Verification of Corrective Actions based on Quantities


Involved In Discussions

My company is working on improving our 8-D process and one area that we feel we need to strengthen is the verification process. When the issue is part quality related, different 8-D team leaders base their closure off of different things. Some use 3 runs to verify (regardless of run size) some use a percentage of parts based off of EAU quantities, some say one run is enough, etc. I am wondering if anyone has any best practices that they could share on this topic?

Thanks in advance.


Involved In Discussions
It depends, there is not a standard quantity that we use to release from containment. We see how serious the issue is for our customer and we base the added inspection quantity/time based on that. Sometimes it is one or 2 runs, sometimes a quantity. If we have had multiple rejections in the past for the same issue (say oily parts) we may change our processes to always wipe parts. Sometimes our customers specify how long or how many pieces. It really depends on the situation, the cost of the rejection and the impact to the customer.


Forum Moderator
While inspection results are definitely acceptable, you are not limited to them. You could also use the results of SPC and capability studies.

Jim Wynne

Don't get me started. Oh, wait--you did get me started. :bigwave: IMHO, the execrable 8D format should be taken out in the woods and put out of its misery. It just begs for ambiguity and subjective reasoning. Some tips:

  1. Be judicious in your use of CA requests. If you require a PPM level (another thing I hate) of your suppliers of say, 200, don't demand CA for an incident that doesn't break the PPM threshold. You can't, on the one hand, say that a certain level of defects is acceptable and then on the other hand demand CA for an obviously one-off incident.
  2. Look carefully at the identification of the root cause and what was done to stop it from happening or somehow neutralize it. That's what you want to know, after all. Does it make sense? Has the supplier demonstrated that the cause can be turned on and off?
  3. Don't make judgments on things you're not qualified to evaluate. If the supplier's response reads like Mandarin to you (and you don't speak or read Chinese), you're probably not qualified to sit in judgment. Get someone who is.
  4. There is no such thing as a "best practice" that will fit all situations. Complex problems require special, knowledgeable appraisal. You can't look in a reference book to find an answer.
  5. If you can't trust your suppliers to do conscientious CA, you need different suppliers or you need to make allowances when you have no influence.

Bev D

Heretical Statistician
Super Moderator
The verification 'quantity' should be based on the defect rate and/or cycle of the excursion. That is really the only way to determine if the Problem has actually been eliminated or reduced.

Other responders are correct that there is no one single mathematical rule that can be applied. We must apply the correct statistical thinking. And the correct physics thinking in understanding the causal mechanism and understanding that the proposed solutions will prevent re-occurrence of that causal mechanism.

A low defect rate will require higher sample sizes.
For defects that cycle, the control chart should demonstrate a minimum of 3 full cycle time periods before closing the action.


The applied corrective action when based on a process change, you will be interested to see the effectiveness of stability of new established process. This may depend on your consumption and the MOQ and your agreement with the part supplier.
If you want him to process the large quantity and make to you several deliveries, then in reality all the deliveries combines to one run. So it comes down to your risk evaluation and time should not be a limiting factor.


Involved In Discussions
Hi Jim,

I agree with your comments. The question is actually regarding our corrective action verifications for issues noted by our customers so we would be the supplier. I think, as stated by others, the verification will be dependent on situation but wanted to see if there is another, more standardized way to look at it.

Thanks to all for the responses!
Top Bottom