Metric Comparison, Combination, & Trade-Off

Project planners must evaluate, compare, and trade-off benefits associated with alternative restoration actions. Broadly defined, a trade-off is giving up one thing to gain another. When metrics are similar, trade-off analysis may be straightforward (e.g., comparing economic benefits and costs in terms of the equivalent unit present day dollars). However, when metrics are dissimilar, trade-offs become less clear. In simple decision problems with few metrics, comparing the benefits and costs of an alternative may be simple and done qualitatively; however, as problem complexity increases, quantitative techniques for metric combination may be required to facilitate decision making.


The ability to compare metrics that measure diverse objectives is crucial to ecosystem restoration decision making. Techniques facilitating metric comparison and combination have been well-studied and may be coarsely divided into the four major categories shown below.

Techniques for comparing and combining dissimilar metrics.
Technique	Description
Narrative Description	For simple decision problems, metric comparison and trade-off may be straightforward, rapid, and require little or no analysis. However, the more metrics one is comparing, the more challenging trade-offs become.
Arithmetic Combination	A variety of simple arithmetic techniques exist for combining dissimilar metrics, such as simple arithmetic, nested combination, conversion to consistent units (e.g., dollars), or transformation to a consistent scale (e.g., 0 to 1).
Multicriteria Decision Analysis	Multicriteria decision analysis uses weighted combination of metrics to capture their relative importance in a decision. Although not explicitly weighted, even the simplest arithmetic combination methods are implicitly weighted; that is, all parameters are assumed to have equal weight and import. Although weights are subjective and derived from expert opinion in MCDA, judgments are collected using visible and traceable methods.
Interdependent Combination	Given the interconnectedness of ecosystems, objectives may be intimately related to other objectives and can be accounted for through simple (e.g., linear dependency of fish passage projects actions on the Truckee River) or complex algorithms.

Selecting a comparison technique
Each metric comparison method is uniquely different in analytical rigor, data resolution, capacity to capture and track uncertainty, time commitment, and cost of application. Thus, selection of a technique is a function of the application at hand. A few key questions to consider when selecting a technique include:

• Are there multiple metrics? Is there a need for combination?
• Do the metrics have equivalent units (e.g., acres, $)?
• Can metrics be combined linearly (e.g., arithmetic mean), or are there non-linear effects associated with combination (e.g., geometric mean)?
• Do metrics need to be transferred to a consistent scale for comparison (e.g., 0 to 1)?
• Are some objectives and metrics disproportionately important?
• Are objectives interdependent?