Forum for Sampling Practices

Scenario 1:
Let's say I have a process in control and the Ppk (long term) is 1.5, what would be a reasonable sampling plan? Assume I have years worth of data showing good performance.


Scenario 2:
A new process has started up and preliminary data shows a Cpk of 1.5, what would be a reasonable sampling plan?

Hi there,

I am actually trying to work on conecting process capability to a sampling plan, but it is difficult. If you have outlayers or the process is not statisticly stable, then you need to test every product/service. If it is statisticly capable.. Well, then you do not need to messure at all. I know this is not much of an answer, but as I am working on it I'll come back with a better one if I can.