The Quality of Audit Process An Empirical Study

TheQualityofAuditProcess:AnEmpiricalStudy

withAuditCommittees

a

RiadhManitaaandNajouaElommalb

Associateprofessor,RouenBusinessschool,BoulevardAndréSiegfried-BP215,

76825Mont-Saint-AignanCedex,France

rmn@rouenbs.fr

b

AssociateProfessor,EMLVléonarddeVinci

PoleuniversitaireLéonardDeVinci92918ParisLaDéfenseCedex,France

njioua@yahoo.fr

ABSTRACT

Therecentfinancialscandals(EnronintheUSA,BataminTunisia)showedtheinabilityofclassicapproachesandtheirmeasureindicatorstoestimatetherelevanceofauditworks.Weconsiderthattheauditqualitycontrolmustresideatthelevelofauditprocess.ThispositionisadoptedbythefinancialsecuritylawinTunisiawhichencouragestheevaluationoftheauditquality,bytheauditcommittee.

Theobjectiveofthispaperistheconstructionofmeasurementscalesoftheauditprocessqualityforauditcommitteesorforothergovernanceorganconcernedwiththeauditquality.TheconceptionandthevalidationofthismeasurementscalesweredoneontheTunisiangroundbyadoptingChurchillapproach.Articulatingqualitativephasesandquantitativephasesbasedontworesearchquestionnaires,thisapproachdevelopedascaleof27qualityindicatorsdistributedonthreestagesoftheprocess.JELClassification:Keywords:

M42

Auditquality;Auditprocess;Auditcommittee;Churchillapproach

88ManitaandElommal

I.INTRODUCTION

Auditconstitutesasolutioninformationasymmetryproblembetweentheleadersandtheshareholdersorbetweentheleadersandtheotherthirdcontractingparties.Asamechanismofgovernance,ithasformainroletoreduceagencycostsandtoreassuretheshareholdersandthirdpartycontractorsconcerningthereliabilityofthefinancialinformationcommunicated.Nevertheless,thequalityofauditisnotuniformandespeciallynotdirectlyapparent.Theauditprocessisverycomplexandhardlyobservablebythirdpartiesandtheauditreport(theresultofanaudit)issostandardisedinitscontentandformatthatitoffersaonlyfewpossibilitiesfordifferentiation(Wooten,2003).Generally,ahigherqualityauditisreliant,aspartofitsmandate,onanabilitytoreduceexistinganomaliesandfailures.However,themajorityofstudiesonauditqualitywerecontentedtoextrapolatethe“auditquality”bythe“auditorquality”(DeAngelo,1981a;CitronandTaffler,1992;Carcelloetal.,1992).Theauditqualityapproachisthereforedeterminedbytheauditor'sability,asmuchintellectuallyasthroughresources,todetectpotentialinadequaciesintheauditsystem(detectionquality)andtogiveanaccountofthe“discoveries”highlightedduringthatwork(discoveryquality)1.Becauseofthedifficultiesinobservingtheauditprocess,themajorityofstudiesconcentrateontheresearchofsubstitutesforauditqualityusinganindirectevaluationapproach.Thesesubstitutesbeingperceivedbythemarketasrelatedtotheintrinsiccharacteristicsofthesetwoconcepts(Venkataraman,Randal;2008;KrishnanandSchauer,2000;Kaplan,1995).

Thereareseverallimitationstothis“indirect”evaluationapproach.Manyresearchershaveemphasisedthepresenceoffailureswhicharebothconceptualsuchastheriskofadverseselection(WattsetZimmerman,1981;Behnetal,1997)andempiricalwithsimplisticmeasuringcriteriagivingrisetocontradictoryresults.Recentfinancialscandals(especiallytheEnronaffair)andthecollapseoftheaccountancyfirmArthurAndersen2haveconfirmedtheinadequacyofthisindirectevaluationapproachincomprehendingauditquality.Forexample,Krishnan(2003)confirmsthatBigFourProxiescannotexplaintheEnronAffair.HedemonstratethatauditpartnersattheAndersenHoustonofficetoleratedmoreaggressiveaccountingpracticesbytheirclientsandwerereluctanttoprevailonthemtoreportbadnewsinatimelyfashion.

InTunisia,weassistedwith(in2002)theequivalenttotheEnroncase.ItinvolvedthecollapseoftheBatamCompany3.ThiscompanycausedacrisisofconfidenceontheTunisianfinancialmarketbyaffectingthereliabilityoffinancialinformationandauditquality.Thesescandalshaveshatteredtheconceptandauditqualitymeasurementwhichuntilthenwasbasedonindicatorsexogenoustotheauditprocess.

Allofthesereportsreinforcetheneed,withpractitionersaswellasacademics,toredefinetheauditevaluationrules.FollowingtheexampleoftheSOXlawintheUSAanewfinancialsecuritylawhavebeenadoptedin2005inTunisia.Thislawrequiresalllistedcompaniestoappointanauditcommitteeandencouragesthiscommitteetomonitorthequalityoftheexternalaudit.

Aftersomeyearsofapplicationofthislaw,itisimportanttoidentifyandtounderstandtheindicatorsusedbytheauditcommitteestoestimateauditquality.Theobjectofthisresearchwastoconceiveandtopropose,fortheauditcommittees,ameasurementscaleoftheauditqualityprocess.

INTERNATIONALJOURNALOFBUSINESS,15(1),2010

BasedonaChurchill’smethodology,ourstudyallowedtheidentificationof27itemsgroupedinthreemainstages.The27identifieditemsconstitutethebaseofanevaluationoftheauditprocessqualityforauditcommittees.Furthertoseveralstatisticalanalysestheseitemsweresummarizedin7keyqualityindicatorsoftheauditprocess.Theselastonesarelinkedtothetechnicalauditprocess,butalsototheauditorqualityandtotheorganizationalcharacteristicsofauditfirms.Thisarticleincludesfourparts.InSectionII,wepresentaliteraturereviewonauditquality.InSectionIII,weexposetheempiricalstepofthesearch(III).WeanalyzetheresultsinSectionIV.Finally,wepresentthediscussingandtheconclusionofourresearch.

II.

AUDITQUALITYMEASUREMENTINDICATORS

AuditqualitywasdefinedbyDeAngelo(1981)as\"theassessmentbythemarketofthecombinedprobabilitythatanauditorwillsimultaneouslydiscoverananomalyorsignificantirregularityintheclientcompany’saccountingsystemandpublishthisanomalyorirregularity\".Manyresearchers(Knapp,1991;Flint,1988;Moizer,1997)selectedthisdoubleapproachtodefineauditqualitybydistinguishingtheauditor’stechnicalcompetence(detectionquality)andindependence(revelationquality).Severalresearchers(KrishnanandSchauer,2000;Kaplan,1995;andLennox,1999)usedthisdefinitionofauditqualitytoidentifysubstitutesofauditqualityperceivedbythemarketorlinkedtotheintrinsiccharacteristicsofthesetwoconcepts(auditorcompetenceandindependence).Severalindicatorsofauditqualityhavebeenidentifiedbyliteraturesuchasthesize(Becketal,1998),fees(Davidetal,2006;Reynoldsetal.200,Chee-YeowLandHun-TongT.,2008)andreputationoftheauditor(McNair,1991;CareyandSimnett,2006),industryspecialisation(Solomonandal.1999,Krishnan2003),Litigationexperience(Plamrose,1988).

However,themajorityoftheseworkswerededicatedtotheanalysisofthe\"qualityoftheauditor\"asanapproximationofauditqualityratherthantheanalysisoftheauditprocessitself.Veryfewstudiesarededicatedtoevaluatingqualityusingtheauditprocess.Initialstudiesconcernedwiththeprocesswerefocusedonabetterunderstandingoftheenvironmentofsubjectiveauditdecisions.Thesestudiesexaminedthedifferentstagesoftheauditprocessbyobservingthefactorsaffectingauditqualityandthedifferentmeasurementswhichcouldbeattributedtothem(GibbinsandWalf,1982,Sutton,1993,Manita2008).Nevertheless,iftheanalysisoftheauditprocesshasbeenusedtoevaluatetheauditprocessforauditfirms,ithasnotbeenusedtodevelopqualityindicatorstohelpauditcommitteememberdirectors.Thesedirectorsdonothavethesameperceptionofauditprocessqualityasauditors.III.

THECONSTRUCTIONOFTHEMULTI-ITEMSMEASURESCALE

ThestudywasconductedinTunisiawithauditcommitteesandadministratorsoflistedcompaniestotheTunisstockexchange.Theinterviewsandquestionnaireswereadministeredbetween2007and2008.Banks’sandinsurances’sauditcommitteeswerespreadgiventhespecificitiesofthesetwosectors.Theimplementationofanevaluationapproachofauditqualityrequirestheconceptionofanempiricalmethodologytoverifythereliabilityandthevalidityofthemeasurescales.ExcepttheworksofSutton(1993),

90ManitaandElommal

Chemangui(2009)andManita(2008),noresearchisinterestedtothedevelopmentofamulti-itemsmeasureabletoarrestthevariousfacetsofauditprocess.Thestudiesarelimitedtotheidentification,basedontheliteratureorexpert’sopinion,thelistofqualityattributesandtheirvalidationbyquestionnaires(BehnandAl,on1997;CarcelloandAl,on1992).Fortheconstructionofthemeasurementscaleoftheauditprocessquality,wehaveadopted,followingtheexampleofManita(2008)andChemangui(2009)aChurchill’sapproach(1979)byconsideringRoehrichadvice(1993).Ourmethodologicalstep,basedonChurchill’sapproach,canbesosummarizedover5stages.A.

TheQualitativeStudywiththeAuditCommitteeMembers

Theobjectiveofthisstagewastoidentifythequalityattributesoftheauditprocessandtopreparethequestionnairefortheadministrators.FromthelistofquotedcompaniestoTunisstockexchange,wespreadbanksandinsurancesandweidentifiedalistof22companieswhichhaveanauditcommitteeatleastsincethepublicationofthefinancialsecuritylaw.

Wecontactedbyphoneallthesecompaniestoestablishalistofauditcommitteemembers.Thenfromalistof30administratorsweobtained10meetings.Amongtheinterviewees,5arepresidentsofauditcommittees.

Aninterviewguidewaspreparedand10semi-directiveinterviewswererealizedtoestablishasampleofitemsmeasuringtheauditprocessquality.Theinterviewguideisdividedintothreeparties.Thepurposeofthefirstpartywastounderstandthevariousrolesplayedbyauditcommitteesintheimprovementoftheauditqualityandtheirrelationshipwiththeinternalandexternalaudit.Thesecondpartystudiedtheattributesoftheauditprocessquality.Firstly,wequestionedtheintervieweesabouttheobservablestagesoftheauditprocess,andthenweidentifiedwiththemthelistofqualitativeattributes.Inthethirdparty,wequestionedtheadministratorsabouttheevaluationmodalitiesoftheauditqualityandabouttheircommunicationmethodswiththeauditors.

Oncethedatawascollectedusingthesemi-directiveinterview,wecompletelytranscribedthem.Then,byinterviewee,wecompiledasummaryoftheinterviewswhichwecouldalsomatchtoareportofeachcasestudy.Thesesummariesweresubmittedtotheparticipantssothattheycouldconductacountervalidation.

Acontentanalysiswasthencarriedoutfromthetranscribeddata.Theanalysiswasconductedintwophases.Athematicanalysisinterviewbyinterviewwasconducted.Thenaverticalandhorizontalthematicanalysisofalltheinterviews(inter-interview)wasconducted.Thisconsistedofgivingeachinterviewamorecomprehensivethematicstructureinitsownright(i.e.verticalanalysis)andcomparingalltheinterviewsontheiroverallthematicstructures(i.e.horizontalanalysis).Thisleadsultimatelynottoconsiderthesingularconsistencyofeachinterviewbutrathertoseekanoverallconsistencyinthebodyofthedataproducedbyallthoseinvolved.

Theyresultinanidentificationofauditqualityindicatorsbystageoftheprocess.Asregardsthevariousprocessstagestobechecked,themembersofauditcommitteeswerealmostconvergentonadivisionoftheauditprocessinthreestages:interimstageofunderstandingofthesystemsandtheevaluationofthecompany’sprocedures,stageofaccountmonitoringandthestageofcertificationandauditreportdrafting.Itwould

INTERNATIONALJOURNALOFBUSINESS,15(1),201091

seemthatthisauditprocessdivisionisconnectedtotheobjectiveandtothepurposeofeverystageconsistinginproducinganevaluationprocessreport,asynthesisnoteofthemission(ortheauditadjustmentsproposition)andareportontheaccountscertification.

Furthermore,withregardsthedeterminingfactorsforauditquality,atthestartweidentified45items.Wethenrefinedthissampleofitemsjudgedintuitivelyredundant,toendupwithanintermediatesampleof34determiningfactorsforquality.Wefinallysubmittedtheseitemstoacontentvalidityprocedurewithfourothermembersofauditcommitteesandtworesearchersinauditinordertoaskthemtoverifythelackofredundancybetweentheitemsaswellastheireaseofcomprehension.Withregardtoredundancybetweentheitems,theauditorswerecalledupontoidentifythoseitemswhichreferredtothesamethemeandwhichriskedcreatingbiasincomprehension.Afterthisstage,wearrivedatafinalsampleof31items.B.

DraftingandPreliminaryTestingoftheQuestionnaire

Onthebasisof31itemsidentifiedinthepreviousstage,aquestionnairewaselaboratedforthememberadministratorsofauditcommittees.Thisquestionnairewasthentheobjectofacontentsvalidationandofapre-testwith3administratorsandtworesearchersinaudittoimprovethedraftingandconsequentlytheunderstandingforauditcommittees.C.

DataCollection

Thequestionnairewassentto68administratorscomposingthelistofauditcommittee‘smembersofcompanieslistedontheTunisianstockexchangeinordertofocustheirappreciationontheimportanceoftheitems.AsperSutton(1993),weselectedaLikert7pointscale.Thischoicefacilitatestheadministrationofthequestionnaireandoffersagoodsensibilityofanswer.Weadministeredthequestionnairebyposttoallthemembersofauditcommittees.Giventherelativelyreducednumberofourpopulation,severalre-contactsproceduresbyphoneandbymailingwerethenmadetoobtainthemaximumofanswers.Thefollowingtablepresentsthedetailoftheadministratorswhorespondedwithregardtothetotalof68auditcommittees’administrators.On39successfulcompletequestionnaires,10emanatedfrompresidentsofauditcommittees.

Asregardsoursecondsample,itwasestablishedwiththeadministratorsofthehighly-ratedcompanies.Fromthelistofthehighly-ratedcompaniesretained,weadministeredoursecondquestionnairetoallthepopulationoftheadministratorsoftheselectedcompanies(2individuals).After-contactingprocedures,weobtained98exploitableanswers.Theanswerrateisabout37%.Ourchoiceisjustifiedbythereducedpopulationofauditcommittee’sadministratorsandbythedifficultytoobtainasecondsignificantsampleofthispopulation.Theuseofalltheadministrators’populationisalsojustifiedbythegoodexperiencethathavewiththeauditors.Fromtheirmeetingwithauditcommittees,theadministratorscanhaveagoodknowledgeofauditprocess.

92ManitaandElommal

D.RefinementoftheMeasurementInstrument

Theobjectiveofthisstageistorefinethemeasurementscalefortheauditprocessqualityandtesttheconsistencyofthescalesmakingupourquestionnaire.Therefinementandthevalidationofthefactorialstructureofscales(exploratoryphaseandvalidationphaseofChurchillapproach)arerealizedfromtwoquantitativestudiesrealizedrespectivelywith39membersofauditcommitteesand98administratorsofthelistedcompanies.Thefirststudyaimsatidentifyingthefactorialstructureofthemeasurescalesoftheauditprocessquality.Thesecondhasforobjectivetotestthestabilityofthisstructureonanadministrators'secondsample.Wesosubjectedthemeasurescalestothereliabilitytest(AlphadeCronbach)andintheexploratoryfactorialanalysis.Accordingtotheresultsobtained,itemsmaybeeliminated(Cattel,1988).E.

ReliabilityandValidityoftheMeasurementInstrument

Toobtainreliableandvalidmeasurementscalesforthequalityoftheauditprocess,weapplied,followingtheexampleofRoussel(1996),astatisticalmethodwhichisconsideredstrongerthanthoseproposedbyChurchill(1979).Soinsteadofapplyingthemultitrait-multi-methodmatrix(MTMM),thedatastemmingfromoursecondquestionnaireweretheobjectoftwotypesofanalysis:aconfirmatoryfactoranalysisandtheRhodeJoreskogcalculation(Roussel,2002).Thefirstaimstoverifythevalidityofthefactorialstructuresproposedintheexploratoryphaseand(usingthestructuralequationsmethod)testthequalityadjustmentofthemeasurementinstrumentwiththeempiricaldata.Thesecondconsistsofverifyingthereliabilityofourmeasurementscalesandstatisticalvalidity.

IV.

A.

THERESULTS

ReliabilityandValidityoftheMeasurementInstrument

Theresultsoftheexploratoryphaseshowthattheauditprocessqualityismultidimensional.Theydisplaysatisfyingoverallconsistencylevelsforallthescales.However,weaknesseswererevealedrelatingtocertainitemswhichwereeliminated.Onthebasisofinternalconsistencyandtheconsistencyofthedifferentscales,thesurveyquestionnairewasreducedfrom31to27items.Differentmeasurementscaleswerethendevelopedforeachstageoftheauditprocess.Generally,thedifferentmeasurementscalespresentverysatisfactoryreliabilityandvalidity.Thedifferentitemsmakingupthescalesdevelopedarewellrepresentedsincetheypresentcommonalitiesgreaterthan0.5.Theyalsopresenthigh“loading”withtheirrelatedfactorialfocuses.Thespecificvaluesassociatedwithallthedimensionsidentifiedexplainmorethan50%ofthetotalvariance.Furthermore,allthedimensionspresentsatisfactoryreliabilitywithCronbachAlphasgreaterthan0.5.Thefollowingtablessummarisetheresultsoftheexploratoryanalysisatthelevelofthethreestagesoftheauditprocessidentified.

INTERNATIONALJOURNALOFBUSINESS,15(1),201093

Table1

Analysisofthescalereliabilityconcerningthedeterminingfactorsinthequalityofthe

interimphaseofauditmission

COMPONENTS

ITEMS

COMMONALITY

Knowledgeofthecompanyanditsenvironment

0,7980,701

Relevanceofthe

risksandsensitivezonesidentified

Organizationandsupervisionofthemission

Composition

and

qualificationofthecontributing

team

Goodunderstandingofthecompanyanditsinformationsystem

Goodknowledgeofthespecificitiesofthecompany,itssectorandtheactivityrisks

Goodunderstandingofthestakesintheaccountingpoliticsofthecompanyandthemanagerrisks

Goodknowledgeoftherisksandsensitivezonesofthecompany.

Coverageofthesystemandsignificantandsensitiveprocessesbytheauditor

Gooduseofinternreports(auditinterns,internalreporting,reportsrealizedbyexpertsforthebenefitofthecompany)inrisksappreciationandtheidentificationofthesensitivezonesRelevanceoftheweaknessesandtenderspotsidentifiedandofthereportofinterimwork

Qualityoforganizationofthemissionandthelevelofsupervisionoftheworksbythemanagersorthepartnerinchargeofthemission

Respectoftheloadplanofthemission

QualityofthemeetingsofinformationanddiscussionwiththeauditcommitteeandthevariousmanagersofthecompanyQualityoffollow-upoftheweaknessesandtheidentifiedrisksyearspast

Levelofexpertise(assessment)oftheteamofauditPhaseconductedbyhighlyqualifiedandexperiencedcontributors

Specific values (λ)ExplainedvarianceCronbachα

0,7170,520

0,5380,6190,5590,6590,7130,6110,7510,5990,7270,5370,555

0,1

0,5820,5420,7480,7940,7380,6160,6100,809

0,6780,7211,7511,65%0,552,7018,03%0,752,4116,08%0,682,4016,02%0,73Table2

Analysisofthescalereliabilityconcerningthedeterminingfactorsofthefinalphaseof

accountsaudit

ITEMS

COMMONALITY

Organizationandsupervisionofthemissionandauditorethics0,7840,7310,7170,7220,9020,7110,8740,7110,876

0,7460,733

2,5925,86%0,81

1,8018,04%0,75

1,7117,14%0,65

1,4414,36%0,56

COMPONENTS

Relevanceand

coherenceoftheCompositionandprogramswithqualificationoftheidentifiedtheauditteam

risks

Communicationlevelofandcollaborationwiththeauditcommittee

WorksOrganizationandlevelofmissionsupervision

Presencetimeandimplicationdegreesofthepartnerresponsibleforthemission

RespectoftheloadplanofthemissionRotationrateofthemanagersandpeopleinchargeofthemission

Coherenceandcredibilityoftheauditprogramswiththerisksandtheweaknessesidentifiedininterimphase

Qualityoffollow-upandassistancetothephysicalinventory

RelevanceofthemissionsynthesisExpertiseLevelofauditteamStabilityoftheauditorsteambetweentheinterimphaseandthefinalphaseofthemission

IndependenceleveloftheteamandthepartnerresponsibleofthemissiontowardsthedirectionCommunicationqualityandthecollaborationleveloftheauditteamwiththeinternsauditorsandtheauditcommitteeSpecific values (λ)ExplainedvarianceCronbachα

0,7840,10,7040,6850,8270,8040,8430,7220,8200,7150,579

94ManitaandElommal

Table3

Analysisofthescalereliabilityconcerningthedeterminingfactorsofthephaseof

accountscertification

COMPONENTIndependenceandauditorethics

0,80,8480,8322,5162,98%0,63

ITEMS

Qualityofauditreports

IndependenceleveloftheauditorintheopinionformulationontheaccountsRespectfortheethicrulesValeurs propres (λ)VarianceexpliquéeαdeCronbach

COMMUNALITES

0,6790,70,795

Table4

Testindicesforthequalitydimensionsofthequalityofauditprocess

INDICES

Absolutemeasurementindicators

χ2GFIAGFIRMRRMSEANFITLICFIχ2norméPNFIRhodejoreskog

Stage136.8580.9290.8520.2910.1380.7440.7880.7923.6850.5960.857Stage2113.8280.9030.8330.0980.0780.8681.8660.5590.98Stage355.5650.9220.8410.1140.0860.9000.9070.9442.0580.5400.907IndicativethresholdsChi-2thesmallestassociatedwithainsignificantprobability

>0.9

>0.8

Aslowaspossible

<0.08>0.9

>0.9>0.9

Between1and5Ashighaspossible(>0.6)

>0.7or>0.8IncrementalmeasurementindicatorsParsimonymeasurementindicatorsReliabilityindices

Theresultsoftheconfirmatoryanalysisappliedtothestructureofeachauditprocessstage,confirmsthevalidityofthefactorialstructurefromtheexploratoryphase.Theyconfirmagoodlevelofreliabilityandvalidityofthedifferentmeasurementscalesfortheauditprocessquality.Theitemswhichmakeupthedifferentmeasurementscalesaresignificant(theCRareallhigherthan1.96).Onthewholethevaluesoftheabsolute,incrementalandparsimonyindicesaresatisfactoryand/oracceptable.TheChi-2textisonthewholesignificant.TheabsoluteindicesGFIandAGFIarealsosignificantbecausetheyexceedtheirrespectivethresholds.Furthermore,theincrementalindices,althoughsometimesslightlylowerthantheir

INTERNATIONALJOURNALOFBUSINESS,15(1),201095

respectivethresholds,canbeconsideredasacceptable,consideringtherelativelylownumberofitemsincertainstagesoftheprocess.Withregardstheparsimonyindices,we note that χ2indexisgenerallysatisfactory.Ontheotherhand,thedifferentscalesdevelopedpresentverysatisfactoryreliabilitiesconsideringthattheRhodeJoreskogvariesbetween0.73and0.92.

V.

DISCUSSIONANDCONCLUSION

Fromarigorousprocessforthedevelopmentofmeasurementscalesoftheauditprocessquality,werealizedqualitativeandquantitativestudiesviaseveralsamplesoftheauditcommitteesmembersandadministrators.Theresultsallowedustoidentify27indicatorsoftheauditprocessqualitygroupedinto3mainstages.

Followingseveralfactorialanalyses(exploratoryandconfirmatory),theseindicatorsweresummarisedinto7qualitydimensions.Amongtheidentifieddeterminingfactors,only4concernedthetechnicalauditprocess:Theunderstandingofthecompanyanditsenvironment;thePertinencerisksandsensitivezonesidentified;TheRelevanceandcoherenceoftheprogramswiththeidentifiedrisksandthecommunicationlevelandcollaborationwiththeauditcommittee.Threeothersdeterminersrelateto:theIndependenceandethicsofthelistener;thecompositionandthequalificationoftheauditteam;theOrganizationandthesupervisionofthemissionandauditorethics.

Thesedeterminingfactorsconstitutethebasisforanevaluationoftheauditprocessqualitybyauditcommittee.Theyshowthattheauditprocessisacomplexprocesswhichneedstobeunderstoodthroughmanydimensions.Theyalsoshowthattheauditprocessqualityisnotonlydependantonthetechnicaldeterminingfactorsrelatedtotheauditorprocess,butalsoontheauditorquality(auditorindependenceandethic),theorganisationalcharacteristicsoftheauditfirm(auditteamandmissionorganisation).TheseresultsconfirmedpartiallythosefoundbyManita(2008).Ourresultsidentified3stagesofauditprocesscontrarytothoseofManita(2008)whoidentified6stages.Besidesthestagesidentifiedbyourresearch,Manita(2008)identifiedthestageofacceptanceofthemission,thegeneralapprehension,andthephysicalinventory.ItwouldseemthattheresultsobtainedbyManita(2008)werebiasedbytheperceptionoftheauditorswhoareinfluencedbytheprocessdecompositionbyauditstandards.Asamatteroffact,inhisstudy,thestagesandthevariousitemswereidentifiedonthebasisofaqualitativestudywithauditors'sampleandnotwithadministrators.Asregardsitemsmeasuringthequalityoftheauditprocess,ourempiricalresultsconfirmlargelyindicatorsidentifiedbyManita(2008)andSutton(1993)4eveniftheevaluationrailingdevelopedbythislastonewasratherintendedfortheauditorsandnotfortheauditcommittees.

ThemaindimensionnotidentifiedbyManita(2008)inthesestagesconcernsthelevelofcommunicationandcollaborationoftheauditorwiththeauditcommittee.Heshowedthattheadministratorsareverysensitivetothewaythattheauditorcommunicateswiththeauditcommittee,ontheidentifiedrisks,ontheauditprogramsdeveloped,ontheresultsobtainedinallthemissionprogressstages.Itseemsthattheauditcommittee’smembersjudgethattheconsiderationbytheauditor,duringhisappreciationoftheauditrisks,oftheunderstandingandknowledgeoftheaudit

96ManitaandElommal

committeeofthevarioussystemsandthecompanyprocessandoftheirworkscontrolresults,isasignofefficiencyandcompetenceoftheauditor.

Thequalityevaluationoftheauditprocessisadifficultandcomplextaskfortheauditcommittee’smembers.Theevaluationscheduledevelopedbyourresearchshouldconsiderablyfacilitatetheirtaskandstrengthentheirinvolvementintheauditprocess.Howeverthismeasurementscalefixesonlytheglobalobjectivesofcontrolbyprocessstage.Itshouldconsequentlybeconsideredasaquestionnaireofqualitycontroltoberefinedandtobedetailedaccordingtothenatureofcompanyactivity,itsinformationsystemanditsspecificities.Thisschedule,doesexemptadministratorsfromhavingtherequiredskillstocomprehendtheauditprocess.Theadministrator’scompetenceevokestheproblemoftheselectionandappointmentprocessforadministrator.TheauditcommitteesforlistedAmericancompaniesgenerallyincludeadministratorselected,notfortheirnetworkofcontacts,butfortheircompetencesinaccountingorfinancialfields.Also,thisschedulemustbecomplementedbytheintegrationofindependentdirectorswithinauditcommitteesoranyboardofdirectorsconcernedwithauditquality.Contrarytowhatisunderstoodbycompanyadministrators,therearesuchindependentprofessionalsinthemarket.Thesemayincludeformerhighlevelauditorsretrainedinanadvisoryrole,businessbankerswhohavebeenconfrontedwithriskevaluationmethodsoracademicsspecialisinginthefield.Theindependenceofthesedirectorscanonlybepreservedbyencouragingcompaniestoequipthemselveswithgovernancechartersorbyfocusingonregulation(intheUSAtheSarbanesOxleylawimposesthisindependence).

Independentlyofobtainedresults,ourstudyhasshownthepossibilitytobuildametricforqualitybasedontheauditprocess.Ourapproachdiffersfromthestandardevaluationapproachesbecauseitdoesnotlookforsubstitutesforauditqualitythroughthequalityoftheauditor,butinsteadattemptstoassessthisqualityfocuseddirectlyonthetechnicalauditorprocess.Obviouslythisworkhasitslimits.Firstly,wewouldpointoutthatforthegenerationofitemsrelativetoourmeasurementscales;weconductedaqualitativestudywithagroupconsistingof10memberadministratorsofauditcommittees.Inadditiontothelimitconnectedtothesizeofoursample,thismethodisbasedontheperceptionofthepracticalusefulnessofourtoolbyauditcommitteesandnotonaneffectivemeasurementofitspracticalrelevance.Wewouldalsopointoutthatgiventheconstraintsconnectedtothesizeoftheauditcommitteemembers’population,wehaduseduringthevalidationofourscalestoadministrators'sample.Evenifthepopulationcharacteristicsareclosetothoseoftheauditcommitteemembers,differencesconnectedtotheexperience,totheknowledgeoftherisksandthecontrolsystemofthecompanycanbiasourempiricalresults.Wewouldfinallypointoutthatinspiteofthemethodologicalstrictnessofourapproachtheresultsofthisstudycannotexceedtheexploratorythresholdandtakeadvantageofanomologicvalidity.Thislimitationisinherenttolittlecomparableresearchinthisdomain,whichdoesnotallowanappropriateconfrontationofourscaleswiththeotherstudiestoidentifythedifferencesandexplainthem.ThislimitdidnotalsoallowustoimplementthelaststageofChruchillapproach(stage8-Developthestandards).Weareindeedsituatedinaresearchdomainwhichhasbeenlimitedlyinvestigatedbytheresearchers,andbesidesthestudyofManita(2008)andSutton(1993),wedidnotfind,toourknowledge,otherworkwithwhichwecouldcompareourempiricalresults.Thisisessentiallythecomparisonoftheconnectionsbetweenresearchvariablesandresultsofearlierstudies

INTERNATIONALJOURNALOFBUSINESS,15(1),201097

whichmayestablishthepredictiveornomologicvalidityofthequestionnaire(Peter,1981).Furthermore,sincethepredictivevalidityandthecontentvalidityhavestrongconnections,wearesatisfiedwiththecontentvalidityanalysis(Evrardetal,1993;IgalensandRoussel,1998).Facedwiththissituation,anewverificationofthereliabilityand,aboveall,thevalidityofthescalesarenecessarybeforeintegratingourmeasurementscalesintotheexplanatorymodelsdeveloped.

ENDNOTES

1.2.3.

Thequalityofanauditapproachisdeterminedbytheauditor'sabilitytodetectpotentialerrorsandanomaliesofthesystem(detectionquality)andtogiveanaccountofthe“discoveries\"highlightedduringhiswork(discoveryquality).

ThesizeoftheauditfirmwasconsideredbothasanindicatorofcompetenceandindependenceandthoughArthurAndersenwentbankruptfollowingtheEnronaffair.

Batamisamassdistributioncompanywellknownlocally,especiallyforelectricalgoodsandquotedontheTunisianstockexchange.Thiscompanywentbankruptwithoutanywarningfromtheauditor.Followingacourtactionagainsttheauditor,hewassentencedtotwoyearsimprisonment.

Withtheexceptionofthosefocusedonanevaluationbytheauditors,notabletobeobservedbythedirectors.Tobespecific,thisisrelatedtotheauditenvironmentand/ortheclientcompanysuchasthecompetenceoftheclient,changeswiththeclient,clientrelationships,thecompletionoftheworkpaper,andthepreparationoftheclient.

REFERENCES

Beck,P.J.,T.J.Freca,andI.Solomon,1998,“AnEmpiricalAnalysisoftheRelation

betweenMASInvolvementandAuditorTenure:ImplicationforAuditorIndependance”,JournalofAccountingliterature,vol7,65-84.

Behn,B.K.,J.V.Carcello,D.R.Hermanson,andR.H.Hermanson,1997,“The

DeterminantsofAuditClientsSatisfactionamongClientofBigSixFirms”,AccountingHorizons,vol11,1,7-24.

Carcello,J.V.,R.H.Hermanson,andN.T.McGrath,1992,“AuditQualityAttributes:

thePerceptionsofAuditPartners,Preparers,andFinancialStatementUsers”,Auditing:AJournalofPracticeandTheory,vol11,1,1-15.

Carey,P.,andR.Simnett,2006,“AuditPartnerTenureandAuditQuality”,The

AccountingReview,vol27,2,125-142.

Cattel,R.,1988,“TheMeaningandStrategicUseofFactorAnalysis,”dansHandbook

ofMultivariateExperimentalPsychology,NesselroadeJetCattelReds.,NewYork:PlenumPress,Chapitre4,131-203.

Citron,D.B.,andR.J.Taffler,1992,“TheAuditReportunderGoingConcern

Uncertainties:anEmpiricalAnalysis”,AccountingandBusinessResearch,vol22,88,337-345.

4.

98ManitaandElommal

Chee-Yeow,L.,andT.Hun-Tong,2008,“NonAuditServiceFeesandAuditQuality:

theImpactofAuditorSpecialisation”.,JournalofAccountingResearch;vol46,1,199-246.

Chemangui,M.,2009,“Propositiond’uneMétriquedelaQualitédel’audit:

ExperimentationdansleCadredesRelationd’agenceInternes”,Comptabilité,Contrôle,Audit,vol1,15,juin.

Churchill,G.A.,1979,“AParadigmforDevelopingBetterMeasuresofMarketing

Constructs”,JournalofMarketingResearch,vol16,-73.David,C.H.,W.R.Knechel,andW.Norman,2006,“AuditFees:AMeta-analysisof

theEffectofSupplyandDemandAttributes”,ContemporaryAccountingResearch,vol.23,1,141–191.

DeAngelo,L.E.,1981,“AuditorSizeandAuditQuality”,JournalofAccountingand

Economics,3,183-199.

Evrard,Y.,B.Pras,andE.Roux,(1993,2000),Market:étudesetrechercheen

marketing,Nathan,2èmeet3èmeédition,Paris.

Flint,D.,1988,PhilosophyandPrinciplesofAuditing,MacmillanEducation,London.Gibbins,M.,andF.Wolf,1982,“Auditors’SubjectiveDecisionEnvironment:theCase

ofaNormalExternalAudit”,TheAccountingReview,vol57,1,105-124.

Igalens,J.,andP.Roussel,1998,Méthodesderechercheengestiondesressources

humaines,Economica,1èreédition,Paris.

Kaplan,S.E.,1995,“AnExaminationofAuditors’ReportingIntentionsupon

DiscoveryofProceduresPrematurelySigned-off”,Auditing:AJournalofPracticeandTheory,vol14,2,90-104.

Knapp,M.C.,1991,“FactorsthatAuditCommitteeMembersUseasSurrogatesfor

AuditQuality”,Auditing:AJournalofPracticeandTheory,vol10,1,615-637.Krishhnan,J.,andP.Schauer,2000,“TheDifferenciationofQualityamongAuditors:

EvidencefromtheNot-profitSector”,Auditing:AJournalofPracticeandTheory,vol19,9-25.

Krishhnan,G.V.,2003,“DoesBig6AudtorIndustryExpertiseConstrainEarnings

management?”.,AccountingHorizons,supplement,vol17,1-16.

Leenox,C.,1999,“AuditQualityandAuditorsSize:anEvaluationofReputationand

DeepPocketsHypotheses”,JournalofBusinessFinanceandAccounting,vol26,7-8,779-805.

Manita,R.,2008“Laqualitédel’auditexterne:propositiond’unegrilled’évaluation

axéesurleprocessusd’audit”,RevueManagementvol11,2,191-210.

McNair,C.J.,1991,“ProperCompromises:theManagementControlDilemmain

PublicAccountingandItsImpactonAuditorBehaviour”,Accounting,OrganizationsandSociety,vol16,7,635-653.

Moizer,P.,1997,“AuditorReputation:theInternationalEmpiricalEvidence”,

InternationalJournalofAuditing,vol1,1,61-74.

Palmrose,Z.,1988,“AnAnalysisofAuditorLitigationandAuditServiceQuality”,

TheAccountingReview,vol63,1,55-73.

Peter,T.,1981,“ConstructValidity:aReviewofBasicIssuesandMarketing

Practices”,JournalofMarketingResearch,vol18,133-145.

INTERNATIONALJOURNALOFBUSINESS,15(1),201099

Reynolds,J.K.,D.R.Dies,andJ.R.Francis,2004.“ProfessionalserviceFeesand

AuditorObjectivity”.Auditing:Ajournalofpracticeandtheory,vol23,1,March,71-93.

Roehrich,G.,1993,Lesconsommateurs-innovateurs:unessaid’identification,Thèse

dedoctoratensciencesdegestion,Grenoble:Ecolesupérieuredesaffaires,UniversitéPierreMendès-France.

Roussel,P.,1996,Rémunération,motivation,etsatisfactionautravail,EdEconomica,

Paris.

Roussel,P.,2002,Méthodedeséquationsstructurelles:recherchesetapplicationsen

gestion,EditionEconomica,Paris.

Solomon,I.,M.D.Shields,andO.R.Whittington,1999,“WhatDoIndustry-specialist

AuditorsKnow?”JournalofAccountingResearch,vol37,1,191-208.

Sutton,S.G.,1993,“TowardanUnderstandingoftheFactorsAffectingtheQualityof

theAuditProcess”Auditing:AJournalofPracticeandTheory,vol24,88-105.Venkataraman,R.,J.Weber,andM,Willenborg,2008,“LitigationRisk,AuditQuality,

andAuditFees:EvidencefromPublicOffrerings”,AccountingReview,vol83,5,1315-1345.

Watts,R.L.,andJ.L.Zimmerman,1981,“TheMarketforIndependanceand

IndependentAuditors”,Unpublishedmanuscript,UniversityofRochester,Rocherster,NY.

Wooten,T.C.,2003,“ResearchaboutAuditQuality”,CPAJournal,vol73,1,48-.

Copyright of International Journal of Business is the property of Premier Publishing and its content may not becopied or emailed to multiple sites or posted to a listserv without the copyright holder's express writtenpermission. However, users may print, download, or email articles for individual use.