Improving the quality of statistics in regulatory ecotoxicity tests.

: The results of an international workshop on the use of statistics in regulatory ecotoxicology are presented. There are currently many errors of omission in the recommendations on statistical analysis given in test guidelines. These are identified and advice is given on how to incorporate best statistical practice. The use of the no observed effect concentration (NOEC) as a summary statistic is questioned, and an alternative is suggested. Several areas of research that would resolve uncertainty in the design and analysis of ecotoxicity tests are also identified.