iSixSigma

Variation (Special Cause)

Definition of Variation (Special Cause):

« Back to Glossary Index

Common cause variations are those variations that happen as part of natural patterns. These kinds of variances are inherent in the process, consistent, and normal. Special cause variations are quite different than these sorts of variances.

Overview: What is Variation (Special Cause)?

These are the variances that are unexpected, non-quantifiable, and unusual. Special cause variations are those that in a process have not been encountered before. They are caused by unpredictable factors. Examples of special cause variations include machine faults, power surges, operator absences, and computer faults.

3 drawbacks of special cause variations

There are some drawbacks to special cause variations that should be acknowledged:

1. They can be difficult to prepare for

Special cause variances can be so random that it can be extremely difficult to adequately prepare for them.

2. They may cause serious problems

Some special cause variations can be eliminated due to them being a technical issue. If preventable, they should be as they can cause major problems for the business.

3. The source can be difficult to find on a control chart

The source of special cause variations can be difficult to spot on a control chart if you are not plotting in real-time. To find the source of the special cause, you would likely have to have annotated data or an exceptionally good memory.

Why are special cause variations important to understand?

For the following reasons, special case variations are important to understand:

1. Statistical instability

If you have a chart that only has common cause variations, it means that your process is likely “statistically stable.” Understanding special cause variations on your chart, helps you recognize the inverse. It means that your processes are “statistically unstable” and that modifications may need to occur.

2. You can spot them on a control chart

By understanding special cause variations, you can probably pretty easily spot them on a control chart. Should the measurements of a process be distributed normally, it is almost guaranteed that a measurement will fall within plus/minus three deviations. Measurements that fall outside these limits are likely to be special cause variations.

3. Understanding them makes them easier to investigate

If you get a measurement on a control chart that appears to be a special cause variation, the expectation is that it will be investigated, there will be a root cause analysis, and appropriate measures will be taken.

An industry example of a special cause variation

A project manager has been leading the test-drilling for a new site where it is believed that there is likely a significant amount of untapped crude oil. The test-drilling was supposed to last for a week, but one of the drills malfunctioned, which caused a delay. Instead, the test-drilling took a total of 30 days once the faulty drill issue was addressed. The malfunction is an example of a special cause variation.

3 best practices when thinking about special cause variations

Here are a few practices to bear in mind when it comes to special cause variations:

1. Countering special cause variations

Contingency plans can be used to counter special cause variations. With this strategy, additional processes are incorporated into operations that prevent or counter a special cause variation.

2. Recognize them on a control chart

Remember that if it falls plus/minus three deviation limits of a measurement on a control chart, it is probably a special cause variation.

3. You can avoid over-tampering

If you do not understand how variation works, you could feasibly over-adjust, which can then lead to even more variation every time there is a process change.

Frequently Asked Questions (FAQ) about Variation (Special Cause)

1. What is assignable cause?

It is another term for special cause variation.

2. Who introduced the idea of special cause variations?

M. Edwards Deming was an engineer and statistician who is considered a founding father of Total Quality Management. He is credited with being the person that came up with the concept of special cause variation.

3. What percentage of issues relate to special cause variations?

Special cause variations account for less than 10%. Common cause accounts for closer to 90% of variances that may occur.

Eliminating special cause variations

While it can be difficult to predict an initial special cause variation, steps can be taken to help ensure that the same problem does not arise again. Technical improvements, proper training, and other strategies can all help protect your company from repeat incidents.

« Back to Dictionary Index