference

46 QP • www.qualityprogress.com
2017 WORLD CONFERENCE ON
QUALITY AND IMPROVEMENT
May 1 – 3, 2017 | Charlotte, NC

Get in Gear and

Get Registered Today!

Early-Bird Pricing Expires March 10, 2017

You won’t believe what’s in store when you attend ASQ’s

2017 World Conference on Quality and Improvement in
Charlotte, NC!

This year’s conference may very well be the best one

yet … inspiring speakers, informative presentations,
and opportunities to connect with leading professionals
to expand your network.

Don’t miss out on all the excitement, enlightenment, and

opportunities for development at ASQ’s 2017 World
Conference on Quality and Improvement.

Visit asq.org/wcqi-qp to learn more and

register today!
 STANDARDS OUTLOOK
Keys to IATF 16949:2016
Understanding important changes to the automotive QMS
Editor’s note: This is the second install-
ment of a two-part series that exam-
ines ISO TS 16949’s recent revision.
Part one appeared in January’s QP.
THE AUTOMOTIVE industry’s revi-
sion of ISO technical specification (TS)
16949—now known as IATF 16949—car-
ries over the requirement for process
effectiveness and efficiency (subclause
5.1.1.2), the latter of which has never
been required by ISO 9001. This sub-
clause now requires process review
activities to be an input to the manage-
ment review process.
There is no definition of the term
“process review” or what this activity
includes, but the International Auto-
motive Task Force (IATF) indicates
process-review activities must include
evaluation methods and implemented
improvements. Top management should
be conducting reviews of the process-
48 QP • www.qualityprogress.com
BY R. DAN REID
specific reviews performed by process
owners.
Process approach
Ever since the 2000 revision of ISO 9001,
implementing and auditing a quality
management system (QMS) has required
a process approach. The most recent
revision (ISO 9001:2015) has not defined
the term, but describes it by using
terms—such as plan-do-check-act cy-
cles—and by providing process models
that show inputs and value-added steps
that transform inputs into outputs.
Having a process owner is a key
success factor in effectively managing
processes. IATF 16949 now requires this
in subclause 5.1.1.3, and it includes en-
suring process owners can perform their
assigned roles. They should have the
necessary competency, responsibility,
and authority for their owned processes’
activities and results. Note that while
the process approach is a more explicit
requirement, ISO 9001:2015 demoted
measuring and monitoring of processes
and products from a subclause level
in the last version to simply embedded
requirements within other clauses this
time. IATF did make measurement of
processes very explicit.
Typically, organizations do a good
job defining and documenting job
responsibilities. Many, however, do not
adequately define and document work-
ers’ authority to take action on their
jobs if things go wrong. For a job to be
done effectively, this guidance should
be included in work instructions. The in-
tent of this addition also is carried over
to subclause 5.3.1 (organizational roles,
responsibilities, and authorities—sup-
plemental) to address the need for orga-
nizations to better document assigned
personnel responsibilities and authority.
ISO TS 16949 required an organization
to address customer requirements, but
this revised subclause now requires the
organization to explicitly meet customer
requirements.
Addressing risk
Risk is still an immature concept in
International Organization for Stan-
dardization (ISO) documents because
there are different definitions of the
term. Risk was explicitly introduced as
a QMS requirement in ISO 9001:2015 and
defined as “effects of uncertainty.”
The new implicit ISO 9001:2015
requirement is for “risk-based thinking”
rather than “risk management,” which
is addressed in ISO 31000:2009—Risk
management—Principles and guide-

lines (ISO, 2009). The revision now
requires an organization to identify
and address risks and opportunities. It
also indicates risk can be positive or
negative. Positive risk is reportedly not
equivalent to opportunity.
In IATF 16949, the definition of risk
is, surprisingly, unimproved from the
ISO 9001:2015 definition. In subclause
6.1.2.1, IATF 16949 explicitly has risk
analysis as a requirement that includes a
periodic review of lessons learned from
product recalls, product audits, field
returns, complaints, scrap and rework.
In clause 6.1.2.2, IATF 16949 carries
over the requirement for preventive
action, which was deleted from ISO
9001:2015. But preventive action is no
longer aligned with corrective action,
which was a concern for ISO 9001 writ-
ers. Preventive action is now aligned
with QMS planning and should be
implemented up front when engineering
or re-engineering a QMS. Similarly, the
contingency plan requirement is carried
over from ISO TS 16949 and aligned
with planning the QMS.
At a minimum, risk should be
analyzed and addressed for products,
processes and an organization’s supply
chain. An effective approach would be
to analyze the risks qualitatively and
quantitatively, and implement actions
to reduce and mitigate the prioritized
risks based on severity and probability
of occurrence.
Objectives—QMS drivers
An organization’s top management
needs a process for setting and
maintaining objectives because they
can drive a QMS’s conformance and
continual improvement. To be effec-
tive, enterprise-level objectives must be
cascaded to relevant levels and func-
tions inside the organization and aligned
with additional objectives or targets set
Having a process owner is a
key success factor in effectively
managing processes.
by middle and lower management as
applicable.
In clause 6.2.2.1, IATF 16949 now
specifies objectives are to be es-
tablished at least annually. To drive
continual improvement, they should
be reset sooner if they are achieved.
Objectives should not be so ambitious
that there is no practical opportunity to
achieve them over time.
Auditor competency
In IATF 16949’s subclause 7.2, there are
expanded and more prescriptive re-
quirements for internal and second-par-
ty auditors. There is now a requirement
for a documented process for compe-
tency of internal auditors. Auditors also
are now categorized as QMS, process
or product auditors with prescriptive
requirements for each category. Most of
these requirements are common to all
categories, such as:
• Auditing according to ISO 19011.
• Understanding the process approach.
• Risk-based thinking.
• ISO 9001:2015.
• The automotive core tools, including
advanced product quality planning,
statistical process control, measure-
ment system analysis, and failure
mode and effects analysis.
Maintenance and improvement
of auditor competence must include
conducting a specified number of audits
annually and knowledge of changes
related to what is referred to as “context
of the organization.”
Embedded software
IATF 16949’s subclause 8.3.2.3 requires
organizations to use a process for the
quality assurance of products with in-
ternally developed embedded software
and to have an appropriate assessment
method to assess their software devel-
opment process.
The software development process
also must be included in the scope of
the internal audit program. Internal
auditors should be able to assess the ef-
fectiveness of the software development
method used. Subclause 7.1.5.2.1 on
calibration and verification records also
includes verification of the software
version used for product and process
control, although it is unfortunate that
the final subclause structure uses up to
five decimal places.
Subclause 8.3.4.2 also has design
validation requirements around embed-
ded software that should be included.
Subclause 8.4.2.3.1 also requires
cascading these software requirements
to suppliers. Software verification and
validation requirements, such as those
found in ISO 13485 for medical devices,
are appearing in sector QMS documents
with increasing regularity.
Purchased product controls
ISO 9001:2015 subclause 8.4 expanded
the scope of these requirements to
include allied and affiliated locations of
the same organization. Quality profes-
sionals probably are still applauding
this change because often the allied
February 2017 • QP 49
 STANDARDS OUTLOOK

At a minimum, risk should be

suppliers are the worst-performing suppli-
ers in terms of quality and delivery. Prior

analyzed and addressed for

to this revision, the quality managers had
no options to remedy the situation.
IATF 16949 adds more prescriptive
requirements for purchased products or
services. Typically, there are three pro-
cesses that are owned by the purchasing
function in organizations:
1. Supplier qualification.
2. Supplier selection.
3. Supplier quality, which can include
monitoring and development.
IATF 16949’s subclause 8.4.1.1 opens
the scope to suppliers of subassembly, re-
work, sequencing and sorting. Subclause
8.4.2.3 requires suppliers to be ISO 9001
certified unless otherwise authorized by
the customer. When compared with prior
requirements, this will require more tier-
two suppliers to be certified.
There is a new and prescriptive sup-
plier selection subclause (8.4.1.2) that
includes elements of supplier qualifica-
tion—such as assessment and supplier
selection processes (quality and delivery
performance). There also is a prescriptive
list of supplier selection criteria. Audi-
tors generally do not have experience in
processes that deal with purchasing, so
this will require more auditor training on
products, processes and an
organization’s supply chain.
that and training on embedded software
and other new requirements.
Next steps
An ISO 9001:2015 certification must be
attained to conform to IATF 16949. IATF
16949 raises the bar on top management
requirements. It’s important to coach and
train top leaders in these requirements for
them to pass IATF 16949 audits. There are
requirements that only top management
will have responsibility for and authority
over so top management must be audited
to obtain objective evidence to determine
whether the QMS is conforming.
Internal and second-party auditors will
need additional training in processes, such
as purchasing, software development,
customer-specific requirements auditing
per ISO 19011 and the automotive core
tools. Organizations should conduct a gap
analysis to determine what must be ad-
dressed from the new standard and imple-
ment actions to address the gaps. QP
BIBLIOGRAPHY
International Automotive Task Force, IATF 16949:2016—
Technical Specification.
International Organization for Standardization (ISO), ISO
9001:2015—Quality management systems—Require-
ments.
ISO, ISO 9000:2015—Fundamentals and vocabulary.
R. DAN REID is the principal con-
sultant with Management Systems
Consulting LLC in Farmington, MI.
He is an author of ISO Technical
Specification 16949, QS-9000, ISO
9001:2000, the first International
Organization for Standardization
international workshop agreement,
the Chrysler, Ford, GM Advanced Product Quality Planning
With Control Plan, Production Part Approval Process and
Potential Failure Modes and Effects Analysis manuals and
the Automotive Industry Action Group’s Business Operating
Systems for Healthcare Organizations. Reid was the first
delegation leader of the International Automotive Task
Force. He is an ASQ fellow and an ASQ-certified quality
engineer.

Are you in a bind at work? Are you looking to clarify

YOURQUALITYADVISOR a term or methodology? Have you run into a problem
where nobody seems to have the answer? Do you wish
you had a quality mentor? Someone you could turn to
when you run into a roadblock?

You do.

QP’s experts will provide answers and insight to your

toughest quality queries. Simply email your situation,
question or problem to editor@asq.org, and QP’s subject
matter experts will offer their sage advice in our Expert
Answers department.

50 QP • www.qualityprogress.com
THE VALUE OF SIX SIGMA
FORUM MEMBERSHIP
When you become a valued member of the ASQ Six Sigma Forum, you join a
distinguished group of problem solvers and analytical thinkers driven by a desire
to achieve breakthrough improvements. Your membership not only exposes you to
a remarkable network of individuals, but also links you to the Six Sigma Body of
Knowledge and supporting resources and case studies. Lastly, you will stay plugged
in to the Six Sigma community through monthly newsletters and members-only online
communities. So don’t delay, call ASQ Customer Care at 1-800-248-1946 and ask
to have the Six Sigma Forum added to your membership.

For more information, visit asq.org/sixsigma.

 UPCOMING
CERTIFICATION
EXAMS
• Calibration Technician
• Pharmaceutical GMP
Professional
• Quality Auditor
• Quality Engineer
• Quality Improvement
Associate
• Quality Process Analyst
• Six Sigma Green Belt
• Software Quality
Engineer

GET YOUR
ASQ CERTIFICATION
Looking for a way to stand out in 2017? ASQ certification can help with just
that. Choose from 19 certifications that are sure to not only differentiate you
from your competition, but can help result in a higher salary and help land
a promotion. It’s time to become a recognized expert in your field.
Apply for the April 2017 exams today!

2017 UPCOMING APPLICATION DEADLINE – March 10

TESTING WINDOW – April 7 – 23

For more information or to apply for an exam, visit asq.org/cert.

Experience

videos.asq.org
 Special Advertising Section

The ASQ 2017

Six Sigma
Resource Guide
Make Six Sigma Work for You

Don’t let bad advice detour your

organization. Turn to these resources
for the tools you need to make your
lean or Six Sigma program a success.

ASQ QI Macros for Excel

600 N. Plankinton Ave. 2696 S. Colorado Blvd., Suite 555
Milwaukee, WI 53203 Denver, CO 80222
Phone: 800-248-1946 Phone: 303-756-9144
asq.org/learninginstitute www.qimacros.com
ASQ offers a variety of both public and on-site Six Sigma and lean Six Sigma You don’t have to be a statistician to start getting immediate results with the
courses. These courses are offered in several formats including instructor-led, QI Macros for Excel! Free 30-day trial: PC or Mac.
classroom-based and self-paced.

EtQ Inc. SigmaXL

399 Conklin St., Suite 208 305 King St., Suite 503
Farmingdale, NY 11735 Kitchener, ON N2G 1B9
Phone: 800-354-4476 Canada
www.etq.com Phone: 888-744-6295
www.sigmaxl.com
EtQ’s quality management software identifies, mitigates and prevents high-risk
events in the quality system. Key functionality includes CAPA, audits, document SigmaXL brings you lean Six Sigma statistical tools, templates and Monte Carlo
control, risk register and more. simulation in the already familiar Excel. Start your free 30-day trial of SigmaXL
or DiscoverSim today!

The National Graduate School

of Quality Management
186 Jones Road
Falmouth, MA 02540
Phone: 800-838-2580, x147
www.ngs.edu
The National Graduate School of Quality Management (NGS) offers accelerated,
accredited degrees in quality systems management. Highly interactive online
formats are designed for adult learners.

To be included in next year’s Six Sigma Resource Guide, contact Lou Brandow, 866-277-5666.
February 2017 • QP 55
 Special Advertising Section

• No entrance or comprehensive
DEGREES IN QUALITY SYSTEMS MANAGEMENT exams
• Small, interactive classes
Bachelor of Science Degree Completion* - 10 courses over 12 months
with Juran Global Lean Six Sigma Green Belt Certification - Classes Start March 21 • Accelerated, accredited degrees
• Innovative, project-based learning
Master of Science Degree - 12 courses over 12 months
with Juran Global Lean Six Sigma Black Belt Certification - Classes Start March 22 • Weekly interaction with faculty and
fellow students
Contact info@ngs.edu or (800) 838-2580 ext. 505 to learn more. . • Financial Aid and scholarships
available for qualified students
Transform yourself. Become a leader in your industry.
NGS offers innovative, flexible degrees designed to support your THE NATIONAL GRADUATE SCHOOL
professional success and development. One-on-one interaction with OF QUALITY MANAGEMENT
faculty and support staff ensure that you won’t get lost in the crowd. Success you can measure

Ask about our special reduced tuition rates for ASQ members. (800) 838-2580 x505 | info@ngs.edu | www.ngs.edu/asq
NGS is accredited by The New England Association of Schools and Colleges (NEAS&C). NGS is a not-for-profit institution.
Certified to operate by State Council of Higher Education for Virginia (SCHEV). *Not available in MA. Not available to MA residents.

Innovative Excel Add-in

Six Sigma Projects in Seconds
with the New Data Mining Wizard 2.0

Download 30-day trial at www.qimacros.com

56 QP • www.qualityprogress.com
 2
THE ASQ
GLOBAL
STATE OF
QUALITY
RESEARCH

READ DISCOVERIES 2016 TODAY

The ASQ Global State of Quality is the only research that creates a baseline of fundamental
quality and continuous improvement practices around the world. The findings and analyses
provide a guide for businesses to improve organizational performances.

Visit globalstateofquality.org to download “Discoveries 2016.”

 QPTOOLBOX
Measuring station
•
Mahr’s measuring station—the MarShaft
Scope 600 Plus 3-D—measures camshafts
completely in 3-D for the first time. The
combination of optical and tactile probes
enables the complete detection of relevant
measuring characteristics, directly in pro-
duction at the processing machine for the
respective processing step.
It is now possible to load the workpieces
by robot and therefore have an inline
measurement. The measuring station is
networked and sends correction data in a
closed loop to the machine tool when there
are deviating tolerances. As a measuring
machine with local intelligence, it has an
optional barcode reader—so the work-
pieces can be detected—and automatically
starts the correct measuring program.
• Call: 800-343-2050.
• Visit: www.mahr.com.
58 QP • www.qualityprogress.com
•
Ultrasonic sensor
SICK has announced the
UD18-2 sensor for double
layer and splice detection
using ultrasonic tech-
nology. The sensor can
determine whether one,
two or no material layers
are present between
its sender and receiver.
The UD18-2 can detect
objects regardless of
material, including paper,
cardboard, shiny metal
and transparent plastic.
The UD18-2 can also teach-in up to four The humid environment versions of the
sensitivity levels, and the sensor can switch robots can be used for water-jet cutting,
between sensitivity levels during operation. machining or industrial parts cleaning. The
This allows the sensor to tackle complex robots also are a good solution for use in
applications and ensure permanent system the food industry where stringent hygiene
availability with consistent production standards apply. They can be cleaned easily
quality. with detergents and hosed off with a water
The UD18-2 is immune to dirt, dust and jet as part of the daily cleaning routine.
humidity, making it well-suited for applica- • Visit: www.staubli.us.
tions in the packaging, print and paper, • Call: 864-433-1980.
electronics and solar, automotive, and
metal and steel industries. Life science app feature
• Email: santi.clarke@sick.com. EtQ launched Traqpath last year as a free
• Visit: www.sickusa.com. app to enable users to conduct simple
quality and compliance tracking activities
•
Robots
Stäubli has introduced a line of
standard robots for industry-specific
requirements. The current Stäubli
robot line is aimed at all packing
technology markets and includes the
Fast Picker TP80 HE and the TX90X-
LHE for humid environments. The
robots have enclosed structures with
all the cables routed internally, which
make them equipped to cope with
extreme conditions.
for events affecting their business by using
either a mobile device or web browser. EtQ
recently announced the addition of Life
Science features in its Traqpath product to
meet the needs of regulatory compliance.
The Life Science features include:
• Forced authentication: Meet the needs
of 21 CFR Part 11 for electronic signa-
tures and records within your Traqpath
workgroup, on web and mobile devices.
• E-Signature categorization: Offers the
ability to apply a meaning to your elec-
tronic signatures to further define what
you are signing.
• Password management: Built-in
password rules help to keep security in
line with regulatory standards, through
automatic password aging and complex-
ity rules.
• Field history and audit trail: Offers the
ability to track the history of all fields—
with each record having its own history.
Users can access and download Traq-
path on their website or on a mobile device
through the App Store or Google Play.
• Call: 800-354-4476.
• Visit: www.etq.com.
•
Ultrasonic flowmeter
Khrone’s ultrasonic flowmeter Optisonic
3400 is available with heat meter approval
according to MID 2014/32/EU Annex VI
MI-004 in accuracy classes 1, 2 and 3.
Target applications include district heating
networks, commercial and industrial heat
metering applications, as well as energy
GOT A QUALITY PRODUCT?
Send your product description and
photo to vellifson@asq.org.
and utility applica-
tions in combined
heat and electricity
generation. To ensure
stated accuracy classes, each heat
meter undergoes a three-point factory
calibration performed with water.
With its direct path configuration, Opti-
sonic 3400 is not susceptible to (magnetite)
scaling-like meters that use reflectors. It of-
fers bi-directional flow measurement and a
fully welded construction without potential
leak points. Optisonic 3400 also features
enhanced diagnostic and status indications
and measurement of the velocity of sound
to gain more information about medium
and process.
• Email: info@krohne.com.
• Visit: www.us.krohne.com.
Metrology tables
•
R&R Fixture metrology tables are designed
for fixturing parts and mounting of portable
coordinate-measuring
machines. They also can be used with
3-D laser scanning machines. With its
aluminum-extruded structure, the metrol-
ogy tables can hold a lot of weight for any
size, material or shape of part you want
to inspect.
Options include either an English or
metric-threaded plate for locating and
elevating your parts off the surface for
full inspection­—whether using touch
probe or laser. R&R also offers a variety
of components to be used with the
metrology tables and are available in
many sizes.
• Visit: www.rrfixtures.com.
• Call: 616-847-6045.
February 2017 • QP 59
 QPREVIEWS
Semi-Markov Processes:
Applications in System
Reliability and Maintenance
Franciszek Grabski, Elsevier, 2015, 270 pp.,
$150 (book).
This is a com-
prehensive and
rigorous research
text focusing on
continuous time
semi-Markov
processes (SMP)
with a discrete
set of states. After
a mathematical
and statistical
introduction to discrete Markov processes
and semi-Markov processes, the author ex-
pands and generalizes into important cases
of these models. Advanced methods are
presented with clear applications to quality
processes—making this text a good refer-
ence for statistical quality professionals.
The first six chapters cover the concepts
and equations of SMPs necessary to intro-
duce more advanced models, specifically
in the area of reliability and maintenance.
These chapters are concise and present the
characteristics of SMPs to enable math-
ematical analyses of the models presented
in later chapters.
The most interesting part of this book
are the processes that have direct implica-
tions for quality-control applications such
as multitask operation and failure rate. In-
cluded are numeric examples that illustrate
the use of the models and descriptions
of the algorithms inherent in the models.
The final chapter, on semi-Markov decision
processes, provides models for the optimal
strategies for maintenance operations and
renewal processes showing algorithms for
60 QP • www.qualityprogress.com
optimization based on discrete state chang-
es with concrete applications in industry.
I. Elaine Allen
San Francisco
Innovating Lean Six Sigma: A
Strategic Guide to Deploying the
World’s Most Effective Business
Improvement Process
Kimberly Watson-Hemphill and Kristine
Nissen Bradley, McGraw-Hill Education,
2016, 304 pp., $35 (book).
This book is down
to earth and
practical with
many charts and
graphics to hold
the reader’s atten-
tion. The authors
draw from their
experience, using
multiple industries
and situations
with differing expectations for needed
changes as examples.
The content is divided into two parts.
Part one is for organizations—in the early
stages of their improvement journey—that
must understand and apply the building
blocks and culture to begin the transforma-
tion process.
Part two is for those already well on
their way to transformation and want to
work on the big problems their organization
is facing. Part two was right on target—fo-
cusing on customers, culture change and
strategy deployment (Hoshin planning) to
align at all levels and applying lean to prod-
uct and process design and redesign.
I really like the last part of chapter
five, which addresses project pipeline
management by referring to Little’s Law:
“The amount of time it takes to complete
any single project is affected by the total
number of projects that are active at any
point in time.” For example, a basic rule of
lean: minimize work in process to reduce
cycle time.
This is a useful book for managers at
all levels, but particularly C-suite leaders
who want to lead positive change in this
competitive, fast-moving world.
Bill Baker
Santa Fe, NM
The Certified Pharmaceutical
GMP Professional Handbook
Mark A. Durivage, ASQ Quality Press, 2016,
516 pp., $89 member, $135 list (second
edition, book).
This book is a re-
markable standout
by itself while
also representing
a significant en-
hancement of the
first edition. This
second version is
not just a simple
“refresh,” but an
overall revision of content and format. It
represents current quality and regulatory
thinking about pharmaceutical quality and
expands the overall body of knowledge for
the certified pharmaceutical good manufac-
turing practices professional (CPGP) exam.
The book incorporates updated sources
and regulatory references—domestically
and internationally.
While serving as the central resource
for ASQ's CPGP exam, it also can be used
by industry and research to help address
theoretical and practical questions regard-
ing pharmaceutical quality, lab operations,
manufacturing and related activities.
The author lists 34 contributors to this
edition and he shapes their efforts into a
comprehensive examination of what con-
stitutes a safe and effective drug product.
The impact of activities—such as
development, technology transfer and
manufacturing—are also examined and
considered in relation to the many types of
drug substances, drug products and related
packaging.
The book's format has changed consid-
erably and now includes eight major parts
broken into 60 chapters. The book serves
as an excellent resource with research
presented through a detailed list of tables
and figures, acronyms and abbreviations, a
preface, acknowledgements, four appen-
dixes, a glossary, a comprehensive list of
references and suggested readings and
then it closes with an index.
The detail and logic of this organiza-
tion allows the reader to understand many
technicalities in a flawless fashion, while
covering a significant number of details,
suggestions and potential problems.
The book is easy to read and follow.
Most everyone will find something valuable
in the writing because it is a useful and
valuable tool for every level of experience.
Frank Pokrop
San Diego
The Art of Integrating Strategic
Planning, Process Metrics, Risk
Management and Auditing
Janet B. Smith, ASQ Quality Press, 2016,
216 pp., $36 member, $60 (book and
CD-ROM).
Between the covers of this book, there is
a course. The course represented by this
book is organized into five parts: strategic
planning, process metrics and risk mitiga- Part seven of the book presents
tion, root causes, auditing, and beyond workshops, or practice exercises, on basic
compliance and sustainability. quality and lean tools. These workshops are
The book has 18 chapters with tools and thinking and activity-based and may be used
templates, plus a CD-ROM with document as a resource for a single reader or a study
files. There is also team—possibly a team of students in a
a chapter offered course or training program. The workshops
as a workshop are interesting and provide further under-
in which tools standing and insight for readers.
and procedures As a resource for leaders tasked with
explored in the delivering quality results, Smith’s book is on
book can be put the recommended list. The book also is a
into practice. resource for the practitioner on the produc-
With the tion line, as well as the program developer,
threepage glos- manager and the assessor of results.
sary and references list, Smith’s book can Two resources in the book may be of
be shelved with the high-value reference value in the broad field of quality. First
collection. are examples of quality processes and
As a resource tool, the book is focused resources presented in the list of examples
on outcomes and results—based on appearing as a section in the table of con-
knowledge and insights of happenings while tents. Second, the book itself is an idea-rich
the processes building results take place. resource for those needing examples of
Though failure is referenced, emphasis is how to present a complex topic. It is full of
on successes, improvements and systems graphics, reference tables, charts and tools
delivering predictable results. for users. This book is a reliable tool for
This book is a recommended resource experienced persons in the field, as well as
for those in the broad, quality field including others new to the field.
government services, education, healthcare Gerald Brong
and fields beyond manufacturing. Ellensburg, WA
Advertisers Index
ADVERTISER PAGE PHONE WEB
CyberMetrics Corp. OBC 800-777-7020 www.cybermetrics.com
Dekra Certification Inc. IBC 800-768-5362 www.dekra-certification.us
EtQ Inc. IFC 800-354-4476 www.etq.com
Minitab 2 800-448-3555 www.minitab.com
The National Graduate School 56 800-838-2580 www.ngs.com
SigmaXL 54 888-744-6295 www.sigmaxl.com
QI Macros for Excel 56 303-756-9144 www.qimacros.com Quality
Council of Indiana 1 800-660-4215 www.qualitycouncil.com
February 2017 • QP 61
 For information on placing an ad,

PROFESSIONALSERVICES contact Media Sales at 866-277-5666.

IMDS Data TQM ASSOCIATES INC.

Here’s a great way to improve
quality or implement a
Quality Management System
and it’s FREE!
qmscapa.net
• We can create your data and submit within “we’re here to support you”
72 hours
• Established in 1994
• We do the work and invoice you
• References found on our website • Women-Owned
• Based in the United States • Quality Assurance Professionals
• Across the U.S. and Worldwide
www.imdsdata.org • 734-205-8874
• Temporary or Permanent
• Source Inspection
Quality Web Based Training • Surveys
JP Russell, Managing Director • Audits
Online Classes – IACET Accredited Provider • Expediting
Quality Auditing CQA Certification Fundamentals
800-424-4729 3990 Old Town Ave. #C109
ISO 9001:2015 Comprehensive Transition Training
Process Improvement Tools/Metrics Fax 619-297-3251 San Diego, CA 92110
www.qualitywbt.com tqmassociates.com email: stephk@tqminc.net

www.amazonconsultinginc.com
Auditing
Training
Consulting

· Hire an eAuditor for your Internal

Audits!
· Join instructor-led certification
prep courses online for:
CMQ/OE, CQA, CQI, CQIA,
SSYB/GB, CQPA We’re a website that’s
Qualified auditors and trainers within reach!
Order at www.amazonconsultinginc.com dedicated to the Quality
Control, Material Test and
Quality & Productivity Inspection industries and
Solutions, Inc. the people that make it run.
Global Consulting, Auditing &
Training (Over 200 topics available)
◆ Mgmt. Systems – All ISO Let us help you find the
◆ Lean Six Sigma
◆ FDA/Regulatory Related personnel you need today!
◆ Agile Project Mgmt.
◆ Certifications (ASQ, APICS, SME, IT, PMI,
Agile/SCRUM & Others) VISIT US TODAY AT:
QPSINC.COM info@qpsinc.com
Inspectionjobs.com
Easy, Fast, eQMS

Q Email to
Sales@inspectionjobs.com
myQualityCloud.com

Looking for Quality

Professionals?
Place a Recruitment Ad in
Quality Progress and reach
more than 100,000 readers
from all over the world!
Contact Media Sales
at 866-277-5666.
Classroom Training Aids… Concierge quality assurance, regulatory
G Quincunx Boards compliance and organizational
G Sampling Bowls excellence coaching for start-ups and
G Catapults established organizations
G Deming Funnels
G And lots more …
QMS Tools, templates and training
materials available for
Visit us at: www.qualitytng.com FDA QSR, ISO 9001, ISO 13485, AS9100
Email sales@qualitytng.com for brochure
Ph: 248-641-7030 Fax 248-641-7031
www.myqacoach.com
PO Box 611 Troy, MI 48099-0611 805-622-3019

62 QP • www.qualityprogress.com
ONE GOOD IDEA
Tips on Effective-Quality
Defining quality throughput using constraint management
WE WERE ASKED to measure the ef-
fectiveness of a large commercial bank’s
audit department. The department’s output
consists of recommendations for action—
scored from one to five according to the
level of potential risk. The common mea-
surement is the number of recommenda-
tions that the department makes during the
year. A better measurement limits the count
to severe recommendations. But these mea-
surements have two weaknesses:
1. They encourage auditors to make more
recommendations—in other words, re-
ducing the department’s signal-to-noise.
2. The measurements ignore the auditors’
actual impact—in other words, the ef-
fect they have on the organization.
It is essential to measure effective-qual-
ity (EQ), which equals throughput divided
by output. Throughput is the effective
output of the system.
Back to the bank’s audit department, the
EQ is measured as the number of recom-
mendations that were implemented by
the organization divided by the number of
recommendations actually made. The EQ—
represented in decimal points—can only
be refined by counting the most important
recommendations (scored four or five).
To evaluate such a department, we use
additional measurements, especially the
absolute throughput of the department. To
assess continuous improvement, however,
we use the EQ measurement.
Making a sale
Salespeople are measured by the amount
of sales made. To focus on improvement ef-
forts, you must identify where waste hurts
an organization. The EQ measurement
highlights areas where a high number of
salesperson meetings with potential clients
BY BOAZ RONEN AND ALEX COMAN
do not result in a sale. This leads to cause-
effect analysis on garbage time.
A study by the pharmaceutical company
Merck shows that if 100 sales reps call on
a physician, 17 depart without seeing the
physician because competitors’ reps are
waiting for the physician and 27 drop off
samples at the receptionist’s desk because
the physician is unavailable. Less than 24
conduct meaningful two-way discussions
with the physician and are remembered
because the average call lasts 4.6 minutes. 1
An EQ of 0.24 is unacceptable, and Merck
urges “new ways of engaging physicians on
their terms … with greater productivity.”2
We want to assess the throughput—the
effective output—and measure its ratio to
the system’s output. We noticed that many
organizations do not even measure the
output of a process, but assume that the
system’s input (budget and person-months,3
for example) accurately reflects its output.
For example, many IT departments in
financial organizations are measured by
person-months invested rather than by
their effective throughput (for example, the
number of features delivered, value added
to the organization and bugs fixed).
Measuring the effectiveness of the New
York Police Department (NYPD) is another
example. Before Bill Bratton’s tenure as the
city’s police commissioner, NYPD seemed
waste prone: widespread patrols of the
subway system, massive paperwork and
attempted arrests during daytime hours.
When measured in terms of input, this seems
reasonable.4
The waste is only salient when effective
actions are measured. Bratton virtually
eliminated wasteful tasks by replacing them
with preventive quality-of-life crimes and
performing arrests during sleeping hours.
The input didn’t change while the EQ rose
significantly.5
EQ is particularly effective at reducing
waste when bidding for new business.
Orion, bidding for heavy-civil marine con-
struction projects, submitted bids to proj-
ects worth $670 million, and it won projects
worth a mere $76 million6—an EQ of 0.068.
Years later, Orion focused its efforts and
bid for projects worth only $429 million.
Less bidding resulted in higher-quality bids.
Orion’s wins rose to $117 million,7 and the
EQ rose to 0.27.
A common mistake senior executives
make is to measure output rather than
throughput (effective output). EQ drives the
organization toward process improvement. A
better process reduces waste—which in turn
improves the quality and value delivered. QP
REFERENCES AND NOTE
1. John Mack, “Merck Rejiggers Its Marketing Mix,” Pharma
Marketing News, Vol. 6, No. 7, 2007, VirSci Corp., www.
news.pharma-mkting.com/pmn67-article01.pdf.
2. Ibid.
3. Businessdictionary.com defines “man month” as a unit of
work representing the productive effort of one person dur-
ing a four-week period. It is also called “labor month.” See
www.businessdictionary.com/definition/man-month.html.
4. W. Chan Kim and Renee Mauborgne, “Tipping Point Leader-
ship,” Harvard Business Review, April, 2003.
5. Ibid.
6. Orion Marine Group, “Orion Marine Group Inc. Reports
Third Quarter 2011 Results,” press release, Nov. 3, 2011,
http://tinyurl.com/orion-pr-2011.
7. Orion Marine Group, “Orion Marine Group Inc. Reports Sec-
ond Quarter 2014 Results; Reports Record Backlog,” press
release, July 31, 2014, http://tinyurl.com/orion-pr-2014.
BOAZ RONEN is a professor (emeritus)
of technology management and value
creation at Coller School of Manage-
ment, Tel Aviv University in Israel.
He has his doctorate in business
administration from Tel Aviv University
and co-authored Focused Operations
Management: Achieving More With
Existing Resources (Wiley, 2008).
ALEX COMAN is a professor of
project management and value
creation at the Academic College of
Tel Aviv-Yaffo in Israel. Coman has
his doctorate in IT from Claremont
Graduate University in California.
February 2017 • QP 63
 BACK TO BASICS
Uncovering the Truth
Understanding sources of variation in statistical analysis
HAVE YOU ANALYZED a data set
with the goal of comparing the standard
deviations of two groups? This common
statistical objective is usually tackled with
tests such as a routine F-test, a more so-
phisticated Levene-Brown-Forsythe test or
Bonnett’s test. Some people may even stop
to check normality and other assumptions
before proceeding with their hypothesis
tests. But could there be more to the data?
For example, I worked on a project that
had a deeper story than just the p-value
from a simple F-test. A parts manufacturer
was developing a new process to make its
product more consistent and hoped this
new method would decrease the variation
seen in the key product-performance met-
ric. Eleven samples were made using the
new experimental process, and they were
compared to available historical data via a
traditional F-test.
The F-test p-value of 0.044 indicates a
statistically significant difference in the
standard deviations with better than 95%
confidence (see Online Table 1, which
can be found on this article’s webpage at
Overlaid histogram
of experimental and
historical data / FIGURE 1
0.09
0.08
0.07
0.06
Density
0.05
0.04
0.03
0.02
0.01
0.00
70 80 90
Y
64 QP • www.qualityprogress.com
BY ARVED HARDING
www.qualityprogress.com). You, therefore,
could conclude that there is sufficient evi-
dence to support the hypothesis that the
new method decreases process variation.
So, what’s wrong with this conclu-
sion? On the surface, it looks correct. The
p-value is less than 0.05. Figure 1 shows a
nice histogram overlaying the data, which
provides a positive testimony for the new
method. But a closer look at the data and
potential sources of variation reveals that
the 32 data points coming from the histori-
cal data set are hiding some important
issues.
Five different sample collection dates
were used for the historical data (see On-
line Figure 1), but only one date was used
during the experimental data. Collecting
data for several different time periods is a
good idea because it allows you to check
for day-to-day variation or even factors
such as variation between work shifts.
The variance component analysis of the
historical data showed that 64.4% of the
variation was due to the sample-collection
date (see Online Table 2). Because the
experimental process was only
run on one day, there is no esti-
mate of the day-to-day variation
for the experimental process. It
is possible that if the experi-
Experimental mental process ran on a few
Historical
different days, you might see
Mean Std. dev. N
83.45 5.549 11 more overall variation. While
85.48 10.27 32
drawing conclusions from our
N = sample size
Std. dev. = initial comparison of standard
standard deviation
deviations may at first appear
to be statistically correct, it
could very well not be the cor-
rect conclusions from the data
100 110 and problem at hand.
What if we took a different analysis
approach and only compared the “within
sample collection date variation” for the
two methods? From just the variance
component analysis of the historical data,
Online Table 2 shows the data pooled
within collection date standard deviation
to be 6.72. With varying sample sizes on
the sample collection dates, the degrees of
freedom for the historical data would be
the sum of the n-1’s for each day or 27.
The 11 samples of the experimental
data were only collected on one day, so
we used all of the data to estimate what
was within data-collection date standard
deviation and 10 degrees of freedom. The
F-test for the pooled summary data (see
Online Table 3) indicated no significant
difference between the two methods (p =
0.537). Using this pooled approach, there
was not strong evidence pointing to the
new method being superior. The team
was advised to collect additional sets of
data from the experimental method and
perform a similar analysis later.
To prevent this issue from recurring,
always ask questions about data-collection
methods—especially if you’re not directly
involved in that step. It is important to
understand potential sources of variation
prior to doing statistical analysis. Remem-
ber the top three rules of data analysis:
Plot the data, plot the data and plot the
data. QP
ARVED HARDING is a senior statisti-
cal associate with Eastman Chemical
Co. in Kingsport, TN, and an adjunct
instructor at Northeast State Com-
munity College in Blountville, TN. He
earned a master’s degree in statis-
tics from Virginia Tech in Blacksburg.
Harding is an ASQ senior member
and an ASQ-certified quality engineer.
 Your One Stop For All
Certification and Training

Looking to get certified or transfer your certification?

Contact DEKRA for a no-obligation quote
DEKRA Certification is one of the world‘s largest
certification bodies
> Over 99% of our clients would recommend us to others
> We are industry leaders, using the newest innovations in the
industry to help your business
> We offer no-cost transfers
DEKRA Certification is a part of DEKRA SE is one of the world’s
leading expert organizations active in over 70 countries through
out the globe. With their newest acquisition, AQS Solutions (for-
merly AQS Management Systems), DEKRA is able to provide one
stop for all certification and training needs.

Our Certification and Assessment Services Include:

ISO 9001 | ISO 14001 | AS9100 | AS 9120 | ISO 13485 | ISO 50001
IATF 16949 | TL 9000 | ISO 27001 | ISO 20000-1 | ISO 22301 | FSSC 22000

Contact us at 1-800-768-5362 or sales.us@dekra.com to get started.

www.dekra-certification.us

Need training?
Established in 1989, AQS Solutions provides training in support of organizational improvement and
implementation of international management system standards. Offering a full range of innovative
training services to industry, government, non-profit, and service in areas of Quality, Environment,
Energy, Food Safety, and Occupational Health and Safety.

> Public and On-Site Training > Integrated Management Systems

> Certified Lead Auditor > Understanding and Implementing
> Certified Internal Auditor > Executive Overview Contact us at 1-800-633-2588 with
> Transition Training > Core Concepts or go to www.aqsperformance.com
to view all our training options.
 The quality of your products
depends on the quality of your measurements.
With a solid 25-year track record and countless auditor recommendations, GAGEtrak calibration management software
has improved quality at over 12,000 facilities worldwide, assuring unwavering accuracy, reliability and standards
compliance. And now, with the addition of GAGEtrak Lite to our product line, we have a software solution to meet every
need and budget. Contact us now to discuss which one is right for your company or download both and see for yourself.

GAGEtrak and GAGEtrak Lite

Two levels of calibration management. One auditor-recommended name.

Download for free at: GAGEtrak.com/downloads 1-800-777-7020 +1 (480) 922-7300