iSixSigma

Relation between Taguchi design and ANOVA?

Six Sigma – iSixSigma Forums Old Forums General Relation between Taguchi design and ANOVA?

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #46375

    Allthingsidiot
    Participant

    Please  explain the  relation if  it  does  exist?thanks

    0
    #153044

    Jim Shelor
    Participant

    The way I understand it, traditional DOE considers only those parameters that are controlled in the process.
    Taguchi designs consider both parameters that are controlled and factors that cannot be controlled, or by choice not controlled.
    Many statisticians do not like Taguchi designs because the statistics of these designs are less rigorous than traditional DOE models.

    0
    #153051

    SS
    Member

    Yes: You are right. These are the Signal factors and Noise factors. Look at the beauty of S/N ratio.
    While one talks about rigour, lets look at the flexibility of Taguchi Designs. How about the Target Performance Measure and the Noise Performance Measure?
    Less rigorous it may be, practical it certainly is – Taguchi techniques.
    For an industrial engineer, Taguchi offers a better and practical chance of doing the experiments compared to the traditional method which might be better for the thorough bred statisticians and academicians.
    Either way, both are just powerful in their own ways.
    Just my opinion.
    SS
     

    0
    #153052

    Robert Butler
    Participant

    A design is just a blueprint for gathering data. ANOVA is a method of data analysis.  The relationship – one can be used to analyze what the other one dictated should be gathered.
    As for the issues concerning traditional DOE vs. Taguchi – they are really more form than substance. If you are interested in understanding the connections between them the thread below may be of some help.
    https://www.isixsigma.com/forum/showmessage.asp?messageID=108236
    The thread started by the post below has additional details about how one would go about extracting the same information from either kind of design.
    https://www.isixsigma.com/forum/showmessage.asp?messageID=113147

    0
    #153053

    Jim Shelor
    Participant

    I agree.

    0
    #153055

    Allthingsidiot
    Participant

    Robert
    Appreciate  your informative  reply  and  the  attached  links,thank  you

    0
    #153056

    Allthingsidiot
    Participant

    Thank You

    0
Viewing 7 posts - 1 through 7 (of 7 total)

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