iSixSigma

MSA Question

Six Sigma – iSixSigma Forums Old Forums General MSA Question

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #38197

    KABKUNG
    Participant

    Hi All,
    Could you please help me to clarify my question?
    1.The recommended discrimination is at most one-tenth of six times the total process standard deviation. Please help to explain in this sentence.
    2. Do we need to perform GR&R for all operetor who use the measurement?
    3. Do we have any method to verify inspector for finding the different between each inspector.
     

    0
    #114047

    Dog Sxxt
    Participant

    Go do your homework first and read about the basic MSA before asking.

    0
    #114053

    miranax
    Participant

    A new trend has arisen in this forum that is bothering me.  I’ve read several un-polite responses following the line of “this is not a place where you ask how questions about homework”.  Pity.  I have posted questions that some all-knowing-blue-bloods may have considererd childish, but I have never, ever received such a harsh response as this one. 
    It has been my experience that on several BB trainings we are given the tools of how to do things, but not the technical explanation as to why certain assumptions must be verified or met.  Not every person that undergoes the BB training has had previous statistical training.  Therefore, I consider that KABKUNG has an honest, legit question.
    Kabkung, I will be addressing your concerns tomorrow.  I’m a bit short on time now.  Don’t worry, most people in this forum are actually very helpful lads.  Jerks are the minority.  And before anybody lashes out at me: The note next to the IP address should be applied to people who treat other participants in the forum as idiots.
    Nevertheless, I do recommend that you make a quick search of previous postings on this forum.  Perhaps you’ll find what you were looking for.

    0
    #114055

    Crunchy
    Participant

    Yeah, it was a pretty easy question dude…Im with Dog S**t on this one.

    0
    #114071

    GGBB
    Participant

    Well “Dog”s generally bark but say very less…
    This is supposed to be a helpful forum. If some of you feel very offended to read very basic or elementary questions, kindly ignore them, you need not respond to every posting. Kindly post if you have an answer or you are willing to give one.
     

    0
    #114081

    Peppe
    Participant

    GGBB agree with you.  There a lot of are very funny chat room in the web that treat different topics.
    Rgs, Peppe
     

    0
    #114083

    Mikel
    Member

    I agree with Dog – go learn something, hell anything, before asking for help.
    i.e. read the MSA manual, do a web search, ….

    0
    #114084

    Peppe
    Participant

    This forum is a web search, maybe not the right one …..
    Rgs

    0
    #114094

    Solo
    Member

    K,
    1.  6 sigma covers 99.99% of the variability… You want a measurement resolution of at least 10 distinct categories within that variability…
    2.  No
    3.  I have no idea what you are tyring to ask.
    -Solo

    0
    #114204

    Oman
    Participant

    Solo and Crunchy you crack me up!  You tell the guy it’s easy and you don’t provide the answer and when you do it’s wrong!!!  Six sigma covers 99.73% of the variation.  AIAG states that you want a minimum of 5 distinct categories not ten.  If 6-sigma = 0.01″ then a tenth of that is 0.001″ for your discrimination or resolution.
    If the number of operators is two then yeah you would be including all operators if you found your measurement system to be suspect.  If you have more than a few operators and are confident in the RR and that all operators could be trained as well as the operators that participated in the RR, then there would be no need to perform the RR on all operators.  Of course it’s really going to be dependent on the amount of resources expended to perform such an RR versus the benefit of the data from including all the operators now isn’t it?
    The third answer is yes, look at the interaction from the parts and operators and see how they performed against each other.  If someone has a problem then you can go back and review the data to try and determine where an errant operator went wrong.
    Oman

    0
Viewing 10 posts - 1 through 10 (of 10 total)

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