|
|
 |
|
|
| |
Communication at the NABS Annual meeting, Keystone, 2000
(2) MIXED MESSAGES: DECISION-MAKING IN BIOASSESSMENTS BASED ON MULTIPLE TEST RESULTS.
R.C. Bailey. Department of Zoology, The University of Western Ontario, London, Ontario, Canada N6A 5B7
Most benthologists realize that many approaches to the bioassessment of aquatic systems can result in a credible diagnosis: the lake or stream site either passes or fails based on comparison of some descriptor of community structure or function to some criterion. We have no single "silver bullet" test to use when evaluating a site. Inevitably, when more than one type of test is carried out (e.g. richness, organic pollution tolerance), there is some degree of inconsistency among the diagnoses generated for a particular site. I will illustrate the meta-analysis of such results, and suggest a gradient of responses appropriate to the suite of diagnoses made for a test site. I will then ponder the power one has to detect failing sites at various points on this gradient, and propose a type of Bonferroni correction to protect against too high a probability of Type II error when one combines bioassessment results in such a manner.
Presented at 1:15 PM on Monday, May 29, 2000 in Bioassessment I
|
|
|
|
|
|
|
