569481_en-US

liliconcom

Confidence Intervals For Cpk

I was trying to replicate the confidence intervals for Cpk in JMP.  I found this page which details the formula:

 

https://www.jmp.com/support/help/en/17.0/#page/jmp/statistical-details-for-capability-analysis.shtml

 

This page says the equation is:

 

So this is the example I am trying to match, 

 

 

 

since it is 95%, the lower confidence interval should be (32 is the sample sample size):

 

 

 

 

This equals, 6.075.  JMP says it is 5.532.  Am I doing something wrong or is this not the correct formula, why is it not matching?

 

When I use this calculation for Ppk, I get the correct answers, but not Cpk.

 

I am on macOS Monterey 12.6 and JMP Pro 16.0.0.

 

Lisa

AnalyticsRe: Confidence Intervals For Cpk

The formula looks correct, or at least close to others that can be found on the internet. Following your values, I get the same results as you both in JMP and with my hand calculations. I also tried several other examples and my hand calculations never match what JMP is providing. I think this might be something that you take to JMP Tech Support to see if they can determine what is going on.

Re: Confidence Intervals For Cpk

Will do.  thank you!

Re: Confidence Intervals For Cpk

You might also take a look at this page for more detail about how the CI is currently calculated. Your result at the top appears to be from a legacy platform that might not be up to date. The key difference is the use of DF that are not necessarily N-1.DF depends on how the sigma is determined. The developer is looking into this matter, too.

Re: Confidence Intervals For Cpk

Thanks. I tried the new calculation and am still getting different results.    I forwarded an email to jmp support.

Re: Confidence Intervals For Cpk

Are there any updates from JMP about this issue?

Thank.