iSixSigma

Sample Size for Quality Check

Six Sigma – iSixSigma › Forums › General Forums › Methodology › Sample Size for Quality Check

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • March 27, 2014 at 8:37 am #54713

    Anoop
    ★ 20 Years ★
    Guest

    Hi – Need inputs on the below:

    Situation:

    Customer requirement : Customer Requires 98% Quality (DPU) on transactoins processed
    First Pass Yield: 95%. So 5% defect leakage happens.
    If i review the transactions again, i would be able to fix any defects in 99% of the cases.
    So the question is what should be the QC sample % that i need to do, to atleast rectify the 3% out of 5% defect leakage so that the end customer requirement of 98% is met.

    Please note that I am not looking at the ‘Quality Audit’ but a second level of verification that I need to build into the process itself.

    Thanks and Regards,

    Anoop

    0
    March 31, 2014 at 8:10 am #196805

    Anoop
    Participant

    Request inputs..Thank you!

    0
    April 2, 2014 at 3:14 am #196821

    Prabhu V
    ★ 10 Years ★
    Participant

    Hi Anoop,

    To answer your question in general, you can apply the sample size calculation using the discrete data method.

    N=(Z/E)^2*p(1-p)

    Where

    N – Sample size
    Z – Constant for confidence level (like 1.645 for 90% confidence, 1.96 for 95% confidence, 2.575 – 99% confidence)
    E – Precision or Error (level of precision desired from the sample in units of proportion)
    p- Proportion

    in your case, like you wish to have the erroneous transaction to 2% from 5% (First pass yield) then, you calculation will be
    N = (1.96/0.02)^2*.05(1-0.05)
    N ~ 456 transactions considering that 95% confidence level being expected from the audit.
    I hope this explanation can help you.

    Pls feel free to post your queries

    All the best!!

    Regards

    Prabhu V.

    0
    April 10, 2014 at 2:19 am #196845

    Marke
    Participant

    If your aim is to calculate a sample size which you can check to prevent defects reaching the customer, then I am afraid you will have to check 100% of your production – which obviously is a waste. So, if the process is generating defects, the way to go is Defects Analysis (top defect types, people contributing to defects, etc.) and then Root Cause Analysis and Solutioning in aspiration of zero defects.

    The above may take a bit of time and if you do not have that.. and.. if you are working on a process with high financial/reputational implications.. you may want to set-up a automated reconcilation/defect identification process or a manual 4-eye check for 100% of the critical transactions. Hope I made sense.

    -marke

    0
    April 10, 2014 at 6:30 am #196848

    MBBinWI
    ★ 10 Years ★
    Participant

    @nairakb – while there obviously are some on this forum who have a need to show how smart they seem to be (emphasis on SEEM), most of those of us who do know something about this stuff don’t do homework. So posting it here, and then imploring us to do it for you, isn’t going to get much response.

    IF you post your query, then posit a solution with your rationale as to how you got to that solution, we MAY respond with feedback to you. Most of us have actual JOBS, so we cannot be sitting on this site day in/day out just waiting to answer your query. (‘cept for @katiebarry – but that is her job – figured since I had already woken you up, a second post wouldn’t hurt so badly).

    0
  • Author
    Posts
Viewing 5 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic.