iSixSigma

Need Help with Ppk Calculation Sample Size

Six Sigma – iSixSigma Forums General Forums Tools & Templates Need Help with Ppk Calculation Sample Size

This topic contains 3 replies, has 4 voices, and was last updated by  GPS1111 3 months, 4 weeks ago.

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • #240438

    Yujie
    Participant

    Hello all,

     

    I am confused about the sample size of Ppk calculation. I check a lot of information on the Internet and books but didn’t find out the definition of sample size for Ppk. Could anyone tell me can I use 30-32 as Ppk sample size( the total number for a lot is 1k-2k)?

     

     

    Thanks,

    Yujie

    0
    #240443

    Darth
    Participant

    @yujie  This is from the Minitab help section:

    Collect enough data to obtain reliable estimates of process capability.  Try to collect at least 100 total data points (subgroup size*number of subgroups), such as 25 subgroups of size 4, or 35 subgroups of size 3. If you do not collect a sufficient amount of data over a long enough period of time, the data may not accurately represent different sources of process variation and the estimates may not indicate the true capability of your process.Data should be collected in rational subgroups, if possible.  A rational subgroup is a small sample of similar items that are produced over a short period of time and that are representative of the process. Observations for each subgroup should be collected under the same inputs and conditions, such as personnel, environment, or equipment. If you do not collect rational subgroups, the variation in the subgroups may reflect special causes rather than the natural, inherent variation of the process.

    You might also consider using confidence intervals around your Ppk calculations so it will reflect any sample size that you select.  You can do those easily in Minitab.

    1
    #240452

    Chris Seider
    Participant

    Nice to see a post @Darth :)

     

    0
    #240564

    GPS1111
    Participant

    Darth is right about collecting sufficient data and about not having enough data.  But not all processes allow large amounts of data to be obtained.  Automotive can develop a lot of data in a very short period of time.  Some aerospace industries cycle so slowly that it could take days or weeks to obtain enough data to be statistically significant.  How long does it take to make your lot sixe of 1K to 2K parts?  Defining some measurement parameters by Control Plan can help.  If it defines taking measurements on five parts per hour and you run three shifts you could have a sample size of 120 in 24 hours.  If running two machines or lines with the same 5 parts per hour on each you can have 240 parts and so on.

    Another thing to consider is what are you trying to show?  Process capability of a specific machine or production line or overall process capability?  Cpk and Ppk are the same but different.  They measure the same things but it is a question of subgroup variation versus overall variation.

    Cp and Cpk are for computing the index with respect to the subgrouping of your data (different shifts, machines, operators, etc.), while Pp and Ppk are for the whole process (no subgrouping).  You could have different Cpk for different machines making the same product.  Cpk will show this individually by line or machine.  Ppk shows it collectively.

    “Ppk produces an index number (like 1.33) for the process variation. Cpk references the variation to your specification limits. If you just want to know how much variation the process exhibits, a Ppk measurement is fine. If you want to know how that variation will affect the ability of your process to meet customer requirements (CTQ’s), you should use Cpk.”  – credit to Michael Whaley.

    Depending on how much data you can get may say calculating by subgroup, or Cpk, may be better than Ppk.   I’ve seen a sample size of 5 “work”.  Does it work well?  Not as well as a sample size of 100 but it may be all you can get in the timeframe you have.  At least creates a starting point.  As more data becomes available process control can improve over time and become Continuous Improvement.

     

    1
Viewing 4 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic.