Measurement and meaningfulness in conservation science.

Incomplete databases often require conservation scientists to estimate data either through expert judgment or other scoring, rating, and ranking procedures. At the same time, ecosystem complexity has led to the use of increasingly sophisticated algorithms and mathematical models to aid in conservation theorizing, planning, and decision making. Understanding the limitations imposed by the scales of measurement of conservation data is important for the development of sound conservation theory and policy. In particular, biodiversity valuation methods, systematic conservation planning algorithms, geographic information systems (GIS), and other conservation metrics and decision-support tools, when improperly applied to estimated data, may lead to conclusions based on numerical artifact rather than empirical evidence. The representational theory of measurement is described here, and the description includes definitions of the key concepts of scale, scale type, and meaningfulness. Representational measurement is the view that measurement entails the faithful assignment of numbers to empirical entities. These assignments form scales that are organized into a hierarchy of scale types. A statement involving scales is meaningful if its truth value is invariant under changes of scale within scale type. I apply these concepts to three examples of measurement practice in the conservation literature. The results of my analysis suggest that conservation scientists do not always investigate the scale type of estimated data and hence may derive results that are not meaningful. Recognizing the complexity of observation and measurement in conservation biology, and the constraints that measurement theory imposes, the examples are accompanied by suggestions for informal estimation of the scale type of conservation data and for conducting meaningful analysis and synthesis of this information.