Welcome to the homepage for FW 544: Quantitative Decision Analysis for Fish and Wildlife Management a graduate-level course offered through the Department of Fisheries and Wildlife at Oregon State University.

This site is intended to beĀ  a repository of materials from previous classes including example decision models, code, and other materials to help provide students and professionals with an understanding of quantitative decision analysis, a.k.a. structured decision making. This site is also linked through the electronic companion to Conroy and Peterson. 2013. Decision Making in Natural Resource Management: a structured adaptive approach.

Course Description
Short story: Would you like to know how to use all that statistics you were taught? This class tells you how and (hint) it has nothing to do with "significance".

More detail: Natural resource managers often are faced with difficult decisions on how to satisfy the socio-economic needs of the public while conserving or restoring ecological systems. To aid in the decision-making process, the decision sciences have developed approaches that allow decision makers to: examine the expected effects of different strategies before implementation; incorporate multiple objectives and values of stakeholders; determine the relative influence of various sources of uncertainty; and estimate the value of collecting additional data. Adaptive management, a special case of decision analysis, is used to reduce uncertainties through monitoring, increasing the value of management. Despite the potential advantages, quantitative decision analysis and adaptive management are not used widely in natural resource management, with the exception of a few notable conservation efforts. To this point, a primary impediment to the broad-scale application of quantitative decision analysis has been a lack of training opportunities for natural resource students and professionals in the concepts and methodology. This course is intended to fill that gap by providing quantitatively- oriented students in natural resources and related fields with the skills needed to interpret and conduct complex quantitative decision analysis for managing animal populations.

Decision analysis allow decision makers to: examine the expected effects of different strategies before implementation; incorporate multiple objectives and values of stakeholders; determine the relative influence of various sources of uncertainty; and estimate the value of collecting additional data. This course provides quantitatively-oriented graduate students in natural resources with in depth overview of decision analysis, emphasizing animal population management. Adaptive resource management (ARM) is introduced as a special case that involves sequential dynamic decision-making, monitoring, and Bayesian updating. Students are taught the steps and techniques used to create and evaluate quantitative decision models. In the laboratory sections, students are taught the methods for parameterizing decision models using techniques that range from simple linear regression to heuristic optimization and are provided with R and WinBugs statistical software code. The course also emphasizes uses of graphical models for communicating complicated models to lay audiences.

The course consists of both lectures and computer laboratory exercises. Each week includes two 90-minute lectures and a 3-hour computer laboratory. Lectures cover the concepts and methods used for quantitative decision analysis.
The computer laboratory is designed to get students familiar with ecological modeling and to develop the skills necessary to build and evaluate quantitative decision models. Students will be required to synthesize and apply their knowledge of quantitative decision analysis by creating a decision model that includes all the required elements. To do so, all students should be able to:
  • Understand and describe the principles of quantitative decision analysis and its uses in natural resource management
  • Perform model selection and multi-model prediction to represent system uncertainty in decision models
  • Estimate the expected error rate of statistical models and incorporate estimates of statistical uncertainty in decision models
  • Use hierarchical models to perform a meta-analysis to combine and incorporate the results of previous studies into decision models
  • Create multi-objective utility functions to incorporate and quantify multiple objectives into decision models
  • Create ecological simulation models to parametrize and solve decision models
  • Perform sensitivity analysis on decision models and interpret and estimate the value of perfect and imperfect information on key decision model components
  • Solve multi-objective decision models to identify optimal decisions using stochastic dynamic programming and similar heuristic optimization techniques
  • Creating a decision model that includes all elements: objectives, decision alternatives, utilities, alternative competing models, and results using a published natural resource management pla


Papers every F & W grad student should read