Incorporation of preferential uncertainty into interval-scale priority functions — a case of multicriteria forestry decision making.

Quantitative decision analysis and respective planning models offer many benefits in forest planning. They are efficient, quick, inexpensive, objective, and repeatable. However, quantitative planning and the respective planning models also include many sources of uncertainty. In this paper our research objective is to analyse decision makers’ preferential uncertainty in the context of multicriteria forestry decision making by using statistical models for so-called interval-scale continuous decision problems. The models are based on 0–1-type utility functions by applying a Finnish school grading system for collecting the preference data. The basic idea is to assess the preferences at some discrete points and then estimate the continuous priority function according to the statistical estimation techniques. An interactive approach for reducing the preferential uncertainty included in the priority model is also developed. In the interactive step, the improvements of the priority model are based on uncertainty measures related to local and (or) global priority models. The role of statistical uncertainty analysis is to make the interactive planning process more efficient and reliable. The interactive and statistical approaches complement each other and promote decision makers’ learning. [ABSTRACT FROM AUTHOR]