![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081014083555im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Aquatic Life Use Support (ALUS)
Biocriteria Links
Additional Information
Performance Characteristics
Perhaps the most important component in making bioassessments useful to water resource programs is the data quality of different assessment methods currently in use and the level of comparability among methods in performing an assessment. The comparability of methods should be judged by the degree of similarity in their performance characteristics (i.e., a performance-based approach) rather than by direct comparison of their respective scores or metric values. To enable a sharing of data and results from various techniques that might be used by different agencies or other groups, some level of confidence in making an assessment must be established for each method based on the quality of data. This performance characteristic is precision, which is dependent upon the sampling methodology and the range in natural variation of the reference condition (note -- use of stream classification will increase precision).
The ability to detect impairment also depends on the sensitivity of the method. In some cases, the desirable sensitivity level depends on how severe or subtle the impairment. For example, it does not require a very rigorous method to detect impairment following an extensive fish kill or algal bloom. It is the subtle impact areas that require some level of rigor that minimizes Type I and Type II errors in a judgment of condition.
Based on preliminary information obtained from bioassessments conducted in Florida, Ohio, and New Hampshire, quantitative criteria for precision and sensitivity can be set conservatively at "high" being less or equal to 20%, "moderate" being between 21 and 49%, and "low" being more or equal to 50%. High precision is equated to having low measurement error (coefficient of variation <20%) and sensitivity is the ability to detect small differences (<20% difference) between reference and the site being assessed.