iSixSigma

Small sample size < 30 for Cpk and Ppk calculation

Six Sigma – iSixSigma Forums General Forums Methodology Small sample size < 30 for Cpk and Ppk calculation

Viewing 2 posts - 1 through 2 (of 2 total)
  • Author
    Posts
  • #256391

    Mariang168
    Participant

    Hi,

    I know that at the minimum I need a sample size of 30 with a normal distribution to calculate Cpk or Ppk.

    What if I only have a sample size < 10? Can I use bootstrapping with resampling to generate a 10000 numbers from the original sample size to calculate my Cpk or Ppk? Sounds so dicey. Is there a methodology that address the small sample size question for Cpk or Ppk calculation?

    Thanks in advance!

    0
    #256394

    Robert Butler
    Participant

    No, you do not need a minimum of 30 samples – Gossett put that bit of nonsense to rest back around 1908 or so. Other than the fact that the basic Cpk calculations require data normality there is no connection between 30 samples and normality. Take what you have and press on with the calculation. If you are really concerned about the structure of your data take a look at it on probability paper and see if it passes the fat pencil test. If you are still worried about violating the basic assumptions surrounding a standard Cpk calculation get a copy of Measuring Process Capability by Bothe through inter-library loan and follow Chapter 8 – Measuring Capability for Non-Normal Variable Data.

    0
Viewing 2 posts - 1 through 2 (of 2 total)

You must be logged in to reply to this topic.