iSixSigma

Validation Sample size

Six Sigma – iSixSigma Forums Old Forums General Validation Sample size

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
    Posts
  • #51937

    Canonman
    Participant

    Hello,
    How would you predict the required sample size to validate a new process when the process produces roughly 25,000 pieces per shift? I’d rather not have a large sample size and I just seem to be stuck.
    Thanks,
     

    0
    #181847

    MM
    Participant

    Hi – this depends on what type of data you are capturing: Continuous or discrete. There are different sample size calculators for this.
    Also, in case you have JMP, it will help you take into account power etc. Though others in the forum will be able to shed more light on this. :)
    Thanks.

    0
    #181848

    Canonoman
    Participant

    Thanks MM,
    I don’t have JMP but I do have MINITAB.

    0
    #181849

    MM
    Participant

    Urw. Also, it will be helpful to know what you want to do with the sample. In case of normal audit, I think the sample size calcs should do.
    In case, you want to do some specific test then use the following:
    In minitab: Stats > Power and sample size
    This may be useful to you. Read the help section if stuck.

    0
    #181854

    Canonman
    Participant

    Thanks Again MM,
    I also forgot to mention the data is continuous.
     
    Canonman

    0
    #181858

    Obiwan
    Participant

    CanonMan
    You will also need to know a few other items…such as…standard deviation of the process, risks you are willing to take, etc.  You stated you have Minitab…use Mini’s help function…it will help immensely.
    When you have specific questions, come back…
    Obiwan

    0
    #181887

    Dr.J
    Participant

    Try to identify first if you want to have a sample size for continuous or proportions data.
    If continuous, need to know the metric, standard deviation and precision. For proportions, same as continuous but need to know the proportion of defectives.

    0
    #181891

    Severino
    Participant

    CanonMan –
    Your organizational procedures should set some form of standard for this.  If your company has not established this, you should solve that problem before you begin worrying about a specific process.  Otherwise, you may end up with lots of different criteria applied to different processes that can leave your validation program in a questionable state (it eliminates the temptation to pick an arbitrary sample size and then attempt to justify it after the fact).
    An example of such a system might be as follows:

    For critical defects/characteristics:  95% confidence/99% reliability
    For major defects/characteristics: 95% confidence/95% reliability
    For minor defects/characteristics: 95% confidence/90% reliability
    Once you have established these guidelines, there is a table in Montgomery that can be used to determine a k factor and sample size for continuous data (assuming a normal distribution) and you can utilize a binomial table for pass/fail data (it is a good idea to copy these into your validation procedures so that everyone in your organization is given the same information). 
    An alternative to the above, comes from Taylor (http://www.variation.com ; visit the tech library, article “Selecting Statistically Valid Sample Sizes”).  If your organization has already established AQLs for your defects/characteristics (i.e. critical – .065%; major – 1.0%; minor – 2.5%, etc.) he advises using your AQL as your LTPD for the first three lots and then transitioning to a less severe sampling plan after validation is completed.  He presents examples using attribute (binomial) data, but the same concept is easily applied to contiuous data.
    The final method your organization may consider is to set Cpk goals for your validations (again these are typically proportional to the criticality of the characteristic) and then use an alpha to beta ratio to determine the sample size necessary to claim you have achieved it as a minimum.  This procedure is also outlined in Montgomery.
    Any of the above approaches would work and can be statistically justified.  Just remember when dealing with continuous data you need to check your distributional assumptions to ensure the model (typically a normal distribution) you used when determining your acceptance criteria and sample sizes were correct.
     

    0
Viewing 8 posts - 1 through 8 (of 8 total)

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