资源描述：
《spss教程_chap_主成分分析与因子分析（spss tutorial _chap_ principal component analysis and factor analysis）》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。
1、spss教程_chap13_主成分分析与因子分析（SPSStutorial_chap13_principalcomponentanalysisandfactoranalysis）Thethirteenthchapter,principalcomponentanalysisandfactoranalysisIntroduce:1,theconceptofprincipalcomponentanalysisandfactoranalysis2,principalcomponentanalysisandfacto
2、ranalysis60%TheconceptofprincipalcomponentanalysisandfactoranalysisNecessityandpossibility:inscientificresearchinallfields,itisoftennecessarytoreflectthingsAlargenumberofvariablesareobserved,andlargeamountsofdataarecollectedforanalysisandfindingregularitie
3、s.Multivariablelargesampleswillundoubtedlyproviderichinformationforscientificresearch,butalsoinacertainprocessTheamountofdatacollectioneffortisincreased,andmoreimportantly,inmostcasesThecorrelationbetweenvariablesmayincreasethecomplexityoftheproblemanalysi
4、sAnalysisbringsinconvenience.Ifeachindexisanalyzedseparately,theanalysismaybeisolated,Notsynthetic.BlindlyreducingindicatorscanlosealotofinformationandispronetoerrorsConclusion.Therefore,itisnecessarytofindareasonablemethodtoreducetheanalysisindexesatthesa
5、metimeReducethelossofinformationcontainedintheoriginalindicator,andmakeathoroughanalysisofthedatacollected.Sincethereisacertaincorrelationbetweenvariables,itispossibletouselessintegratedvariablesAllkindsofinformationareintegratedineachvariable.Principalcom
6、ponentanalysisandfactoranalysisIssuchawaytoreducethedimension.PrincipalcomponentanalysisandfactoranalysisconvertmultiplemeasuredvariablesintofewunrelatedonesMultivariatestatisticalanalysismethodforcomprehensiveindexofLinearsyntheticindicatorsareoftennotdir
7、ectlyobservable,buttheyaremorereflectiveofthingsQuality.Therefore,inthefieldsofmedicine,psychologyandeconomics,aswellassocializedproductionBewidelyused.60%Theconceptofprincipalcomponentanalysisandfactoranalysis(Continued)Duetothefactthatthemeasuredvariable
8、shavesomecorrelation,Therefore,itispossibletocombinethemwithasmallnumberofsyntheticindicatorsThecombinationofallkindsofinformationineachvariableIndicatorsarenotrelatedtoeachother,thatis,theindicatorsrepresent
