Pass / Fail Confidence Interval

[Aurora Tiffany-Davis]

Hi,

A coworker recently asked me to investigate a problem she ran into. I've never used Weibull software for such a problem and I'm curious if the software has a tool for this.

She is running 100 "one-shot" pass / fail trials (1 trial per sample). She'd like to know how to calculate confidence intervals for the results. That is to say, should 3 parts of 100 fail this test, she would like to express results as, "With X% confidence, we can state that the failure rate is greater than Y%, but less than Z%".

[Pantelis]

Yes... use the DRT tool.

See http://www.weibull.com/LifeDataWeb/demonstration_test_design.htm for more details.

[Aurora Tiffany-Davis]

Thanks! I got the results I needed by using the Non-Parametric Binomial Tab of the DRT.

I'm also interested in the formula used to calculate these results - the link in your previous post results in a "404 Not Found" error, and I've been unable thus far to track down the formula(s) used in the SW help files or on the website - any suggestions?

Thanks again,

A. Tiffany-Davis

[Arai.M]

Hello,
I just tried that link and I had no problems with it. Would you mind trying again? there might have been something else going on at the time of your attempt.
Please do let us know if it still doesn't work.
Regards,
Arai

[Arai.M]

Actually, how about this one http://www.weibull.com/LifeDataWeb/demonstration_test_design.htm
Thanks,
Arai

[Arai.M]

Hmm, aparently the message cuts off the text of the link but if you click on the link it should work. If you would like to cut and paste, use the following: http://www.weibull.com/LifeDataWeb/demonstration_test_design.htm

[Aurora Tiffany-Davis]

Got it, thanks!