Data collection, data analysis, and data interpretation go to the heart of the scientific method. The scientific method is an iterative inductive-deductive process for approaching scientific and engineering problems that involves a constant interaction between the physical universe and abstractions of that universe through mathematical/empirical models. The proper planning of all data collection is essential for the success of this approach.
This talk reviews the scientific method and casts it as an iterative approach for scientific and engineering discovery. It then outlines that the proper approach to experimentation within this method must be sequential in nature. Such an emphasis on sequential experimentation stands in stark contrast to the current literature on experimental design with its focus on one-shot, variance optimal designs. The talk concludes with an outline for future steps to reconcile these two different perspectives.
|