Special Cause Variation

Special cause variation is one of the two main categories of variation. Common cause, the other type, is the consistent, recurring fluctuation within a system, sometimes referred to as “noise”.

Special cause variation, in layman’s terms, are the spikes that are caused by problems outside of those that regularly affect a process.

While most people look at special causes as being something different than common causes, in truth, the only real difference is mathematical.

What that means is that how a cause is categorized really has more to do with its magnitude and frequency rather than the actual underlying characteristics of the problem.

Take, for example, a glitch in a machine that causes it to drill holes incorrectly. If the glitch is frequent, it will show up as part of the natural fluctuation of a process. Statistically, that means that the variation is within a control limit that is defined by the standard deviation of a process.

Now, if that same glitch happened infrequently, and the process was stable relatively, it would fall outside of the control limit. It would then be categorized as a special cause.

In practice, though, special causes are good continuous improvement targets because they generally require unusual activities to deal with the problem. That disrupts flow. Responses to common causes, because they are more pervasive in the system, are generally managed by the normal process, and tend to be handled in stride.

Words of Warning

• Don’t fall into the trap that special causes are worse than common causes. The overall impact of common causes, because they happen regularly, can be much higher than that of special causes.
• Don’t categorize special causes by the type of issue that is occurring. Special cause variation is a function of where its impact falls relative to the ongoing performance of a process. If you eliminate one common cause, the standard deviation of a process will get smaller. That means that another common cause, if it falls outside of that tighter control limit, can suddenly become a special cause without anything about it changing.

• Don’t fall into the trap that special causes are worse than common causes. The overall impact of common causes, because they happen regularly, can be much higher than that of special causes.
• Don’t categorize special causes by the type of issue that is occurring. Special cause variation is a function of where its impact falls relative to the ongoing performance of a process. If you eliminate one common cause, the standard deviation of a process will get smaller. That means that another common cause, if it falls outside of that tighter control limit, can suddenly become a special cause without anything about it changing.