When it comes to design verification testing of a medical devices, a justification of sample size is required. If I need to verify outer diameter for example, typically I would use a minimum sample size of 15 and claim a 95% confidence level with a 90% reliability/conformance level of variable data (assuming normal dist.). But what if we are measuring 100% of the samples in production to verify that the diameter is with in a specification of 2.0mm +/- 10%. When production gives me the samples for testing I already know they are in specification. So do I still have to do design verification testing using relaibility statistics? Doesn't seem to make sense. What type of analysis of the data set do I say I'm going to perform in my protocol? The testing of the diameter is still required according to the regulations.