a question about process capability for nonlinear data

Iheartjmp

Hello JMP community,

 

I am a bit confused about process capability. I thought for a process to be controled, it should have a normal distrabution. However, I have read that is not always true? https://community.jmp.com/t5/JMPer-Cable/Process-Capability-Analysis-for-nonnormal-data/ba-p/38112 The link mentions "several methods have been developed that create analogous process capability indices for nonnormally distributed data" can someone point to these methods?

 

Can a process be non-linear and be in control? If a process is non-linear and not in control is it appropratie to draw conclusions about process capability?

 

 

txnelson

Process data in the real world is not always in a normal distribution.  I suggest that you study the Distribution Platform, and the Process Capability Platform, about how to deal with Process Capability for nonnormal data.  They both have the capability to calculate Cp and Cpk for normal and several of the nonnormal distributions.  The Distribution Platform also has the ability to try to determine what the actual distribution of your data is.