iSixSigma

Clarification on AGRR

Six Sigma – iSixSigma Forums Old Forums General Clarification on AGRR

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #38606

    Rambad
    Participant

    Hi,
    Can anyone suggest the time gap required between the two tries that have to be conducted for a AGRR study. Is it necessarily 2 / 3 days or can it be less also?
    Tks and regards,
    Rambad

    0
    #115736

    RK
    Member

    The purpose of repeated trials in a GRR study is to detect the contribution equipment variation. An R&R study is conducted in a single session and not over 2 or 3 days. The parts that are used to perform a GRR study need to represent the variation in the process that manufactures them. These can be collected over 2 to 3 days so as to have an opportunity to capture the process variation.
    Regards

    0
    #115806

    Rambad
    Participant

    RK,
    I fully agree with you. But what I wanted to know is about Attribute Gage R&R wherein the measurement result is Yes or No. When I got my GB training it was told that we should maintain a gap of 2 – 3 days between Try 1 & Tr 2. Hope u understand my query.
    Rgds,
    Rambad
     

    0
    #115816

    RK
    Member

    Rambad,
    I do not know the reason why 2 trials of an attribute GRR study should be done on two different days. If you do not want the appraiser to ‘remember’ the previous trial decision you can always randomize the trials. If you have adequate number of parts and randomize the trials, the chances that the appraiser will identify the part and remember the previous trial result are not high.
    What was the reason given by your GB trainer when he asked you to maintain a 2 – 3 days gap? I would be interested in understanding.
    Thanks

    0
    #115819

    Rambad
    Participant

    RK,
    Ur argument is correct if we have identical outputs. What do we do for un identical output from a process? How do we check the Repeatability?
    Rgds,
    Rambad

    0
    #115821

    RK
    Member

    Identical outputs? I don’t understand.
    The same set of parts are to be measured by all appraisers across all trials. For an attribute gage study, the parts should be a mix of good parts and rejected parts. Select parts close to the Upper and the Lower limits of the attribute gage so that they have a good opportunity to demonstrate R&R. To get a mix of good and rejected parts, the parts collection can be done over 2 -3 days as you suggest. Analysis of R&R can be done by computing Coehn’s kappa parameter, or using the signal detection method detailed in the AIAG MSA manual (3rd edition). This strictly doesn’t isolate Repeatability.
    To compute repeatability alone, you could use the analytic method where you plot the Gage Performance Curve. The method and formula for calculating Repeatability of an attribute gage is given in the AIAG MSA manual.
    Regds

    0
    #115822

    Rambad
    Participant

    RK,
    I think I’ve not communicated properly. The set of parts that will be used for measurement will be same for all the appraisers. But each of the part will be different from one another. In such a case the chance of remembarance is very high and hence we leave atleast 3 days gap before we conduct the 2nd Try.
    Rgds,
    Rambad
     

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

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