SREL Reprint #2468

Strong inference, science fairs, and radioecology

Tom Hinton

Savannah River Ecology Laboratory, University of Georgia, Drawer E, Aiken, SC 29802, USA

Introduction: The concept of hypothesis testing and the scientific method is typically presented to students within the United States for the first time when they are in the fourth grade. Students learn the fundamentals of the scientific method as they develop a research project, conduct an experiment that tests their hypothesis, and present their results in a poster format at a Science Fair. Posters are judged at regional Science Fairs based on their adherence to the scientific method. Cash prizes and ribbons are awarded to encourage the students in their first rigorous attempt at science.
Recently, I had the opportunity to be a judge at one of the science fairs, and the thorough approach and professionalism of some of the projects impressed me. I could not view the posters, however, without occasionally breaking into a large smile over the naivete and candor of these fledgling 10-year-old scientists. I recall the young child whose hypothesis was that mold grows faster in warm environments. He tested his hypothesis by determining how long it took bread to turn moldy when stored under his bed, compared to bread kept in a refrigerator. Pictures of the refrigerator and his bed, proudly displayed on his poster, made me wonder what his mother's first response must have been when he asked to convert the dark, warm, underside of his bed into a fungal factory not to mention her tolerance of what he wanted to do to her clean refrigerator.
Yet, as I searched for a properly stated hypothesis within each poster, I began to think of my own research and the rigor, or lack thereof, with which I apply hypothesis testing to the problems I address. On further reflection, I sensed that a lack of hypothesis testing is perhaps more common in radioecology than those of us deeply rooted in the discipline might want to admit. It seems that we more often report yet another concentration ratio, or invoke yet another correlation coefficient, neither of which delineates causation or gives much aid to our understanding of the mechanisms involved. . . .

SREL Reprint #2468

Hinton, T. 2000. Strong inference, science fairs, and radioecology. Journal of Environmental Radioactivity 51:277-279.

This information was provided by the University of Georgia's Savannah River Ecology Laboratory (srel.uga.edu).