• ## Follow Blog via Email

Join 777 other followers

## 5th Edition PMBOK Guide–Chapter 8: Control Charts

1.  Introduction–Control Limits vs. Specification Limits

The control chart is an important tool of quality control.   The PMBOK Guide definition is “A graphic display of process data over time and against established control limits, which has a centerline that assists in detecting a trend of plotted values toward either control limit.”

The definition implies that on either side of the centerline, there is a control limit; one is called the upper control limit, and the other is called the lower control limit.

As I mentioned in my previous post, the control limit exists so that the process stays within the specification limit.    The analogy that comes to mind is that of the lane marker on a highway and the guard rail at the side of the road.    The guard rail is like the specification limit, in that if you go outside that limit, you are definitely “off road” and this is definitely a situation to avoid.    But in order to stay within these outer limits, the control limits in this case are like the lane markers, assuming that there is some sort of shoulder between the lane marker and the guard rail.    If you can steer your car so that you are within the lane markers (control limits), you will never have to worry about hitting the guard rails (specification limits).

2.   Assignable or special cause vs. random variation

A measurement will have some sort of variation which is due to random variables.   In practical terms, this means that the measurement will be moved off the centerline in one direction as often as it is moved in the other direction.     A common cause is the usual variation in a system, the “background noise”, if you will.

However, if there is a special or assignable cause, it will move the measurement off the centerline in a way that is not random.    The practical upshot of this is that there are certain rules about control limits that tell if a process is being affected by a special or assignable cause in such a way that it is out of control.

3.   Control chart rules

One of the first rules is that no measurement should be outside the upper control limit or the lower control limit.   This one is fairly obvious; the reason why the control limits are put in place is precisely because going outside of them indicates the process is out of control.

The second rule is the “rule of seven,” which says that if seven consecutive points are above or below the centerline, this must be due to some process which is skewing the values to one side of the center line.

The third rule is that a certain number of points are outside certain standard deviations of the measurement.    One variation of this rule is:

• one point over 3 sigma
• 2 out of 3 points over 2 sigma
• 4 out of 5 points over 1 sigma

The idea here is that the number of points over 1 sigma, for example, should be only 31% or roughly 1 out of 3.   If 4 out of 5 points are over 1 sigma, or around 80%, then the number of points over 1 sigma is higher than it should be, and there is probably a special or assignable cause for this.

These rules of thumb give a project manager a way to look at control charts to detect if there is a special or assignable cause.   If the control charts indicate there is some sort of special or assignable cause, the next task is to find out what that cause is.   There are other quality tools for that.    So the control chart is the beginning of the story of shifting from monitoring to controlling quality, not the end of it.

This is the last post discussing Chapter 8 of the guide on Quality Management.   The next post will start discussing Chapter 9 on Human Resource Management.