调查问卷的信度与效度的评价方法研究

*:(:2007A08)(:06C474)1(410205)2*李灿1,2辛玲2(),,,:?,,(reliability)(validity),SPSS(SPSS130)1,,2,3,;,,,,,,,,,1,,,1,A,,C,B,,,D,A,1,,,,,X,T,E,X=T+E,2X=2T+2E,,541200810255:RX=2T/2X=1-(2E/2X)1,,,,:(1)(testretestreliability),,,,,:,,;,,,,,,2~4,,Kappa;,ICC(intraclasscorrelationcoefficient)075,0404,(2)(parallelformsreliability),,,(equivalencereliability),,,(3)(splithalfreliability),,(C.E.Spearman),,,(roe),,-(SpearmanBrownformula):r=2roe1+roe:;;;,:;,,(4)(internalconsistencyreliability),,,kk-1,k,k:=kk-11-ki=1s2is2T(k,s2ii,s2T),Cronbachs,,,,080(5)(scorerreliability),,,,,,()PearsonKendallSpearman,Kendall2SPSSSPSSAnalyzeReliabilityAnalysis,1,ReliabilityAnalysisScaleifitemdeleted,correcteditemtotalcorrelationAlphaifItemDeleted542ChineseJournalofHealthStatistics,Oct2008,Vol.25,No.51AnalyzeScaleReliabilityStatisticsIntraclasscorrelationcoefficientICC()AnalyzeDescriptivestatisticsCrosstabsKappaKappa()AnalyzeCorrelationbivariatePearsonKendallstaubSpearmanSpearsonKendallSpearmanAnalyzeScaleReliabilityModelSplithalfSpearmanBrownGuttmanSplitHalfAnalyzeScaleReliabilitymodelAlphaCronbachsAlphaAnalyzeScaleReliabilityStatisticsFriedmanchisquareKendallAnalyzeCorrelationbivariatePearsonKendallstaubSpearmanSpearsonKendallSpearman,,:Vx=2Tx/2x=1-(2T0+2E)/2x,T=TX+T0,TX,T0RX=2T/2X=1-(2E/2X),2Tx2T,,1,,:(1)(FaceValidity),,,,,,(2)(contentvalidity),,2,X1X,X,,2(3)(criterionvalidity,),,(predictivevalidity),(),,,020~060,070,04~08(concurrentvalidity),,,04~08,,,(4)(constructvalidity),,,,,,,(convergentvalidity),(discriminatevalidity):,,,,,KMO(KaiserMeyerOlkin),KMO,,Kaiser(1974)KMO05,KMO09,:,54320081025540%;(04),,,;04,40%2SPSSSPSS,,Analyze,2,,,,:12AnalyzeCorrelationbivariatePearsonKendallstaubSpearmanSpearson()KendallSpearman()AnalyzeDataReductionFactor(ExtractionRotationPrincipalcomponentsVarimax);,,2,;,;,;,3,4,,,,1,.:,2004,192280.2,.,2000,213258.3,SPSS.,2007,168300.4BaileyKD.MethodsofSocialResearch.4thedition.NewYork:TheFreePress,1994,115165.5..,1997,(4):174177.544ChineseJournalofHealthStatistics,Oct2008,Vol.25,No.5