iSixSigma

Attributive Gauge RR

Six Sigma – iSixSigma Forums Old Forums General Attributive Gauge RR

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #29955

    Nitin Agrawal
    Participant

    Can anybody help me understand how do we conduct an Attributive Gauge R&R Study and how do we ascertain the different contributing factors.

    0
    #77613

    Erik L
    Participant

    Nitin,
    There are two methods that could be utilized to analyze a Measurement system that is functioning with Attribute data.  If you have a process that is generating nominal data, you can use Kappa (K) as a methodology to determine the degree of agreement within operators and then across all of the operators in a study.  If your process is generating ordinal data, then you could use Intraclass Correlation (ICC) as a means of achieving the same end.  The thing to remember is that with Kappa technique, that misclassifications are all treated with the same severity so if there is a scale to the measure (e.g. ordinal) then you want to look at ICC as a methodology. 
    At its highest level, K is the proportion of agreement between raters after agreement by chance has been taken out of the picture.  Kappa values can range from -1 to 1.  If a measurement system comes out lower than .7 it might warrant some attention.  This typically will lead to the identification of training and/or poor opertional definition of the measure the operator is to be taking.  Kappa takes on a Chi-square look with the data going into contingency tables.  The math gets pretty messy beyond a 2×2 matrix, but Minitab has Kappa capability. 
    ICC has several different variants that need to be decided on to generate the right components of variation.  These primarily deal with single vs multiple raters and if these are to be considered fixed or random effects. 
    There are some pretty good templates out there that cover both of these techniques.  Let me know which variant you’re looking at for the MSA.
    Regards,
    Erik 

    0
    #77614

    Sambuddha
    Member

    Nitin,
    Minitab version 13.32 has an Attribute Gage R&R included in the Stat>Quality Tools> Attribute Gage R&R study included as a tool. Read the statguide for more info.
    Here is a tutorial. It includes 2 examples :

    Service Example
    Manufacturing Example
    Kappa statistic is used as a measure of “chance agreement”. It is a pretty controversial one, (more in the field of linguistics, bio-metrics etc.) so please use it with caution.
    Hope that helps.
    Best,
    Sambuddha

    0
    #77615

    Ron
    Member

    The Attribute Gage R&R in minitab is all you need to complete a detailed  attribute analysis.
    Of course you can manually duplicatethe assumptions if you must.

    0
    #77622

    Nitin Agrawal
    Participant

    Thank you all for giving me an idea of how to go about it.
    Dear Erik,
    Could you please provide me with templates for both the techniques? I also want to teach this to my colleagues who might require one of the techniques in their area. You can send it to me on my email address ‘[email protected]’.

    0
    #77641

    Khandekar
    Participant

    Nitin, AIAG MSA manual (third edition) has detailed procedures on Attribute GR&R.
    Rgds,
    -Atul.

    0
    #77650

    Nitin Agrawal
    Participant

    Eric,
    Could you please tell me more on ICC. I got some details on Kappa from Minitab.

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

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