I work in plastics, and we run "test runs" of many samples to be used for various tests. Most of the these tests are pass/fail, defect or no defect. How do I use binomial distribution in JMP to make sure I am testing enough samples? For example, instead of running 75 samples for what we call a cold drop test (freeze containers and drop them from a height, no damage allowed to occur), how do I satistically prove that 3 samples tested from a lot of 75 would work, or something like that?
