iSixSigma

Stable Process

Six Sigma – iSixSigma Forums Old Forums General Stable Process

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
    Posts
  • #53288

    Doa
    Participant

    Hi To All,
    Although technically unstable: If one has 30 sub groups of 3 ea, and runs X-bar and R charts, and two points are beyond UCL/LCL on the X-bar chart, one which is shifted in a positive direction, that is towards the target and the second point is shifted below the LCL, is it reasonable to label the process stable. The Range chart is fine.
    Thanks  To All
    Marty

    0
    #189467

    Mikel
    Member

    No

    0
    #189469

    Mike Carnell
    Participant

    You went rail to rail in 90 parts what would make you think that process was even close to being stable? Forget the rules just use some basic logic.
    Just my opinion.

    0
    #189474

    Doa
    Participant

    Thanks for response Mike, a little more insight, my course instructor intimated that when a data point is lets say above the UCL and shifted towards the desired or target mean…..not much concern should be given to this point. 
    Given the minitab output and test that were set, I know the point or points by definition make the process “unstable”, I was looking at this from the standpoint of accurately identfying the process as stable or not then on to capability…the instructor was suggesting that at some point judgement calls can be made 
    Thanks again..

    0
    #189476

    Mikel
    Member

    Your instructor is an idiot

    0
    #189478

    Mike Carnell
    Participant

    optomist1,
    I agree with Stan in terms of your instructor and Stan was actually being kind.
    Regarless of who says what rules are coming into play or whatever rational they want to use you simply have yo look at what is happening in a process and realize that when a mean is bouncing around like that there is something in your process that is really screwed up. The good part is that you have a mean that is moving and you can generally look at knob type variables to see what is driving that. That was generally – no guarantees so all you people that are going to post you one example that you saw 10 years ago don’t need to bother.
    Just my opinion

    0
    #189479

    Doa
    Participant

    Hey Mike,
    No arguement from me….thats what this process is all about.  One data point out regardless of where and the process is deemed unstable.
    Thanks to you and Stan.  After all it is Statistics and we must be exacting in our endeavor….
    Regards,
    Marty

    0
    #189481

    Mike Carnell
    Participant

    Marty,
    If you just make stuff better you are going to do what 90% od the people trying to do aren’t getting done and the guys that write your paycheck don’t care how you did it. They will reward you for doing it though.
    Good luck

    0
Viewing 8 posts - 1 through 8 (of 8 total)

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