Control charts are used to monitor processes and ensure the processes are in control, but all processes are subject to degrees of variation. Most of the variation is inherent in the process and the surrounding environment and is referred to as common cause variation. Variations not normal to the process are referred to as special cause variation. To better understand special and common cause variation, they are applied to an everyday occurrence.
One example is of a morning commute to work. Most people know when to leave their house in order to get to work on time, but the amount of time needed for the commute varies slightly from day to day. If the average commute is 50 minutes, maybe the fastest it can be done is in 46 minutes, while the slowest normal trip takes 57 minutes.
This illustrates an example of natural variation because everyone goes through this exercise to get to work on time. This kind of variation is called common cause variation because all the things that cause the variation, such as road conditions, traffic patterns and distance are inherent in the process; they affect everyone in the process the same way.
Then one day the trip takes 74 minutes. What happened? Maybe there was a major accident that blocked all lanes going both ways for 30 minutes. Maybe a freight train or bad weather slowed the commute, or maybe the car had a flat tire. These are examples of special cause variation.
These types of variation are intuitively evaluated when it comes to the morning commute. If one lives in Chicago and the morning commute includes the Dan Ryan and Edens Expressways, chances are the commuter expects a traffic snarl because accidents on these highways happen frequently and have a common impact on the morning commute.
But someone living in a small town probably does not plan to encounter a major traffic accident on the way to work. A 20-minute delay because of a traffic accident would be a special cause for this system.
The response to these two types of variation is critical; the reaction for each should be different. If a traffic snarl because of an accident is common cause variation, the response should be something like, “I need to leave early enough so that I get to work on time in spite of the accidents that are going on all around me. In fact, accidents are more likely when people hurry, so I’ll leave early enough to reduce my risk of being involved in the accident.”
On the other hand, if the accident is a special cause event, because, for example, one lives in a small town where major accidents are rare, the best thing to do if caught in a traffic snarl is to help whoever is in the accident and then apologize when showing up late for work. It would be foolish to leave earlier every day in anticipation of coming upon a major accident.
There are appropriate responses for common cause variation and appropriate responses for special cause variation. The irony is, however, that the two responses are diametrically opposed. The correct response for common cause variation makes special cause variation worse, and the correct response for special cause variation makes common cause variation worse.
Applying common cause solutions to special cause problems, such as leaving early to avoid accidents when living in a small town, inevitably drives up costs and probably does not prevent the special cause from recurring. By the same token, applying a special cause solution to a common cause problem, such as driving recklessly down the Dan Ryan at rush hour because one didn’t leave enough time for the commute, creates more special cause problems.