iSixSigma

Z-value counterpart

Six Sigma – iSixSigma Forums Old Forums General Z-value counterpart

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #40499

    anton
    Participant

    When describing or measuring your current process using continuous data, one may opt to use z values to describe how many units are possibly within or outside  spec limits.    On the other hand, if you are dealing with attributes data, how is this  estimated or best handled?
    I would appreciate your thoughts on this. Thanks.
    Anton
     
     
    /
     
      

    0
    #125860

    “Ken”
    Participant

    Anton,
    Consider using the ratios:
    # Defects / Units Inspected
    # Defectives / Units Inspected
    Ken

    0
    #125862

    anton
    Participant

    Ken, thanks for your inputs. 
    I have been reading your comments lately and I think that with your sharing, you must have helped a lot of 6 sigma practitioners.  Your comments has so much depth and so well put.
    Keep it up Ken.
    Anton
     
     
     

    0
    #125864

    MAK
    Participant

    calculate DPMO (defects per million opportunities) and see corresponding Z value

    0
    #125869

    anton
    Participant

    Mak, thanks.  That looks neat!
    Anton
     

    0
    #125879

    Ken Feldman
    Participant

    Anton, you can also consider using the normal approximation to the binomial and poisson.  This allows the calculating of a z value for both of these discrete distributions.  There are some assumptions involved so be careful.  This lets you use the Z table for equivalent computations so you don’t just have to settle just for DPMO.

    0
    #126035

    Anton Javier
    Participant

    Hi Darth, thanks for the input.  I will try your suggestions later this evening and see if I can crack it. 
    If not, I will post another message to seek help and advice.  I hope that this is ok.
    Anton
     
     

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

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