ABSTRACT
Fewlaboratorymanagerswilldisagreewiththestatement thatthegreatestprobleminprovidingaccuratedataiswith thesamplesthemselves.Errorsthatresultfromnonrepresentative,contaminated,orunstablesamplesfarexceedthose causedbytestmethoddeficienciesorlackofanalyticalskill. Unfortunately,thelaboratoryoftenhaslittlecontroloverthe samplingandisexpectedtoworkwithwhatitreceivesfrom itsclients.Thus,althoughtheanalyticaldatamaybecorrect withrespecttothelaboratorysample,theclientmayquestion itsaccuracyor,ifnot,acceptitastrulyrepresentativeofthe materialofinterestandcometoacompletelyerroneous conclusion.
