Questioning Data

Six Sigma – iSixSigma Forums Old Forums General Questioning Data

Viewing 1 post (of 1 total)
  • Author
  • #52720


    Background for a question:
    For 114 different mixtures made at this plant, I have developed IMR charts for two measures for each mixture.  In practice, due to our system, I can only monitor the individual measures, not the moving range.  For the individuals I took the UCL and the LCL and printed them on the mixing instructions, and when the mixing is done the operator takes the two measures and compares that to an “Acceptance Range” (The LCL to the UCL).  Each mixture is not made at a consistent rate, some are once a month, others multiple times per week.  The number of data points used to develop the UCL and the LCL varies per the mixture, as data gathering is limited by how often the mixture is made.   For mixtures for which the number of data points is >40, the data appears to be normally distributed.  Only about 5 mixtures appear to not represent a stable process.
    The question:
    I review around 30+ mixtures (60+ measures) per day, and it has been about 3 weeks since a value has not been recorded as in the acceptance range.  My suspicions are up — that I should have had an outlier (outside of the acceptance range) out of the roughly 900+ measures (two measures per mixture, but the different measures seem to be highly correlated).  Am I right?
    Thanks in advance for your thoughts.

Viewing 1 post (of 1 total)

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