Special Cause Variation

Last Updated on 13 September 2023

There will frequently be variation in your processes. There are two types of variation; common cause and special cause variation. Special cause variation is large variation away from the ‘normal level’ of variation that is the common cause variation. These are the ones that are worthwhile removing as part of a process improvement project.

Special Cause Variation is the variation in the system that is worth investigating.

What is special cause variation?

You will have an allowable band in your measurement of ‘normal’ variation that is to be expected.

The special cause variation are those that lie outside those bounds, which are from causes other than standard variation. These causes should be removed from Six Sigma projects, reducing the variation down to just being common cause variation.

Control charts

This concept is used in control charts. These have a bar of expected variance, which is the ‘common cause’ variation. It’s not usually worth investigating the ones within the bounds, as they are of minimal effect, and it isn’t cost effective to investigate.

The measurements outside the control bar bounds are unexpectedly large variations. These are the special cause variation that need investigating.

A process is considered to be in control if variation is predictable, rather than there is no variation. This means that you only need to tackle special cause variation to get your process ‘in control’.