Comparing normative and descriptive methods for multi-criteria decision analysis: A case study evaluating wetland restoration opportunities in the Chesapeake Bay watershed, USA

Abstract

Normative and descriptive multi-criteria decision analysis methods were compared to understand their differences in providing decision insights, with a wetland restoration planning study, Chesapeake Bay watershed, USA. Excessive nutrient and sediment runoff is causing impaired water quality and degradation of aquatic habitat in the Chesapeake Bay. A team at The Nature Conservancy evaluated a decision to prioritize large-scale wetland restoration opportunities. A total of 964 potentially restorable wetland alternatives were delineated using spatial analysis. The alternatives were evaluated on seven water quality and climate resilience criteria. Normative methods included ranking the alternatives based on scaling and averaging criteria values per alternative using value and distance functions. Descriptive methods included grouping the alternatives based on their statistical similarities and dissimilarities using principal component analysis. Sensitivity analysis incorporated different data transformations based on risk and riskless decision contexts, different criteria weighting scenarios, and two formats for grouping alternatives. The descriptive method provided inherent information about tradeoffs that normative methods did not, whereas the normative methods provided a similar type of information about tradeoffs. Inherent tradeoff information can provide useful decision insights without complicated criteria scaling and weighting methods, which can be favorable to researchers that desire a more objective analysis. However, descriptive methods do not guarantee favorable results and should be considered along with preferences and judgments. Multiple methods can guide a more comprehensive approach to thinking about tradeoffs in other similar contexts. Data AvailabilitySupporting data are available on request: David.Martin@tnc.org.