培训课件-SIGMA.pptVIP

* For DoE analysis, I feel that it is acceptable to use 0.1 as alpha risk as it is still a considerable “significant” factor of effect. * Be prepared to review this and other session window output. * * * Note that if we were to base our choice of settings for A C on only main effects, we would choose A=High, C=High. However, the best response (yield = 87.25%) occurs at A=Low, C=High. * * Model reduction is an iterative process. By removing one insignificant factor at a time, arrive at a model wherein all factors are significant. Note that if a single factor (main effect) is not significan