A Summer School: Uncertainty and Variability in Ecological Inference, Forecasting, and Decision -- An Introduction to Modern Statistical Computation

Duke University, Durham NC

Investigators

Abstract

A Summer School: Uncertainty and Variability in Ecological Inference, Forecasting, and Decision -- An Introduction to Modern Statistical Computation Clark, James S. Duke University A two-week, graduate/post-graduate level summer school is proposed to introduce ecologists and earth scientists to modern statistical computation techniques that have emerged over the last decade. Ecological inference and forecasting are limited by large and diverse sources of variability that operate at a range of scales. Hierarchical Bayes and Markov chain Monte Carlo simulation provide powerful tools for analyzing processes characterized by multiple sources of uncertainty and variability. However, adoption of these techniques in ecology has been hindered due to limited training options. In the proposed summer school, leading statisticians and ecologists will provide day-long presentations and hands-on training with computation techniques. Students will include advanced graduate students and postdoctoral associates selected by an open application process. They will participate in small working groups that will each produce a chapter to be included in a published volume, together with lecture notes. By training a select group of young, quantitative ecologists and earth scientists, who will in turn train others as well as utilize this training in their research, this summer school will expedite the dissemination of modern statistical computation techniques. In addition, the volume that will be published as a result of this course will serve as a useful reference for a much broader audience. Adoption of these modern methods by ecologists and earth scientists will strengthen the field of ecological forecasting, enhance its reputation, and increase it usefulness to resource managers and policy makers.