当代市场调研10E全套配套课件中英文PPTch16.ppt 23页

ChapterSixteen16-1 EvaluatingDifferencesandChangesTheissueofwhethercertainmeasurementsaredifferentfromoneanotheriscentraltomanyquestionsofcriticalinteresttomarketingmanagers.Seethetextforsomespecificexamplesofmanagers’questionsrelatingtoevaluatingthesedifferencesandchanges.16-2KeyTerms&Definitions MathematicalDifferences:Bydefinition,ifnumbersarenotexactlythesame,theyaredifferent.Thisdoesnot,however,meanthatthedifferenceiseitherimportantorstatisticallysignificant.StatisticalSignificance:Adifferencethatislargeenoughthatitisnotlikelytohaveoccurredbecauseofchanceorsamplingerror.StatisticalSignificance16-3KeyTerms&Definitions ImportantDifferencesforManagers:Onemustbeabletodistinguishbetweenmathematicaldifferencesandstatisticallysignificantdifferencesinusingthedataanalysisinmanagerialdecisionmaking.Hypothesis:Assumptionortheorythataresearcherormanagermakesaboutsomecharacteristicofthepopulationunderstudy.StatisticalSignificance16-4KeyTerms&Definitions StepOne:Statingthehypothesis“NullHypothesis”–Thehypothesisofstatusquo,nodifference,noeffect.“AlternativeHypothesis”-anotheralternativeisproventobetrue.StepTwo:ChoosingtheappropriateteststatisticTestofmeans,testofproportions,ANOVA,etc.StepThree:DevelopingadecisionruleDeterminethesignificancelevel.Neededtodeterminewhethertorejectorfailtorejectthenullhypothesis.TheKeySteps:HypothesisTesting16-5KeyTerms&Definitions StepFour:CalculatingthevalueoftheteststatisticUsetheappropriateformulatocalculatethevalueofthestatistic.StepFive:StatingtheconclusionStatedfromtheperspectiveoftheoriginalresearchquestion.TheKeySteps:HypothesisTesting16-6KeyTerms&Definitions StatisticalTestsandTheirUses16-7KeyTerms&Definitions StatisticalTestsandTheirUsesContinued16-8KeyTerms&Definitions Rejectionofthenullhypothesiswhen,infact,itistrue.Acceptanceofthenullhypothesiswhen,infact,itisfalse.Testsareeitheroneortwo-tailed.Thisapproachdependsonthenatureofthesituationandwhattheresearcherisdemonstrating.One-Tailed:“Ifyoutakethemedicine,youwillgetbetter.”Two-Tailed:“Ifyoutakethemedicine,youwillgeteitherbetterorworse.”OneandTwoTailedTestsHypothesisTesting–ErrorTypes16-9KeyTerms&Definitions TypeIandTypeIIErrors16-10KeyTerms&DefinitionsRejectionofthenullhypothesiswhen,infact,itistrue.Acceptanceofthenullhypothesiswhen,infact,itisfalse.TypeI:TypeII: IndependentSamples:Samplesinwhichthemeasurementofavariableinonepopulationhasnoeffectonthemeasurementofthevariableintheother.RelatedSamples:Samplesinwhichthemeasurementofavariableinonepopulationmayinfluencethemeasurementofthevariableintheother.DegreesofFreedom:Numberofobservationsinastatisticalproblemthatarefreetovary.Isequaltothenumberofobservationsminusthenumberofassumptionsorconstraintsnecessarytocalculateastatistic.CommonlyUsedStatisticalHypothesisTests16-11KeyTerms&Definitions GoodnessofFitChi-SquareTest:TestofthegoodnessoffitbetweentheOBSERVEDdistributionandtheEXPECTEDdistributionofavariable.1.Chi-SquareTestofaSingleSample2.Chi-SquareTestofTwoIndependentSamples:Marketingresearchersoftenneedtodeterminewhetherthereisanyassociationbetweentwoormorevariables.16-12KeyTerms&Definitions ZTest:Hypothesistestusedforasinglemeanifthesampleislargeenoughanddrawnatrandom.Usuallyforsamplesofabout30andabove.tTest:HypothesistestusedforasinglemeanifthesampleistoosmalltousetheZtest.Usuallyforsamplesbelow30.Hypothesistestingthatteststhedifferencesbetweengroupsofdata.AboutOneMean:AboutTwoMeans:TypesofHypothesisTests16-13KeyTerms&Definitions ZTest-Example16-14KeyTerms&Definitions ZTest–ExampleContinued16-15KeyTerms&Definitions ZTest–ExampleContinued16-16KeyTerms&Definitions HypothesisAboutTwoMeans-Example16-17KeyTerms&Definitions HypothesisAboutTwoMeansContinued16-18KeyTerms&Definitions ProportioninOneSample:Testtodeterminewhetherthedifferencebetweenproportionsisgreaterthanwouldbeexpectedbecauseofsamplingerror.TwoProportionsinIndependentSamples:Testtodeterminetheproportionaldifferencesbetweentwoormoregroups.TypesofHypothesisTests16-19KeyTerms&Definitions AnalysisofVariance(ANOVA)ANOVA:Testforthedifferencesamongthemeansoftwoormoreindependentsamples.Whenthegoalistotestthedifferencesamongthemeansoftwoormoreindependentsamples,analysisofvariance(ANOVA)isanappropriatestatisticaltool.Althoughitcanbeusedtotestdifferencesbetweentwomeans,ANOVAismorecommonlyusedforhypothesistestsregardingthedifferencesamongthemeansofseveral(C)independentgroups(whereC>or=3).ItisastatisticaltechniquethatpermitstheresearchertodeterminewhetherthevariabilityamongoracrosstheCsamplemeansisgreaterthanexpectedbecauseofsamplingerror.16-20KeyTerms&Definitions FTestFTest:Testoftheprobabilitythataparticularcalculatedvaluecouldhavebeenduetochance.16-21KeyTerms&Definitions pValuesandSignificanceTestingpValue:Exactprobabilityofgettingacomputedteststatisticthatisduetochance.Thesmallerthepvalue,thesmallertheprobabilitythattheobservedresultoccurredbychance.16-22KeyTerms&Definitions 16-23KeyTerms&DefinitionsStatisticalSignificanceHypothesisNullHypothesisDecisionRuleTypeIErrorTypeIIErrorIndependentSamplesRelatedSamplesDegreesofFreedomLinksandbuttonareactivewhenin“SlideShowMode”KeyTerms&DefinitionsChi-squareTestZTesttTestZTestExampleHypothesisAboutTwoMeansHypothesisTestofProportionsAnalysisofVariance(ANOVA)FTestpValue