Statistical Methods for Environmental Scientists and Managers
(4.0 cr; =[ANSC 2211, STAT 3011, STAT 5021]; Prereq-Two yrs of high school math; A-F or Aud, spring, every year)
This course focuses on both the foundations of statistical methods (the mathematical principles that underlie the methods) and the application of those methods. It is unlike similar courses in the emphasis it places on context. Methodological approaches will be motivated using applications from environmental science and management. With that as background, we will be able to more meaningfully study the principles, theory and foundations of the methods, including important theorems and proofs. The end result will be that you will possess: (1) a more complete understanding of assumptions made in deriving methods (and therefore the limitations of those methods) and (2) a better ability to extend and adapt methods as particular problems require it. An early example of this is the emphasis placed on randomization theory as it leads us to proper approaches for data collection. A second notable difference between this and similar courses is the emphasis on regression modeling for description and prediction using observational data as opposed to the confirmatory objectives of analysis of variance for designed experiments. We will cover regression very early in the course as a means of describing bivariate data and return to the topic in more detail once we have established the foundational principles that underlie our ability to do more than describe with regression. Reports on studies applying statistical methods abound in the popular press, including newspapers. We will draw on such studies regularly to initially illustrate proper, and often, unfortunately, improper application of methodology and the drawing of conclusions; we can then move on to the more complex challenges we face with data from the fields of environmental science and management.