seems all our experiment are somehow related to aviation. this time paper helicopters. was to keen on the factors, and i guess the ratio approach sort of made the problem more complicated. didn't quite get the matrix right, however we were lucky that this effect was rather limited, since the interaction did not significantly change the result predicted. perhaps should sort out ABC first in the right order next time instead of jumping into calculations.
robustness, i think i must didn't quite understand the word before. lots of things became more clearer today for me. why the products of the company we worked for was performing fantastically in Germany, but failing at a crazy rate in china. my initial thought was that the product was not that stable, and it is perhaps the MD covered something up to save faces. on the contrary, the product was really good on the other side of the world. what I assumed wrong was they are performing under the same sort of condition, which is not. this difference had a great impact. the noise produced by the environment must had significant impact on the product at the tuning stage, as the product is constantly adjust towards that environment, and made it less robust. I think we must had a wide variation as well, that's perhaps was the reason why we need to adjust the product to fit the specification under certain environmental conditions.