ebook img

SPSS Inc. SPSS Regression 17.0 PDF

67 Pages·0.569 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview SPSS Inc. SPSS Regression 17.0

i SPSS Regression 17.0 FormoreinformationaboutSPSSInc.softwareproducts,pleasevisitourWebsiteathttp://www.spss.comorcontact SPSSInc. 233SouthWackerDrive,11thFloor Chicago,IL60606-6412 Tel: (312)651-3000 Fax: (312)651-3668 SPSSisaregisteredtrademarkandtheotherproductnamesarethetrademarksofSPSSInc.foritsproprietarycomputer software. Nomaterialdescribingsuchsoftwaremaybeproducedordistributedwithoutthewrittenpermissionofthe ownersofthetrademarkandlicenserightsinthesoftwareandthecopyrightsinthepublishedmaterials. TheSOFTWAREanddocumentationareprovidedwithRESTRICTEDRIGHTS.Use,duplication,ordisclosureby theGovernmentissubjecttorestrictionsassetforthinsubdivision(c)(1)(ii)ofTheRightsinTechnicalDataand ComputerSoftwareclauseat52.227-7013. Contractor/manufacturerisSPSSInc.,233SouthWackerDrive,11th Floor,Chicago,IL60606-6412. PatentNo. 7,023,453 Generalnotice:Otherproductnamesmentionedhereinareusedforidentificationpurposesonlyandmaybetrademarks oftheirrespectivecompanies. WindowsisaregisteredtrademarkofMicrosoftCorporation. Apple,Mac,andtheMaclogoaretrademarksofAppleComputer,Inc.,registeredintheU.S.andothercountries. ThisproductusesWinWrapBasic,Copyright1993-2007,PolarEngineeringandConsulting,http://www.winwrap.com. PrintedintheUnitedStatesofAmerica. Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,ortransmitted,inanyformorbyanymeans, electronic,mechanical,photocopying,recording,orotherwise,withoutthepriorwrittenpermissionofthepublisher. Preface SPSSStatistics17.0isacomprehensivesystemforanalyzingdata. TheRegression optionaladd-onmoduleprovidestheadditionalanalytictechniquesdescribedinthis manual. TheRegressionadd-onmodulemustbeusedwiththeSPSSStatistics17.0 Basesystemandiscompletelyintegratedintothatsystem. Installation ToinstalltheRegressionadd-onmodule,runtheLicenseAuthorizationWizardusing theauthorizationcodethatyoureceivedfromSPSSInc. Formoreinformation,seethe installationinstructionssuppliedwiththeRegressionadd-onmodule. Compatibility SPSSStatisticsisdesignedtorunonmanycomputersystems. Seetheinstallation instructionsthatcamewithyoursystemforspecificinformationonminimumand recommended requirements. SerialNumbers YourserialnumberisyouridentificationnumberwithSPSSInc. Youwillneedthis serialnumberwhenyoucontactSPSSInc. forinformationregardingsupport,payment, oranupgradedsystem. TheserialnumberwasprovidedwithyourBasesystem. CustomerService Ifyouhaveanyquestionsconcerningyourshipmentoraccount,contactyourlocal office,listedontheWebsiteathttp://www.spss.com/worldwide. Pleasehaveyour serialnumberreadyforidentification. iii TrainingSeminars SPSSInc. providesbothpublicandonsitetrainingseminars. Allseminarsfeature hands-onworkshops. Seminarswillbeofferedinmajorcitiesonaregularbasis. Formoreinformationontheseseminars,contactyourlocaloffice,listedontheWeb siteathttp://www.spss.com/worldwide. TechnicalSupport TechnicalSupportservicesareavailabletomaintenancecustomers. Customersmay contactTechnicalSupportforassistanceinusingSPSSStatisticsorforinstallation helpforoneofthesupportedhardwareenvironments. ToreachTechnicalSupport, seetheWebsiteathttp://www.spss.com, orcontactyourlocaloffice, listedonthe Websiteathttp://www.spss.com/worldwide. Bepreparedtoidentifyyourself,your organization,andtheserialnumberofyoursystem. AdditionalPublications TheSPSSStatisticalProceduresCompanion,byMarijaNorušis,hasbeenpublished by PrenticeHall. A new versionofthisbook, updatedfor SPSSStatistics17.0, is planned. The SPSS Advanced Statistical Procedures Companion, also based on SPSS Statistics 17.0, is forthcoming. The SPSS Guide to Data Analysis for SPSS Statistics 17.0 is also in development. Announcements of publications available exclusively through Prentice Hall will be available on the Web site at http://www.spss.com/estore(selectyourhomecountry,andthenclickBooks). iv Contents 1 Choosing a Procedure for Binary Logistic Regression 1 2 Logistic Regression 3 LogisticRegressionSetRule.. ... ... ... ... ... ... ... ... ... ... ... ... 6 LogisticRegressionVariableSelectionMethods.. ... ... ... ... ... ... ... 6 LogisticRegressionDefineCategoricalVariables. ... ... ... ... ... ... ... 8 LogisticRegressionSaveNewVariables . ... ... ... ... ... ... ... ... ... 9 LogisticRegressionOptions .. ... ... ... ... ... ... ... ... ... ... ... .. 11 LOGISTICREGRESSIONCommandAdditionalFeatures... ... ... ... ... .. 12 3 Multinomial Logistic Regression 13 MultinomialLogisticRegression.. ... ... ... ... ... ... ... ... ... ... .. 15 BuildTerms.. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 16 MultinomialLogisticRegressionReferenceCategory. ... ... ... ... ... .. 17 MultinomialLogisticRegressionStatistics ... ... ... ... ... ... ... ... .. 18 MultinomialLogisticRegressionCriteria.. ... ... ... ... ... ... ... ... .. 20 MultinomialLogisticRegressionOptions.. ... ... ... ... ... ... ... ... .. 21 MultinomialLogisticRegressionSave. ... ... ... ... ... ... ... ... ... .. 23 NOMREGCommandAdditionalFeatures.. ... ... ... ... ... ... ... ... .. 24 v 4 Probit Analysis 25 ProbitAnalysisDefineRange . ... ... ... ... ... ... ... ... ... ... ... .. 28 ProbitAnalysisOptions... ... ... ... ... ... ... ... ... ... ... ... ... .. 28 PROBITCommandAdditionalFeatures... ... ... ... ... ... ... ... ... .. 29 5 Nonlinear Regression 31 ConditionalLogic(NonlinearRegression). ... ... ... ... ... ... ... ... .. 33 NonlinearRegressionParameters ... ... ... ... ... ... ... ... ... ... .. 34 NonlinearRegressionCommonModels .. ... ... ... ... ... ... ... ... .. 35 NonlinearRegressionLossFunction.. ... ... ... ... ... ... ... ... ... .. 36 NonlinearRegressionParameterConstraints. ... ... ... ... ... ... ... .. 37 NonlinearRegressionSaveNewVariables... ... ... ... ... ... ... ... .. 38 NonlinearRegressionOptions ... ... ... ... ... ... ... ... ... ... ... .. 39 InterpretingNonlinearRegressionResults ... ... ... ... ... ... ... ... .. 40 NLRCommandAdditionalFeatures... ... ... ... ... ... ... ... ... ... .. 40 6 Weight Estimation 42 WeightEstimationOptions ... ... ... ... ... ... ... ... ... ... ... ... .. 45 WLSCommandAdditionalFeatures .. ... ... ... ... ... ... ... ... ... .. 45 7 Two-Stage Least-Squares Regression 46 Two-StageLeast-SquaresRegressionOptions ... ... ... ... ... ... ... .. 48 2SLSCommandAdditionalFeatures .. ... ... ... ... ... ... ... ... ... .. 49 vi Appendix A Categorical Variable Coding Schemes 50 Deviation . ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 50 Simple ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 51 Helmert .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 52 Difference ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 52 Polynomial ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 53 Repeated . ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 54 Special... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 55 Indicator.. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 56 Index 57 vii Chapter 1 Choosing a Procedure for Binary Logistic Regression BinarylogisticregressionmodelscanbefittedusingeithertheLogisticRegression procedureortheMultinomialLogisticRegressionprocedure. Eachprocedurehas optionsnotavailableintheother. Animportanttheoreticaldistinctionisthatthe LogisticRegressionprocedureproducesallpredictions,residuals,influencestatistics, andgoodness-of-fittestsusingdataattheindividualcaselevel,regardlessofhow thedataareenteredandwhetherornotthenumberofcovariatepatternsissmaller thanthetotalnumberofcases,whiletheMultinomialLogisticRegressionprocedure internallyaggregatescasestoformsubpopulationswithidenticalcovariatepatterns forthepredictors,producingpredictions,residuals,andgoodness-of-fittestsbasedon thesesubpopulations. Ifallpredictorsarecategoricaloranycontinuouspredictorstake ononlyalimitednumberofvalues—sothatthereareseveralcasesateachdistinct covariatepattern—thesubpopulationapproachcanproducevalidgoodness-of-fittests andinformativeresiduals,whiletheindividualcaselevelapproachcannot. LogisticRegressionprovidesthefollowinguniquefeatures: (cid:132) Hosmer-Lemeshowtestofgoodnessoffitforthemodel (cid:132) Stepwise analyses (cid:132) Contraststodefinemodelparameterization (cid:132) Alternativecutpointsforclassification (cid:132) Classification plots (cid:132) Modelfittedononesetofcasestoaheld-outsetofcases (cid:132) Savespredictions,residuals,andinfluencestatistics MultinomialLogisticRegressionprovidesthefollowinguniquefeatures: (cid:132) Pearsonanddeviancechi-squaretestsforgoodnessoffitofthemodel 1 2 Chapter 1 (cid:132) Specificationofsubpopulationsforgroupingofdataforgoodness-of-fittests (cid:132) Listingofcounts,predictedcounts,andresidualsbysubpopulations (cid:132) Correctionofvarianceestimatesforover-dispersion (cid:132) Covariancematrixoftheparameterestimates (cid:132) Testsoflinearcombinationsofparameters (cid:132) Explicitspecificationofnestedmodels (cid:132) Fit1-1matchedconditionallogisticregressionmodelsusingdifferencedvariables

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.