Page 176 - The Handbook for Quality Management a Complete Guide to Operational Excellence

162 P r o c e s s C o n t r o l Q u a n t i f y i n g P r o c e s s Va r i a t i o n 163

process has not changed: it is stable; it is in statistical control. No one
changed the bucket: it is the same bucket, with the same percentage of
red beads.
Needless to say, not all phenomena arise from constant systems of
common causes. At times the variation is caused by a source of variation
that is not part of the constant system. These sources of variation were
termed assignable causes by Shewhart; Deming calls them special causes of
variation. Experience indicates that special causes of variation can usually
be found and eliminated.
The basic rule of statistical process control is:
Variation from common cause is indicative of the system. Changing the
amount of common cause variation requires fundamental changes to the sys-
tem itself. Special causes of variation can be identified by the control chart,
and often quickly eliminated.

The control chart in Fig. 9.5 illustrates the need for statistical methods
to determine the type of variation. The control charts answer the question:
Are these variations built into the system, or are they indicative of a change to the
system?
Variation between the control limits desig nated by the UCL and LCL
lines are considered variation from the common cause system. Any vari-
ability beyond these limits is associated with special causes of variation.
Any system exhibiting only common cause variation is considered statisti-
cally controlled. It must be noted that the con trol limits are not defined by
management or customers: they are calculated from the data using statis-
tical theory. A control chart is a practical tool that provides an operational
definition of a special cause. That is, we cannot determine which causes of
variation in the process are special causes until the control chart identifies
them as such.
There are two broad categories of control charts: those for use with
variable or contin uous data (e.g., measurements) and those for use with
attributes data (e.g., counts).
The basis of all control charts is the rational subgroup. Each plotted point
on the chart contains the data from a single subgroup. Subgroups are con-
sidered rational subgroups if “all of the items are produced under condi-
tions in which only random effects are responsible for the observed
variation” (Nelson, 1988). We often form rational subgroups using con-
secutive items from the process, or items that are representative of the pro-
cess during a short time period. This reduces the likelihood of special
causes of variation occurring within the subgroup. The within-subgroup
variation is used to define the control limits, which provide our estimate of
the common cause variation in the process (i.e., the longer-term between-
subgroup variation), so it is important to exclude special causes of varia-
tion in this estimate.

09_Pyzdek_Ch09_p151-208.indd 163 11/21/12 1:42 AM