iSixSigma

Test for Equal Mean

Six Sigma – iSixSigma Forums Old Forums General Test for Equal Mean

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #27794

    wei
    Member

    I want to perform a test of equal mean (t-test) to three incoming lots from vendor. I know that I can use Two-Sample T-Test for test equivelency for two lots, then what about three lots? Is there a particular test for three lots? Thanks!

    0
    #68512

    Ken Myers
    Participant

    Wei,
    To compare more than two means you could use a statistical test called Analysis of Variance or ANOVA.  It’s a very standard test and is explained in most stat texts, as well as supported by most stat software.
    Good luck,
    Ken

    0
    #68515

    DEVASHISH DEY
    Participant

    You can also use paired t-test
     

    0
    #68516

    Riaz Rasul
    Member

    Hi,
    You can do this by ANOVA. This will basically compare multiple number of means.
    Thanks
    Riaz
     

    0
    #68536

    Shree Phadnis
    Member

    Use Analysis of variance

    0
    #68538

    Ken Myers
    Participant

    A paired t-test looks at differences between TWO groups subjected to the same treatment.  This test is not appropiate for more than TWO groups.

    0
    #68552

    Thom
    Member

    Ken is absolutely correct.  If you used paired comparisons for all possible comparisons, you really hurt yourself for your Type I error rate.  Use ANOVA, but be cognizant of the assumptions.  If these do not pan out, consider the nonparametric alternatives (like Kruskall-Wallace).
    Now, some folks struggle with the results from the ANOVA.  Okay, I have concluded that there is a significant difference between treatment means.  Where are these differences?  To answer this, you then utilize multiple comparisons.  There are quite a few to chose from, and you have to understand what you are actually looking for in relation to your Types I and II errors.  These include Tukey, Bonferonni, Scheffe, Duncan’s Multiple Range Test, Student-Newman-Keuls (SNK).
    BUT, with any test/procedure, know what you are testing for, understand the assumptions associated with the test/procedure, understand what diagnostics you have to make (as well as remedial actions).

    0
    #68556

    Tommy
    Member

    Test the variances between the three as an indication of process stability for each supplier, I think this is now called a test for equal variances in MiniTab.

    0
    #68567

    kbower50
    Participant

    You’re right – it used to be called the homogeneity of variance test, but was changed to “test for equal variance” due to customer demand.  I’ve a paper on using MINITAB to do ANOVA, along with relevant references, reproduced from SC&I at http://www.minitab.com/company/VirtualPressRoom/Articles/SCIFeb2000ANOVA.htm
    Hope it helps.  There’s also a paper on the paired t-test through http://www.minitab.com/company/virtualpressroom/Articles/index.htm

    0
Viewing 9 posts - 1 through 9 (of 9 total)

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