基于两阶段风险的定制类信息系统开发

37420077()JournalofZhejiangUniversity(HumanitiesandSocialSciences)Vol.37,No.4July2007[]20070116[][](70571072)[]1.(1971),,,,,;2.(1945),,,,,,(,310058)[],,,;,,,[];;;;[]C931.6[]A[]1008-942X(2007)04-0134-10TheDevelopmentandEmpiricalAnalysisonaTwo2stageRiskBasedPerformanceModelofCustomizedInformationSystemsDevelopmentProjectsWANGQiu2zhen,MAQing2guo(DepartmentofManagementScienceandEngineering,ZhejiangUniversity,Hangzhou310058,China)Abstract:Withinformationtechnologyplayinganincreasingroleineconomy,companieshavegrownmoreheavilydependentonthesuccessfuldeliveryofinformationsystems(IS).However,thefailurerateofISdevelopmentprojectsishigh.Advocatesofsoftwareriskmanagementclaimthatakeyquestiontoreducethechanceoffailureofaprojectishowtodealwiththeuncertaintiesofsoftwaredevelopment.Thispaperdevelopsatwo2stageriskbasedperformancemodelofISdevelopmentprojectstostudytherelationshipbetweenprojectinherentuncertainty,managementpracticesanddifferentdimensionsofprojectperformance.Two2stageriskreferstoperformanceriskindevelopmentstageandperformanceriskinimplementationstage,whichrespectivelydenotesthedifficultiesofestimatingprojectoutcomesduringthedifferentstages.Fourobjectiveuncertaintyfactorssuchasrelativeprojectsize,technologycomplexity,developmentteamskill,andclient/userexperienceconstituteprojectinherentuncertainty.Inthispaper,twodimensionsofprojectperformance:processperformanceandproductperformanceareconsidered.Basedonthedatasetcollectedfrom181customizedISdevelopmentprojectsofsoftwarehousesinHangzhouCity,thispapercarriesoutanempiricalanalysisontheresearchmodelwithLISREL8.7,astructuralequationsmodelingpackage.Theresultsrevealthatperformanceriskinthedevelopmentstageandperformanceriskintheimplementationstagearetwoimportantinterveningvariables,whichmediatetheeffectsofprojectinherentuncertaintyandmanagementpracticesonbothprocessandproductperformance.Whileprojectinherentuncertaintyincreasesperformanceriskindifferentstages,managementpracticessuchasprojectplanningandcontrol,internalintegration,userparticipationandtopmanagementsupportreducesuchrisk,assuggestedbytherisk2basedapproaches.Moreover,theresearchfindsthatprojectplanningandcontrolandinternalintegrationhaveadirectpositiveimpactonprocessperformance,andinternalintegration,userparticipationhaveadirectpositiveimpactonproductperformance.Further,topmanagementsupporthasastrongpositiveeffectonuseparticipation.Inaddition,thekeyfactorswhichinfluenceperformanceriskarefoundtobedifferentindifferentprojectstages.Forperformanceriskindevelopmentphase,keyinfluentialfactorsareprojectplanningandcontrol,internalintegrationanduserparticipation.Forperformanceriskinimplementationphase,keyinfluentialfactorsareuserparticipationandtopmanagementsupport.However,wehavenotfoundevidenceforsignificantpositiveinfluenceofuserparticipationonprocessperformance.Thelackofsupportforasignificanteffectofprocessperformanceonproductperformanceindicatesthatprojectswhichcomeinwithintimeandcostbudgetsmaydeliverpoorsystems.Theimplicationsofthesefindingsarediscussedalongwithdirectionsforfutureresearch.Keywords:informationsystemsdevelopmentproject;two2stagerisk;inherentuncertainty;projectperformance;structuralequationmodeling,,,StandishGroup,70%,54%,30%,34%[1],,20%,,,,,[2],(),,[3],()[4-8],,,5314,:PintoMantel,,[9],,[3,10-11],Nidumolu,[4]195,,Nidumolu,,,,,,(CCWResearch),2006,82.9,40.1%,,,,,,,()1.,,(SIZE)(COMP)(SKILL)/(CEXP),,,,;,;;/2.,,(PPC)(INTE);(),(SUPP)(UPAR),;;;3.(DRISK)(SRISK)Nidumolu,,,631()37[4],Nidumolu,:,;4.:(PROC)(PROD),(),,,[4],[12],,,:H1:,H2:,WhiteLeifer,[13],,,,,,,,,,,:H3:,H4:,H5:,H6:,H7:,H8:,,[2]50%,40%-50%[11],[14],[4-7],Kwan2Sik,IT,[15],,:H9:,H10:,[16-17]7314,:[17-18][19][4,20],:H11:,H12:,H13:,H14:,H15:,,,,;,,[17]Nidumolu,,[4]:H16:,Nidumolu,,,,,,,[4],,,[15-16],:H17:,,1:1(),831()37,,,,Likert7,71-7,,Barki[3],Wallace[16],/Moynihan[21],Nidumolu[4],Barki[8],Aladwani[22],Nidumolu[6]RaiAl2Hindi[23]Wallace[16](),,600,20053102005831,190,181,30.2%,LISREL(),?,Boomsma100,200[24];AndersonGerbing100-150[25];BentlerChou()5[26];Marsh/,34,100,6,50[27],,181,,,()LISREL8.7,AndersonGerbing[25],(),Bollen,,,[28],2df(RMSEA)(goodnessoffitindex,GFI)(adjustedgoodnessoffitindex,AGFI)(normedfitindex,NFI)(non2normedfitindex,NNFI)(comparativefitindex,CFI)2df15,RMSEA0.08,GFINFINNFICFI0.90,AGFI0.80():(9314,:/)()()(),(ConfirmatoryFactorAnalysis,CFA),,,,/,2,2,,[25],,Cronbach,0.70T,GerbingAnderson[29],2(23.84),(),,43,,,GFI[30],CFINFINNFIRMSEA2/df22,,,(P0.05),:60%70%48%59%,,,:2/df=1720.07/834=2106,RMSEA=0.069,NFI=0.92,NNFI=0.95,CFI=0.95,,UNCEPROCUNCEPRODPPCSRISKINTE041()373P0.05,33P0.01SRISKSUPPDRISKPPCPROD,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,H16LeishmanCook,[31],,,,,,,:,,,;,,,,,,,,[18-19],,,,1414,:Nidumolu[4]Deephouse[32],,,,,,,,,,,(lognitudinaldata)(SEM),,,,,,,,,,,,,,IT,,,,[][1]StandishGroup.TheChaosReport[EB/OL].[2]McFarlan,F.W.PortfolioApproachtoInformationSystems[J].HarvardBusinessReview,1981,59(5):142-150.[3]Barki,H.,Rivard,S.,Talbot