ABSTRACT
Statistical experiment design can be used to efficiently select experimental runs to achieve a given experimental purpose. However, uncertainty is a fact of life in experimentation. The experimenter is faced with uncertainty in inputs, uncertainty in outputs from both random variability and uncertainty in measurement processes, as well as uncertainty about the underlying model structure of the phenomenon under investigation. In the face of all this uncertainty, the experimenter must try to collect and analyze data that will address questions of scientific interest.
ACKNOWLEDGMENTS
The author acknowledges George Box for his many insights on statistical theory and practice and the role of uncertainty in experimentation. The author also thanks John Weigle, Los Alamos National Laboratory, for providing information on the polymers experiment example and the accompanying figure and photograph.
Notes
From Wendelberger (Citation1991).