Laserfiche WebLink
<br />December 2003 <br /> <br />3.0 Ranking and Weighting Exercises <br /> <br />After a review of the consequence table, stakeholders participated in a structured values <br />elicitation process. There are three main reasons to use structured methods to elicit <br />stakeholder values and preferences: <br />to increase the accuracy and consistency of individual stakeholder judgments; <br />to provide focus for constructive deliberations and refinement of the options; <br />to increase the accountability and transparenc.y of decisions by making the trade-offs made <br />by stakeholders explicit. <br />The purpose is not to prescribe an answer. The goal of decision modeling, as in ecological <br />modeling, is to provide insight as an aid to decision making. Stakeholders are still responsible <br />for making difficult value based trade-offs and choices. <br /> <br />There are many ways to elicit values. Different methods usually produce different results; no <br />method is necessarily right. The use of multiple methods provides insight to the decision by <br />thinking about it in different ways. By examining choices from different perspectives, <br />stakeholders will have more confidence that their choices reflect their values, and are not the <br />result of methodological bias. <br /> <br />In this exercise, we used two methods: a) Direct Ranking and b) Swing Weighting. In Direct <br />Ranking, stakeholders were asked to rank and then score each management option directly, <br />based on a review of the consequence table. In swing weighting, they were asked to rank and <br />weight each attribute. The term "swing" weighting is used because decision makers are asked <br />to say which attribute they would most want to "swing-up" from its worst to its best value. <br />This is important because in some cases an attribute may be important in a general sense, but <br />the actual change in the attribute value that results from the choice among management <br />options may be relatively insignificant (i.e., it is not partic.ularly sensitive to the option set); <br />this should affect the weight assigned to it, as we are weighting the importance we assign to <br />the attribute in this specific decision context. <br /> <br />For swing weighting, attribute weights are entered into the following equation that computes <br />an overall score for each option: <br /> <br />SCORE(a) = W1 (x1a) + WI(xla) + ...... <br /> <br />Where: <br /> <br />SCORE(a) <br />W1,WI'" <br /> <br />x1' XI'" <br /> <br />= the calculated score for a management option (e.g. 'a') <br />= the weight of an attribute <br /> <br />= the scaled impact of a given option on each attribute <br /> <br />Ranks for each management option for each stakeholder are then derived. <br /> <br />The ranking and weighting questionnaires are shown in Appendix Co <br /> <br />8 <br />GCDAMP MATA: <br />December 2003 Workshop Report <br />