|
Abstract:
|
Today's environmental chemicals of interest include complex mixtures of pharmaceuticals (antibiotics, antidepressants), personal care additives and synthetic hormones at nanogram per liter levels in rivers and drinking waters. Their effects include subtle and cumulative disruptions in the brain or reproductive cycles. Environmental scientists today need statistical training relevant to their current work. Their needs have moved beyond computing n, assuming a normal distribution, and confidence intervals. They require better training in robust statistics, censored data techniques, permutation tests, Bayesian methods and experimental design for multivariate contaminants. Statisticians must work alongside environmental scientists to use and develop new tools for interpreting the results of complex studies of trace-level contaminants.
|
