iSixSigma

Cpk for non-normal data

Six Sigma – iSixSigma Forums Old Forums General Cpk for non-normal data

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #44260

    Fontanilla
    Participant

    Everytime I try to do a capability analysis on my non-normal data in Minitab vs 14 I only receive Ppk, not Cpk.  Why is this?
     
    Please respond.

    0
    #141536

    Brit
    Participant

    A Cpk value/analysis assumes normal data is being used.

    0
    #141537

    Fontanilla
    Participant

    Thank you.  That is what I thought.  A black belt told me I can still use a Cpk even though my data is not normal.

    0
    #141576

    Savage
    Participant

    Your black belt is correct. You can still calculate Cpk when data is not bell-shaped. Besides, you will never get a perfectly normal distribution anyway. So if this is true, then aren’t all distribution non-normal? But I digress …
    Anyhow, you can use other techniques to get a Cpk from data that is not expected to be bell-shaped. I.e. Pearson curve-fitting routines and Johnson data transformations are two such techniques.

    0
    #141582

    Brit
    Participant

    As was mentioned by Matt – you can still calculate the Cpk (and right – data is normally (he he) not normal.  However, it was designed specifically for normal data.  The problem is the accuracyof the value.  You can use Cpk as a sandy check, but I wouldn’t stake my life on it or report the value to customers if they require one.  I would use more of a Ppk number or go with the suggestions by Matt.
    Also – as Matt ponted out, since most data is non-normal, a lot of suppliers who provide a Cpk value may be doing so erroneously – just a thought.

    0
    #141592

    Tony Bo
    Member

    What are your thoughts on using discrete process capability to calculate your sigma on non-normal data? 
     N total sample, X outside specs, X opportunity….etc….

    0
    #141647

    Tony Bo
    Member

    What are your thoughts on using discrete process capability to calculate your sigma on non-normal data? 
     N total sample, X outside specs, X opportunity….etc….

    0
Viewing 7 posts - 1 through 7 (of 7 total)

The forum ‘General’ is closed to new topics and replies.