In Lean Six Sigma, a control chart can display two types of variation, specific cause variation (points outside of the limits or a nonrandom pattern) and common cause variation.

When the control charts for a process measure show a random pattern or variation, it is considered a common cause variation. All points within the control limits are included. A process measure is considered to be stable or statistically controlled if it shows common cause variation on a control chart.

If you want to reduce variation due to common causes, fundamental changes must be made to critical components of your process, such as machine, material, people, environment, measurement, and personnel.

Three benefits of common cause variation in Lean Six Sigma

If the process is under statistical control

1. In the short-term, the process outcome is predictable

This is a great benefit! You can predict the outcome of a process if it is predictable.

2. Process capability can be evaluated

Process capability Compares common cause variations from your process measure with the specifications and target of the customer or engineer for that measure.

If the process is stable and the common cause variation falls within the limits of the specification, the process is capable.

3. The process can be improved

Stability is the prerequisite for a competent process. You can reduce the variation of a process measurement by fundamentally changing its process components once it is stable. We conducted an experiment to show that certain machine settings can reduce process measure it.

What is the significance of common cause variation?

Tampering is when you treat variation from common causes like it’s from a particular cause. Tampering is not a good idea. It is a waste of time and resources that will increase the variability in the process.

Common cause variation is when the process measurement moves beyond the calculated control limits. To reduce this variation, you will need to make fundamental changes to the process components.

Consider, for example, a stable control chart that shows the number of scratches on a casing. It has a maximum limit of 6 scratches, and a minimum of zero scratches. Over three samples, the average scratch rate was between 3 and 2 to 5.

If you asked your workers “What happened?” when their error rate rose to 5 they would likely reply “nothing new!” because they are tampering. An error rate of 5 is normal as it is within control limits.

Instead, improve the polishing process to make it more efficient. This will reduce variation and eliminate waste from tampering.

Three best practices for common cause variation

It’s a process that’s under statistical control or stable. Common cause variation is used to:

1. Calculate the expected range of values from your process measurement

A control chart of yield, for example, is operating at an average yield of 90% and has a 92% upper limit and an 88% lower limit.

2. If the process requirements are not being met, make fundamental changes to key process components.

Our manager wanted to get a higher yield from her process so she established a more rigorous mechanical maintenance schedule in order to make sure that the machines are operating as they should.

3. Continue plotting the control chart following the change

Our control chart showed that our manager made a change. It showed a shift in the yield (our yield increased to 95%) and/or a decrease in variation (our yield variation remained at +/- 2%) after the change.