Munoz, B., Savage, R., & Baker, V. (2018). General issues in statistical analysis of RAMs. In J. Dorney, R. Savage, R. Tiner, & P. Adamus (Eds.), Wetland and Stream Rapid Assessments : Development, Validation, and Application (1 ed., pp. 251-258). Article 10 Elsevier Publishing.
Decisions regarding management, regulation, and protection of water bodies often rely on results from RAMs. Those results may include indices, metrics, and/or scores. Because the results sometimes have implications for policies and regulations, their credibility can be strengthened by correctly applying appropriate statistical tests at several points during the development or modification of a RAM, and in some cases using those results to refine the RAM further. This part of Section 3 describes statistical methods available for calibrating and then validating indexes, scores, and metrics. The described statistical methods are standard in the calibration and validation of models in many research areas. The goal of this chapter is to guide the identification and implementation of appropriate statistical methods as applied to RAM data for streams and wetlands as well as to aid in the interpretation of results. These methods are also appropriate for calibrating and validating indexes (e.g., Biotic Index) and components of longer-term assessments. At the end of this chapter, an example demonstrates the calibration and validation process used by one categorical RAM (NC Wetland Assessment Method (NC WAM)—see Chapter 4.3.1). This could provide an analysis template for use in calibrating and validating similar RAMs. The example also includes programming code and annotated output.