CHAPTERIII METHODOLOGY Basedonchaptertwoliteraturereview,thethirdchapterofthisresearchmainlyfocusesonthe appropriateresearchmethodusedtoanalyzeTheImpactofProductQuality,Price,Promotion andBrandImageonConsumerSatisfactionandPurchaseDecisionofLocalMoisturizerAmong GenZinIndonesia.Themethodologythatisdiscussedinthischapterincludesthedata collectionprocessallthewaythroughtherecommendationsbasedonthedataanalysis. 3.1ResearchDesign Researchdesigninaninvestigativeplan,roadmap,andblueprintstrategytoacquireanswersto researchquestions(Kothari,2004citedinSitompul,2021).Theobjectiveofincludingresearch designinanykindofresearchistominimizeerrorswhenconductingtheresearch(Myersetal., 2010citedinMalikahasri,2023).Inthisresearch,variouslocalmoisturizerbrands’product quality,price,promotionandbrandimageimpactonconsumersatisfactionandpurchase decision.Infigure3.1thefirststepinthisresearchistoidentifytheproblem,asstatedin Chapter1.Followingproblemidentification,preliminaryresearchisconductedtofurthercheck andvalidatetheproblemindicatedinChapter1.Thesecondchapterismadeupofliterature reviewsthatwereconductedbyassessingavarietyofliterature.Chapter3coversdatagathering, whichinthiscasewasdoneusingquantitativemethods.Afterdatacollection,thedataacquired wasfurtherevaluatedinChapter4.Theauthorthencametoaconclusionandmade recommendationsinChapter5basedonthedatathatwasexaminedinChapter4. 33 Figure3.1ResearchDesignFlowchart 3.2ResearchDesignApproach AccordingtoCreswell(2015),aresearchapproachisacollectionofplansandproceduresfor conductingresearchthatincludesrangingfromgeneralassumptionstodetaileddatagathering, analysis,andinterpretationtechniques.Theycontainbothbroadstrategiesandspecific techniquesforconductingresearch.Thisstudyemploysaquantitativeapproachsinceitismore suitedtoansweringtheresearchquestionanddeterminingthecorrelationsbetweeneach variable.Thequantitativeapproachisdescribedasfollows: 3.2.1QuantitativeApproach Asstatedby(Saunders,Lewis,&Thomhill,2003,ascitedinAlyaningtyas,2019),aquantitative approachisdescribedastheuseofscientificormathematicalfactstoanswerquestionsorsolve problems.AccordingtoCresswell&Cresswell(2018),quantitativeresearchisanapproachto 34 researchthatusestechniqueslikesurveysandexperimentstocollectdatawithpresettools, producingstatisticaldatainordertoanalyzerelationshipsbetweenvariables.Malhotra,Nunan, andBirks(2016)relyonaquantitativeapproachtoaddressresearchconcernsandcoretheories. Thequantitativeapproachisutilizedinthisstudybecausetheresearcherwantstoknowmore abouttherelationshipbetweentheindependentanddependentvariables. 3.2.1.1SurveyMethod Theresearcherutilizessurveyresearchasoneofthemethodsusedinthisresearch. Questionnairesareemployedasaresearchapproach,withapredeterminedcollectionof questionstosupplementsurveyadministration.AsdefinedbyKelleyetal.(2003),theterm 'survey'isusedinavarietyofcontexts,butgenerallyreferstotheselectionofarelativelylarge sampleofpeoplefromapredeterminedpopulation(the'populationofinterest';thisisthelarger groupofpeopleinwhomtheresearcherisinterestedinaspecificstudy),followedbythe collectionofarelativelysmallamountofdatafromthoseindividuals.Dataarecollectedina specificform.Surveysareintendedtoprovide"asnapshotofhowthingsareataspecifictime" (Denscombe,1998). 3.2.1.2DataCollection a.PopulationandSample AsstatedinChapter1,thisresearchfocusedonGen-ZwholiveinIndonesia,specifically JavaIslandaged12-27yearsold,thathavepreviouslyorcurrentlyusedlocalmoisturizer products.Respondentsarefrom:personalnetworksandsocialmedia.However,sincethe populationofthatparticularpopulationisundetermined,thereforeaminimumsample sizeof200isadvisedformarketingteststudies,includingthisresearchthatisbasedon Malhotraetal.(2016). b.SamplingTechnique Thisresearchutilizesamethodofnon-probabilitysampling,whichMalhotra(2010) describestherearefourtechniques.However,purposivesamplingwasselectedbecause accordingtoMalhotra(2010),Purposivesamplingisparticularlyeffectivewhenthe 35 researcherneedstoobtaininsightsfromspecificindividualswhohavetheexperienceor knowledgenecessarytocontributetotheresearchquestions.Purposivesamplingwas chosenforthisstudytoensurethatthesampleincludesindividualsfromGenerationZ whohavesubstantialexperiencewithlocalmoisturizerbrands.Thisapproachenablesthe researchertofocusonrespondentswhocanprovidedetailedandrelevantinsights. c.QuestionnaireDesign Theresearchutilizesquestionnairestogatherprimarydata.Respondentsaregivena questionnairetofilloutinordertoassesswhichkeyfactors(ProductQuality,Price, PromotionandBrandImage)influenceGenZ'sConsumerSatisfactionandPurchase DecisionforLocalMoisturizers.ThequestionsonthequestionnaireasseeninTable3.1 wereadoptedfromexistingliteraturesquestionnairesforeachvariableswhichincludes (Noerindah&Bernarto,2022)fortheProductQualityvariable,basedonthestudyof (Muthmainnahetal.2023)thequestionsofpriceareadapted,promotionandbrand imagequestionsareestablishedaccordingto(Qotrunnada&Marsasi,2023;Davisetal., 2009).Thequestionsrelatedtothevariableofpurchasingdecisionandconsumer satisfactionareaccordingtotheresearchofMuthmainnahetal.(2023). ThequestionnairewaswritteninBahasaIndonesiatoeasetherespondentsin understandingandansweringthequestionnaire.Allquestionsrelatedtothevariablesare close-ended.Inthisstudy,Theclosed-endedquestionsaredevelopedusingthe5point LikertScale,namely:1)StronglyDisagree;2)Disagree;3)Neutral;4)Agree;5) StronglyAgreetomeasureandevaluatethelevelofagreementforeachquestion..The researcheralsoincludedquestionsabouttherespondents'socio-demographicsand behaviorsinthisquestionnaire.Toprovidecontextforthelocalmoisturizerbrands,the authorfirstaskstherespondentsaboutthelocalmoisturizerbrandtheynowuseasthe mainsubjectofthequestions. 36 Table3.1Demographic-BehaviouralQuestionnaire DemographicQuestions 1Gender -Male -Female 2Age -12-16yearsold -17-21yearsold -22-27yearsold 3Domicile -Jakarta -Bandung -Semarang -Surabaya -Tangerang -Yogyakarta 4Occupation -Middle/HighSchoolStudent -CollegeStudent -Fulltime/parttimeemployee -Entrepreneur -Freelance -Others(pleasespecify) 5MonthlySpendingon Skincareproducts -<Rp300.000 -Rp300.000-Rp500.000 -Rp500.000-Rp1.000.000 ->Rp1.000.000 BehaviouralQuestions 1Howoftendoyou purchaselocal moisturizerproducts. -Onceaweek -2-3timesamonth -Onceamonth -Onceevery2-3months -Rarelyornever 2Wheredoyouusually purchaselocal moisturizers. -Localmarkets -Skincare/beautystores -Onlineplatforms -Supermarket/Conveniencestore -Others(pleasespecify) 3Whatisthemost importantfactoryou considerwhen -ProductQuality -Price -Brandreputation 37 purchasingalocal moisturizerproduct. -Discounts/Promotions -Positiveproductreviews -Recommendationsfromfriends/family -Recommendationsfromsocialmedia -Others(pleasespecify) 4Whatsourcesof informationdoyourely onwhenpurchasinga localmoisturizer. -Onlinereviews -Brandsocialmedia -InfluencersReviewinSocialmedia -In-storerecommendations -Websitesearching -Consultationwithbrand -Others(pleasespecify) 5Whatbrandsoflocal moisturizerdoyou currentlyuse. -Somethinc -Wardah -Emina -Azarine -MsGlow -Hanasui Furthemore,thefollowingtable3.2consistsofasetofmeasurementsfortheoperational variables: Table3.2OperationVariableQuestionnaire QuestionsbasedonOperationalVariables No VariableLabel Indicators References 1.ProductQualityPQ1Ifeelcomfortableusingproductsfromthe localmoisturizerbrandthatiuse. Noerindah& Bernarto(2022) PQ2Ibelievethatthelocalmoisturizerbrandthat iusearegoodforbeautycare. Muthmainnahet al.(2023) PQ3Thelocalmoisturizerbrandproductsiuseis suitableformyskincondition. Noerindah& Bernarto(2022) 2.Price PE1ThepriceofthelocalmoisturizerbrandI useisaffordableforme. Muthmainnahet al.(2023) PE2ThequalityofthelocalmoisturizerbrandI useisrelativelyinaccordancewithitsprice. 38 PE3ProductsfromthelocalmoisturizerbrandI usedarepricedaccordingtothebenefits. 3.PromotionPN1Ifindthelocalmoisturizerbrand’s promotions,suchasdiscountsandspecial offers,compellingandinteresting. (Qotrunnada& Marsasi,2023) PN2Thepromotionalmaterialsforthelocal moisturizerbrandprovideuseful informationaboutthebenefitsandusageof theproducts. PN3Thepromotionalactivitiesforthelocal moisturizerbrandiusearewell-tailoredto meetcustomerneeds. 4.BrandImageBI1Afterusingthelocalmoisturizerproduct,I believethebrandIuseisagoodbrand. (Daviset al.,2009) BI2Afterusingthelocalmoisturizerproduct,I believethebrandIuseissafetouse. BI3Afterusingthelocalmoisturizerproduct,I believethebrandIuseoffershighquality. 5.Purchase Decision PD1Ifeltconfidentinmydecisiontobuythis localmoisturizerfromthisbrandafter learningabouttheproductinformation. Kotler& Armstrong(2008) PD2Idecidedtopurchasethislocalmoisturizer becauseitismypreferredbrand. PD3Ipurchasedthislocalmoisturizerfromthis brandbecauseitfitsmydesiresandneeds. 6.Consumer Satisfaction CS1Basedonmyexperienceusinglocal moisturizerfromthisbrand,Iamvery satisfied. Muthmainnahet al.(2023) CS2Mysatisfactionwiththislocalmoisturizer fromthisbrandmakesmeinterestedin purchasingitagain. CS3Iamsatisfiedwiththislocalmoisturizer fromthisbrandthatIrecommendittomy friendsandrelatives. 39 3.3DataAnalysis a.DescriptiveStatistics Descriptiveanalysisisanessentialstageindataanalysis,providingimportantsummaries andinsightsintothedata.Itinvolvesassessingcentraltendencyandvariability,bothof whicharenecessaryforunderstandingthedataset'soverallpropertiesand distribution.AccordingtoCooperandSchindler(2014),descriptiveanalysisisthe foundationofquantitativeresearchsinceitsummarizesthekeypropertiesofadataset andprovidesbasicsummariesaboutthesampleandmeasures.Thedatacollectedfrom theprecedingquestionnairewasanalyzedusingdescriptivestatisticstoanalyzemore deeplythecharacteristicsofrespondentsfromeachconstruct.Descriptiveanalysisisan essentialcomponentofquantitativeresearch,commonlyusedtocharacterizelargedata setsbypresentingstatisticalmeasuresthatdemonstratecentraltendenciesand variabilities(Miksza&Elpus,2018).Thistypeofstatisticalanalysisassistsin understandingthegeneralbehaviorofdatabyprovidinginformationontheaverage, variance,andoveralldistributionofdatapoints(Bhandari,2023). b.PartialLeastSquare(PLS-SEM)Analysis AccordingtoHairetal.(2019),PLS-SEMisaneffectivestatisticalapproachfordealing withcomplexvariableassociationsaswellasscenarioswheretheresearchfocusison constructpredictionandexplanation.Unlikestandardcovariance-basedSEMmethods, PLS-SEMcanhandlesmallersamplesizesandmakesfewerassumptionsaboutdata distribution,focusinginsteadonoptimizingtheexplainedvarianceofdependent constructs(Hairetal.,2017).Sarstedtetal.(2022)arguethatPLS-SEMisusefulfor developingtheoriesinearly-stageresearch.Itsupportstheevaluationofcomplicated models,includingthehandlingofbothobservableandlatentvariables,makingit appropriateforawiderangeofapplicationsineconomics,finance,andmanagement. c.ReliabilityTest AshighlightedbyHairetal.(2018),theconceptofreliabilityinresearchfocusesonthe consistencywithwhichamethodmeasuressomething.Ifthedataisinconsistent,itis 40 mostlikelynotreliable(Huck,2007).AccordingtoWong(2013),TheSmartPLS programassessesindicatorreliabilitybycalculatingthesquareofeachouterloadandthe compositedependabilityscorefordetermininginternalintegritymustbeatleast0.7. d.ValidityTest CronbachandMeehl(1955)defineconstructvalidityaswhetheratestreflectsthe theoreticalfeaturesitisintendedtomeasure.Forinstance,Hubleyetal.(2013)and Messick(1995)examinecriterionvalidity,whichentailsdeterminingiftheresultsofa testcorrelatewithotherknownmeasuresoftheconstruct.AccordingtoHairetal. (2014),thisisusuallyachievedbyassessinganumberofcriteria,includingtheAverage VarianceExtracted(AVE)forconvergentvalidity,whichindicatesthatsignificantportion ofthevarianceinobservationsisduetothehypothesizedfactors.AccordingtoBagozzi andYi(1998),whentheaveragevarianceextractedfromtheconstruct(AVE)is0.5or above,convergentvalidityisachieved.ThesquarerootoftheAVElatentvariablemust begreaterthanthelatentvariables'correlation(Wong,2013). e.CollinearityTest Hairetal.(2014)defineacollinearitytestinstructuralequationmodeling(SEM)as determiningwhetherpredictorsinthemodelareoverlyintercorrelated,whichcanaffect theanalysis'sreliability.KnowingthevalueofVIFmakesitpossibletodetermine whethertheformativeindicatorsaremulticollinear.Toreducecollinearityconcerns,the acceptableVIFvalueshouldbesmallerthan5(Wong,2013).Toavoidcollinearity, recommendedthateachindicator'svarianceinflationfactor(VIF)scorebelessthan5. f.StructuralPathSignificanceinBootstrapping AsstatedbyHairetal.(2014),thebootstrappingtechniqueisusedtotestthestatistical significanceofpathcoefficientsinaPartialLeastSquaresStructuralEquationModeling (PLS-SEM)framework.Bootstrappingisanonparametricresamplingprocessthatuses replacementfromtheoriginalsampletoestimateanestimator'ssamplingdistribution. ThistechniqueisespeciallybeneficialinPLS-SEMbecauseitdoesnotpresumedata normality,makingitsuitableforsmallerornon-normalsamples.Hairetal.(2014) 41 describePLS-SEM,whichdeterminesthesignificanceofstructuralroutesbygenerating t-statisticsandcorrespondingp-valuesforeachpathcoefficientviabootstrapping.Ifthe scoreexceeds1.96,T-statisticsshouldbeusedwithatwo-sizet-testata5%significant level(Wong,2013). g.CoefficientofDeterminants(R²)andStone-Geisser(Q²)Test Hairetal.(2014)usetheCoefficientofDetermination(R²)inPLS-SEMtodeterminethe amountofvarianceindependentvariablescanbeexplainedbyindependentvariables. Hairetal.(2011)andChin(1998)definesignificantR²valuesas0.67orhigher, moderateR²valuesas0.33,andweakR²valuesaslessthan0.19.TheStone-GeisserTest (Q²),ontheotherhand,isusedtoevaluatethepredictiverelevanceofthemodel.Wong (2013)definesaQ²valueof0.02aspoorpredictiverelevance,0.15asmediumpredictive power,and0.35orhigherasstrongpredictiverelevanceofanexogenousconstructto endogenousvariablesinthemodel. h.FSquare(F2)EffectSize TheFSquare(F²)EffectSizeevaluateshowapredictorvariableaffectsthevarianceofa dependentvariableinastructuralequationmodel.AsindicatedbyHairetal.(2014),F² measurestheimpactofindependentvariablesonadependentvariableinPLS-SEM models.Hairetal.(2014)proposethataF²valueof0.02indicatesaminorimpact,0.15 isamediumeffect,and0.35isasignificanteffect. i.TotalIndirect,TotalEffect,andMediationResults Thiseffectincludesalltheindirectpathsbywhichanindependentvariablemight influenceadependentvariableviaoneormoremediatorvariables(Chen&Hsiang-Chin, 2016).CheungandLau(2008)defineitasthecollectionofalldistinctindirectpathways thatmediatethelinkbetweenvariables.Thismetriciscriticalindetermininghow effectivemediatorsareinconveyingeffectsfrompredictorstooutcomes. TheTotalEffecttakesintoaccountapredictor'sdirectandindirectinfluenceonan outcomevariable.KennyandJudd(2014)underlinethatthetotaleffectequalsthesumof 42 thedirectandindirecteffects,whichisimportantforunderstandingthewhole relationshipdynamicsinSEM.Hairetal.(2017)underlinetheimportanceofknowing thetotaleffectsinceitprovidesaholisticviewofanindependentvariable'sinfluenceon adependentvariable,takingintoaccountallmediatedanddirectpathways. Mediationanalysisisastatisticalapproachthatanalyzesandquantifiesthemechanisms bywhichindependentfactorsinfluencedependentvariablesthroughoneormore intermediaryvariables,knownasmediators.Hairetal.(2017)underlinethenecessityof assessingmediationtodeterminewhetherandhowmediatorsconveytheeffectsof independentvariablestodependentvariables.Thisiscrucialforunderstandingthe underlyingmechanismsofamodel.Moreover,Zhao,Lynch,andChen(2010)divide mediationintotwocategories:partialandfullmediation.Inpartialmediation,boththe directpathfromtheindependentvariabletothedependentvariableandtheindirectpaths throughthemediatorarestatisticallysignificant,implyingthatthemediatorcontributesto butdoesnotfullyaccountfortheindependentvariable'seffect.Fullmediation,onthe otherhand,occurwhenthemediatorispresentandthedirectpathfromtheindependent variabletothedependentvariableisnolongerrelevant,implyingthatthemediatorfully accountsfortheconnectionbetweenthem. 43 Thispageisintentionallyleftblank. 44.