简介：Amajorroleofecologicalriskassessment(ERA)hasbeentoprovidescientificguidanceonwhetherafuturehumanactivitywillcauseecologicalharm,includingsuchactivitiesasreleaseofageneticallymodifiedorganism(GMO),exoticspecies,orchemicalpollutantintotheenvironment.Thisrequiresthedeterminationofthelikelihoodsthattheactivity:wouldcauseaharm,andwouldnotcauseaharm.Inthefirstcase,thefocusisondemonstratingthepresenceofaharmanddevelopingappropriatemanagementtomitigatesuchharm.Thisisusuallyevaluatedusingstandardhypothesisanalysis.Inthesecondcase,thefocusisondemonstratingtheabsenceofaharmandsupportingadecisionofbiosafety.WhilemostERAresearchershavefocusedonfindingpresenceofharm,andsomehavewronglyassociatedthelackofdetectionofharmwithbiosafety,anovelapproachinERAwouldbetofocusondemonstratingdirectlythesafetyoftheactivity.Although,someresearchershavesuggestedthatretrospectivepoweranalysiscanbeusedtoinferabsenceofharm,itactuallyprovidesinaccurateinformationaboutbiosafety.Adecisionofbiosafetycanonlybesupportedinastatisticallysoundmannerbyequivalencetests,describedhere.Usinga20%ecologicalequivalencestandardinGMOexamples,weillustratedtheuseofequivalencetestsfortwo-sampleswithnormalorbinomialdataandmulti-samplenormaldata,andprovidedaspreadsheetcalculatorforeach.Insixoftheeightexamples,theeffectsofCrytoxinsonanon-targetorganismwereequivalenttoacontrol,supportingadecisionofbiosafety.Theseexamplesalsoshowedthatdemonstrationofequivalencedoesnotrequirelargesamplesizes.Althoughmorerelevantecologicalequivalencestandardsshouldbedevelopedtoenableequivalenceteststobecomethemainmethodtosupportbiosafetydecisionmaking,weadvocatetheiruseforevaluatingbiosafetyfornon-targetorganismsbecauseoftheirdirectandaccurateinferenceregardingsafety.