结构方程模型原理及其应用注意事项

##(261042)秦浩陈景武(StructuralEquationModel,SEM),(CovarianceStructureModeling,CSM),2080,,112,,,,,12-62,,()(),(measurementmodel),(measurementequation),;(latentvariablemodel),,(structuralequation),::y=+yG+E(1)x=+xN+D(2):G=BG+„N+F(3)(1)(),ypp@1,Gm()m@1,+yyGp@m,Epp@1(2)(),xqq@1,Nn()n@1,+xxNq@n,Dqq@1(3)Bm@m,,„m@n,NG,Fm@1:E(F)=0,E(D)=0,E(E)=0,E(N)=0,E(G)=0;NF,DN,EG,FDE;B0,(I-B)1728:+y,+x,B,„,,7,(E(D4,N,7F,(E(DED,+y,+x,B,„,,7,(E(D(),,,,8,4,,,:p=m=0,3:+x(D,;m=pn=q+y+x,(E(D,3:B„7,;,B=0,11(),,,8,,,#367#20068234,,,,,(),,,;,,,tMIMIC21,()/0/0/0,,:TSLS(two-stageleastsquares)ULS(un-weightedleastsquares)ML(max-imumlikelihood)GLS(generalizedleastsquares)WLS(generallyweightedleastsquares)DWLS(diagonallyweight-edleastsquares)ML,ML,,ML;TSLS,,(),,,,,,,(),,,,,;,,,,,:(1);(2);(3);,:(scale-invariantmodel);(scale-freeparame-ters),,,;,,18231,,:(-1+1);;,;,,:,,(Mt)(Ms),V2RMSEA(RootMeanSquareErrorofApproximation,)SRMR(StandardizedRootMeansquareResidual,)GFI(GoodnessofFitIndex,)AGFI(AdjustedGoodnessofFitIndex,)V2,,V2,V2,RMSEA,,,,,RMSEA011,SRMR0~1,,0108GFIAG-FI,,,(Mt)(Mn)NNFI(non-normedfitindex,)NFI(normedfitindex,)CFI(comparativefitindex,)NNFI0~1,,019,,0~1NFINNFI,0~1,,CFI0~1,,019,CFI,,,,182,,,41#368#ChineseJournalofHealthStatistics,Aug2006,Vol.23,No.4,,:(1);(2)(MI);(3);(4)(coefficientofdetermination)R2;(5)1196192,:(1),,(2),MI:MI,,,MI,,,MIMI31846163,MIV2,N,,MI182MI,(expectedparameterchange,EPC),,,EPC(standardizedEPC,SEPC)(3),:,;,,2192(4),:,,,,,SEM,LISREL(LinearStructuralRela-tion),AMOS(AnalysisofMomentStructures),EQS(E-quations)Mplus,,11AndersonJC,GerninDW.Structuralequationmodelinginpractice:Are-viewandrecommendedtwo-stepapproach.PsychologicalBulletin,1998,103:411-423.21,..,2003,11(3):221-222.31,,..,2004,18(1)19-23.41,..,2003,4(4):261-263.51,,..,2003,17(1):31-35.61,,,...2004,35(1):130-131.71,,..,2004,21(2):70-74.81,,,..,2004,7.91,..,2002,4.#369#20068234