Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: A Multi-Attribute Tool for Decision Support: Ranking a Finite Number of Alternatives

Author
item Yakowitz, Diana

Submitted to: Soil and Water Conservation Society Proceedings
Publication Type: Proceedings
Publication Acceptance Date: September 10, 1996
Publication Date: N/A

Interpretive Summary: Many environmental management decision making deal with the problem of deciding among a number of alternatives based on a number of criteria. In this paper, a method is described for ranking the alternatives in order of preference. The range from best to worst possible scores for each alternative are computed by solving two mathematical programs and can be displayed by simple bar graphs. The method can be easily incorporated into a computerized decision support system for a wide variety of applications including environmental decision making or as an enhancement to existing systems and methods. Adaptions to the method are also developed. An example application and suggested display of the information are provided.

Technical Abstract: Many decision making problems can be expressed as the problem of deciding among a finite number of alternative management systems based on a finite number of criteria or attributes (impacts on the environment) of the alternatives. In this paper, a method is described for ranking the alternatives in order of preference without specifying weights on each criteria. The range from best to worst possible scores, given a priority order (ordinal) of the criteria, are computed by solving two simple linear programs and can be displayed by simple bar graphs. Strong or absolute dominance of one alternative over another can often be discerned immediately. The method can be easily incorporated into a decision support system for a wide variety of applications including environmental decision making or as an enhancement to existing systems and methods. Adapting the method to include the consideration of a hierarchy of the criteria and the possibility of different states of nature with some probability are also developed. Use of the model within a sampling scheme to predicted unsampled subdivisions is also suggested. An example application and suggested display of the information are provided.

Last Modified: 12/22/2014
Footer Content Back to Top of Page