Key Points

• This type of cause can be applied to data points outside of established control limits.
• It can be a beneficial or negative cause, depending on circumstances.
• Learning how to account for and use this cause is just part of standard operational procedure.

Assignable cause, also known as a special cause, is one of the two types of variation a control chart is designed to identify. Let’s define what an assignable cause variation is and contrast it with common cause variation. We will explore how to know if your control is signaling an assignable cause and how to react if it is.

Overview

A control chart identifies two different types of variation: common cause variation (random variation resulting from your process components or 6Ms) and assignable or special cause variation.

Assignable cause variation is present when your control chart shows plotted points outside the control limits or a non-random pattern of variation. Since special cause variation is unexpected and due to some factor other than randomness, you should be able to assign a reason or cause to it. 

When your control chart signals assignable cause variation, your process variable is said to be out of control, or unstable. Such variation signals can be identified by the use of the Western Electric rules, which include:

•   One point outside of the upper control limit or lower control limit
•   A trend of 6 or 7 consecutive points increasing or decreasing
•   A cycle or repeating pattern
•   A run of 8 or more consecutive points on either side of the average or center line.

These variations can be attributed to a defect, fault, mistake, delay, breakdown, accident, and/or shortage in the process. Or it can be a result of some unique combination of factors coming together to improve the process.

When these causes are present, your process is unpredictable. The proper action and response is to search for and identify the specific assignable cause. If your process was improved as a result of your assignable cause, then incorporate it so that the cause is retained and improvement maintained. If your process was harmed by the cause, then seek to eliminate it.

Expecting the Unexpected

None of us can tell the future, at least, not that I know of. But as it stands, there are going to be circumstances out of your control in any given situation. Power outages, changing suppliers, machines breaking down, and so forth can all be valid reasons for this cause. However, positive results are also applicable, like products becoming a smash hit due to an unintended functionality.

3 Benefits

Assignable causes can be good or bad. They are signals that something unexpected happened. Listen to the signal.

1. Signals Something Has Happened 

Special or assignable cause variation signals that something unexpected and non-random has occurred in your process.

2. Specific Cause

By investigating and identifying the specific cause of your signal, you can narrow in on your next steps for bringing the process back into control.

3. Can Become a Common Cause Variation 

Good news! You found that your cause for lowered production was a power outage. Unfortunately, you may not be able to stop power outages in your community. If nothing is done, your assignable cause becomes a common cause. 

You might not be able to stop power outages, but could you install a backup generator? Then, if the generator doesn’t kick on, you will have an assignable cause you can do something about.

Why Is It Important to Understand?

Interpreting what an assignable cause tells you is important to understand. 

Provides Direction for Action

Since this cause can be a signal of something good or bad, you need to understand the different actions. Don’t ignore special or assignable causes. 

Not Every Unusual Point Has an Assignable Cause

While at your favorite casino, you may throw a pair of dice at the craps table. Is there an assignable cause for throwing an 11 or a 10, or is it a random variation? No, you would expect the process of rolling a fair pair of dice to show 10s and 11s.

What about a 13? That would be unexpected and probably the result of something unusual happening with the dice. The same is true for your process. Don’t assume an assignable special cause unless your control chart signals it. 

Useful for Determining Whether Your Improvements Worked

When you improve the process, your control chart should send signals of special cause variation — hopefully in the right direction. If you can link that signal to the specific assignable cause of your improvement, then you know it worked. 

An Industry Example

The accounts receivable department of a retail chain started to get complaints from its customers about overbilling. Fortunately, the manager of the department had participated in the company’s Lean Six Sigma training and had been using a control chart for errors.

Upon closer review, she noticed that errors seemed to occur more on Fridays than the rest of the week. The chart showed that almost every Friday, the data points were outside the upper control limit. She was concerned that nobody was identifying that as a signal of special cause.

She put together a small team of clerks to identify why this was happening and whether there was an assignable reason or cause for it. The assignable cause was determined to be the extra workload on Fridays. 

The team recommended a change in procedure to better balance the workload during the week. Continued monitoring showed the problem was resolved. She also held an all-hands meeting to discuss the importance of not ignoring signals of special cause variation and the need to seek out a cause and take the appropriate action.

3 Best Practices

Signals of special cause variation require you to search for and identify the cause.

1. Document Your Search

If you’ve identified the assignable cause, document everything. If this cause happens again in the future, people will have some background to act quickly and eliminate/incorporate any actions.

2. Quickly Identify the Cause 

Time is of the essence. If the cause is resulting in a deteriorating process, act quickly to identify and eliminate the cause. The recommendation is the same if your cause made the process better, otherwise, whatever happened to improve the process will be lost as time goes by.

3. Don’t Ignore Signals of Assignable Cause

Even if you get a single signal of a special cause, search for the assignable cause. You may choose not to take any action in the event it is a fleeting cause, but at least try to identify the assignable cause.

Other Useful Tools and Concepts

Are you hungry for more? While we’ve discussed the unexpected, what about testing for it? The Kruskal-Wallis test is a non-parametric test that allows for fast results with data points that fall outside of your normal ranges.

Additionally, if you’re looking to implement product improvements fast, you might consider the Kaizen Blitz. This is a breakneck approach to process improvement and is highly situational. However, in the right hands, it’ll have you right where you need to be.

Final Thoughts

All processes will exhibit two types of variation. Common cause variation is random, expected, and a result of variation in the process components. Special cause variation is non-random, unexpected, and a result of a specific assignable cause. 

If you get a signal of special cause variation, you need to search for and identify the cause. Once found, you will either seek to incorporate or eliminate the cause depending on whether the cause improved or hurt your process.

The image featured at the top of this post is ©Juice Flair/Shutterstock.com.