iSixSigma

Experiment with Hard to Change Factors

Six Sigma – iSixSigma Forums Old Forums General Experiment with Hard to Change Factors

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • #52895

    Saherngu
    Participant

    An experiment is run to test an electical part.  252 parts are used altogether.  There are two factors – Voltage (3 levels) and Electrical Pulses (3 levels).  These 252 are broken into 3 groups of 84 parts.  Each part of the 1st group of 84 is screened to 1700 volts / Each part of the 2nd group of 84 is screened to 1900 volts / Each part of the 3rd group of 84 is screened to 2100 volts.  With each of these groups, the parts are screened all at once (nosetup variation or variation due to equipment or operator here).
    The 1700 volt screened group is then split into 3 groups (28 parts each) – the groups are pulsed either 50, 150, or 500 times.  As with the voltage screening, each group (of 28) is tested at the same time.
    The 1900 volt screened group is then split into 3 groups (28 parts each) – the groups are pulsed either 50, 150, or 500 times.  As with the voltage screening, each group (of 28) is tested at the same time.
    The 2100 volt screened group is then split into 3 groups (28 parts each) – the groups are pulsed either 50, 150, or 500 times.  As with the voltage screening, each group (of 28) is tested at the same time.
    Then the strength of each part is tested for.
    My question: what is the proper way to analyze this experiment??  It seems that the degrees of freedom must be changed from the error to the factor terms (as in a split plot experiment), but I’m not sure. 

    0
    #186693

    Björn
    Participant

    Hi James,
    It sounds like a strip plot experimental problem, i.e an experiment performed in multiple stages. So, you should check that litterature for advice.
    Kind regars
    /Björn 

    0
    #186696

    Saherngu
    Participant

    Thanks Bjorn. 
    I’m somewhat familiar with split plots, but not strip plots.  Of what I know, there would be a whole plot for each of my 2 factors (Voltage and Elecrical Pulses).
    Do you know of sources for more information (preferably with a few examples carried out)?  It’s always nice to see an example – especially done with a software (Minitab for me).

    0
    #186701

    Saherngu
    Participant

    Does anyone know how Strip Plot experiments are handled differently from Split Plot experiments?

    0
Viewing 4 posts - 1 through 4 (of 4 total)

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