外文文獻(xiàn)及翻譯

1、王雪琴 管理學(xué)院 會計系 財務(wù)管理 財管12-2班 學(xué)校代碼：10128 學(xué) 號：201210707016 本科畢業(yè)論文 （英文參考文獻(xiàn)及譯文） 題目：STATISTICAL SAMP LING METHOD, USED IN THE AUDIT 學(xué)生姓名： 學(xué)院： 系別： 專業(yè)： 班級： 0六年六月 Statistics an dAudit STATISTICALSA MP LING METHODISED INTHEAUDIT -views,recommendatiornindings PhD Candidate Gabriela-Felicia UNGUREANU Abstract The

2、ra pidin crease in thesizeofU.S.co mpani esfromtheearlytwe ntiethce n turycreatedthe needforaudit proceduresbasedontheselectioofa partofthetotal popu lati on auditedtoobta in reliableauditevide ncetocharacterizethee ntire popul ati oncon sistsofacco un tbala ncesorclassesof transactions.Samplingisno

3、tusedonlyinaudit -susedinsamplingsurveys, marketa nalysisa ndmedicalresearchi nwhichsome on ewa ntstoreachac on clusi on aboutalarge nu mberofdatabyexam iningonlyap artofthesedata.Thediffere nceist he “ population” fromwhichthesampleisselected,iethatsetofdatawhichisintende dtodrawac on clusio n.Audi

4、tsa mp li ngapp lies on lytocertai nty pesofaudit procedur es. Keywords: samp li ng,sa mpl erisk ,popu lati on,samp li ngun it,testsof con trols,substa ntive procedures. Statistical sam pling Committeestatisticalsa mp li ngofAmerica nin stituteofCertifiedPublicA ccountantsof(AICPA) issuedin1962aspec

5、iaireport,titled“ Statistigalsamplin gpoofe dindependentauditors whichallowedtheuseof statisticalsam pli ngmethod,i naccorda ncewithGe nerallyAcce ptedAuditi ngSta n dards(GAAS).Duri ng1962- 1974,theAICPApublishedaseriesofpapersonstatisticalsampling, “ Auditor Ap s proachtoStatisticalSampling” forus

6、eirgorotfeasonaleducationofacc oun ta nts.Duri ng1962-1974, theAICPApublishedaseriesofpapersonstatisticalsampling, “ Auditor Approac htoStatisticalSampling” ,foruseincontinuingprofessionaleducationofaccounta nts.In 1981,AICPA issuedtheprofessionalstandard,“ AuditSampling ” ,whichpro videsge neralgui

7、deli nesforbothsa mp li ngmethods,statisticala ndnon-statistical. Earlieraudits in cludedchecksofalltra nsacti onsin the periodcoveredbyth eauditedfi nan cialstateme nts.Atthattime,theliteraturehas no tgive np articularatte n ti on tothissubject. Onlyin1 971,a naudit procedures progra mprin tedi nth

8、e“ Feder alReserveBulletin(FederalBulletinStocks)” includedseveralreferencestosampl in gsuchasselect in gthe“ fewiteventory.of Thep rogramwasdevel op edbyas pecialcommittee,whichlaterbecametheAICPA, thatofCertifiedPublicAcc ountan tsAmerica nin stitute. In thefirstdecadesoflastce nturytheauditorsoft

9、e napp liedsa mp li ng,buts ampi esizewas notin relatedtotheefficie ncyofi ntern alc on trolofthee ntity. In 1955,America nIn stituteofAcco untan tshas publishedastudycaseofexte ndingt heauditsa mp li ngsummariz in gaudit programdevel op edby certified publicacco untan ts,toshowwhysa mp li ngis nece

10、ssarytoexte ndtheaudit. Thestudywasi mp orta ntbecauseis on eoftheleadi ngjournalo n sam plin gwhichrecog nizearelati on shi pofdependen cybetwee ndetaila ndreliabili tytesti ngofi nter nalc on trol. In 1964,theAICPA AuditingStandardsBoardhasissuedareportentitled “ Therelati on shi pbefree nstatisti

11、calsa mp li ngan dGe nerallyAcce ptedAudit ingS tandards(GAAS) ” whichillustratedtherelationshipbetweentheaccuracyandrelia bilityi nsam plingandp rovisio nsofGAAS. In 1978,theAICPA publishedtheworkofD on aldM. Roberts, “ StatisticalAuditirwhichexpiainstheunderlyingtheoryofstatistic al sam pling in a

12、uditing. In 1981,AICPA issuedthe pro fessi on alsta ndard ,n amed“ AuditSa mp li ng ” ,whichprovidesguidelinesforbothsamplingmethods,statisticalandnon statistical. An auditordoes no trelysolely on theresultsofas in gle proceduretoreachac on clusi ononan acco un tbala nce,classoftra nsact ionsorop er

13、ati on aleffective nesso fthec on trols.Rather,theauditfi ndin gsarebased on comb in edevide ncefromsevera lsources,asac on seque nceofa nu mberofdiffere ntaudit procedures.Whe nan audit orselectsasa mpl eofa popu lati on ,hisobjectiveistoobta inarep resentativesam ple ,iesamplewhosecharacteristicsa

14、reidenticalwiththepopulation characteristics. Thismea nsthatselecteditemsareide nticalwiththoseremai nin goutsidethesa mp le. Inp ractice,auditorsd onotkno wforsureifasa mpl eisre prese ntative,eve na fterco mpl eti on thetest,butthey“ mayin creasethe probabilitythatasa mpl eisre pres en tativebyacc

15、uracyofactivitiesmaderelatedtodesig n,sampl eselect ionan devalu ati on ” 1.Laclscpfecificityofthesa mpl eresultsmaybegive nbyobservati on error san dsa mp li ngerrors.Riskst op roducetheseerrorsca nbeco ntrolled. Observationerror (riskofobservati on)app earswhe ntheaudittestdid not iden tifyexisti

16、ngdeviati onsin thesa mpl eorus inganin adequateaudittech niq ueorb yn eglige nceoftheauditor. Samp lingerror (sa mp li ngrisk)isa nin here ntcharacteristicofthesurvey, whichresultsfromthefactthattheytested onl yafract iono fthetota Ipopu lati on. Sam p li ngerroroccursduetothefactthatitis po ssible

17、for theauditortoreachac on clusio n,based on asam plethatisdiffere ntfromthec on clusi on whichwouldbereachedifthee ntire popu lati on wouldhavebee nsubjecttoaudit p rocedureside ntical.Sa mp li ngriskca nbereducedby adjusti ngthesa mpl esize,de pendingon thesizea ndpopu lati on characteristicsa ndu

18、 singanapprop riatemethodofselectio n.l ncreasi ngsa mp lesizewillreducetherisk ofsa mp li ng; asa mple ofthe all popu lati on will prese nt an ullriskofsa mp li ng. AuditSam plin gisamethodoftesti ngforgathersufficie nta ndapprop riatea uditevide nceforthe purpo sesofaudit.Theauditormaydecidetoa pp

19、i yauditsa mpli ngonan acco un tbala nceorclassoftra nsacti on s.Sam pli ngaudit in cludesaudit proc edurestolesstha n100%oftheitemswithi nan acco un tbala nceorclassoftra nsactio n s,soallthesa mpl eabletobeselected.Auditorisrequiredtodeterm ineapprop riatew aysofselect in gitemsfortesti ng.Auditsa

20、 mp li ngca nbeusedasastatisticala pp roach andanon- statistical. Statisticalsam plingisamethodbywhichthesa mp leismadeso thateach un itc on sistsofthetota Ipopu lati on hasa nequa Ip robabilityof bei ngin cludedi nthesa mp le,methodofsa mp leselectio nisra ndom,allowedtoasse sstheresultsbased onp r

21、obabilitytheorya ndriskqua ntificati onofsamp li ng.Choosi n gthea pprop riate popu lati onm akethatauditor ndin gsca nbeexte ndedtothee nti repopu lati on. Non- statisticalsam plingsamethodofsa mp li ng,whe ntheauditoruses professio nal judgme nt to select eleme nts of asampi e.Si nee the purpose o

22、fsa mp li ngistodrawc on clusi on saboutthee ntire popu lati on ,theauditorshouldsel ectare prese ntativesa mpl ebychoos in gsa mpleun itswhichhavecharacteristicst yp icalofthat popu lati on .Resultswill no textra po latethee ntire popu lati on asthesa mpl eselectedisre prese ntative. Audittestsca n

23、bea pp lied on thealleleme ntsofthe popu lati on, whereisasm all pop ulatio noronanunrep rese ntativesam pl e,wheretheauditork nows the particularitiesofthe popu lati on tobetesteda ndis abletoide ntifyasmall nu mberofitemsofi nteresttoaudit.lfthesa mp lehas notsimila rcharacteristicsfortheeleme nts

24、ofthee ntire popu lati on ,theerrorsfo undin theteste dsa mpl eca nno textra po late. Decisi ono fstatisticalor non statisticala pproachdependsontheauditor professionaljudgmentwhichseeking sufficie ntapprop riateauditsevide nceon whichtoco mpl eti on itsfi ndin gsaboutthe audit opinion. Asastatistic

25、alsam plingmethoctefertothera ndomselectio nthata nypos siblecomb in ati ono feleme ntsofthecom mun ityisequallylikelytoe nterthesa mple. Simpleran domsa mp li ngisusedwhe nstratificati on was no ttoaudit.Us in gra ndoms electio nin volvesusi ngrandomnu mbersge neratedby aco mpu ter.Afterselect in g

26、ara ndomstart ingpoin t,theauditorfo un dthefirstra ndo mnu mberthatfallswithi nthetestdocume ntnu mbers. Onl ywhe nthea pp roachhast hecharacteristicsofstatisticalsa mp li ng,statisticalassessme nts ofriskarevalidsa mp li ng. Inano thervaria ntofthesa mp li ngp robability, namelythesystematicselect

27、 ion( alsocalledra ndommecha ni cal)eleme ntsn aturallysucceedi no ffices paceorti me;theauditorhasa prelim in arylist in gofthe popu lati onan dmadethedecisi ononsa mpl esize.“ Theauditorcalculatedaco un ti ngste p,an dselectsthesa mpl eeleme ntm ethodbased on ste psize.Ste pcountin gisdeterm in ed

28、bydividi ngthevolumeoftheco mmu ni tytosa mp lethe nu mberofu nitsdesired.Adva ntagesofsystematicscree nin gareits usability .Inm ostcases,asystematicsa mpl eca nbeextractedquicklya ndmethodaut omaticallyarrangesnumbersinsuccessiveseries.” 2. Selectionb yp robability prop ortionaltosize isamethodwhi

29、chemphasizesthose population units recordbigher values. The samp leisco nstitutedsothatthe probabilityofselecti ngany give neleme ntofthe pop ulati oni sequaltotherecordedvalueoftheitem; Stratifiedselection- isamethodofe mp hasisofu ni tswithhighervaluesa ndisregisteredi nthestratificatio nofthepopu

30、 lati onin sub popu lati on s.Stratificati onp rovidesaco mpl ete pictureoft heauditor,whenpopulation(datatabletobe analyzed)isnot homogeneous.Inthis case,theauditorstratifies apopu lati on bydividi ngthem in todisti nctsub popu lati on s,whichhavecom monch aracteristics, predefined. “Theobjectiveos

31、tratificationistoreducethevariabilityofelementsineac hlayera ndthereforeallowareduct ioninsampi esizewithouta prop orti on ate in creas eintheriskofsampling.” 3IfpopulatidSinOiQtiisdoneproperly,theamount ofsa mpl esizetocomelayerswillbelesstha nthesa mpl esizethatwouldbeobta in eda tthesamelevelofri

32、sk give nsampi ewithasa mpl eextractedfromthee ntire popu lati on .Auditresultsa ppii edtoalayerca nbedesig nedonlyon itemsthatare partofthatlayer. lapp reciated asuseful someviews onno n-statistical sam pling methods,whichi mp liesthat guidedtheselectionofthesa mpl eselect in geachelem en taccordi

33、ngtocerta in criteriadeterm in edbytheauditor.Themethodissubjective; becausetheauditorselectsi nte ntio nallyitemsco ntainin gsetfeatureshim. Theselectionoftheseriesd on ebyselecti ngmulti pleeleme ntsseries(su ccessive).Us in gsa mp li ngtheseriesisrecomme nded onlyi fareas on able nu mberof setsus

34、ed.Usi ngjustafewseriesthereisariskthatthesa mp leis notre prese ntative.Th istyp eofsa mp li ngca nbeusedi nadditi on toothersa mpl es,wherethereisahigh prob abilityofoccurre nceoferrors. Atthearbitraryselecti on,n oitemsareselected preferablyfromtheauditor. RevistaRoma n? deStatistic? nr. 5/ 2010

35、thatregardlessofsizeorsourceorcharacteristics.Is no ttherecomme ndedmethod,b ecauseis no tobjective. Thatsamplingisbasedontheauditor professionaljudgment,whichmay decidewhichitemsca nbepartorno tsa mpl ed.Becauseis no tastatisticalmethod,itc anno tcalculatethesta ndard error.Althoughthesa mp lestruc

36、tureca nbeco nstructedtore producethe pop ulatio n, thereis no guara nteethatthesa mpl eisre prese ntative.lfomittedafeaturethatwould bereleva nti nap articularsituati on ,thesa mpl eis no tre prese ntative. Samp li ngapp lieswhe ntheauditor plan stomakec on clusi on sabout popula tion ,based on ase

37、lect ion. Theauditorc on siderstheaudit programa nddeterm in esau dit procedureswhichmaya pp lyra ndomresearch.Sa mp li ngisusedbyauditorsanin ternalcontrolsystemstesting,andsubstantivetestingofo perationshege ner alobjectivesoftestsofc on trolsystema ndop erati on ssubsta ntivetestsaretoverifyth ea

38、pp licati ono fpre- defi nedc ontrolpro cedures,a ndtodeterm in ewhether op erati onscontainm ateriale rrors. Controltestsare inten dedt op rovideevide nceofop erati on alefficie ncyan dcon trolsdesig norop erati onofacon trolsystemt op reve ntordetectmaterialmissta teme ntsinfinan cialstateme nts.C

39、 on troltestsare necessaryiftheauditor plan stoass essco ntrolriskforassertio nsofma nageme nt. Con trolsarege nerallyex pectedtobesimilarlya pp liedtoalltra nsact ion sco veredbytherecords,regardlessoftra nsact ion value.Therefore,iftheauditorusessa mp li ng,itis no tadvisabletoselect on lyhighvalu

40、etra nsactio ns.Sa mp lesmustbecho sen soastobere prese ntative popu lati onsample. An auditormustbeawarethata nen titymaycha ngeas pecialco ntrolduri ngt hecourseoftheaudit.lftheco ntrolisre placedbya no ther,whichisdesig nedtoachiev ethesames pecificobjective,theauditormustdecidewhethertodesig nas

41、am pl eofal ltra nsact ion smadeduri ngorjustasa mpl eoftra nsact ionscon trolledaga in.Appropr iatedecisi ondependson theoverallobjectiveoftheaudittest. Verificatio nofi nter nalc on trolsystemofa nen tityisi nte ndedto providegui danceon theide ntificati ono freleva ntcon trolsa nddesig nevaluati

42、on testsofc on trol s. Othertests: In test ingintern ale on trolsystema ndtest ingop erati on s,auditsa mpl eisuse dtoestimatethe prop ortio no feleme ntsofa pop ulatio neon tai ning a characteristicorattribute analysis.Thisproportioniscalledthe frequencyofocc urrenceor p ercentageofdeviationa ndise

43、qualtotheratioofeleme ntsco ntai ninga ttributes pecifica ndtota Inu mberof popu lati on eleme nts.Weight deviati onsin asa mpl earedeterm in edtocalculatea nestimateofthe prop orti onofthe tota Ipopu lati on deviati ons. Riskassociatedwithsam pling referstoasa mpl eselecti on whichca nno tbere pres

44、e ntativeofthe popu lati on tested.I no therwords,thesam pl eitselfmayc ontainm aterialerrorsordeviati on sfromtheli n e.However,issu in gac on clusi on based on asa mpl emaybediffere ntfromthec onclu Sion whichwouldbereachedifthee ntire popu lati on wouldbesubjecttoaudit. Types of risk associated w

45、ith sam pling: Con trolsaremoreeffectivetha ntheyactuallyareorthatthereare no tsig nifi can terrorswhe ntheyexist- whichmea nsaninapprop riateaudito pin io n.Con trolsarelesseffectivetha ntheyact uallyarethattherearesig nifica nterrorswhe nin facttheyare not- thiscallsforadditio nalactivitiestoestab

46、lishthati nitialc on clusio nswerei ncorrect. Attributestesting - theauditorshouldbedefi nin gthecharacteristicstotesta ndcon diti on sformisc ondu ct.Attributestesti ngwillmakewhe n requiredobjectivestatistica Ip roject ionson variouscharacteristicsofthe popu latio n. Theauditormaydecidetoselectite

47、msfroma popu lati on based on itsk no wledgeab outthee ntitya nditse nviro nmen tco ntrolbased on risk an alysisa ndthes pecificcharacteristicsofthe popu lati on tobetested. Popu lationisthemassofdataon which theauditorwishestoge neralizethefi ndin gsobta inedon asa mpl e.P opu lati on willb edefi n

48、edco mp lia nceauditobjectivesa ndwillbeco mpl etea ndcon siste nt,becauser esults of thesample can be designed only for the population from which thesa mpl ewasselected. Sampling unit - a unit ofsampling may be, for example, anin voice,a nen tryorali neitem.Eachsa mpleun itisa neleme ntofthe popu l

49、ati on.Th eauditorwilldefi nethesam plinguni tbased on itsco mp lia ncewiththeobjectivesofa udittests. Sam plesize todeterm in ethesa mpl esizeshouldbec on sideredwhethersa mp li ngriskisreducedt oan acce ptablemi nim umlevel.Sa mp lesizeisaffectedbytheriskassociatedwithsa mp li ngthattheauditoriswi

50、lli ngtoacce ptit.Theriskthattheauditoriswilli ngtoacce p tlower,thesa mpl ewillbehigher. Error - fordetailedtesti ng,theauditorshould projectm on etaryerrorsfo undin thesa mplep opu lati onan dshouldtake in toacco un tthe projectederror on thes pecificobjectiveof theaudita ndotherauditareas.Theaudi

51、torprojectsthetotalerroro nthe pop ulatio nto getabroad pers pective on thesizeoftheerrora ndco mparin gitwithtolerableerror. Fordetailedtest in g,tolerableerroristolerablea ndmisre prese ntati ons willbeavaluelesstha norequaltomaterialityusedbytheauditorforthe in dividualcla ssesoftra nsactio nsorb

52、ala ncesaudited.lfaclassoftra nsactio nsoracco un tbala nces hasbee ndividedi ntolayerserrorisdesig nedse paratelyforeachlayer.Desig nerrors an di neon siste nterrorsforeachstratumarethe ncombi nedwhe neon sideri ngthe pos sibleeffect on thetotalclassesoftra nsact ionsan dacco un tbala nces. Evaluat

53、ionofsam pleresults* theauditorshouldevaluatethesa mp leresultstodetermi newhetherassessi ngreleva n tcharacteristicsofthe popu lati oniscon firmedor needstoberevised. Whe ntestingcontrolsa nunexp ectedlyhighrateofsa mp leerrormaylead toanin crease in theriskassessme ntofsig nifica ntmisre prese nta

54、ti onuni essitobta in edadditi on alauditevide ncetos upp ortthe in itialassessme nt.Forc on troltests,a nerr orisadeviati onfro mthe performa nceofcontrolp rocedures prescribed.Theauditor shouldobta in evide nceaboutthe naturea ndexte ntofa ny sig nifica ntcha nges inin ter n alc on trolsystem,i nc

55、ludi ngthestaffestablishme nt. Ifsig nifica ntcha ngesoccur,theauditorshouldreviewthe un dersta nding ofin ternalco ntrole nviro nmentandcon sidertesti ngtheco ntrolscha nged.Alternati vely,theauditormaycon siderp erformi ng substa ntivea nalytical proceduresortestsofdetailscoveri ngtheaudit period.

56、 In somecases,theauditormight notn eedtowaitu ntilthee ndaudittoformac on clusi on abouttheeffective nessofo perati on alc on trol,tos upp ortthec on trolriska ssessme nt.l nthiscase,theauditormightdecidetomodifythe plann edsubsta ntivete stsaccord in gly. If testingdetailsa nunexp ectedlylargeam ou

57、n toferrori nasa mpl emayca usetheauditortobelievethataclassoftra nsact ion soracco un tbala ncesisgive nsig ni fica ntlywr ongin theabse nceofadditi on alauditevide ncetoshowthatthereare notm aterialmisre prese ntatio ns. Whe nthebestestimateoferrorisveryclosetothetolerableerror,theauditor recog ni

58、 zestheriskthata no thersa mpl ehavediffere ntbestestimatethatcouldexcee dthetolerableerror. Conclusions Followinganalysisof samplingmethodsconcludethatallmethodshavea dva ntagesa nddisadva ntages.Buttheauditorisi mp orta ntin choos in gthesa mp li ng methodisbased onpro fessi on aljudgme ntan dtake

59、 in toacco un tthecost/be nefitrati o.Thus,ifasa mp li ngmethod provestobecostlyauditorshouldseekthemostefficie n tmethodi nviewofthema inandsp ecificobjectivesoftheaudit. Theauditor shouldevaluatethesa mp leresultstodetermi newhetherthe pr elim in aryassessme ntofreleva ntcharacteristicsofthe popu

60、lati onm ustbec on firme dorrevised.lftheevaluati onsampi eresults in dicatethatthereleva ntcharacteristics ofthe popu lati onn eedsassessme ntreview,theauditormay:requirema nageme ntto in vestigateide ntifiederrorsa ndlikelihoodoffutureerrorsa ndmake necessaryadju stme ntstocha ngethe nature,timi n