D&M信息系统成功模型

TheDeloneandMcLeanModelofInformationSystemsSucess:ATen-YearUpdateAuthor-WilliamH.DeloneandEphraimR.McleanWilliamDeloneABOUTD＆MWilliamDeloneisanAssociateProfessorofInformationSystemsandChairoftheInformationTechnologyDepartmentattheKogodSchoolofBusinessatAmericanUniversityinWashington,DC.ProfessorDeLone’sprimaryareasofresearchincludetheassessmentofinformationsystemseffectivenessandvalue,theimplementationanduseofinformationtechnologyinsmallandmedium-sizedbusinesses,andtheglobalmanagementofinformationtechnology.EphraimR.McleanEPHRAIMR.MCLEANisaRegents’ProfessorandGeorgeE.SmithEminentScholar’sChairinInformationSystemsintheRobinsonCollegeofBusinessatGeorgiaStateUniversity,Atlanta.PriortocomingtoGeorgiaStateUniversityin1987,hewasonthefacultyofUniversityofCalifornia,LosAngeles(UCLA)for18years.Dr.McLean’sresearchfocusesonthemanagementofinformationservices,thevalueofISinvestments,andcareerissuesofISprofessionals.AbstractTenyearsago,wepresentedDeLoneandMcLeanInformationSystems(IS)SuccessModelasaframeworkandmodelformeasuringthecomplexdependentvariableinISresearch.Inthispaper,wediscussmanyoftheimportantISeffortsthatapply,validate,challenge,andproposeenhancementstoouroriginalmodel.Basedonourevaluationofthosecontributions,weproposeminorrefinementstothemodelandproposeanupdatedDeLoneandMcLeanISSuccessModel.Wediscusstheutilityoftheupdatedmodelformeasuringe-commercesystemsuccess.Finally,wemakeaseriesofrecommendationsregardingcurrentandfuturemeasurementofISsuccess.CONTENTSD＆MISSucessModelModelanalysisandModelissuesModelmodificationandextensionsTheupdatedD＆MISSucessModelE-CommerceSucessSummaryandConclusionModeltheoreticalbasisD＆MISSucessModel1ModelIssues3Modelanalysis2Modelmodification4ModeltheoreticalbasisTheShannonandWeavermodel1TheinformationinfluencetheoryofMason2MISresearchstudy3ModeltheoreticalbasisShannonandWeaverdefinedthetechnicallevelofcommunicationsastheaccuracyandefficiencyofthecommunicationsystemthatproducesinformation.Thesemanticlevelisthesuccessoftheinformationinconveyingtheintendedmeaning.Theeffectivenesslevelistheeffectoftheinformationonthereceiver.IntheD&MISSuccessModel,”systemquality”measurestechnicalsuccess;“informationquality”measuressemanticsuccess;and“use,usersatisfaction,individualimpacts,”and“organizationalimpacts”measureeffectivenesssuccess.D＆MISSucessModelTheprimaryconclusionsoftheoriginalpaper1.ThemultidimensionalandinterdependentnatureofISsuccessrequirescarefulattentiontothedefinitionandmeasurementofeachaspectofthisdependentvariable.Itisimportanttomeasurethepossibleinteractionsamongthesuccessdimensionsinordertoisolatetheeffectofvariousindependentvariableswithoneormoreofthesedependentsuccessdimensions.2.Selectionofsuccessdimensionsandmeasuresshouldbecontingentontheobjectivesandcontextoftheempiricalinvestigation;butwherepossible,testedandprovenmeasuresshouldbeused.3.DespitethemultidimensionalandcontingentnatureofISsuccess,anattemptshouldbemadetoreducesignificantlythenumberofdifferentmeasuresusedtomeasureISsuccesssothatresearchresultscanbecomparedandfindingsvalidated.4.Morefieldstudyresearchshouldinvestigateandincorporateorganizationalimpactmeasures.5.Finally,”hissuccessmodelclearlyneedsfurtherdevelopmentandvalidationbeforeitcouldserveasabasisfortheselectionofappropriateISmeasures”.Systemuse-individualimpactsSystemusewastypicallyvoluntaryandwasmeasuredasfrequencyofuse,timeofuse,numberofaccesses,usagepattern,anddependency.Individualimpactsweremeasuredintermsofjobperformanceanddecision-makingperformance.Modelanalysis2SystemQuality-IndividualImpactssystemqualitywasmeasuredintermsofease-of-use,functionality,reliability,flexibility,dataquality,portability,integration,andimportance.Individualimpactsweremeasuredasqualityofworkenvironmentandjobperformance.InformationQuality-IndividualImpactsInformationqualitywasmeasuredintermsofaccuracy,timeliness,completeness,relevance,andconsistency.Individualwasmeasuredintermsdecision-makingperformance,jobeffectiveness,andqualityofwork.OtherLinksWithoneexception,alltheotherlinksorassociationsintheD&MISSuccessModelwereempiricallyvalidated.[Theassociationbetweensystemuseandorganizationalrevenuesandprofitabilitywasnotsignificant]ModelIssues3ProcessVersusCausalModelsTheD&MISsuccesstaxonomyanditssixsuccesscategoriesarebasedonaprocessmodelofIS.Inaddition,wearguethatthesixdimensionsareinterrelated,resultinginasuccessmodelthatindicatesthatcausalityflowsinthesamedirectionastheinformationprocess.However,citinganearlierpaperbyNewmananRobey,Seddonarguesthat“theboxesandarrowsinvariance-andprocess-modeldiagramsrepresentquitedifferentconceptsandcan’tbecombinedmeaningfullyinonemodel….Unfortunately,combiningvarianceandprocessmodelsisexactlywhat[DeloneandMcleanhave]attemptedtodo”.ModelIssues3SystemUseasaSuccessMeasureSeddonfurtherarguesfortheremovalof“systemuse”asasuccessvariableinthecausalsuccessmodel,claimingthatuseisabehavior,appropriateforinclusioninaprocessmodelbutnotinacausalmodel.Hearguesthatusemustprecedeimpactsandbenefits,butitdoesnotcausethem.Wedisagree,Webelievethatsystemusageisan