REVISED APPROACH TO TOXICITY TEST ACCEPTABILITY CRITERIA USING A STATISTICAL PERFORMANCE ASSESSMENT

Current acceptability requirements for toxicity tests are often more r estrictive than necessary. They focus primarily on response in a contr ol and generally ignore what a test was ''designed'' to detect as a si gnificant difference from that control. An approach is presented that takes into account the performance of an entire test and the magnitude of the deviation from the current acceptability requirements. The pro cedure is based on analyzing the past statistical performance of a tes t method (i.e., what kind of difference from the control was the test designed to detect). It takes into account traditonal control acceptan ce criteria, but adds a requirement for selecting a difference from th e control desired to be detected as statistically significant (a thres hold value). Choice of statistical procedure is not relevant to the ap proach. The proposed method allows a sliding scale of acceptance. The greater the deviation of mean control response below current requireme nts, the less likely a test is to be accepted. An example is presented using data from a 10-d sediment test using the marine amphipod Ampeli sca abdita. Use of the proposed acceptability criterion will reduce th e frequency of required retesting without sacrificing defensibility of data. Using the old acceptability criterion, 19% of the samples in th e amphipod data set would require retesting. The proposed criterion re duces the potential percentage of retests to 9%.