A supervisor once told me while developing a new method that my GC run did not have enough area counts... How many is enough and how is the official way to know??? Maybe this is related.. but also what is the official/accepted value on reproducibility of 10 injections... 90%,95% relative accuracy? I'm thinking shoot 10 reps, take ave. for "true value", 3std dev will give me a range. However if my range is half of the "true value" isn't that a bad method? If my true is 10ppm but my reproducibility an analysis could be a 5ppm or 15ppm that sounds useless. Please site a source for the truth.
Thanx.