作者:标准资料网 时间:2024-05-18 10:06:19   浏览:8616   来源:标准资料网

【英文标准名称】:StandardPracticeforStatisticalAssessmentandImprovementofExpectedAgreementBetweenTwoTestMethodsthatPurporttoMeasuretheSamePropertyofaMaterial
【原文标准名称】:测量同一物质相同属性的两种测试方法间预期协定的统计评定与改进标准实施规程
【标准号】:ASTMD6708-2008
【标准状态】:现行
【国别】:美国
【发布日期】:2008
【实施或试行日期】:
【发布单位】:美国材料与试验协会(US-ASTM)
【起草单位】:D02.94
【标准类型】:(Practice)
【标准水平】:()
【中文主题词】:
【英文主题词】:Agreement(testingmethods);Bias;Cross-methodreproducibility;Interlaboratorytesting;Linearbiascorrection;Quantifieddata;Statisticalmethods--petroleum;Testingmethods
【摘要】:Thispracticecanbeusedtodetermineifaconstant,proportional,orlinearbiascorrectioncanimprovethedegreeofagreementbetweentwomethodsthatpurporttomeasurethesamepropertyofamaterial.Thebiascorrectiondevelopedinthispracticecanbeappliedtoasingleresult(X)obtainedfromonetestmethod(methodX)toobtainapredictedresult(x02C6;Y)fortheothertestmethod(methodY).Note78212;Usersarecautionedtoensurethatx02C6;YiswithinthescopeofmethodYbeforeitsuse.Thebetweenmethodsreproducibilityestablishedbythispracticecanbeusedtoconstructanintervalaroundx02C6;YthatwouldcontaintheresultoftestmethodY,ifitwereconducted,withabout95%confidence.Thispracticecanbeusedtoguidecommercialagreementsandproductdispositiondecisionsinvolvingtestmethodsthathavebeenevaluatedrelativetoeachotherinaccordancewiththispractice.1.1Thispracticecoversstatisticalmethodologyforassessingtheexpectedagreementbetweentwostandardtestmethodsthatpurporttomeasurethesamepropertyofamaterial,anddecidingifasimplelinearbiascorrectioncanfurtherimprovetheexpectedagreement.ItisintendedforusewithresultscollectedfromaninterlaboratorystudymeetingtherequirementofPracticeD6300orequivalent(forexample,ISO4259).Theinterlaboratorystudymustbeconductedonatleasttenmaterialsthatspantheintersectingscopesofthetestmethods,andresultsmustbeobtainedfromatleastsixlaboratoriesusingeachmethod.Note18212;ExamplesofstandardtestmethodsarethosedevelopedbyvoluntaryconsensusstandardsbodiessuchasASTM,IP/BSI,DIN,AFNOR,CGSB.1.2Thestatisticalmethodologyisbasedonthepremisethatabiascorrectionwillnotbeneeded.Intheabsenceofstrongstatisticalevidencethatabiascorrectionwouldresultinbetteragreementbetweenthetwomethods,abiascorrectionisnotmade.Ifabiascorrectionisrequired,thentheparsimonyprincipleisfollowedwherebyasimplecorrectionistobefavoredoveramorecomplexone.Note28212;Failuretoadheretotheparsimonyprinciplegenerallyresultsinmodelsthatareover-fittedanddonotperformwellinpractice.1.3Thebiascorrectionsofthispracticearelimitedtoaconstantcorrection,proportionalcorrectionoralinear(proportional+constant)correction.1.4Thebias-correctionmethodsofthispracticearemethodsymmetric,inthesensethatequivalentcorrectionsareobtainedregardlessofwhichmethodisbias-correctedtomatchtheother.1.5Amethodologyispresentedforestablishingthe95%confidencelimit(designatedbythispracticeasthebetweenmethodsreproducibility)forthedifferencebetweentworesultswhereeachresultisobtainedbyadifferentoperatorusingdifferentapparatusandeachapplyingoneofthetwomethodsXandYonidenticalmaterial,whereoneofthemethodshasbeenappropriatelybias-correctedinaccordancewiththispractice.Note38212;Inearlierversionsofthisstandardpractice,thetermx201C;cro......
【中国标准分类号】:A41
【国际标准分类号】:75_080
【页数】:14P.;A4
【正文语种】:英语