As a part of validation, should i include the acceptance criteria of CPk index >1.33?

ifserav

Involved In Discussions
Dear all

I have a question, as a part of validation, should i include the acceptance criteria of CPk index >1.33?
Only in this way can i conclude a process is validated?

Are there any other way to show it?

Thanks in advance

Sergio Ávila:thanks:
 

Bev D

Heretical Statistician
Leader
Super Moderator
Re: As a part of validation, should i include the acceptance criteria of CPk index >1

How about with actual data? I know some FDA reviewers are now asking for Cpk values but these are fraught with so many issues.

Validation of a process is intended to demonstrate that with the inputs properly controlled within known tolerances the output will be acceptable and stable. OQ verifies that process yields acceptable results across the allowable range of input conditions. PQ verifies that the process can create acceptable sequential lots under normal production operating conditions. The combination of these two studies demonstrates the validation and capability of the process. to reduce this data to a single index is silly.

And what will you do if you have a non-homogenous process? it may be perfectly stable and capable but the Cpk will be artificially low...

and what about your specifications? were they engineered to avoid tolerance stackups? so that if a part is in spec then it will perform perfectly well? Or were they created extra tight to provide you margin? how does a Cpk>1.33 help you?

The original intent of Cpk was NOT based on a prediction of defect rates. It was intended simply to provide an indication of how many parts were near the specification limits (back when spec limits were not well engineered). the idea was to drive reduced variation. Some specs are properly derived, some are not. Some characteristics are subject to stackups, some have a continual loss of 'quality' the further they get form the target (like visual characteristics), some aren't. which do you have? Does Cpk >1.33 make sense for your situation or is it just a convenient rubber stamp?
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
Re: As a part of validation, should i include the acceptance criteria of CPk index >1

Not only that, Cpk only applies to process outputs (not print specifications) that have target values centered within a bilateral tolerance with an expected normal distribution (and it not NOT expected from all processes.) Do you know by looking at your print what the process distribution is? No. So, how can you demand that as a requirement?
 

Golfman25

Trusted Information Resource
Re: As a part of validation, should i include the acceptance criteria of CPk index >1

So where do you guys learn the finer points of spc/gage RR and cpk? I need reference materials to show people who attend a 2 hour class and suddenly think they are "experts."
 

Bev D

Heretical Statistician
Leader
Super Moderator
Re: As a part of validation, should i include the acceptance criteria of CPk index >1

So where do you guys learn the finer points of spc/gage RR and cpk? I need reference materials to show people who attend a 2 hour class and suddenly think they are "experts."

1. Experience - we have used these methods for years, figured how they are valuable and not valuable. we don't treat them like some cookbook approach
2. Thinking - we THINK about what is actually happening vs what the 'theory' is
3. Reading - we research this stuff because we do 1&2 and realize we don't understand yet. then we repeat 1 and 2.

Its hard to find the good stuff for all of the **** out there. 90% of what we find through the web is just the same pabulum surface silly stuff over and over again. But we've been able to understand and synthesize what we have researched and what we've experienced.

a short list of GOOD references for these topics is below.
hope this helps.

REFERENCES
SPC
Anything by Donald Wheeler (his articles in Quality Digest, Quality and on his web site are free and his references and experience go back to the origins of these things; he skips those people who have misinterpreted the original intent of SPC)
Wise, Stephen A.; Fair, Douglas, C., Innovative Control Charting, ASQ Quality Press, 1998

MSA (guage R&R)
Donald J Wheeler, Craig Award Paper, “Problems With Gauge R&R Studies”, 46th Annual Quality Congress, May 1992, Nashville TN, pp. 179-185.
Youden, William John, “Graphical Diagnosis of Interlaboratory Test Results”, Industrial Quality Control, May 1959, Vol. 15, No. 11
Donald S. Ermer and Robin Yang E-Hok, “Reliable data is an Important Commodity”, The Standard, ASQ Measurement Society Newsletter, Winter 1997, pp. 15-30.
Donald J Wheeler, “An Honest Gauge R&R Study”, Manuscript 189, January 2009. https://www.spcpress.com/pdf/DJW189.pdf

Cpk/Ppk
Sullivan L. P., “Reducing Variability – A New Approach to Quality”, Quality Progress, July 1984
Gunter, Bert, “The Use and Abuse of Cpk, Parts 1, 2,3 and 4”, Quality Progress, January 1989, March, 1989, May 1989, July 1989
Pignatiello, Joseph J., Ramberg, John S., “Process Capability Indices: Just Say “No!””, ASQC Quality congress Transactions – Boston, 1993
Kane, Victor E., “Process Capability Indices”, Journal of Quality Technology, Vol. 18, No. 1, January 1986
 
Last edited:
Top Bottom