deepikanegi
Starting to get Involved
If a process has High Pp and Low Ppk? What does that mean?
Ppk and Pp
- Thread starter deepikanegi
- Start date
The "k" refers to the "centering" of the process......(I am NOT a statistician, so someone will jump in and give a more academically accurate answer (I hope
Essentially if Pp is "good" then your process has a minimal amount of variation....i.e. a high potential. However, if the Ppk is "low" then the process is skewed from center. Something that can often be adjusted/corrected.
Hope this helps.
Be well.
Last edited:
I am not a big fan of the various process capability indexes, but here is a reasonable link:
Process Capability & Performance (Pp, Ppk, Cp, Cpk) (sixsigmastudyguide.com)
According to this and related links in the article, "k" does not have to do with kurtosis, but how centered the data are between the specification limits.
Process Capability & Performance (Pp, Ppk, Cp, Cpk) (sixsigmastudyguide.com)
According to this and related links in the article, "k" does not have to do with kurtosis, but how centered the data are between the specification limits.
Well, this conversation has me rethinking/investigating the source/definition of those individual letters. I found this interesting information...
(broken link removed)
Be well.
(broken link removed)
Be well.
We have discussed this several times. See this explanation.
Now my obligatory rant: What difference does it make? The whole capability index thing is just statistical alchemy. It is nothing more than an attempt to reduce variation to a single number for lazy people. It destroys knowledge and insight. In the very beginning it had a modicum of usefulness. Then the hacks took hold of it and conflated it with defect rates and tried to add in inferential and distributional statistics which are enumerative at best and have no place in industrial quality. The hacks then said you could understand your process capability and performance with just 30 samples. Too many people keep chasing this evil little number that they have no time left to truly understand their processes and improve quality. 
<rant off>
Now my obligatory rant: What difference does it make? The whole capability index thing is just statistical alchemy. It is nothing more than an attempt to reduce variation to a single number for lazy people. It destroys knowledge and insight. In the very beginning it had a modicum of usefulness. Then the hacks took hold of it and conflated it with defect rates and tried to add in inferential and distributional statistics which are enumerative at best and have no place in industrial quality. The hacks then said you could understand your process capability and performance with just 30 samples. Too many people keep chasing this evil little number that they have no time left to truly understand their processes and improve quality. <rant off>
Yes!!!!We have discussed this several times. See this explanation.
In the simplest terms, it probably means that your process is good enough to make the part to tolerance, but you need to center it IN the tolerance. Which is the easy fix: usually a programming offset in CNC machining.
Similar threads
