Cumulative Confidence Levels - Design Verification and Validation Sampling

#12
THE MEDICAL DEVICE VALIDATION HANDBOOK 2ND EDITION--MAX SHERMAN H/C 9781596934221 | eBay
Thanks for recommending such a good book.

I went through the chapter 14 (which is about sample sizes) easily as I am very familiar with the topic but didn't find any references to either the cumulative confidence formula for independent runs or the aproach of overall confidence a whole protocol. Am I missing something?

PD: Are you the same Steven who wrote this chapter?:unsure:
 

Statistical Steven

Statistician
Staff member
Super Moderator
#13
Thanks for recommending such a good book.

I went through the chapter 14 (which is about sample sizes) easily as I am very familiar with the topic but didn't find any references to either the cumulative confidence formula for independent runs or the aproach of overall confidence a whole protocol. Am I missing something?

PD: Are you the same Steven who wrote this chapter?:unsure:
I did in fact write the chapter. I would recommend you look at Bonferroni confidence intervals for sequential testing. The reason it is not covered in the chapter is that one can assume that if the runs are independent, we can pool across runs and get a single decision (not sequential decisions). For example, if my overall sample size is 30, if I sample 10 for each run, and make my inference on the 30, I do not need to adjust the alpha level as I am only making a single comparison. If instead you want to sample from each of 3 runs, where each run must pass, then the alpha level for each run would be set at 0.017 to ensure an overall alpha level of 0.05.
 


Top Bottom