Data Handling, Simultaneity, and Rare Events
PETER C. WILLIAMS · Journal of the Air Pollution Control Association · 1984
This paper is directed to those who design experiments and/or handle nonrandom data, e.g., environmental impact statements, historical data, models, scientific studies, and the design of equipment related to discharges. In recent years, statistical distributions have been used to describe physical data which result from many factors, some of which are unknown. Variations in these descriptions represent natural fluctuations rather than experimental error and the impact of dependence on data cannot be ignored. This paper provides general principles for handling these kinds of data.