iSixSigma

is this outlier in grr data?

Six Sigma – iSixSigma Forums Old Forums General is this outlier in grr data?

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #31419

    Wing
    Member

    Hi all,
    I have a question. We performed a GRR study on an ATE test system which is used to measure DC voltage output (15VDC nominal) of our product (power supply). 10 parts were selected randomly and measured, one of them was failed in the test without output. If we evaluate the ATE system with this failed part, the %GRR is very low, but if the study is done excluding this failed part as an outlier, the   %GRR is very high (up to 70% of total variation). What is the correct result? Thanks.

    0
    #82839

    MMBB
    Participant

    The GR&R assumes all the units to be from a stable common distribution. Is that one failed unit defective? Investigate it and see if it is truely from a different population (of defects). Then you can feel comfortable leaving it out.

    0
    #82844

    Whitehurst
    Participant

    A good GR&R software package will also include SPC control chart plotting capabilities.  If the data point is out of control, you can delete it.  Of course, you can do the calculations manually also.  It sounds like you want to keep the flyer because it is giving you a low GR&R result.  First of all, this sounds illogical.  A flyer should increase the variation in the data and result in a higher GR&R, not lower.  So, I would examine the data and calculations for an error.  Secondly, if it is a 3 sigma flyer (only 3 chances in a thousand that it belongs with the rest of the data) why would you want to include data that has a probability of 99.73% of being erroneous?

    0
    #82846

    sivaram reddy
    Member

    Can some body provide me in calculating Gage R&R with an example for descrete data. I believe the Gage R&R = summation of (Repeatability and Reproducipility). Also, I need and example for continous data.
    Thanks in advance
    Siva.

    0
    #82850

    Anup
    Participant

    For measuring Gage R&R for Discrete data you need to use Discrete Data Analysis techniques.

    0
    #82852

    reddy
    Participant

    Thanks for responding my question.
    I knew that, I have to use descrete data analysis technique to measere Gage R&R. But I am looking for an example in calculating the same.
    For example: Lets take an simple example. If two persons had spent some time on identifying the defective items from the list
    Items.           Person# 1               Person# 2
    1                     Defective             Non-defective
    2                     Defective             Defective
    3                     Non-Defective    Defective
    4                     Non-Defective    Non-Defective
    5                     Defective            Defective
    What’s the repeatablity of this measurement here. Also, if the same person had done the same thing in the second time.
    Items.           Person# 1 (Trail#1 )    Person#1 (Trail#2)
    1                     Defective                  Non-defective
    2                     Defective                  Defective
    3                     Non-Defective          Defective
    4                     Non-Defective          Non-Defective
    5                     Defective                 Non-Defective
    What’s the reproducibility ? Also, what’s gage R&R for this measuring system?
    Thanks,
    Reddy.

    0
    #82854

    MMBB
    Participant

    The Automotive Industry Action Group’s Measurement System Analysis Reference Manual 3rd edition (usually referred to as AIAG MSA manual) describes two kinds of attribute data analyses:
    One basically compares results between multiple operators and with a reference (known) value. This method uses Kappa statistics to quantify agreement. MINITAB Release 13 can do this method.
    Another (some call this the long method) uses a form of probit analysis to study attribute data. Basically the method tracks several parts that have differing probabilities of acceptance. The user creates a normal probability plot of these probabilities and then uses that to estimate a repeatability (it doesn’t est. reproducibility). MINITAB 13 does not do this method.
    To purchase the AIAG MSA manual, to to http://www.aiag.org/publications/quality/msa3.asp .

    0
    #82863

    Tierradentro
    Participant

    Help me understand what value you used for the one part. You said there was no output, so what number was put into the analysis?

    0
    #82869

    Chip Hewette
    Participant

    It appears that you are reviewing the GR&R output from a statistical package, and looking at the components of variation attributed to (a)repeatability, (b)reproducibility, and (c)part-to-part.  When the failed part is measured at zero volts, and nine other parts are measured at 15 vDC, the variation in the measurement system attributed to part-to-part variation is quite high.  Of course, the variation for R&R then goes down, as all must sum to 100%.  A GR&R at 70% with the failed part removed indicates to me that you have not put the specification for voltage in the software, and that you are seeing only the proportion of observed variation, not the percentage of the specification width absorbed by measurement system variation.
    To evaluate the ATE, one must acquire a set of parts having a reasonable part-to-part variation.  If ZERO volts is reasonable, in that your component failure rate is such that the ATE would see ZERO, then you can include this type of failed part in the study.  However, in so doing you are reducing the likelihood of either a repeatability error or a reproducibility error, as most measurement systems can reliably detect a ZERO.  This is not studying the discrimination of the ATE properly.  The ability to detect a failure must be evaluated against a binomial distribution, and not against some variable measurement where the output is far above zero.  Therefore failure measurements (ZERO vDC) should be tested using another method.  A good Quality Manager would have two numbers to describe the ATE.  One would describe the GR&R for the 15 vDC specification.  One would describe the gage’s ability to detect a failure 100% of the time.
    If you are following the traditional method of ten good parts evaluated by three operators randomly, then you can plug and chug through the variation estimates.  The percent GR&R reported is only valuable with respect to a specification width…e.g.15 vDC plus or minus 0.25 vDC.

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

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