TITLE:
The Power of DOE: How to Increase Experimental Design Success and Avoid Pitfalls
AUTHORS:
Bruno G. Rüttimann, Konrad Wegener
KEYWORDS:
Design of Experiments, Inference, Power, Beta Risk, Noise, Approach, Algorithm
JOURNAL NAME:
Journal of Service Science and Management,
Vol.8 No.2,
April
24,
2015
ABSTRACT: Personal empirical experience when lecturing and consulting shows that not only students, but also experienced engineers familiar with DOE, show much more interest in the modeling of a process than to statistical inference, neglecting attention to “boundary conditions” of the process. But exactly the observation of ancillary boundary conditions of experiments, such as minimizing Beta-risk and noise, is determinant for the efficient execution of an experimental design and the effective application of DOE derived models. This essay focuses attention to the must-dos in the DOE statistics approach in order to avoid research pitfalls by presenting a fail-proof 14-step approach when applying DOE modeling.