'Doing DOE' Versus Designing Experiments Well

摘要

Test and Evaluation is an integral part of any weapon system acquisition process, and the percentage of lifecycle costs devoted to testing is substantial. While every test is in some sense a "designed experiment," the methodology of "Design of Experiments (DOE)" is a special class of combinatorial tests having unique properties that will make the evaluation of a test clean and straightforward. Thus, DOE makes the linkage between the activities of Test and Evaluation clear. While testers are faced with what may seem to be an insurmountable task in evaluating many variables or factors in a test, some of which can be controlled and some of which cannot, DOE brings a disciplined approach to the entire test and evaluation process that will ultimately save money and reduce testing time for those who choose to use it. The principles of DOE need to be part of every tester's mindset, because designing experiments well is just as much a thought process as it is a toolset. This article focuses on the difference between "doing DOE" and applying the necessary thought processes associated with DOE to designing experiments well.