iSixSigma

Linearity and Bias study

Six Sigma – iSixSigma Forums Old Forums General Linearity and Bias study

Viewing 3 posts - 1 through 3 (of 3 total)
  • Author
    Posts
  • #46560

    Mares
    Participant

    I have ten parts that i have manipulated so that they measure at different specs withing the set tolerance. My question is, in order to conduct a linearity and bias study, “how many times should i measure each part to get good data to run the study?” i figured measuring each part 5 times randomnly would do it, but id rather get a second opinion on the matter.

    0
    #154097

    Jim Shelor
    Participant

    Breyfogle, Chapter 12, uses 16 measurements per part.  Although he never actually states this as a recommented minimum, he implies that a large number of measurements is needed for the stability of the analysis.
    The stability being referred to can be likened to the recommendations for X-bar/R and I/mR control charts where 25 and 30 points respectively are recommended before establishing control limits.
    I would recommend a minimum of 10 readings per part.

    0
    #154098

    Doc
    Participant

    The AIAG Measurement System Analysis Reference Manual calls for at least 5 parts whose measurements cover the operating range of the gage, and then to measure each part at least 10 times by an operator who normally uses the gage.

    0
Viewing 3 posts - 1 through 3 (of 3 total)

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