iSixSigma

MSA Question

Six Sigma – iSixSigma Forums Old Forums General MSA Question

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #42478

    ROSS
    Member

    In Measure….if I am the only one doing the data collection (no one else is involved)….do I need to do an MSA?Β  If so…just for repeatibility?….my data collection will involve cycle time ofΒ certain steps in a process.Β 
    Thank you in advance for the help…!!Β  Β 

    0
    #134067

    Mr IAM
    Participant

    Tony,
    I would say yes – there are still MSA issues that can occur.Β  You mentioned repeatability for one, but there are also accuracy issues, linearity, and stabilityΒ (although linearity and stability may applyΒ not for this situation).Β  How many of these contributors could apply in your situation it is hard to say.Β  But, I woudn’t discount measurement error simply because your the only one making the measures.
    I would work on a clear “start time” definition and a clear “end time” definition and at a minimum compare your measurements to someone elses measurements for the same events.Β  You could use a paired T-test for analysis.
    I think repeatability error could be hard to get at because you may not be able to measure the exact same event multiple times.Β  I.e.. it’s not like you’re in a situation where you can measure the diameter of a pipe multiple times to get a measurement of repeatability error.
    Will you be using a stop watch, or some other gage?
    I will be interesting to see what some others have to say but, those are my initial thoughts….
    Cheers!

    0
    #134071

    ROSS
    Member

    Thank you for the response….can you tell me….what tools to use….gage r&r ?

    0
    #134088

    Reath
    Participant

    I would agree with your analysis and the questions you ask. It would be hard to do a fullblown G R&R with cycle times unless the gage (stop watch??) would an issue.

    0
    #134097

    Craig
    Participant

    Tony,
    My initial thought is that an MSA study would be a mis-application of a six sigma tool. One could argue that you are actually measuring something, so why not do a study? But….how could you decipher between the measurement variation and “Part” variation. In this case, your cycle time events would be like the parts. How would you go back and measure the same event again? Look at it from a discrimination standpoint. If you are measuring cycle times in minutes and you have a second hand on your timer…you are OK.Β  {Unless you fall asleep while you are measuringΒ  :)Β Β  ]
    Best of Luck!

    0
    #134298

    Ruddy
    Participant

    Then, does anybody can give a clear answer?

    0
    #134300

    Mike Carnell
    Participant

    Michael,
    The answer from Mr. IAM is a pretty clear answer. The tool applies. With one person doing measurements it is repeatability only. There are other issues in a MSA study as he pointed out bias, linearity and stability that need to be addressed.
    If you are looking for a way to do repeates on the study you simply videa tape the operation and run the operator (person doing the time measurement) past them in a random order. Easy enought to do with todays software.
    Just my opinion.
    Good luck

    0
    #134376

    Scott
    Member

    MR lam gives a good response. An MSA is always required if you consider the purpose of an MSA is simply to ensure truth in measurements. An MSA may be as simple as subjective validation that the measurements are believable and acceptable amount of risk of being wrong.
    Gage R&R is the more comprehensive MSA dependent on continuous data and very useful in you are concerned about variationΒ from poor repeatability or reproducibility. Sometimes, two people clocking events is enough to judge truth. Think about the precision to tolerance when making this judgment. Is my potential for error much less than the tolerance I need to consider?
    Be practical. If in doubt, test it out.
    Β 

    0
    #134377

    Craig
    Participant

    The clear answer is to take a step back for a minute and be objective. Are you measuring a process that takes a few minutes? Are you using a digital stopwatch?Β Β Are the start and end points easily recognizable?
    If yes to all the above, then don’t worry about an MSA study.Β 
    When you are looking at a measurement process like clocking the 50 yard dash and you have to discriminate between 1/100ths of seconds, then you have to worry about your measurement process.Β  If you are a timer on a tournament waterski boat, you have to worry about the measurement process. The driver has an extremely tight spec for going through the course at 36 MPH. The timer has to press start as soon as the boat goes by the first gate buoy and then hit stop when the same point on the boat passes by the end gate buoy.
    If you are measuring a PM cycle time, and you are looking at time between PM start and the time of the first good part out after the PM, then no need to worry about the measurement system (Unless Maintenance fudges the time for PM start!)
    Good Luck!
    Β 

    0
    #134394

    Arun Kumar
    Participant

    All,
    I think very valuable points have been raised.
    I just want to add a few things:
    1. The golden rule of Tolerance and Least Count of the gage
    2. Operational definition of what is getting measured – end points, time of day/ month etc. So all about when , what, how, where to measure.
    3. Sample size for data collection
    4. Again, randomising events and re-assessing them perhaps using the measurement work instructions and another person for reference
    5. Of course, if you use a gauge – things like bias.
    rgds,
    arun
    Β 

    0
Viewing 10 posts - 1 through 10 (of 10 total)

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