Quote:
Originally Posted by mjohnson
I'd expect that this book (haven't seen but I'm interested) would go well past the first layer of screening tests. If you are digging into some identified mechanism I think that it gets more interesting and useful than your 6-S standard training.
One problem is that everything is estimates of estimates of a population and things get wonky when you deviate from standardized PDFs. Even more so when you're looking to understand the likelihood of rare things in the tails.
|
mostly the thing i am pointing out is if you are looking at all sorts of nonlinear, dependent and interaction effects in your system, like something you need a really good DOE to figure out ... then you dont really know how your system is working or why.
im less mocking DOEs, and more pointing out that when you need a good DOE, you are deep **** and not having fun. life is hard when you dont have a fundamental understanding of your system.