目錄一、考情分二、專業(yè)詞抽樣風險：Sampling非抽樣風險：Non-sampling信賴不足風險：Underreliancerisk信賴過度風險：Overreliancerisk控制測試：Testsofcontrol控制有效性：Effectivenessof重大錯報：Material實際水平：Actual審計結(jié)果：Audit誤拒風險：Rejectionrisk誤受風險：Acceptancerisk細節(jié)測試：Testofdetails樣本量：Samplessize具體目標：Specific統(tǒng)計抽樣：Statistical成本效益：Cost的充分性：Sufficiencyofthe可接受的抽樣風險：Acceptablesamplingrisk可接受誤差（可誤差）：Acceptableerror認定層次：Assertionlevel總體變異性：Population預計的總體誤差：Expectedpopulation總體規(guī)模：Population專業(yè)判斷：Professional上限：Upper風險系數(shù)：Risk可偏差率：Tolerabledeviation修正重大錯報風險評估程度：Modifymaterialmisstatementriskassessment實質(zhì)性程序：Substantive調(diào)整賬目：Adjustbook賬面金額：Book實際金額：Actualmonetaryunit零余額：Zero負余額：Negative屬性抽樣：Attribute變量抽樣：Variable三、重點、Samplingriskandnon-sampling抽樣風險（由于采用抽樣而引起的風險）Samplingrisk（riskcausedbySamplingrisk:itistheriskarisingfromthepossibilitythattheCPA’sconclusion,basedonasample,maybedifferentfromtheconclusionthatwouldbereachediftheentirepopulationweresubjectedtothesameauditprocedure.ControlofsamplingSamplingriskisthepossibilitythatsamplesselectedmaynotcompleyrepresentthepopulation.Relationshipbetweensamplesizeandsamplingriskisopposite,thelargerthesamplesize,thelowerthesamplingriskis.Therefore,samplingriskalwaysexistsbecauseofsampling.抽樣風險的類型Typesofsamplingofcontrol.Overreliancerisk:theexpectedeffectivenessofcontrolishigherthantheactuallevel,riskassessedofmaterialmisstatementsislowerthantheactuallevel,therebyinfluencingtheauditeffectiveness（quality）.Underreliancerisk:theexpectedeffectivenessofcontrolislowerthantheactuallevel，assessedriskofmaterialmisstatementsishigherthantheactuallevel,therebyinfluencingtheauditefficiency.②CPAismainlyfacedwithrejectionriskandacceptanceriskwhenconductingdetailedAcceptanceriskrefersthatCPAinfersthatthematerialmisstatementisnotexistedbutactuallyitisnot,therebyinfluencingtheauditeffectiveness（quality）.RejectionriskrefersthatCPAinfersthatthematerialmisstatementisexistedbutactuallyitisnot,therebyinfluencingtheauditefficiency.Non-samplingriskrefersthepossiblefalsedecisionmadebyCPAbasedonreasonsnotrelatedtosamples.Followingcircumstancesmaycausenon-samplingriskin（1）會計師選擇了不適于實現(xiàn)特定目標的審計程序CPAperformsauditprocedureswhicharenotsuitableforthespecific（2）會計師選擇的總體不適合于測試目標ThepopulationselectedbyCPAisnotsuitableforthetestmisstatement）,whichmakesCPAunabletofindthedeviationormisstatement;（4）CPAdoesnotadequayevaluatetheauditNon-samplingriskcannotbefied,butnon-samplingriskcanbereducedtoanacceptablelevelthroughapplicationofappropriatequalitycontrol&proceduresandproperguidance,supervisionandreviewofaudit,aswellastheappropriateimprovementtoCPApractice.CPAcanalsocarefullydesigntheirauditprocedurestominimizenon-samplingStatisticalsamplingandnon-statisticalSelectionbetweenstatisticalsamplingandnon-statisticalsamplingCostbenefitsaremainlyconsideredwhenchoosingbetweenstatisticalsamplingandnon-statisticalsamplingmethod.Disadvantagesofstatisticalsampling:statisticalsamplingmayincurextra（2）Advantagesofstatisticalsampling:samplingriskcanbemeasuredobjectively,andactuallycontrolledbyadjustingthesamplesize,whichisthemostimportantdifferentfromnon-statisticalsampling.StatisticalsamplinghelpsCPAtoefficientlydesignsamples,measurethesufficiencyoftheevidencesobtained,andevaluatethesamplingresults2.AuditsamplingandprofessionalProfessionaljudgmentisrequiredforCPAwhendesigning,implementingandsamplesnomatterthestatisticalsamplingornon-statisticalsamplingisIfthedesignofnon-statisticalsamplingisappropriate,itcanprovidethesameeffectiveresultasstatisticalsamplingdesignedproperly.Whenusingnon-statisticalsampling,CPAmustconsiderthesamplingriskandreduceittotheacceptablelevel,however,thesamplingriskcannotbeaccurayestimated.AttributesamplingandVariableAttributeAttributesamplingisastatisticalsamplingmethodusedtodrawconclusionsabouttheincidenceofaneventinthepopulation.2.VariableVariablesamplingisastatisticalsamplingmethodusedtodrawconclusionsaboutthepopulationamount.一般而言，屬性抽樣得出的結(jié)論與總體發(fā)生率有關(guān)，而變量抽樣得出的結(jié)論與總體的金額有關(guān)。Generallyspeaking,theconclusiondrawnfromattributesamplingisrelatedtotheincidencerateofthepopulation,andtheconclusiondrawnfromvariablesamplingisrelatedtothepopulationamount.However,thereisanexception,thatis,monetaryunitsamplinginvariablesampling,andtheconclusiondrawnusingtheprincipleofattributesamplingisreflectedatamount.3.可偏差3.TolerabledeviationIncontroltesting,tolerabledeviationratereferstotheratethatdeviatesfromproscribedinternalcontrolsetbyCPA.Inanotherwords,itmeanstheumdeviationacceptedbyCPA.Ifthedeviationexceedsthatnumberthenitwouldreduceoreliminatetherelianceontheinternalcontrolprocedures.ApplicationofauditsamplingFactorsaffectthesamplesize:AcceptableoverrelianceoverrelianceriskacceptedbyCPA,thelargerthesamplesizeisneeded.可偏差2.TolerabledeviationTolerabledeviationrateisinoppositerelationshipwithsamplesize.WhendeterminingTheestimatedpopulationdeviationratechangesinthesamedirectionwiththesamplethesamplesizeisrequired.3.3.EstimatedpopulationdeviationForcontroltesting,whenconsideringthepopulationcharacteristics,CPAshouldofasmallnumberofitemsinthepopulation.4.總體規(guī)4.PopulationUnlessthepopulationsizeisverysmall,ithasnoinfluenceonthesampleThemorecomplexthecontrolprocedureis,themoresamplesareThesamplesizealsodependsonthetypeofcontroltested,usuallymoretestingareneededformanualcontrolthanthatofautomatedcontrol.5.其他因素?？刂七\行的相關(guān)期間越長，需要測試的樣本越多，因為會計師需要對整個擬信賴5.Otherfactors.Thelongertheperiodofcontroloperationis,themoresamplesneedtobetested.BecauseCPAsneedtoobtainevidenceforthecontrolvalidityofthewholeperiodthatnnedtobetrusted.Ⅳ.ApplicationofauditsamplingincontrolStatisticalsamplingmethodsusedincontroltestusing–resultsofevaluationTheuseofstatisticalsamplingincontroltestingtoevaluatesamplingPopulationdeviationrate＝samplingdeviationrate＝samplingdeviation÷sampleConsideringsamplingUsingformula：upperlimitofpopulationdeviation＝riskcoefficient÷sampleUsingtable:usingthesampleresultsevaluationtableEvaluationandconclusionofstatisticalsampling.Theupperlimitofpopulationdeviationrateislessthanthelimitoftolerabledeviationratesopopulationisacceptable.Theupperlimitofpopulationdeviationrateislowerthanbutclosetotolerablescope;Theupperlimitofpopulationdeviationrateisequaltoorhigherthantolerablelevel,andincreasethenumberofsubstantiveprocedures;ortestothercontrols,tosupportthennedriskassessmentlevelofmaterialmisstatements.Innon-statisticalsampling,theCPAwillalsousethedeviationrateofsamplesasthebestestimateofthedeviationrateofthepopulation.可偏差率相比較，以判斷總體是否可以接受。Butinanon-statisticalsampling,samplingriskcannotbemeasureddirectly.TheCPAusuallycomparesdeviationrateofsamples（i.e.,thedeviationrateofthepopulationUsingnon-statisticalsamplingmethodintestofcontrolstoevaluatesamplingPopulationdeviationrateofthepopulation＝deviationrateofthesamples＝deviationamountofthesamples/samplesize非統(tǒng)計抽樣的結(jié)果評價與結(jié)論（非統(tǒng)計抽樣無法量化抽樣風險Evaluationandconclusionofnon-statisticalsamplingresult（thesamplingriskofnon-statisticalsamplingcannotbefied）:總體“可以接受”；樣本偏差率“低于但接近”可偏差率，總體“不可接受”；Ifdeviationrateofthesamplesishigherthanthetolerabledeviationrate,thepopulationcannotbeaccepted;ifdeviationrateofthesamplesissignificantlylowerthanthetolerabledeviationrate,thepopulationcanbeaccepted;ifdeviationrateofthesamplesislowerthanbutclosetothetolerabledeviationrate,thepopulationcannotbeacceptedDeviationrateofthesamplesislowerthanthetolerabledeviationrate,butthedifferenceisneithertoobignortoosmall,considerwhethertoacceptthepopulation;considerexpandingthesamplesizeinordertocollectfurtherevidence.V.ApplicationofauditsamplingintestofDeterminationofthesampleRegardlessoftheuseofstatisticalsamplingornon-statisticalsamplingmethod,size,usingprofessionaljudgmentandexperiencetodeterminethesamplesize.Usingthedeterminationtableofsample-5879SampleSizeofMonetaryUnitSamplinginTestofEstimatedmisstatementofthe/TolerableTolerablemisstatement/Bookvalueof-5879Usingthe錯報之比為0.20，會計師確定樣本規(guī)模為：Forexample,theriskofincorrectacceptancedeterminedbyCPAis10%,theratioofofthepopulation/tolerablemisstatementis0.2,thesamplesizedeterminedbyCPAis:Usingtablemethod:Usingformulamethod:AssurancecoefficientwhendeterminingsamplesizeinmonetaryunitEstimatedmisstatementofthe/TolerableriskofincorrectUsestatisticalsamplingmethodintestofdetailstoevaluatesamplingDeducemisstatementofthepopulation（traditionalvariable①均值法MeanMeanofsampleapprovedamount＝Approvedamountofsample/SampleEstimationofpopulationvalue＝Meanofsampleapprovedamount×PopulationMisstatementofpopulationdeduced＝Estimationofpopulationvalue-Bookvalueofthe②差額法DifferenceMisstatementofpopulationdeduced＝Averagemisstatementofsample×PopulationEstimatedpopulationamount=Bookvalueofpopulation–Misstatementofpopulation③比率法RatioRatio＝Approvedamountofsample/BookvalueofEstimatedpopulationamount＝Bookvalueofthepopulation×RatioMisstatementofpopulationdeduced＝Bookvalueofthepopulation-EstimatedpopulationEvaluationandconclusionofthestatisticalsampling可錯報，總體“可以接受”。Ifupperlimitofthepopulationmisstatementexceedsorisequaltotolerablemisstatementamount,thepopulationcannotbeaccepted;Ifthetransactionoraccountbalancetestedhasmaterialmisstatement,CPAshouldsuggesttoinvestigatethemisstatementandadjustbookrecordwhennecessary；Ifupperlimitofthepopulationmisstatementislessthantolerablemisstatementamount,thepopulationcanbeacceptedVI.MonetaryunitAdvantagesofmonetaryunitaccountanttocalculatethesamplingsizeandassessthesampleresults.Therefore,itisusuallyeasiertobeusedcomparedwithtraditionalvariablesampling.Whendeterminingthesamplesizeofmonetaryunitsampling,itisnotnecessarytoconsiderthecharacteristicsofpopulation（e.g.populationdeviation）.Becauseeachmonetaryunitinthepopulationhasthesamesize,whilethesamplesizeintraditionalvariablesamplingiscalculatedbasedonthedeviationorstandarddeviationofcommoncharacteristicsofamounthasaproportionalrelationshipsothereisnoneedtoreducedeviationthroughstratification.Whileintraditionalvariablesampling,itisusuallynecessarytostratifypopulationtoreducesamplesize.Whenusingsystemsamplingmethodtoselectsampleinthemonetaryunitsampling,automaticallyidentifyallindividualmajoritems,namelytheitemwillbecertaintobesizeofmonetaryunitsamplingissmallerthanthatoftraditionalvariablesampling.obtainingcompletedpopulation.DisadvantagesofmonetaryunitMonetaryunitsamplingisnotsuitablefortheunderstatementofpopulation,becausetheprobabilityofbeingchosenforasmallbookvaluebutgrosslyunderstatedprojectislow.Ifunderstatementisdiscoveredinmonetaryunitsampling,certifiedpublicaccountantsneedspecialconsiderationwhenevaluatingsamples.Whenmisstatementsareidentified,iftherisklevelisfixed,monetaryunitsamplingmayoverestimatetheeffectofsamplingriskwhenevaluatingsamples,whichmayleadtothatCPAmayrefuseanacceptablebookvalueofthepopulation.CPAneedstoaccumulatetotalamountonebyoneinmonetaryunitsampling,inordertodeterminewhetherthepopulationiscompleteandconsistentwiththefinancialstatementsofauditing（5）Whenthevolumeofmisstatementsinthepopulationincrease,thesamplingsizeofmonetaryunitsamplingwillalsoincrease.Underthesesituations,thesamplesizeofmonetaryunitsamplingmaybegreaterthanthatoftraditionalvariablesampling.CalculatetheupperlimitofpopulationmisstatementsinmonetaryunitIfnomisstatementisdiscoveredinthesample,theupperlimitofpopulationmisstatements＝Assurancecoefficient×Samplinginterval.Forexample,riskofincorrectacceptanceis5%,samplingintervalis3000Yuan（thesamebelow）,CPAshaven'tfoundany×3000）,whichiscalledbasicIfamisstatementisfoundinalogicalunitthatislargerthanorequaltothesampleinterval,nomatterwhetherthepercentageofthemisstatementis100%,theupperlimitofpopulationmisstatements＝factualmisstatement＋basicaccuracy.Forexample,CPAfoundonemisstatementinsample,thebookvalueoftheitemis5000Yuan,approvedamountis4000Yuan,theupperlimitofpopulationmisstatementsis10000YuanIfamisstatementwith100%misstatementpercentageinsample（excludinglogicalunitthatislargerthanorequaltothesampleinterval）isfound,theupperlimitofpopulationmisstatements＝Assurancecoefficient×Samplinginterval.Forexample,CPAfoundonemisstatementinsample,thebookvalueoftheitemis20Yuan,approvedamountis0Yuan,thentheupperlimitofpopulationmisstatementsis14250Yuan（4）Ifamisstatementwiththemisstatementpercentagelowerthan100%in（excludinglogicalunitthatislargerthanorequaltothesampleinterval）isfound,CPAshouldfirstcalculateprojectedmisstatements,thenafrrangingprojectedmisstatementsindescendingorderaccordingtoamount,multiplytheincrementaltyofcorrespondingassurancecoefficientrespectively（i.e.Underthegivenlevelofriskofincorrectacceptance,thedifferencebetweentheassurancecoefficientcorrespondingspecifiednumberofoverstatementmisstatementsandtheassurancecoefficientofaboveline）,plusbasicaccuracy,calculatetheupperlimitofpopulationmisstatementsatlast.Theupperlimitofpopulationmisstatements＝projectedmisstatements×incrementaltyofassurancecoefficient＋basicaccuracy.Forexample,riskofincorrectacceptanceis5%,samplingintervalis3000Yuan,CPAfoundtwomisstatementsinsample,thebookvalueofaccountAis2000Yuan,approvedamountis1500Yuan,projectedmisstatementis750Yuan（500÷2000×3000）；thebookvalueofaccountBis1000Yuan,auditedvalueis200Yuan,projectedmisstatementis2400Yuan（800÷1000×3000）.將推斷錯報按金額降序排列后，由表可知，在5%的誤受風險水平下，賬戶A對應的保證系數(shù)增量Afterarrangingprojectmisstatementsindescendingorderaccordingtoamount,fromthetable,underthe5%levelofriskofincorrectacceptance,theincrementaltyofcorrespondingassurancecoefficientofaccountAis1.55,theincrementaltyofcorrespondingassurancecoefficientofaccountBis1.75.Therefore,theupperlimitofpopulationmisstatementsis14363Yuan（750×1.55＋2400×1.75＋3×3000）百分比低于100%的錯報。Iftherearebothlogicalunitswithbookvaluegreaterthanorequaltothesamplinginterval,andlogicalunitswithbookvaluelessthanthesamplinginterval,andinthelogicalis100%isdiscoveredandmisstatementwithmisstatementpercentagelessthan100%isalsodiscovered.Howtosolve?CPAcoulddivideallthesampleitemsintotwoThefirstgroupismadeupoflogicalunitswithbookvaluegreaterthanorequaltothesamplinginterval.CPAcouldcalculatethefactualmisstatementofitemsinthisgroup.第二組是賬面金額小于選樣間隔的邏輯單元，無論該組項目的錯報百分比是否為100%，會計Nomatterwhetherthemisstatementpercentageofitemsinthisgroupis100%,CPAshouldfirstcalculatetheprojectedmisstatementofeachitem,thenafterarrangingalltheprojectedmisstatementsindescendingorderaccordingtoamount,multiplytheincrementaltyofcorrespondingassurancecoefficientrespectively,andaddupthecalculationresults.Usethiscumulativeresulttoaddbasicaccuracy,andaddthefactualmisstatementinthefirstgroup,thenistheupperlimitofpopulationmisstatements.However,CPAsshouldalsosummarytheprojectedmisstatements（excludingthefactualmisstatementswhichhavebeenalreadycorrectedbytheauditee’smanagement）withotherfactualmisstatementsandprojectedmisstatements,toassesswhetherthefinancialstatementsasawholemaybematerialmisstated.四、同步系Incase（1）,thestatement“CPAcanuseauditsamplinginallproceduresoftestsofcontrols”isinappropriate.Onlywhencontroloperationleavestrack,thenCPAcanperformauditsamplingtotesttheeffectivenessofinternalcontroloperation.第（2）項，“會計師將總體分為三層”不當。在進行控制測試時無需對總體進行分層Incase（2）,thestatement“CPAshoulddividethepopulationinto3levelssub-populations”isinappropriate.Nopopulationstratificationisrequiredwhenperformingtestsofcontrols.Incase（3）,thestatement“thesamplingmethodhaseffectonthenatureofauditproceduresperformed”isinappropriate.Theauditproceduresperformedinsamplesselectedareunrelatedtothesamplingmethodused.Incase（4）,thestatement“thelargerthepopulationis,themoresamplesarerequired”isinappropriate.Intermsofoveralllargepopulation,thepopulationsizehasnoeffectonsamplesize.Incase（5）,thestatement“thelowertheacceptablesamplingriskis,thelargerofsamplesisrequired”is率上限低于但接近可偏差率時，應考慮總體是否可以接受，并考慮是否需要擴大測試范圍。Incase（6），thestatement“whentheupperlimitofpopulationdeviationrateislowerthanbutclose