iSixSigma

Proportion for three level

Six Sigma – iSixSigma › Forums › Old Forums › General › Proportion for three level

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • May 15, 2003 at 3:31 am #32242

    ROSS
    ★ 10 Years ★
    Member

         I’m chosing the best technical method(total three technical methods) for the production yield, as we know, we can use two-proportions test to compare the  two level proportion or chose the better one, but could someone tell us how can we compare the proportion or choice the best one in three level?

    0
    May 15, 2003 at 2:00 pm #85917

    Zilgo
    ★ 20 Years ★
    Member

    The easiest way to accomplish this task is to perform the 2 Proportion test multiple time.  Since there are only three comparisons to perform (1v2, 1v3, 2v3) this isn’t really very tedious or time consuming, as it would be if you had like eight proportions to compare.  This will at least give you the three confidence intervals you need to compare.  Hopefully you will be able to make a decision from those results.

    0
    May 15, 2003 at 3:03 pm #85923

    Carl H
    ★ 20 Years ★
    Participant

    Good question.
    I have used 1 way anova in Minitab and used  one of the optional comparisons methods for the groups.  This provides all the combinations of group comparisons, comparisons with the “control” group or with the “best ofhte rest” (maybe what you are after).
    Carl

    0
    May 16, 2003 at 12:56 am #85950

    ROSS
    ★ 10 Years ★
    Member

    Zilgo Carl:
       Thanks for your answering frist.
       We can perform the 2 Proportion test multiple time(for three levels, it is only three time), but it will be some problems if the lever is more than 3. So, I think , it should has some better tools in Minitab.
       Carl: you suggest me to using the ANOVA for this issue, but I have some confusion, as I know, ANOVE require the data must be the continuos and normal distribution, can ANOVA used in proportion test?

    0
    May 16, 2003 at 8:19 am #85951

    Sigma Singh
    ★ 10 Years ★
    Member

    Tony
    you can use chi sq test in this case. In Minitab this test is availabe in stat menu under tables.
     
    Sigma Singh

    0
    May 16, 2003 at 9:51 am #85955

    ROSS
    ★ 10 Years ★
    Member

    Sigma Singh£º
         I have ever try the chi-suqare test , but still has some problem.
        Supposing my data is as belowing:
       Method    Total   Defect
       1              500      20
       2              400      17
       3              700      35
       How can I use chi-square test to chose the best one?
      
     

    0
    May 16, 2003 at 11:04 am #85959

    Sigma Singh
    ★ 10 Years ★
    Member

    Tony
    You may refer to “help” in the chi2 test.
    In your case enter data in 3 columns:
    Metod 1   method2     method3
    480       383             665     (this data represents ok pcs)20         17                 35     (this refers to non conforming pcs)
    go to stat/tables/chisq test.
    look at the p value (gloden rule never changes – making life simpler for lesser mortals like us)
    In this case it is 0.685. So there is no difference between three methods
     
    Sigma Singh

    0
    May 16, 2003 at 12:42 pm #85968

    ROSS
    ★ 10 Years ★
    Member

    Sigma Singh:
        Thanks for your help.
        For chi-square test, we can get the information that method is the significant X to yield(Y) or not, but what I want to know is which method is the best. Can chi-square test reach the purpose?

    0
    May 17, 2003 at 6:49 am #85993

    Sigma Singh
    ★ 10 Years ★
    Member

    Tony
    If chisq test says is significant – different methods give different yields, probably you can look at values of proportions to know which is largest / smallest.
    Sigma Singh

    0
  • Author
    Posts
Viewing 9 posts - 1 through 9 (of 9 total)

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