iSixSigma

NONPoission Distribution

Six Sigma – iSixSigma › Forums › Old Forums › General › NONPoission Distribution

Viewing 13 posts - 1 through 13 (of 13 total)
  • Author
    Posts
  • February 16, 2004 at 7:37 am #34597

    littleann
    ★ 10 Years ★
    Participant

    thank Tim, Isabel, Andy ‘s reply to my question about Poission distribution.
    I tried Chi squre test, and get rejected.  and mean and variance are not even closed to  equal.  My data is about nonforminities–defects of product, so they are supposed to distribute  poission.
    Now i just dont know what to do. can i still use c chart or u chart on this data?
     
     

    0
    February 16, 2004 at 1:37 pm #95549

    Anonymous
    ★ 20 Years ★
    Guest

    Littleann,
    I believe that what you really need a physical model. If you can provide me with an email address, I’ll describe a simple experiment from which you can derive a statistical model for %defective, DPU, and DPO; and how to distinguish between random and non random defects.
    Cheers,
    Andy

    0
    February 16, 2004 at 2:58 pm #95553

    Savage
    ★ 10 Years ★
    Participant

    Andy,
    If it’s not too much trouble, I would like to see an example of the model you described.
    Thanks,
    [email protected]

    0
    February 16, 2004 at 3:08 pm #95554

    Dr. Steve W.
    ★ 10 Years ★
    Participant

    If so, it is called over dispersion. There are a lot of text books out there that talks about how to handle overdispersion. Overdispersion is very common situation for sount data.

    0
    February 16, 2004 at 3:11 pm #95555

    Dr. Steve W.
    ★ 10 Years ★
    Participant

    long as your subgroup size is 4 or more. The C chart is a little bit more sensensive to deviation from Poission distribution.

    0
    February 16, 2004 at 6:48 pm #95577

    Anonymous
    ★ 20 Years ★
    Guest

    No problem .. give me a couple of days.

    0
    March 8, 2004 at 8:50 am #96545

    littleann
    ★ 10 Years ★
    Participant

    Dr. Steve W.
    Thanks. you help a lot.
    If  the variance is much larger than mean, we may use square root of variance instead of square root of mean when we calculate sigma in u chart.
    do you think it is a good idea?

    0
    January 20, 2007 at 7:04 pm #150865

    Anonymous
    ★ 10 Years ★
    Participant

    Dr. Steve,The post is 3 years old, but I would be interested in your opinion of Laney’s P’ and U’ charts to handle overdispersion:Laney, David B., P-Charts and U-Charts Work (But Only Sometimes),  Laney, David B., Improved Control Charts for Attribute Data, Quality Engineering 2002;14:531–7.. 

    0
    March 22, 2007 at 6:20 pm #153770

    Martínez
    ★ 10 Years ★
    Participant

    Have you incorporated the Laney P’ chart? Would like to see how you applied it.
    I was looking into the Laney P’ chart as well. If anyone can suggest a place to get more information, please let me know.

    0
    March 23, 2007 at 10:40 am #153808

    Anonymous
    ★ 10 Years ★
    Participant

    MBB, Here are some references:Laney, David B., P-Charts and U-Charts Work (But Only Sometimes),  Laney, David B., Improved Control Charts for Attribute Data, Quality Engineering 2002;14:531–7. M A Mohammed and D Laney, Overdispersion in health care performance data: Laney’s approach, Qual. Saf. Health Care 2006;15;383-384. Laney’s P’ and U’ have been implemented in SigmaXL Version 5.1.

    0
    March 23, 2007 at 10:55 am #153809

    Anonymous
    ★ 10 Years ★
    Participant

    Sorry for the duplicate information.  I did not read my earlier post.
    The health care paper is a good example of use.

    0
    March 23, 2007 at 2:36 pm #153823

    Sigmordial
    ★ 10 Years ★
    Member

    Howdy Dr. Steve,
    Interested in your thoughts relative to the overdispersion Little Anne is encountering and if it is due to a non-homogeneous Poisson process.  If so, could she revisit her subgroup composition in the rational subgrouping sampling strategy to remedy this for her control charts?

    0
    March 24, 2007 at 6:11 pm #153853

    Ken Feldman
    ★ 10 Years ★
    Participant

    Is this the old Dr. Steve formerly of BOA affectionately known as JarJar????

    0
  • Author
    Posts
Viewing 13 posts - 1 through 13 (of 13 total)

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