Selection of Relevant Effect Levels for Using Bioequivalence Hypothesis Testing

Abstract

The use of classic hypotheses tests based on a null hypothesis of equal means frequently promote the occurrence of false positives and false negatives in ecotoxicological assays results. The use of criteria or appropriate statistical analyses is recommended to prevent these occurences and to guarantee the quality of the results from the biological as well statistical point of view. Therefore, the relevant effect levels were established for ecotoxicological assays with Daphnia similis, Ceriodaphnia dubia, Danio rerio, Hyalella azteca, H. meinerti, Lytechinus variegatus and Mysidopsis juniae. Such effect levels were estimated on the basis of the 75th percentile of the Minimum Significant Difference (MSD) of a historical series of analytical results. The estimated values were used to evaluate the allowable variability of the analytical results. It was demonstrated that the integration between the relevant effect levels and the bioequivalence hypothesis testing minimizes the occurrence of false positive results relative to those observed using traditional hypothesis testing.