Page 403 - Six Sigma Demystified

Part 3 S i x S i g m a To o l S 383

beads is different every time. The control limits tell us that we should expect
anywhere between 0 and 21 percent red beads on a given sample of 50 beads.
When the process is in control, we can use the control limits to predict the
future performance of the process. For example, we might use the chart to
predict the error levels, which tell us something about the cost to produce the
product or deliver the service. If the process variable is the time to deliver the
product or service, we can use the chart to plan resources, such as in a doctor’s
office, or deliver product, such as in a warehouse.
Another fundamental benefit of a control chart is its ability to provide an
operational definition of a special cause. Once we accept that every process
exhibits some level of variation, we might consider how much variation is natu-
ral for the process. If a particular observation seems large, is it unnaturally large,
or should an observation of this magnitude be expected? The control limits
remove the subjectivity from this decision and define this level of natural pro-
cess variation.
When a sample subgroup is beyond the control limits, we say that it is a
result of a special cause of variation. In other words, something unexpected
happened at this point in time (or just prior to this point in time) to influence
the process differently from what we’ve seen before.
Although we may not be able to immediately identify the special cause in
process terms (e.g., cycle time increased owing to staff shortages), we have
statistical evidence that the process has changed. This process change can occur
as a change in process location or a change in process variation.
A change in process location is sometimes simply referred to as a process shift.
For example, the average cycle time may have changed from 19 to 12 days.
Process shifts may result in process improvement (e.g., cycle time reduction)

or process degradation (e.g., an increased cycle time). Recognizing this as a
process change rather than just random variation of a stable process allows us
to learn about the process dynamics, reduce variation, and maintain
improvements.
The variation in the process also may increase or decrease. This is evidenced
by a change in the width of the control limits. Generally, a reduction in varia-
tion is considered a process improvement because the process then is easier to
predict and manage.
In the absence of control limits, we assume that an arbitrarily “different”
observation is due to a shift in the process. In our zeal to reduce variation, we
adjust the process to return it to its prior state. For example, we sample the
value highlighted in the leftmost distribution in Figure F.49 from a process that