Week-3.1b-Mixed-Methods

MIXEDMETHODSDPUIC–May2015NilaM.WieseCharacteristicsofMixedMethodsInvolvecollectionandanalysisofbothquantitativeandqualitativedataProceduresforbothtypesofdataneedtobeconductedrigorouslyTwoformsareintegratedintotheanalysis(e.g.,merged,connected,embedded)Datacollectionproceduresareincorporatedintoaclearlydefineddesign,thatspecifiestiming(sequentialorconcurrent)andemphasis(equalorunequal)foreachdatabaseMixedMethods:TheIntroductionTheintroductiontoamixedmethodsresearchapproachmustdefineanddescribetoreaderswhatMixedMethodsis.Researchapproachshouldbejustified:Whyisthisapproachneededorbetterinlightoftheresearchquestionsbeinginvestigated?IdentifythetypeofmixedmethodsdesignIndicatetheoreticalperspectiveguidingstudyNotechallengesposedbymixedmethodsresearch(e.g.,cost,time,skillsetofresearcher)TheValueofMixedMethodsAllowtoleveragestrengthsofqualitativeandquantitativemethods,whileminimizingtheirweaknesses/limitationsAllowtheexplorationofresearchquestionsinmoredepth,moreholistically,frommultipleperspectivesEnhancedvalidity:BytriangulatingonesetofresultswithanotherComplementarity(seekingelaboration,clarification)Initiation(discoveringparadoxes/contradictionsthatleadtoresearchquestionsbeingreframed)Development(usingonemethodtoinformtheuseofanother)Expansion(extendbreadthandrangeofinquiry)UseofTheoryinMixedMethodsTheoryinmixedmethodscan:Beuseddeductivelytotesttheoretically-basedresearchquestionsandhypothesesBeusedinductivelytogeneratenewtheoreticalperspectivesGuidetheentireprocessofdatacollection,analysis,andinterpretation:SocialScienceTheoryFrameworkTransformativeTheoryFrameworkTypesofMixedMethodsDesignConvergentParallelExplanatorySequentialExploratorySequentialTransformativeEmbeddedMultiphaseConvergentParallelDesignDefinition/PurposeDataiscollectedusingbothmethods,resultsarecomparedtoseeiffindingsconfirmordisconfirmeachother.Assumption:Differenttypesofinformation,buttogethertheyyieldresultsthatshouldbethesame.DataCollectionAllmethodsofdatacollectioncanbeusedDataiscollectedonthesamevariables,constructs,orconceptsSamplesizesmallerforQual.study.Qual.sampleincludedinQuantsampleAnalysis&InterpretationAnalysisisdoneseparately,andthenresultsaremergedSide-by-sidecomparison:Analysisofonemethod’sresultsfollowstheotherDatatransformation:QualdataisturnedintoquantvariablesandincorporatedintoQuantdatabaseJointdisplayofdata:TowformsofdataarepresentedintableorgraphExplanatorySequentialDesignDefinition/PurposeTwophaseproject:QuantitativeresultsinformthetypesofparticipantstobepurposefullyselectedforthequalitativestudyGoal:QualdatahelpexplaininmoredetailtheinitialquantresultsAppealstoindividualswithastrongquantitativebackground,orfromfieldsrelativelyknewtoqualdesignsDataCollectionRigorousquantitativesampling;purposefulsamplingforqualphaseSameindividualsshouldbesampled(Qualisasub-sample)Analysis&InterpretationDatabasesareanalyzedseparately,andlaterconnectedDiscussionthenspecifieshowthequaldatahelpedexplaintheresultsfromthequantdataExploratorySequentialDesignDefinition/PurposeStartwithaQualphaseandfollowupwithquantstudyGoal:TouseQualstudytodevelopbettermeasurementswithspecificsamplesofpopulationsDataCollectionTwophasesofdatacollection.InformationfromQualstudyisusedto:DevelopinstrumentwithgoodvalidityandreliabilityDevelopnewvariables,identifytypesofscales,orformcategoriesofinformationtobeusedinquantphase;ORDeveloptheoreticalmodeltotestinquantphaseSamplefromQualphaseshouldnotbeusedinquantphase,butsamplesforbothstudiesshouldcomefromsamepopulationAnalysis&InterpretationDatabasesareanalyzedseparately,andconnectedintheDiscussionOtherMixedMethodsDesignsEmbeddedMixedMethodsNestsoneormoreformsofdata(quant,qual,both)withinalargerdesign(e.g.,narrative,experiment).Ex:collectingqualitativedataduringanexperiment(convergent)orbefore/aftertheexperiment(sequential).TransformativeMixedMethodsResearcherincorporateselementsoftheconvergent,explanatorysequential,andexploratorysequentialdesignsintoastudythatusesasocialjusticetheoryasaframework.MultiphaseMixedMethodsResearchersconductseveralmixedmethodsprojects,sometimesincludingmixedmethodsconvergentorsequentialapproaches,sometimesincludingonlyquantitativeorqualitativestudiesinalongitudinalstudywithafocusonacommonobjectiveforthemultipleprojects.ThreatstoValidityinMixedMethodsConvergentParallelDesignSamplesofdifferentsizesLackoffollowupduringdiscussionwhenresultsdivergeExplanatorySequentialDesignResearcherdoesnotfollowuponallrelevantaspectsfromquantstudyresultsInvalidateresultsbydrawingondifferentsamplesforthetwophasesSamplesizesforeitherstudymaybeinadequateExploratorySequentialDesignDevelopmentofmeasurementmaynotfollowgoodpsychometricproceduresInstrumentdevelopmentmaynottakeadvantageoftherichnessofqualdataduetolackofrigorinqualstudyanalysisSampleinqualphaseshouldnotbeincludedinquantstudyChoosingaMixedMethodsDesignChoicebasedon:Outcomesexpected(Table10.3)Howthedatawillbeusedtogether(orintegrated)TimingofthedatacollectionEmphasisplacedonea