Understanding control chart and measurement capability


Starting to get Involved
I thought I had a pretty good handle on this but after trying to explain to others the risks of control charting without practical understanding, I realize I need more help to explain clearly:

1) Thinking about the real values. Is there really a difference to the customer/product/process if the difference is X? Just because the chart tells you that, think about what that means...

2) This is kind of a subset of 1): If you can only measure to half a gram and the control chart range is 3 grams should you be concerned when the measurement is just out of range because of the uncertainty of the measurement?
Where can I find information to better understand this (measurement uncertainty and impact on SPC)?

3) Reacting without using rules, "yes that is the highest value for the year, but within the control chart - normal variation vs. special circumstance." This does not seem to come across when I try to explain it.

I really appreciate your help!


Forum Moderator
I am not certain what you are asking in question 1.

Regarding your second question, the control limits include the measurement variation. Therefore, a point outside the control limits has exceeded both the common cause variation of the process AND the measurement variation.

For question 3, there is a good illustration by Deming on the impact of "tampering", which is reacting to a process that is still in control, and how this actually increases variation.

Bev D

Heretical Statistician
Super Moderator
I would add to point 3 to not forget that if your process is 'in control' but still giving you 'bad' results you can and should investigate to understand the cause of all of the ("common cause" variation (as opposed to just one result, aka "assignable cause")

..and clarity around your first question will help us
Top Bottom