Six Sigma special cause variation is intermittent variation attributed to assignable events. Control charts are often used to distinguish between special cause variation and common cause variation. If the process measurement control chart shows a plotted point or non-random pattern, it is called special cause variation.

A process measure is considered unstable or out of control when a control chart shows signs of special cause variation. The following are common special cause variation signals:

  • An exception to the upper limit or below limit
  • A trend: 6 to 7 points increasing/ decreasing
  • A repeating or cycle
  • A run is 8 points or more on either side the average

Special causes of variation can be attributed to any defect, fault, mistake or delay in the process. Process quality can be unpredictable when special causes are present.

Drawbacks to special cause variation

If you don’t plot the control chart in real-time, it can be difficult for you to identify the source of a particular cause. Control charts made from historical data will not help you in your search for the cause of the special causes unless you have good memories and annotated data.

It is almost certain that a process measure will be out-of-control if it has not been charted before. You will often see many special causes when you first begin to study a process using a control chart. Start a study on the critical process components to find the source.

If a source of six sigma special variation cause cannot be found, it will become a common problem. The special causes become more common and less rare over time. These causes then multiply the variation of common or natural causes.

What is the significance of six sigma special cause variation?

Quality is defined as the minimum deviation from a target. A control chart can be used to determine if process variation is increasing, or decreasing, or if the center of the target is shifting over time.

A process component that is or has been changing is a special cause. The source of the special cause can be investigated:

  1. You will know when it is time to improve or adjust the process.
  2. Avoid missing an opportunity to improve the process. You will never know what to do if the same problem is repeated.
  3. Give data to evaluate or suggest process improvements.

If there is no cause for the variation, it is safe to leave the process alone. Tampering is the act of making process changes without a special cause is called Tampering. It can increase variation and lower the quality of the process.

Three best practices for thinking about special cause variation

You need to take action to bring your process under control.

Identify the source

If you notice a six sigma special cause for variation, it is important to quickly identify the source. It is a good idea to see if any process components have changed since the special cause occurred. You could also ask process experts for reasons the special cause samples weren’t under control.

A gage that is not calibrated could cause a rise in screw thickness.

Improve the source

Make improvements to the source of variation due to special causes. Watch what happens when the thickness samples are plotted after you have fixed the problem (such as recalibrating the gage). You know that you have found the problem when the plot moves toward stability.

Everything should be documented

You should identify any recurring special causes, and the sources. Then create a control plan to help process operators know what they can do if they encounter the same special cause again.

The control plan could direct the worker to calibrate the next time the screw thickness trend up. This would send the process back towards stability.

Final thoughts about special causes

Variability is a part of every process measurement. You will never achieve zero variability. It is essential to be able to recognize the nature of variability in order to improve and control your processes.

You can use the six sigma special cause variation signal to identify the critical process components that are causing variation and need improvement. To unlock the path to process control, use special cause variation