 |
Understanding Variation
By Stephen Halliday
“Variation there will always be, between people, in output,
in service, in product. What is the variation trying to tell us?”
(Deming)
“Of all the issues that an Engineer has to resolve, dealing with
variation is the most difficult.”
(T. Davies, Ford)
A key element in achieving Operational Excellence is to identify sources
of variation that affect a process, product or service. Once identified
the key sources can then be earmarked for attention. The current
focus on Six Sigma, as a methodology to achieve Operational Excellence,
has at its roots a focus on variation.
“The objective in driving Six Sigma performance is to reduce or
narrow variation to such a degree that six sigmas – or standard
deviations – can be squeezed within the limits defined by the customer’s
specifications.”
(Pande)
An important part of the Six Sigma model is the measurement phase. This
produces the data that shows how much variation is present in the process.
Once identified, the Six Sigma tools are then used to reduce this variation.
There are two basic sources of variation:
• Common Cause variation
• Special Cause variation
Common Cause variation is created by many factors, that
are commonly part of the process, and are acting totally at random and
independent of each other. Their origin can usually be traced to the key
elements of the system in which the process operates. (Materials, Equipment,
People, Environment, Methods). If only common causes of variation are
present, the output of a process forms a distribution that is stable over
time.
Special Cause variation is created by a non-random event leading to an
unexpected change in the process output. The effects are intermittent
and unpredictable. If Special Causes of variation are present, the process
output is not stable over time and is not predictable. All processes must
be brought into statistical control by first detecting and removing the
Special Cause variation.
Dr. W. Edwards Deming introduced the terms Common Cause and Special Causes
to business, although he assigns the first usage to W. Shewhart. An
understanding of the type of variation being observed is important if
the correct course of action is to be taken.
“A fault in the interpretation of observations,
seen everywhere, is to suppose that every event (defect, mistake, accident)
is attributable to someone (usually the nearest at hand), or is related
to some special event. The fact is that most troubles with service and
production lie in the system. We shall speak of faults of the system
as common causes of trouble, and faults from fleeting events as special
causes.”
(Deming)
Many involved in Six Sigma recognise the importance of this distinction.
“One of the most important things Deming taught me is that management
must master the concept of common cause versus special cause variation.”
(Eckes)
Not everyone makes this distinction. It is interesting that Mikel Harry
in his book “Six Sigma” does not use the terms in identifying
sources of variation, implying by omission that distinguishing between
Common Cause and Special Cause variation is unnecessary.
It is important to distinguish between the two types of variation because
they require different approaches to deal with them.
“There are different improvement strategies depending on which type
of variation is present (common cause or special cause), so it is important
for a team to know the difference.”
(M.L. George)
“This type of variation (common cause) requires a long-term strategy
to identify, understand and reduce it through on-going process management
and improvement. Special causes of variation, inherently unnatural in
origin, require quite different and immediate action.”
(G. Tennant)
If the type of variation is not identified then it is highly likely that
the wrong tools will be used. Although many advocates of the Six
Sigma methodology state that different tools are required to deal with
the two types of variation, few make any attempt to explain the difference.
Hence the danger, and this is especially true for Six Sigma, is that Common
Cause tools will be used to deal with a Special Cause.
“In analysing variation there are two kinds of mistakes we could
make:
1. We could mistake the cause of variation as being special in nature,
when in fact it is random and caused by the system (common cause)
2. We could mistake the source of variation as being systemic in nature
(common causes), when in fact it is special in nature (special cause)
and can and should be identified and, if possible eliminated”
(R. Aguayo)
There is no debate that the tools taught in Six Sigma can be used to significantly
reduce Common Cause variation but what of Special Cause?
Some teaching ignores Special Cause and treats all variation the same
– a serious mistake as mentioned by Aguayo. Others limit the appropriate
tools to the “five whys” and a Cause and Effect diagram. Although
useful, they have limited success in rapidly identifying the root cause
of special variation.
The five whys suggest that by continually asking “why”, it
is possible to move from a symptom towards root cause. If the situation
is fully understood then the root cause itself may be identified. However,
in many Special Cause situations, the answer to at least one “why?”
is “don’t know”. At this point the Cause and Effect
diagram is used. It is not difficult in such situations to generate
10 to 20 ideas and may even reveal more than 100. With so many possible
causes the most likely causes are chosen and tested. The process at this
stage relies on the root cause being identified through experience and
can result in much testing.
A more robust methodology for dealing with Special Cause
variation is available. Such methodologies have been in existence long
before Six Sigma appeared on the scene. These are often referred to as
Special Cause Problem Solving or simply Problem Solving methodologies.
Typical approaches to dealing with Special Causes are:
• Problem Analysis (Kepner Tregoe)
• Global 8D (Ford)
• R.I.S.E. (wdp consulting)
Their vocabulary may differ but their basic approach is similar. A key
part of the methodology is to identify changes, a point recognised by
George without any methodology to investigate the situation.
“To eliminate a special cause of variation, you need to investigate
what is different or changed in the process – something that isn’t
always evident.”
The model for Special Cause Problem Solving has similarities to the DMAIC
model of Six Sigma, however as mentioned previously the tools differ.
Example: R.I.S.E.
Recognition – identify the situation (Define)
Investigate – gather facts, identify and verify
root cause (Measure, Analyse)
Solution – identify and implement the best solution
(Improve)
Evaluate – evaluate effectiveness and transfer
knowledge (Control)
The tools used in Special Cause Problem Solving are Sub-dividing and stair-stepping,
Problem Profile Analysis, Differences and Changes, Test Matrix, Selection
Analysis and Planning and Problem Prevention.
Within these methodologies is a term that can cause confusion and that
is the word “Problem”.
Problem: - a deviation from expectation with cause unknown.
This very definition of problem places this methodology
in the Special Cause arena. Obviously, if the root cause of the Special
Cause is known then the methodology is not required since all that is
needed is to verify that the root cause has been found and identify solutions.
Unfortunately, Six Sigma itself is referred to as a
problem solving methodology.
“The ‘problem-solving’ methods of
Step 4 – Process Improvement and process Design / Redesign –
are most powerful when driven by a wider focus and long-term perspective.”
(Pande)
As can be seen from the above quote, Six Sigma is a
process improvement tool and not, in its purest form, a problem-solving
methodology. That is, Six Sigma deals with Common Cause variation. However,
no process can be improved if Special Cause variation remains present.
Thus, there is a need for an understanding of both Common Cause and Special
Cause tools. Unfortunately, in many Six Sigma training courses, this latter
issue is ignored.
Conclusion
Understanding variation is important if a process is to be improved. It
is critical that the two sources of variation, Common Causes and Special
Causes, are identified and handled using the appropriate tools. Six Sigma
tools can be used to reduce Common Cause variation but not before the
Special Cause Problem Solving methodology has been used to remove Special
Cause sources of variation.
A final word from Deming.
“Confusion between common causes and special causes leads to frustration
of everyone, and leads to greater variability and to higher costs, exactly
contrary to what is needed.”
References
Peter S. Panda, Robert P. Neuman and Ronald R. Cavanagh,
The Six Sigma Way (McGraw-Hill 2000)
Geoff. Tennant, Six Sigma: SPC and TQM in Manufacturing
and Services (Gower 2000)
W. Edwards Deming, Out of the Crisis (Massachusetts
Institute of Technology, Centre for Advanced Engineering Study 1986)
George Eckes, The Six Sigma Revolution (Wiley 2001)
Rafael Aguayo, Dr Deming (Fireside 1990)
Michael L. George, Lean Six Sigma (McGraw-Hill
2002)
top of page |
 |
