iSixSigma

Attribute Gage RR- Correct Confidence Intervals?

Six Sigma – iSixSigma Forums Old Forums General Attribute Gage RR- Correct Confidence Intervals?

Viewing 2 posts - 1 through 2 (of 2 total)
  • Author
    Posts
  • #34825

    SSBB
    Member

    One calculation, which I came across has made me post this question.
    I need to know the way to get the 95% upper and lower confidence limits on the calculated score during attribute gage RR study. This 95% CI represents the range within which the true Calculated Score lies given the uncertainty associated with limited sample size.
    Lets assume that calculated score of one operator is 96.7%.
    Sample size is 30. Number of matches = 29. There is one mismatch.
    Distribution – Binomial.
    What I observed in published article was:
    Sample Size = 30
    # Matches = 29
    95% UCL = 99.9 % ???
    Calculated score = 96.7 %
    95% LCL = 82.8 % ???
    This does not match with my calculations.
    I got proportion mismatches = p-hat= 0.033 (propor. that matched = 0.967, i.e. 96.7% of the time the operator could match the Attr. value)
    Z_alpha/2 = Z_0.025 = 1.96.
    Using formula for 95% 2-sided CI  for p-hat, i.e.
    p-hat – Z_0.025*SQRT(p-hat(1-(p-hat))/n) < p < p-hat + Z_0.025*SQRT(p-hat(1-(p-hat))/n)
    Using the results from the above equation I got CI as follows:
    LCL = 90.2 %
    UCL = 100 %
    Would appreciate the correct method to arrive at published LCL= 82.8 % and UCL= 99.9 %. Or are these CIs false?
    Thanks
     
     

    0
    #96571

    Statman
    Member

    SSBB,
    You are using a normal approxamation to calculate the confidence intervals for a binomial distribution.  The published article has used the exact calculations from a binomial. 
    When  np and n(1-p) are both greater than or equal to five and n is greater than 30, then the normal approximation is pretty good.  If not, as in your case (n(1-p) =0.987), the approxamation is not so good.
    Minitab will calculate the exact binomial CIs for you in the 1-proportions test.
    Statman

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

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