共同方法偏差的统计检验与控制方法-周浩

~AdvancesinPsychologicalScience942*B841.2commonmethodbiases40[1]20040616*70171046027-62532443E-mail:[e-mail]126-943-[2,3]1[4,5]Harman[1][6]OlkinFinn[7]Williams1996[8]Lindell[9,10]-944-2004[1]SPSS[11]AndersonWilliams1992[12]reciprocalrelationship[12][13]126-945-Williams[8][14][1][13]multitrait-multimethod,MTMM11(Podsakoff2003)a2/m2a3/m3c2/m2b1/m1b2/m2b3/m3c1/m1c3/m3a1/m1-946-2004[15]correlateduniquenessmodel[16][17](Podsakoff2003)a2/m2a3/m3c2/m2b1/m1b2/m2b3/m3c1/m1c3/m3a1/m1126-947-[13](directproductmodel)[1]3LISRELEQSAMOS[18](Podsakoff2003)m1/a1m3/b3m1/b1m2/a2m2/b2m3/a3-948-20044HarmanHarmanHarman126-949-MTMM[17]MTMMMTMMMTMM[19][1]PodsakoffPM,MacKenzieSB,LeeJYetal.Commonmethodbiasesinbehavioralresearch:acriticalreviewoftheliteratureandrecommendedremedies.JournalofAppliedPsychology,2003,88(5):879~903[2]LivingstoneLP,NelsonDL,BarrSH.Person-environmentfitandcreativity:anexaminationofsupply-valueanddemand-abilityversionoffit.JournalofManagement,1997,23(2):119~146[3]EbyLT,DobbinsGH.Collectivisticorientationinteams:anindividualandgroup-levelanalysis.JournalofOrganizationalBehavior,1997,18:275~295[4],,..,2003,34(2):183~191[5]HarrisSG,MossholderKW.Theaffectiveimplicationsofperceivedcongruencewithculturedimensionsduringorganizationaltransformation.JournalofManagement,1996,22(4):527~547[6]BurkeMJ,BriefAP,GeorgeJM.Theroleofnegativeaffectivityinunderstandingrelationsbetweenself-reportsofstressorsandstrains:Acommentontheappliedpsychologyliterature.JournalofAppliedPsychology,1993,78:402~412[7]OlkinI,FinnJD.Correlationredux.PsychologicalBulletin,1995,118:155~164[8]WilliamsLJ,GavinMB,WilliamsML.Measurementandnonmeasurementprocesseswithnegativeaffectivityandemployeeattitudes.JournalofAppliedPsychology,1996,81:88~101[9]LindellMK,BrandtCJ.Climatequalityandclimateconsensusasmediatorsoftherelationshipbetweenorganizationalantecedentsandoutcomes.JournalofAppliedPsychology,2000,85:331~348[10]LindellMK,WhitneyDJ.Accountingforcommonmethodvarianceincross-sectionaldesigns.JournalofAppliedPsychology,2001,86:114~121[11].-SPSS.:,1999.87~114[12]AndersonSE,WilliamsLJ.Assumptionaboutunmeasuredvariableswithstudiesofreciprocalrelationships:Thecaseofemployeeattitudes.JournalofAppliedPsychology,1992,77:638~650[13]BagozziRP,YiY.Assessingmethodvarianceinmltitrait-multimethodmatrices:Thecaseofself-reportedeffectandperceptionsatwork.JournalofAppliedPsychology,1990,75:547~560-950-2004[14]WilliamsLJ,AndersonSE.Analterativeapproachtomethodeffectsbyusinglatent-variablemodel:Applicationsinorganizationalbehaviorresearch.JournalofAppliedPsychology,1994,79:323~331[15]SpectorPE,ChenPY,O`ConnellBJ.Alongitudinalstudyofrelationsbetweenjobstressorsandjobstrainswhilecontrollingforpriornegativeaffectivityandstrains.JournalofAppliedPsychology,2000,85:211~218[16]MarshHW,BaileyM.Confirmatoryfactoranalysesofmultitrait-multimethoddata:Acomparisonofalterativemethods.Appliedpsychologicalmeasurement,1991,15:47~70[17]ScullenMJ.Usingconfirmatoryfactoranalysisofcorrelateduniqunesstoestimatemethodvarianceinmltitrait-multimethodmatrices.OrganizationalResearchMethod,1999,2:275~292[18]BeckerTE,CoteJA.Additiveandmultiplicativemethodeffectsinappliedpsychologicalresearch:Anempiricalassessmentofthreemodels.JournalofManagement,1994,20:625~641[19]..,2001,23(5):26~30StatisticalRemediesforCommonMethodBiasesZhouHao,LongLirong(DepartmentofPsychology,CentralChinaNormalUniversity,Wuhan430079)Abstract:Theproblemofcommonmethodbiaseshasbeinggivenmoreandmoreattentioninthefieldofpsychology,butthereislittleresearchaboutitinChina,andtheeffectsofcommonmethodbiasarenotwellcontrolled.Generally,therearetwowaysofcontrollingcommonmethodbiases,proceduralremediesandstatisticalremedies.Inthispaper,statisticalremediesforcommonmethodbiasesareprovided,suchasfactoranalysis,partialcorrelation,latentmethodfactor,structuralequationmodel,andtheiradvantagesanddisadvantagesareanalyzedseparately.Finally,suggestionsofhowtochoosetheseremediesaregiven.Keywords:commonmethodbiases,statisticalremedies,psychologicalmethod.