Post on 24-Jan-2021
transcript
Value Based Segmentation of Luxury Consumption
Behavior among Iranian Employed Women
Sara Mehrabi, Roxana Zahedi
Business Administration, masters level 2016
Luleå University of Technology Department of Business Administration, Technology and Social Sciences
1
ACKNOLEDGMENTS
Firstofall,weareextremelygratefultooursupervisor,Dr.AnneEngstöm,forherconsistent,patientandgenerousguidanceandsupervisionofourresearchfromthebeginningtotheend.
Additionally,wewouldliketothankeveryfacultyandstaffmemberindepartmentofBusinessmanagementinLuleaUniversityoftechnologyandalsoBaharBusinessManagementSchoolinIran.
2
ABSTRACT
Luxurymarkethasacontinuousgrowththatprovestheincreasingdemandofluxuryaroundtheworld despite recession or other similar problems. For this reason luxury concept has beenstudied from several aspects in different regions around the world; but as the concept ofmarketingandconsumerbehaviorcanbecustomizedbasedoncultureandpeoplesbeliefsandvalues in every region of the world; local studies are necessary for marketers and businessownersinordertopenetrateindifferentmarkets.In this study the side of the consumer segments and perception of luxury is discussed. Thepurpose of this study is to explore and describe the luxury value perception among Iranianemployedfemalepopulationandtodescribehowthevalueperceptioncanbeusedasabaseforsegmentationorinbriefitisgoingtobediscussed:
1. How value perception of Iranian employed females in purchasing from luxury brandscanbedescribed?
2. WhatsegmentscanbecreatedbasedonIranianemployedfemales’valueperception?
ByknowingmoreaboutIranianconsumers’valueperception,betterstrategiescanbedevisedandplanned;and the resultsmaybeuseful for the luxurybusinessowners tomakedecisionaboutbasedonwhichvaluestheycanattractmorecustomers.This study is performed based on a cross cultural framework which was suggested andexaminedbyWeidman,HennigsandSiebels (Wiedmann,Hennigs,Siebels,2007) (Wiedmann,Hennigs, & Siebels, 2009) and therefore 8 hypotheses were presented in this study to beexamined.Quantitativeapproachwaschosenasresearchmethodologyof thisstudyandDatacollectionwasmadebyonlineandprintedquestionnaires;thenumberofusablerespondsinthisresearchis972.Factoranalysiswasusedtoextractthetargetfactorsoutofdefinedvariablesinthisstudyandfinally6factorswereextractedthatpresentperceivedvaluesoftargetrespondents.Inorder toachieverelevantsegmentationof theresulted factor in thisstudy,hierarchical,k-mean and two-Step clustering process has been conducted based on the factors’ accepteditems.The finding of this research shows that the luxurymarket of Iranian working women is nothomogeneousand3segmentswerefoundbasedonperceivedluxuryvalues.
3
TableofContents1. Introduction.......................................................................................................................................5
1.1. Backgroundresearch...................................................................................................................5
1.2. Definitionofluxuryconcept........................................................................................................6
1.3. LuxuryIndustryanditsMarket....................................................................................................7
1.4. LuxuryBrands..............................................................................................................................7
1.5. Researchtrendsinluxuryfield....................................................................................................8
1.6. ConsumersPerceptionofLuxuryBrands.....................................................................................9
1.7. Problemdiscussion......................................................................................................................9
1.7.1. Iran’sdemographicandfinancialstatusstatistics.............................................................11
1.8. WomenasTargetMarketSegment...........................................................................................12
1.9. ThepurposeofThesis................................................................................................................12
1.10. Delimitations.............................................................................................................................13
1.11. ExpectedContribution...............................................................................................................13
2. LiteratureReview.............................................................................................................................14
2.1.WhatIsLuxury?..........................................................................................................................14
2.2. LuxuryConsumers.....................................................................................................................14
2.3. ValueandConsumerPerceivedValue.......................................................................................15
2.4. ValueDimensions......................................................................................................................15
2.5. Frameworks...............................................................................................................................16
2.5.1. BrandLuxuryIndex............................................................................................................16
2.5.2. Wiedmann,Hennigs&SiebelsConceptualModel............................................................17
2.5.2.1. FinancialValue...............................................................................................................18
2.5.2.2. FunctionalValue............................................................................................................19
2.5.2.3. Individualvalue..............................................................................................................19
2.5.2.4. Socialvalue....................................................................................................................20
2.6. MarketSegmentation................................................................................................................21
2.6.1. GenderSegmentationinMarketingStudies......................................................................21
2.7. ResearchHistoryBasedonWiedmannandHennigsConceptualModel...................................22
2.8. SelectedFramework..................................................................................................................22
3. ResearchMethodology....................................................................................................................25
3.1. Introduction...............................................................................................................................25
4
3.2. TypeofResearch........................................................................................................................25
3.3. ResearchApproach....................................................................................................................26
3.4. SampleSelection........................................................................................................................26
3.5. DataCollection...........................................................................................................................27
4. EmpiricalDataandAnalysisProcess................................................................................................28
4.1. DatascreeningandCleaning.....................................................................................................28
4.1.1. OnlineQuestionnaire.........................................................................................................28
4.1.2. PrintedQuestionnaire........................................................................................................29
4.2. Descriptiveanalysis....................................................................................................................30
4.2.1. ValidityTest.......................................................................................................................32
4.2.2. ReliabilityTest....................................................................................................................33
4.3. Discussion..................................................................................................................................34
4.3.1. Factoranalysis...................................................................................................................34
4.3.2. Clustering...........................................................................................................................39
4.3.3. NumberofClusters............................................................................................................39
4.4. IntroducingtheClusters............................................................................................................45
4.4.1. Cluster1-RationalHedonists.............................................................................................45
4.4.2. Cluster2-OstentatiousIndividualists................................................................................46
4.4.3. Cluster3-QualitySeekers..................................................................................................46
5. Conclusion........................................................................................................................................48
6. Furtherresearchandmanagerialimplications................................................................................50
Appendices...........................................................................................................................................51
OnlinePersianQuestionnaireLink....................................................................................................51
Englishquestionnaire........................................................................................................................51
GeneralQuestions.............................................................................................................................54
References............................................................................................................................................56
5
1. IntroductionIn thischapter,generaloverviewabout luxuryconceptand luxury industryalongwithabriefhistoryof luxury indifferentaspects isgiven;thentheobjectivesandpurposeofthethesis isdiscussed.
Thisintroductioncoversthebackgroundresearch,problemstatement,aimandpurposeofthisresearch.
1.1. BackgroundresearchNowadaystheluxuryconceptisnolongerlimitedtothewesterncountries.Asiancountriesareidentifiedasabigmarketforluxurybrandstoenterandgrow;andeventheyarebecomingthelargestmarketforfashionindustrygoods.(Holmsäter,2009)
The global growth of luxurymarket from €77 billion in 1985 to €223 billion in 2014 as it ispresented in the following figure (Statista, 2015) shows the considerable demandingmarketgrowthofluxuryproducts.
Figure1-1:ValueofthePersonalLuxuryGoodsMarketWorldwide(inEuro)
Source:Statista.com;Valueofthepersonalluxurygoodsmarketworldwidefrom1995to2014
6
Accordingtothecontinuousgrowingnatureofthismarket,itisimportantforluxurymarketerstoknowabouttheconsumers’reasonsforbuyingluxury,theiropinionsandbeliefsaboutwhatluxury is and in what way consumers’ perception of luxury value can affect their buyingbehavior (Wiedmann, Hennigs, & Siebels, 2009). Consumer behavior changes in luxuryconsumptiontimetotime;forexampleglobalfinancialrecessionmadetheconsumerschangetheir behavior andpreferences toward luxury in differentways such as renting luxury goodsinstead of buying them; or by growing the information technology and easy access to theinformation,newtermsandexpressionsarealignedwithluxurysuchaseco-awareness,greenconsumption,ethicallifestyle,etc….(Yeoman,2011).Insuchadynamicmarket,itseemsthatidentification and profiling of different segments of the luxury consumers by performingconsumer based researches in timely manners can be regarded as a major and necessarystrategyformarketers.(Anderson&He,1998)
1.2. DefinitionofluxuryconceptAlthoughtheword“luxury”isclearlyexpressedinseveraldictionariessuchas:“aconditionorsituation of great comfort, ease, and wealth” or “something that is expensive and notnecessary” (Merriam-Webster online dictionary, 2014), however from the perspective of themarketingandbusiness“luxury”isnotclearlydefinedanditdoesnothaveasingledefinition(KapfererJ.-N.,2010);orintheotherwordluxuryiselusiveinmeaningasitchangesbasedonseveral key components such as human involvement, recognition of key values and limitedsupply(Cornell,2002).
Luxurycanberegardedasan“absoluteconcept”bywhichelitesandpowerfulpeoples’livesareimagined (Kapferer J.-N. , 2012). In the far past centuries the signs of luxury glittered inchurches, temples and similar glorious buildings bywhich people tried to takemercies fromGod(s) and later luxurywas the sign of rank in aristocratic societies. (Podolny, 2010). In thepast, social stratification was the cause of luxury life while nowadays luxury causes socialstratification(Kapferer&Bastien,2009).
Generally thepeople’s ideaabout“luxury”conveys lifestyleelements thana specialproduct;forexampleprivateairplanes,privateislandsforspendingholidays,etc.(KapfererJ.-N.,2012);butwhenitcomesto“luxurybrands”theyclearlypointtofamousvaluablebrandssuchasLuisVuitton, Prada, Gucci, etc. This shows that the concept of luxury is absolute and needs nospecialbrandorproducttobeexpressed(KapfererJ.-N.,2012).
Howeverinordertobemorepreciseinthisresearch,“Luxury”isdefinedasthe“highestlevelof prestigious brands” in which different kinds of psychological and physical values exist(Wiedmann,Hennigs,Siebels,2007).
7
1.3. LuxuryIndustryanditsMarketLuxury industry isnotconsideredasahuge industryregardingthenumberofcompaniesandbrands;howeverthescaleof itsmarketisconsiderablylargeinwhichbillionsofdollarsexists(Tovikkai&Jirawattananukool,2010).Normallythebestqualityinalldimensionsofproduction,fromdesigntopackaging,makesaluxuryproductmoreattractiveforconsumersoverEuropeandU.SaswellasemergingmarketssuchasChina,Brazil,IndiaandRussia(Ko&M.Megehee,2012). Even Despite the effects of recession on people’s income during 2007 to 2010, thedemand of luxury products increased and themarket continues its positive growth. (VerdictMarketReport,2013)
Thegrowthrateinthisareaisattractiveenoughforproducers,businessmakersandscientists(Tovikkai & Jirawattananukool, 2010). But despite the numerous scientific researchesperformedinthisregion,therearestillsomeambiguitiesdefiningthemainconcepts(Dubois,Laurent,&Czella,2001).Asanexample,consumersconfusingperceptionofluxuryandprestigecanbementioned(Dubois&Czellar,2002);thereforeitseemsthatsmallnumberofscientificresearchesexistinthisfieldwhiletheremustbemoreinordertocovertheambiguities.
1.4. LuxuryBrandsBrandisidentificationofproductwhichcandifferentiateitselffromcompetitorsbytheuseoflogoanddesignandothervisualsignsandsymbols(Heding&Knudtzen,2008).
AsAmericanMarketingassociationdescribes:“AbrandisaName,term,design,symbol,oranyotherfeaturethatidentifiesoneseller'sgoodorserviceasdistinctfromthoseofothersellers.”(AMA,2014)
Thereisnotauniqueandclarifieddefinitionof“luxurybrand”;ortherearevariousdefinitionsfor it;howeveramoreprecisedefinitionstates thata luxurybrand isexclusivelydifferentbyowning the attributes of rareness, sophistication and high quality. (Chevalier & Mazzalovo,2008).
Tosumuptheconstituentsbywhomaluxurybrandisidentified,thefollowingcomponentsareconsidered(Phau&Prendergast,2000):
• Aluxurybrandelicitsexclusivity.• Itownsanoutstandingbrandidentity.• Brandawarenessandperceivedqualityisinhighlevel.• Itretainssaleslevelandcustomers’loyalty.
Inordertobemorefamiliarwithcurrentglobal luxurybrands,thefollowingtablerepresentstheglobal top10 luxurygoods companieswhich carry themostwell-knownbrandnamesallovertheworldin2012fiscalyear.(DTTL,2014)
8
Table1-1:GlobalTop10LuxuryGoodsCompaniesFY12
luxury
good
ssales
rank
Companyname Countryof
OriginFY12Luxury
goodsales(US$mil)
FY12totalrevenue(US$mil)
FY12Luxury
goodsalesgrowth
FY12netprofitmargin
FY12returnonassets
FY10-12luxurygoodssalesCAGR
1LVMHMoëtHennessyLouisVuittonSA
France 21,060 36,143 18.2% 13.9% 7.8% 18.6%
2CompagnieFinancièreRichemontSA
Switzerland 12,391 13,078 13.9% 19.8% 13.8% 20.4%
3TheEstéeLauderCompaniesInc.
UnitedStates 10,182 10,182 4.8% 10.1% 14.3% 7.5%
4 LuxotticaGroupSpA Italy 9,113 9,113 13.9% 7.7% 6.5% 10.6%
5 TheSwatchGroupLtd. Switzerland 8,319 8,319 15.3% 20.6% 14.3% 13%
6 KeringS.A. France 7,990 12,522 26.3% 11.1% 4.3% 24.5%
7 L’OréalLuxe France 7,161 7,161 16.0% 14.2% 17.2% 11.2%
8 RalphLaurenCorporation UnitedStates 6,945 6,945 1.2% 10.8% 13.8% 10.8%
9 ShiseidoCompany,Ltd Japan 5,522 8,200 -0.7% -1.9% -1.8% 1.5%
10 RolexSA Switzerland 5,122 5,122 n/a n/a n/a n/a
Source:AdaptedfromDeloitteToucheTohmatsuLimited,2014
1.5. ResearchtrendsinluxuryfieldThehistoryofresearchinluxuryfieldisdividedintoseveralcategoriessuchasluxurydefinition,luxurybrands,luxuryproducts,consumerattitudestowardsthisconcept,etc…someresearchespointsouttotherelationshipamongart,necessity,moralityandluxuryandsomeotherdiscussaboutluxurymarket.(Ciornea,Pop,Bacila,&Drule,2012)
Luxury consumers, their motivation, their attitude and behavior toward luxury andconsumption models are other subjects based on which several researches have beenconducted.(Ciornea,Pop,Bacila,&Drule,2012)
9
Other concepts considering consumptionof luxury, thegrowingnatureof luxurymarket, thenecessitytobeawareofconsumers’perceptionof luxuryforbusinessowners,marketersandresearchers, luxury consumers’ behavior and the reasonsbasedonwhich theyprefer tobuyluxury.(Ciornea,Pop,Bacila,&Drule,2012)(Wiedmann,Hennigs,&Siebels,2009)
Othermaterials in this subject focus on strategic aspectswhich include general strategies orstrategiesregardingspecialproductsorbrandsinluxurydomain,internationalizationandbrandcounterfeits.(Ciornea,Pop,Bacila,&Drule,2012)
1.6. ConsumersPerceptionofLuxuryBrandsConsumershavedifferentperceptionofluxurybrandsthatperhapstheyarerelatedtocultureandthepeopleconcerned(Wiedmann,Hennigs,Siebels,2007).
Themotivationofluxurybrandconsumptionandperceptionofluxuryvalueisnotjustrelatedtosocialvaluessuchasimpressingothers,showingoffthesocialstatus,etc…;butalsorelatedtootherutilitiesofluxurybrandsinindividual,financialandfunctionalaspects.Forthisreasonit is important to synthesize all relevant cognitive and emotional value dimensions in amultidimensionalmodel(Wiedmann,Hennigs,Siebels,2007).
1.7. ProblemdiscussionTheconceptofluxuryhasbeenstudiedformanyyearsindifferentaspectsanditseemsthatthegrowingnatureofluxurymarketmakesthisfieldattractiveenoughforscientistsandresearcherstoextensivelystudyluxuryconceptinalldimensions;howeverthereisalackofscientificstudyinthisareawhichmaybecausedbysomereasonsasfollows(Ciornea,Pop,Bacila,&Drule,2012):
• Ambiguitiesindefiningmainconceptssuchasdefiningluxurygoods(Vickers&Renand,2003)createdifficultiesforperformingacertainandobviousresearch.(Dubois,Laurent,&Czella,2001)
• Majorchangeshappenedinluxurymarketovertimewhichmayaffectthismarketinthefuturesignificantlyandthereforemorestudiesseemstobeneededinthatregard;Forexample:
§ In contradiction with the past, luxury consumers are not limited to affluents;thereforethedemandofluxurygoodsishigherandtheconsumerbehaviorhasbeenmodifieddrastically.(Silverstein&Fiske,2008)(Twitchell,2003)
§ Newmore affordable luxury products emerged in thismarket, the number ofluxury brands increased and some changes happened in products’characteristics.(Silverstein&Fiske,2008)(Vickers&Renand,2003)
§ Thesizeofinternationalmarket(Cohen,2007)andthegrowthrateinemergingmarkets(PWC,2012)forluxurygoodsandservicesislarge(Cohen,2007).
10
§ Thesymbolsthat luxuryconsumptioncreates inthesociety(Sekora,1977),thedegreeofflexibilityofthemarketversuscrisissuchasrecession,scandals,etc…,thestabilityoftheluxurybrandsinthemarketforalongtimemakethismarketcertain enough to growth and apparently cause less interest for scientificresearchesandthestudiesseemstobelimitedtopractitioner-orientedjournals(Reyneke,2011).
Therewithal there are valuable market researches with regard to European and Americanconsumersanditseemsthattheycanbeusedasapatternforothermarkets,howeverothersimilar researches show that theremaybedifferentperceptions, values, cultural factors andexpectations that affect consumer behavior in each market (Teimourpour & HeidarzadehHanzaee,2011)(Tovikkai&Jirawattananukool,2010).
Asitisdiscussedinthefollowingparts,Iran’smarketcanbebeneficialandattractiveforluxurymarketersandbusinessownerstopenetratein.ButasnoneofthemostpopularluxurybrandsareoriginallyIranian,morecomprehensiveacademicstudiesinthiscountrywhosemarketandconsumerbehaviorespeciallyinluxuryconsumptionconceptisalmostunknowntonon-Iranianmarketersseemsnecessarytoovercometheprobableissuescausedbyculturaldifferences.
ByknowingmoreaboutIranianconsumers’valueperception,betterstrategiescanbedevisedandplanned;and the resultsmaybeuseful for the luxurybusinessowners tomakedecisionabout based on which values they can attract more customers. (Heidarzadeh Hanzaee &Teimourpour,2012)
Luxury market has developed to “new rich” market nowadays; that means it does not justbelongtowesternordevelopedcountries(Bian&Forsythe,2012).Iranasavastcountryisnotapartfromthegrowthof luxurymarket,despitethedamagedeconomyandsanctionsinpastyears.For instance, Iranhasexperiencedthe luxurycarbooms inrecentyears (NYDailyNews,2012); also it is estimated that Iranians spendabout $12billion annually on imported luxurygoods,accordingtotheIranianIndustryBusinessMinister(RFE/RL'sRadioFardaandAP,2014).Thebusinessopportunitiesofthismarket isgettingmoreclarifiedbyJean–ChristopheBabin,thechiefexecutiveofBulgari:“It’sverywealthyandyouhaveapopulation in Iranwhichhasbeenusedtoluxury.IranwillbethenextbigthingintheMiddleEast…WewereactiveinIranyearsago.Soweknowthemarket,andwehaveplansforthatmarket.”(Flanagan,2015)
Iranmarketisfloodedbyluxuryproductssuchascars,shoes,clothesandfurnitureandthereisahightaxationforimportingthosegoods;sopeoplewouldbuythemeveninhigherpricesthanusual(Sinaiee,2008).“Everydayweseebigstoreswithluxurydecorationappearsineverypartof big cities in Iran. Some wealthy Iranians would pay up to $360,000 to buy Porsche car.
11
Nowadays inTehranstreetsandsomeothercitiesof Iran,wecanseepeoplearedrivingnotonly BMW and Mercedes-Benz which would cost them from $30,000 to $370,000 but alsoMaseratiandPorschethatcouldbefrom$220,000to$470,000”(Nahavandian,2013).
With the growing demand of Iranians for luxury items, it would be a great opportunity forLuxurybrandstoenterandgrowtheirbusiness.
Anyway it must be noticed that even if the luxury brands can appear in the same way indifferent regions of theworld, it doesn’tmean that local consumers buy from them for thesame reasons. The preferences, culture and consumers value perceptions can affect theirdecisionmaking (Heidarzadeh Hanzaee & Teimourpour, An empirical analysis of consumers’luxuryvalueperception,2012).
TheresultofthisresearchwouldhelpcompaniestodeterminethemosteffectivestrategiestopenetrateIran’sluxurymarket,identifyandsegmentconsumersinwomensectorandkeepingtheloyaltyofcustomers.
1.7.1. Iran’sdemographicandfinancialstatusstatistics
Irancontains5or6millionaffluentswhosewealthisnotlessthanrichAmericansaccordingtothe Iranian Minister of Welfare. Iranian elites spend approximately $7,000 in average forrestaurantsandhotelsandnearly$11,000forwaterandelectricityeachyear.SimilarlyCentralBankofIranstatedthattheaveragegrosscostofoneurbanfamilyinIranis$9,200(eachfamilyunitcontains3.47peoplebyaveragein2012).ThegrosscostoflivingforrichestIranianswasabout$380,000in2012.(Tabnak,2014)
Iranpopulationin2012-13wasapproximately77,890,000(numbersareroundedup)inwhichabout25,700,000womenand26,100,000menliveinurbanareas(Iran,2014).
Based on the current declaration of Iran’s Ministry of work, the total number of employedpeoplein2012-13isapproximately21,340,000inwhichabout3,144,000womenexist.(Aftab,2014)
Thefollowingfigurerepresentsthedistributionofthe Iranianfamilymembersbasedontheiractivitiesandgenderin2012.(CentralBankofIran,2012-2013)
12
Figure1-2:DistributionofIranianFamilyMembers;Source:CentralBankofIran,2012-2013
1.8. WomenasTargetMarketSegmentInmostcaseswomenpaymoreattentiontotheirphysicalappearanceandthoughtheyhavemorepositivetrendstowardluxurybrands(Stokburger-Sauer&Teichmann,2013).Iranisavastcountrywiththepopulationofabout75,000,000(Wikipedia,2011)andwomencompriseabout49 percent of the whole population (Trading Economics, 2014). Nowadays in the world ofeconomywomenhavea considerableplacewhich isestimated tobeevenhigher than someemergingmarkets.Thiswonderfulgrowingmarketcannotbeignoredbybusinessownersandmarketersandmostofthecompaniestrytofindaneffectivestrategyinordertopenetrateintowomensegmentmarket(Silverstein&Sayre,2009)(Flanagan,2015).
Womenareselectedastargetsegment inthisresearchfortheycanplayabigrole inmakingbuyingdecisions (Miglaniet. al,2011)and the industries inwhich theynormally spendmoremoneyare:apparel,beauty,fitness,financialservices,foodandhealthcare(Silverstein&Sayre,2009)andthisshowstheopportunitiesforluxurybrandstodeviseastrategyforwomensector.
1.9. ThepurposeofThesisThepurposeofthisstudyistoexploreanddescribetheluxuryvalueperceptionamongIranianemployedfemalepopulationandtodescribehowthevalueperceptioncanbeusedasabaseforsegmentation.Thereforethefollowingtworesearchquestionsarestated:
13
1- HowvalueperceptionofIranianemployedfemalesinpurchasingfromluxurybrandscanbedescribed?
2- WhatsegmentscanbecreatedbasedonIranianemployedfemales’valueperception?
Thedetailedhypothesesarepresentedinchapter2.
1.10. DelimitationsInthisstudythefollowingspecificdelimitationsaredefined:
1- ThestudyfocusesonIranianwomenandmorespecificallyIranianworkingwomen,insteadofthewholepopulation.
2- Asitwasdescribedinthepurposeoftheresearch,theachievementofthisstudyistomakeasegmentationbasedonfinancial,functional,individualandsocialvalues.
3- Thescopeofluxurybrandscoveredinthisstudyislimitedtoluxurybrandsintheretailsectorsuchasbags,shoes,pants,belts,andleatherproducts.
1.11. ExpectedContributionTheachievementsof this thesiswouldbeuseful to thecompanies/retailerswhoproduce/sellluxury brands to better understand the perceived values of luxury products among workingladiesinIranandcustomizetheirstrategiesinordertoattractmorecustomersinthismarket.
14
2. LiteratureReviewThe previous studies and researches on the subject of “consumers’ luxury brands valueperception” are reviewed in this chapter and based on their achievements and results thesuitableframeworkforthisstudyischosen.
2.1.WhatIsLuxury?An old common idea states that luxury or status goods are the ones that belong to specialbrandsandregardlessofthefunctionality,dedicateprestigetotheowner(Husic&Cicic,2009).Some similar opinions also state that if thingswereonly valuedbasedon their functionality,logically there would be no place for luxury and this topic would becomemeaningless; butwhensocialrelationshipsandtheirconsequencesisconsidered,theirrelateddynamicssuchasconspicuous consumption, comparing oneself to others in the social groups makes luxurymeaningful and dedicate a social functionality to it (Kapferer J.-N. , 2010). Beside all thedefinitions and expressions about luxury, Bernard Arnault, the Founder of famous LVMH,represents a modern definition of luxury as: “The ordinary of extraordinary people andextraordinaryofordinarypeople.”(KapfererJ.-N.,Luxuryafterthecrisis:Prologoornologo.,2010)
2.2. LuxuryConsumersAccordingtocontinuousgrowingnatureofluxurymarket,itseemsconsiderablynecessaryforbusiness owners, marketers and researchers to achieve precise information about luxuryconsumers and their perception, their behaviors and attitudes toward luxury concept andcostumersegmentsinthismarket(Wiedmann,Hennigs,&Siebels,2009).DuboisandDuquesnealsobelievethatinordertoachieveabetterunderstandingofluxurymarket,therelationshipbetweenconsumersandbrandsshouldbeidentified.(Vigneron&Johnson,2004)
Asitwasmentionedbefore,luxuryishardtobedefinedanddespitethegeneralideabasedonwhichaluxuriousitemisdefinedaccordingtoitsprice,themoreprecisedefinitionisinferredbyfollowingmaincomponents:(Vigneron&Johnson,2004)
• Humaninvolvement• Limitednumberofproduceditems• Recognitionoftheirvaluebypeople
Thephrase“moreluxurious”isinterestinglyusedinluxuryworldthatshowstheslipperynatureofluxurydefinition(Vigneron&Johnson,2004);forexampleintheworldofluxuriousfashionaLuis Vuitton hand bagmaybe consideredmore luxurious than other similar luxurious brandssuchasPrada.
15
Luxurybrandsaredefinedasthosebrandsinwhichcomparedtotheotherbrandsthepriceofgoodsandtheirqualityishigherbutwithregardtothefunctionalitycomponent,theyarenotnecessarilythebest.(Wiedmann,Hennigs,Siebels,2007)
Althoughtheluxuryitemswereassumedtobeconsumedbyelites,howeveranewprogressof“democratization”hasstartedsince30yearsagobasedonwhichluxurybrandsproducemoreaffordableitemsforthemiddleclasstobepurchased(Vigneron&Johnson,2004);thatshowsluxuryconsumptiondoesnotbelongonlytorichpeople.
2.3. ValueandConsumerPerceivedValueSimilartothemaintopicofthisresearch,“Luxury”,scientificstudiescouldnotdefine“value”and“consumerperceivedvalue”clearlyandbecauseofthatvariousapproacheswerechosentodefinetheseconcepts(Sánchez-Fernández&Iniesta-Bonillo,2007).Howevermarketershavefoundoutthatrecognitionof“values”and“consumers’perceivedvalues”canbesignificantlyimportant in strategicmanagement (Mizik& Jacobson, 2003); They can create a competitiveadvantage for business owners (Wang, Po Lo, Chi, & Yang, 2004) and therefore bring themcustomerloyaltyandprofits(Khalifa,2004).
As a whole, the outcome of variety of definitions presented by researchers and scholars,definesvalueas the resultof an “evaluative judgment”whichnaturally is cognitive-effective,perceptional and preferential (Sánchez-Fernández & Iniesta-Bonillo, 2007) (Holbrook, 1986)(Zeithaml,1988)(Babin&Attaway,2000).
According to thedefinitionof “value”, the interactionbetweena customerandaproductorservice can be implied which is called “consumer perceived value” (Payne & Holt, 2001)(Sánchez-Fernández&Iniesta-Bonillo,2007).
2.4. ValueDimensionsSeveralstudiesrelatedtoluxuryconsumersandtheirbehavioridentifythatluxuryvalueshaveseveral dimensions in both personal and situational aspects. These dimensions includeconspicuousness and snobbery that is considered as interpersonal dimension, hedonism andperfectionism that are related to personal aspect and other components such as social andpoliticalaspectsthatarerelatedtosituation.(Wiedmann,Hennigs,&Siebels,2009)
According to micro-economic consumer theory, the patterns defined in conspicuousconsumptioncancategorizeconsumersintwodifferentgroups:“Conformists”and“Snobbish” (Dubois,Laurent,&Czella,2001).Themainreasonbywhichconformistsaremotivatedtobuyaproductisthatotherpeoplebuythesamething;whilesnobbishtendtobuytheproductthatislesspurchasedbyotherpeople.Anywaythereasonofthetwobehaviorsisthattheydesiretobeconsideredashigh-classandrichandnotaspoor(Dubois,Laurent,&Czella,2001).
16
Thevaluedimensionsextractedfrommostofearlierstudiesinbothpersonalaspect(Vigneron& Johnson, 1999) and interpersonal aspect (Mason, 1992) are conspicuousness, uniqueness,quality,hedonicandextended-selfwhichisdescribedinnextparts.
2.5. FrameworksThe earliest framework suggested by Vigneron and Johnson named as “Prestige-SeekingConsumerBehavior”(Vigneron&Lester,1999)inwhichthebehaviorofconsumerswithregardtopersonalperceptionswasdescribed.ThisframeworkthatwasbasicallyinspiredbyManson’s“status seeking behavior” framework (Mason, 1992) was based on interpersonal aspect ofconsumers’perceptions.
Basedupontheresultsofprecedingstudiesmorecomprehensiveframeworksweresuggestedasshowninnextsubsections.
2.5.1. BrandLuxuryIndexIn2004amorecompleteframeworkwassuggestedbyVigneronandJohnsonnamed“BrandLuxury Index” with regard to consumers’ decision making process based on previousaccomplished researches. This framework is constructedon consumers’personal andnon-personalperceptions.Non-personalperceptionscontain3components(Vigneron&Johnson,2004)(Figure2-1):
• PerceivedConspicuousness:Peoplemayconsumeluxurybrandstofindsocialstatus.Forthesepeoplethehighpriceofanitemisasignofluxury(Vigneron,2000)
• PerceivedUniqueness:Thisdimensionassumesthatrarityandexclusivityofaluxurybrand makes it desirable for people; and the more expensive the brand, thedesirability increases. The perceived uniqueness of a luxury item can reduce itsavailability in themarket and the reason could be for its limited edition (Vigneron,2000).
• Perceived Quality: This component states that people may perceive luxury brandshavebetterqualitycomparedtootherbrands(Vigneron,2000).
Whilepersonalperceptionsincluded2componentsas:
• PerceivedHedonism:Luxury itemscouldbeconsumedfor itcanbringthefeelingofjoyandpleasuretotheconsumer.Inthiscaseluxurybrandconsumercaremoreaboutthesubjectiveemotionalbenefitsthanfunctionalityofluxuryitem(Vigneron,2000).
• Perceived Extended self: people may desire to imitate the affluent life style ordistinguish themselves from non-affluent by consuming luxury brands. Thereforesocial referencing or construction of oneself can be considered as a determinant inluxuryconsumption(Vigneron,2000).
17
Figure 2-1-VigneronandJohnsonFrameworkofBrandLuxuryIndex
Source:Vigneron&Johnson,Measuringperceptionsofbrandluxury,2004
2.5.2. Wiedmann,Hennigs&SiebelsConceptualModelWiedmann,Hennigs&Siebelsin2007extendedbrandluxuryindexframeworkbyconsideringother dimensions and making some modifications in it. In the new framework, dimensionsextendtofourvaluesinwhichdetailedvalueswereincluded.(Figure2-2)(Wiedmann,Hennigs,&Siebels,2009)
18
Figure 22-WiedmannandHennigsConceptualModel;Source:Wiedmann,Hennigs,&Siebels,2009
Wiedmanncrossculturalframeworkhasfourmaindimensionsinsidewhichthereareseveralvalues.
2.5.2.1. FinancialValueAs the name of this dimension implies, this component cares about themonetary aspect ofconsumers’ perception. Theonly elementwhich is demonstrated in this dimension is “Price”that means the value of the item in a valid monetary unit such as dollars, Rials, Euro, etc.(Wiedmann,Hennigs,Siebels,2007)
Someauthorsbelievethatthehighertheitem’spriceis,themoreitattractsluxurycustomersbycreatingapositiveperceptionofthequality intheirmind.(Wiedmann,Hennigs,&Siebels,2009)
19
Forsomeoftheconsumers,highpriceofaluxuriousgoodindicatestheprestigeoftheowner.Inthisregard,prestige-pricinghasbeendefinedassettingahighpriceforaniteminordertodemonstrateahigherstatusorqualityforit.(Wiedmann,Hennigs,&Siebels,2009)
2.5.2.2. FunctionalValueIn general, this dimension describes howmuch functional an item is to the consumer. It isincludedofthreevaluesasshowninFigure1:
• UsabilityvalueWhichmainlyfocusesonthecorebenefitsoftheproductoritemandthedegreethatitcancoverthecustomers’needs.Itcanbeunderstoodbyexamininghoweasytouseanitemcanbe.Accordingtoprimaryusage,consumersexpectanitemtoworkproperlyinlongperiodoftimeandmeanwhilehaveagoodlook. (Wiedmann,Hennigs,&Siebels,2009)
• QualityvalueStates the consumers’ intention to buy a luxury good is that the brand itself whichguarantees the superior quality of the items it produces; that means sometimesconsumers relate the high brand quality to luxury items which shows they perceivemorevalueofaluxuryitem(Wiedmann,Hennigs,&Siebels,2009).
• UniquenessvalueBasedonanassumption, the rareran item is, the consumer find itmoredesirable topurchase.Inaddition,theconsumers’desireincreaseswhenthepriceoftheluxuryitemishigher.To sumup thevalueofuniqueness in consumer’seye, it canalsobe statedthat by purchasing an expensive luxury good, the consumers perceive themselves asmore different and unique in the society. (Wiedmann, Hennigs, & Siebels, 2009) Theperceiveduniquenessvalueiscalledalsoassnobeffect.(Husic&Cicic,2009)
2.5.2.3. IndividualvalueThis dimension addresses consumers’ personal perception regarding luxury consumption andthekindsofpersonalvalue(s)theyperceivebyconsumingluxury.(Teimourpour&HeidarzadehHanzaee, 2011) (Wiedmann,Hennigs,& Siebels, 2009) Themostprominentperceived valuesareasfollows:
• Self-identityvalue
20
Based on the definition in the theory of consumer’s purchasing behavior, self-congruityreferstotheprobabilityof“comparingoneselftootherobjectsorstimuli”(Liu,Li,Mizerski,&Soh,2012)and in thesubjectofbrandpurchasingbehavior it isnamedas“BrandSelf-Congruitywhichismadeofthreemaincomponents:
1. BrandPersonalityCongruity,2. BrandUserImageryCongruity3. BrandUsageImageCongruity
BrandUserImageryandpersonalityCongruityaddresstothebrandconsumer’sperceptionbywhich they feel themselves as a typical user of the brand or by purchasing from thebrand, they match their own personality to the perceived brand personality. (Liu, Li,Mizerski,&Soh,2012)
• Hedonicvalue
According to definition, hedonic products are the ones whose consumption give theconsumer a feeling of fun, fantasy, pleasure, enjoy and excitement. (Zhong & Mitchell,2010)Thushedonismpointsout totheself-pleasureand internalsatisfactionachievedbyconsumingaluxuryproduct.(Wiedmann,Hennigs,&Siebels,2009)Thereforeluxurygoodscanberegardedashedonicproductstowhichtheconsumersareemotionallyattractedandtheymayfindself-pleasurebyusingthem(Wiedmann,Hennigs,&Siebels,2009).
• MaterialisticvalueMaterialismisdefinedasthedegreeofimportancethatpeoplegivetothematerialsratherthanspiritualthings(Merriam-WebsterOnlineDictionary,2014).Theliteratureofresearchwiththesubjectof“Materialism”alsosuggestsaframeworkwithwhichitcanbeshownthelink between individuals and their material possessions. Materialistic consumers prefermoretoconsumeluxuryproductstorepresenttheirownsocialstatustotheotherpeopleortoevaluatetheirownorotherpeople’sprestigebythat.(Vigneron&Johnson,1999)
2.5.2.4. SocialvalueSocialvalueisdiscussedasanotherimportantdimensionofthisconceptualmodelassomeconsumersusesluxuryproductstoimpressotherpeopleandtoincreasetheireffectinthesociety(Husic&Cicic,2009)ortolinkthemselvestotheirdesiredsocialgroups(Kastanakis&Balabanis, 2012). This perceived social value is alsonamedasbandwagoneffect. (Husic&Cicic,2009)
Accordingtothemodel,twokeyantecedentconstructsofsocialdimensionare:
21
• Conspicuousnessvalue:Incontrastwithtraditionalbeliefwhichforalongwhilestatedbuyingdecisionmakinghappensprivatelyandregardlessofother factors, for the firsttime Veblen noted (Veblen, 1899) that as the wealth and richness increases in thesociety,peopleintendtofollowhighersocialclassbehaviorandthedesireofbelongingtoahighersocialclassleadthemtoconspicuousconsumption.(Kastanakis&Balabanis,2012)Theperceivedconspicuousvalue isalsonamedasVebleneffect. (Husic&Cicic,2009)
• Prestige:Based on definition prestige is “a subjective evaluative judgment about thehigh social status of people or inanimate objects such as brands.” (Dubois& Czellar ,2002)Thedefinitionofprestigemaydifferbasedonpeople’ssocialoreconomicstatus,thuspeople’s perceptionof a product or brandprestige level differs (Vigneron& Johnson,1999)andingeneral luxuryandprestigedefinitionsareoftenusedinterchangeablybymarketers.Howeverconsumingluxurybrands,thatcanbecalledasprestigiousbrandseither,dedicatessocialvaluestoconsumersandaswellasinterpersonaleffectssuchasostentation for Veblenian consumers, non-conformity for snob consumers andconformityforbandwagonconsumers(Dubois&Czellar,2002).
2.6. MarketSegmentationForthefirsttimemarketsegmentationconceptwasintroducedin1956statingthatinsteadofmassmarket,amuchbettermarketcanbefoundforthegoodsbyrecognizingtherealneedsindifferentmarketsegments.(Foedermayr&Diamantopoulos,2008)
A similar definition states that market consists of heterogeneous customers with differentwants.Customerspaydifferentlyformeetingtheirneedsandwantsandtheyhavedifferentlifestyle, therefore theybuydifferentquantities. Amarket segment is a customergroup that issimilarinspecialcharacteristicsthataresignificantformarketers.(Doyle,1998)
Inferredbyallabovedefinitionsmarketsegmentationcanbedefinedasasubdividingprocessinwhichcustomerswithsimilarneedsandbehavioraregatheredinseparatesubsetswhichiscalled“segment”.Eachsegmentcanbeconsideredastargetmarketforcompaniesaccordingtotheirstrategies(Bennett,1995).
2.6.1. GenderSegmentationinMarketingStudiesMarket segmentation based on gender is a popular form which is commonly used bymarketers. Specifically “female” segment canbe identified as an important segment in themarketasthelevelofimpulsebuying,hedonicconsumptionandbrandcommitmentishigher
22
in women than men (Tifferet & Herstein, 2012). Therefore women can be regarded asimportanttargetsinluxurymarket.
Nowadayswomenare inseparablepartsofdifferent industriesof theworld, including Iran,and they play different roles as employees in the society. The share of employment forIranianwomen is 23.4% in agriculture 25% in industry and 51.6% in service industry (Irna,2014).Astheemploymentresultstomoreindependencyinfinancialissues,itprovidesmorechoicesforwomentobuytheitemstheydesire.Inthisstudy,wearegoingtocategorizeandmake segments of Iranian employedwomen, basedon their perceived value in purchasingfromluxurybrands.
2.7. ResearchHistoryBasedonWiedmannandHennigsConceptualModelIn2009,Wiedmann,HennigsandSiebelsmadearesearchinordertofindoutwhyconsumersareinterestedtobuyfromluxuriousbrandsandtorecognizethetypesofconsumersbasedondimensionsofthemodel.TheoutputofthisresearchsuggestedfourclustersofconsumersasMaterialists, Rational Functionalists, Extravagant Prestige-Seekers and Introvert Hedonists(Wiedmann,Hennigs,&Siebels,2009).Twoother researchesweredirectlyperformed in IrantheinthefirstonesimilartotheoriginalresearchIranianconsumersweresegmentedandtheresultedclusterswereasNon-Luxury,Rational,Social-ValueseekerandMaterialisticconsumers(HeidarzadehHanzaee, Teimourpour,& Teimourpour, 2012). Another researchwasmade bythe same group in order to investigate the impact of cultural factors on Iranian consumersaccording to purchasing from luxury brands and the dimensions of the framework(Teimourpour&HeidarzadehHanzaee,2011).
2.8. SelectedFrameworkThepurposeofthestudyistoexploretheluxuryvalueperceptionsamongIranianemployedfemalepopulationanddescribehowtheycanbesegmentedbasedonthementionedvalueperceptionsorbrieflythisstudyisgoingtodescribeandexplore:
What are the perceived values among Iranian working women in purchasing from luxurybrandsandwhatsegmentscanbecreatedbasedonthementionedvalues?
To answer the above question a comprehensive framework including different valuedimensions was required and therefore Wiedmann and Hennigs (2007) framework wasselected as the main foundation of the current research, since it was found morecomprehensive framework inwhichall precedingvaluesdescribed in literaturearediscussedandalsootherstudiesandpapersreferredtothisframework.
23
As itwasmentioned inprevious section,WiedmannandHennings (2007)modelwasused insomeIranianstudies.Foritwasnecessarytotranslateandlocalizethequestionnairetobeusedin Iran, the questionnaire of Iranian studywhichwas based on the selected frameworkwasusedasthebaseofthisthesis(HeidarzadehHanzaee,Teimourpour,&Teimourpour,2012).InHeidarzadeh&Teimourpour research thequestionsweremodifiedbasedon Iranian culture,passedallthetranslationstandardsandapprovedbytheministryofscience.
In this study as it was done in Iranian research (Heidarzadeh Hanzaee, Teimourpour, &Teimourpour, 2012), the price value is described considering quality, prestige andconspicuousnessvalues.
Hereisthehypothesisofourresearchbasedontheframeworkwechose.
Hypothesis1TheperceivedlevelofusabilitywithregardtosuperiorfunctionalvalueisasuitablecriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis2TheperceivedlevelofqualitywithregardtosuperiorfunctionalvalueisasuitablecriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis3TheperceivedlevelofuniquenesswithregardtosuperiorfunctionalvalueisanappropriatecriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis4Theperceivedlevelofself-identityintermsofsuperiorindividualvalueisasuitablecriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis5 TheperceivedlevelofhedonismtowardaluxurybranditemwithregardtosuperiorindividualvalueisarightfulcriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis6Thelevelofrequirementtomaterialsintermsofsuperiorindividualvalue,isarightfulcriterionforvaluebasedsegmentationofIranianemployedwomen.
Hypothesis7TheperceivedvalueofconspicuousnessofaluxurybranditemintermsofsuperiorlevelofsocialvalueisanappropriatebasisofvaluebasedsegmentationofIranianworkingwomen.
24
Hypothesis8TheperceivedlevelofprestigeofaluxurybranditemintermsofsuperiorlevelofsocialvalueisanappropriatebasisofvaluebasedsegmentationofIranianworkingwomen.
25
3. ResearchMethodologyIn thispart, the researchpurpose, researchapproach,data sampling,datacollectionmethodandanalysisalongwithresearchstrategyisbeingdescribed.
3.1. IntroductionTheanalyses,whicharebeingused,areFactoranalysisandclustering.Factoranalysisisusedtodeterminethebasisforclassifyingconsumers’perceptionofluxuryvalue.
Factoranalysis isamultivariatestatisticaltechniquebywhicha largenumberofvariablesaresummarized into a smaller set of variables or factors (Hair JR,Wolfinbarger Celsi, H.Money,Samouel,&J.page,2011).Infactoranalysisthenumberofitemsandthestrengthoftheinter-correlations among the items are important issues. The factorability of the data is beingassessedbytwostatisticalmeasuresinSPSS.Bartlett’stestofsphericityandKaiser-Meyer-Olkin(KMO) measures of sampling adequacy (Hair JR, Wolfinbarger Celsi, H.Money, Samouel, &J.page,2011).
In the cluster analysis, classifying the respondents is the purpose. Cluster analysis is amultivariate statistical approach that combines items into homogenous groups or classifygroups into the clusters that objects are similar to each other and are different from otherobjects inothergroupsorclusters. (Hair JR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011)
3.2. TypeofResearchTherearethreetypesofresearchdesigns:
• Exploratory• Descriptive• Casualdesign
When there are no clear questions and little theory available orwhen researchers find it sohard to formulate the basic research problem exploratory research is used to develop anunderstandingofresearchquestions.(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011)
When we would like to describe a situation by using descriptive statistics we would usedescriptive research.Usuallydescriptivestatistics talkabout frequencycounts (howmany)oruse mean and mode and variations and so forth. (Hair JR, Wolfinbarger Celsi, H.Money,Samouel,&J.page,2011)
Themostcomplexoneiscasualresearchdesign,inwhichifoneeventhasinfluencetheotheristested.(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011)
26
This study is based on previously tested model. The framework belongs to Wiedman andHennigsframework(Wiedmann,Hennigs,&Siebels,2009)butthequestionnairebelongstotheTeimourpour and Heidarzadeh research (Teimourpour & Heidarzadeh Hanzaee, 2011) whichwasdonebasedonWiedmanandHannigsstudies.
For Teimourpour and Heidarzadeh study the questionnaire that was derived fromWiedmanand Hannigs was content validated by 15 marketing experts and face-validated by 65respondentsinidentifyingthevagueitemsandreducingthetotalnumber.Sofinallytherewasa questionnaire of 63 Items, including the demographic items and 53 items, excluding thedemographicquestions.
3.3. ResearchApproach There are two possible research approaches; one is qualitative and the other one isquantitative. Effectivedecision-makingrequiresqualitativestudytodevelopideas,whichcanbetestedbyquantitativeapproaches.Quantitativedataistalkingaboutmeasurementswhichnumbersareuseddirectlytoshowthecharacteristicsofsomething.However,qualitativedatadescribethingswithoutassigningnumberstoitdirectly.(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011)Here in this thesis, quantitative approach is used to test hypothesis by applying statisticalcriteriatothemeasureand3stepswerefollowedtoanalyzethedataviaSPSSver.21:
1- Thepoolofitemsisbasicallycreatedbasedonthedimensionsofthededicatedmodel.2- Exploratory factor analysis and reliability testwere conductedbywhich firstly validity
and reliability of data is assured and secondly the factors into which the predefinedvariables can be reduced and combined, becomeuncovered and accept or reject thementionedhypotheses.
3- BasedonthefactorsHierarchicalandK-meansclusteringisconductedinordertoreachtotargetedsegmentsofemployedwomen.
3.4. SampleSelectionItisidealtohavedatafromallpeopleunderinvestigation,butinmostsituationsitisimpossiblesoasampleofthepopulationisused.Asampleisasubsetofpopulation.
Sample isdrawnusingeitherprobabilityornonprobabilityprocedure.Probability sampling isused in quantitative approach in research. This consists of the random selection from thepopulationtoassuretheobjectivityinselectingdata.Samplesizeisalsoveryimportantinbothqualitative and quantitative approaches. The sample should reflect the characteristics of thepopulationsoweneedtohavetheminimumerrorsinhavingthesample.(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011)
27
Thesamplinghas5procedures:(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011)
1) Definingthetargetpopulation2) Choosingthesamplingframe3) Selectingthesamplingmethod4) Determiningthesamplesize5) Implementingthesamplingplan
Probabilitysamplingisonewayofsamplingwhichisbasedonarandomprocedurethateachsamplehasknownandnonzerochanceofbeing selected.Thismethod isunbiased if isdonecorrectlyandproperly.
Some researchers believe that a sample size of 500 is useful for showing the opinion of thetarget population of 15 million people (Hair JR, Wolfinbarger Celsi, H.Money, Samouel, &J.page,2011).Howeverthenumberofcollecteddatainthisthesisreachedtomorethan1000whichmaketheresultsmorereliable.
SincewewouldliketohaveworkingwomenastargetsegmentofourthesissomeelementaryinformationwithregardtopopulationofemployedwomeninIranisdedicatedbelow:
BasedonthelatestreportofMinistryofwork,thenumberofworkingforceinIranin2014is21,337,478 people; inwhich the number ofmen andwomen are 18,193,544 and 3,143,932respectively.Inaddition,amongallworkingpeopleinIran15,100,061arelivinginurbanareas.Theworkingwomenpopulationwho live in urban areas is 2,183,000people. (Tabnak, 2014)(CentralBankofIran,2012-2013)
3.5. DataCollectionData is very important regardless of whether the research approach is qualitative orquantitative.Data canbe gained in variousways. This study is descriptive andneeds a largeamountofdatagatheredthrough large-scalesurvey.Nowadays informationtechnologyhelpsresearchertoconveytheirsurveyfast(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011);sotheonlinemethodsalongwithself-completionsurveywereusedasanefficientwaytogatheringdatainthisstudy.
The questionnaire used in this study, belongs to Teimourpour and Heidarzadeh research(HeidarzadehHanzaee&Teimourpour, 2012).However theoriginal questionnairebelongs toWiedman and Hennigs study (Wiedmann, Hennigs, & Siebels, 2009) based on whichTeimourpour and Heidarzadeh customized and translated the needed questions for theirresearch inFarsi. In fact there isnosignificantdifferencebetweenmentionedquestionnaires
28
except that Teimourpour and Heidarzadeh omitted some of the questions for localizationreasons.
Thequestionnaireconsistsof63items.Thefirst53itemsarerelatedtovaluedimensionsandhypotheses and the rest are general demographic questions. The previously mentioned 53itemswhicharerelatedtotheresearchvaluedimensionsarescoredona5pointLikertscaleas:
1- StronglyDisagree,2-Disagree,3-Neutral,4-Agree,5-StronglyAgree
Each respondent can choose one answer based on their own desire and personal opinions.Also,thereare10regardingthedemographicinformationofpeoplewhoareaskedtofillinthequestionnaire.
Theonline formofquestionnairewaspreparedandbecameavailable tobesubmittedonlinethrough google drive service. People were invited to fill the form through popular socialnetworkssuchasFacebookandpopularcommunicationtoolssuchasViberandWhatsApp.andalsotheinvitationsweresentbyemailstothefriendsandpeopletowhomwewerepermittedtosendemail.
Theprintedquestionnairesweredistributedinsomecompaniesandalsogetofferedtopeopleinbigluxuriousmalls.
AstheresearchisnotlimitedtoTehran,wetriedtohaverespondentsfromallworkingwomeninIran’surbanareas.Toovercometheproblemoftimelimitationfortraveling, inadditiontoprinted forms,we found internetandonline formasa successfulmedia toattractour targetrespondentsfromalloverthecountry.
4. EmpiricalDataandAnalysisProcessIn this chapter the methods of data gathering and some statistics about participants aredescribed.Thereliabilityandvalidity,Factoranalysisandalsoclusteringarebeingdiscussedinthischapter.
4.1. DatascreeningandCleaning
4.1.1. OnlineQuestionnaireAsitwasmentionedpreviously,onlineformwasdesignedandpreparedthroughgoogledriveservicestoletthetargetrespondentssubmittheinformationonline.Thelinktothementionedform was introduced on well-known social networks and famous social communicationapplications and also on theweb siteswhichwere specializedmostly forwomen under the
29
subjects such as women fashion, motherhood, women health, women legal rights, etc. thenumberofrespondsthroughthismethodreachedto618,inwhich9respondswereincompleteand 22 respondentsweremen or unemployedwomen,whichwere out of the range of thetarget segment of this study and therefore they get eliminated from final database. The netrespondsachievedbythismethodwas587.
4.1.2. PrintedQuestionnaire750printedquestionnairesweredistributedamongcompaniesandinluxuriousbrandstoresofTehran.Inthismethod119questionnaireswerelost,62formsreturnwithnoanswerand184formswerefilledwithincompleteorinvaliddata(men,unemployedwomen).Thenetrespondsachievedinthismethodwas385.
Thereforethenumberofusablerespondsinthisresearchis972asthefollowingtableshows:
Table4-1:DataGatheringResults
CaseSummaries
AgeinYears
Province Mean N
Tehran 31.95 613
EastAzarbyjan 30.05 41
Ardebil 33.11 9
Ispahan 35.16 19
Alborz 36.47 19
Boushehr 30.82 17
Charmahal-Bakhtiari 31.00 4
RazaviKhorasan 26.77 35
Khouzestan 34.10 61
Sistan-Baluchestan 30.62 13
Kordestan 38.67 15
Kermanshah 37.50 14
Kohgilooyeh-BoyerAhmad 33.95 19
Guilan 34.59 29
Lorestan 36.13 15
Mazandaran 25.00 5
Markazi 28.30 40
Hamedan 27.00 4
Total 32.10 972
30
Ascanbeseeninfromtheabovetable,morethan63%ofrespondentswerefromTehranandcompletedemographicinformationcanbefoundintable4.1-4.
4.2. DescriptiveanalysisAccordingtothequestionnairewehave53itemsorvariablesrelatedtothehypotheses,whichmight need to be summarized to have smaller set of components or factors to bediscussedeasier.
Factoranalysis isamethodbywhichasmallersetof factorscanbeextractedoutofasetofvariablesinordertobetterdescribetheinterrelationshipamongthemandthemostcommonmethodstoperformitare:principalcomponents,principalfactors,imagefactoring,maximumlikelihood factoring, alpha factoring, unweighted least squares and generalized least squares(Pallant, 2010). Themost commonmethodwhich is used inmost of the studies is “PrincipalComponents”withVarimaxrotation(Pallant,2010).
Therearetwomaintechniquesinfactoranalysis:exploratoryandconfirmatory.
Exploratoryisoftenusedatthebeginningpartoftheresearchtogatherinformationabouttherelationshipamongvariables.However confirmatory isused in laterpartsof research to testhypothesis.(Pallant,2010)
Thisstudyhastwomainstages.Thefirststageistofindoutthebasisforclassifyingconsumers’perceptionofluxuryvalues.Inordertoreachthispurpose,factoranalysiswasperformedon53items related to the chosen framework’s value dimensions out of 63 items which the restbelongtogeneraldemographicquestions,inordertosummarizerelatedvariablesintosmallersetoffactors.Therefore the various dimensions underlying luxury value perception were uncovered by afactoranalysisusingtheprincipalcomponentmethodwithVarimaxrotation.In order to reach to satisfactory number of factors, it is recommended to use one of thefollowingmethods:Kaiser’scriterion,screetestandparallelanalysis(Pallant,2010).Kaiser’s criterion or eigenvalue rule, shows the amount of total variance explained by thatfactor (Hair JR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011), is themostcommonmethod by which the factors with an eigenvalue equal or more than 1 are kept to beinvestigatedmoreintheresearch.(Pallant,2010)This study is originally based on Wiedmann and Hennings (2007) conceptual framework inwhich 9 value dimensions were examined; but as the tested and localized framework byHeidarzadeh&Teimourpour(2012)andtheirquestionnaireexamined8valuesbyomittingthe“Price”valuestatingthisvalue isdescribedconsideringquality,prestigeandconspicuousness
31
values(HeidarzadehHanzaee,Teimourpour,&Teimourpour,2012);inthisstudy“price”valueisomittedinasimilarwayinordertoavoidmajormodificationsinstandardizedquestionnaire;alsopeople’smotivationtokeeptheirrealincomelevelhiddenduetolowrateofemployment,unstableincomeinthesocietyandothersimilarmotivations(Mohanty&Misra,2000)andalsowithregardtosomefactorssuchasIran-Iraq8yearsofwarandU.S.sanctions,Iran’seconomyperformancewasunderitsfullpotentials(Valadkhani,2003)itseemsthatIranianslessintendto talk about their true income which is necessary for evaluating “Financial” dimension.Howeverwithregardtothequestionsinwhichpriceandfinancialvaluesofluxuriousitemsareincluded, this dimension is indirectly considered. Therefore the ideal number of extractedfactorsiseightthatiscompletelyadaptedtotheoriginalframework.
At the first step the factor analysis was performed using Eigenvalue extraction method
(Eigenvalues>1)andfifteenfactorswereextractedandsuggestedintheresult.Asthisstudyis
basedonTeimourpourandHeidarzadehconceptualframeworkinwhich8factors(values)are
giventobeexamined,againfactoranalysiswasdoneusing“Fixednumberoffactors”setto8
and using this solution 52.3% of the total variance in 53 itemswas gained. The number of
factorsarefixedbasedonpreviousstudieswhichwasmentionedaboveandalsobasedonthe
rule that states thenumberof total variancewhich shouldn’tbe less than50%of thewhole
population(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011).
Despite itwasexpectedthatrelated itemsofeachfactorappear inside itsownrelatedfactor
withhighfactorloading,someextractedfactorscomprisenon-relateditemswithhigherfactor
loadings.Thisphenomenoniscalled“crossloading”andwhenithappensitdoesn’tmeanthat
factoriswrongoritcannotbeused(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,
2011).
Thereforeamongextracted8factors,2ofthem(prestigeandself-identity)couldnotbeprovedsince their items were totally diffused in different factors and nothing could be interpretedfromtheirfactorloadingabove0.3andthereforethehypothesesno.4(Theperceivedlevelofself-identity in terms of superior individual value is a suitable criterion for value basedsegmentation of Iranian working women.) and no. 8 (The perceived level of prestige of aluxurybrand item in termsofsuperior levelof socialvalue isanappropriatebasisofvaluebasedsegmentationofIranianworkingwomen.)Couldnotbeproved.
32
As the variables achieved by factor loadingwhich are not interpretable and are useless and
their factor loading are less than 0.3 can be ignored, (Hair JR,Wolfinbarger Celsi, H.Money,
Samouel,& J.page, 2011) the variables belong to prestige and self-identity factorswere also
eliminated.Consequentlyfactoranalysiswasperformedagaintocheckifeliminationofthose
variables has any influence on others. Finally, some other variable among the rest were
eliminated,onwhichwehadcross loading.Finally6 factorsachievedusing thismethod.This
time each itemwas placed in relevant factor and was easy to interpret. Thementioned six
factorsexplain57.21%ofthetotalvariance(Table4.6).
4.2.1. ValidityTestValiditytestis“accuratemeasurementofaconceptorconstruct”;(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011)andthetoolsbywhichthistestisperformedareasfollows.
TheKMOstatistic issamplingAdequacymeasurement.TheKMOshowshowsmallthepartialcorrelations are. If variables have common factors then correlation should be small and theKMOshouldbecloseto1.0.KMOvaluemorethan0.8isconsideredgoodwhichshowsthatthecomponent for factor analysis is useful for variables. KMO value less than 0.6 would needremedialactions. Itcouldbebetterbyomittingoffendingvariablesoraddingothervariables.(HenryKaiserandBarbaraA.Cerney,1977)
ThecalculatedvalueofKaiser-Meyer-Olkinmeasureofsamplingadequacy(KMO)inthisstagemustbemorethan0.6andthevalueofsigforBartlett’stestofSphericitymustbesocloseto0to prove the value of Chi square parameter is significant and therefore it proves the intercorrelationmatrixofvariablesaresuitableforfactoranalysis.InthisstudythevalueofKMOis.849>0.6 for 53 variables and .852 for 28 variables and the sig value for Bartlett’s test ofSphericityis0.000,sofactoranalysisisappropriatetechniqueandthoughthevalidityofdataisproved.
Table4-2:KMOandBartlett'sTestfor53variablesand8factors
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .849 Approx. Chi-
Square 25976.426
Bartlett's Test of Sphericity Df 1378 Sig. .000
33
Table4-3:KMOandBartlett'sTestfor28variablesand6factors
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.
.852 Approx. Chi-Square 10663.151 Bartlett's Test of Sphericity Df 378 Sig. .000
4.2.2. ReliabilityTestReliability test is a common method by which the internal consistency or reliability ofquestionnaire items,usually in the formof Likert, isdeterminedand it isnormallymeasuredwithCronbach’salpha(Laerd,2014).
Cronbach’salphameasurestheinternalconsistency,whichmeanshowmuchasetofitemscanbeasagroup.Itcanbeameasureofscalereliability.“Cronbach'salphaisnotastatisticaltest-itisacoefficientofreliability(orconsistency).” Here is the formula for Cronbach’salpha.Nisthe number of items and c-bar is the average inter item covariance and V-bar is averagevariance. Here can be seen that if the number of items increase, α is increased as well. Ifaverageinter-itemcorrelationislow,alphawillbelowandviceversaiftheaverageinter-itemcorrelationincrease,alphawillincrease(ifweholdthenumberofitems)(Idre,2015)
Asaruleof thumb,thevalueofChronebach’salphashouldbemorethan0.7 inordertobeacceptableandthevalueof0.9provesexcellentstrengthofassociation.(HairJR,WolfinbargerCelsi,H.Money, Samouel,& J.page, 2011) .As table4-2 indicates, the valueofChronebach’salpha resulted by reliability test in this study is 0.910 that confirms the strong internalconsistencyofitems(Table4-4).
Table4-4:ReliabilityStatistics(step1)
Cronbach'sAlpha Cronbach'sAlphaBasedon
StandardizedItems
NumberofItems
.910 .909 53
Anotherreliabilitytestwasperformedafter2valuedimensions(Prestigeandself-identity)andtheirrelatedquestionnaireitemswereremovedastheirreliedhypotheses(4and8)couldnot
34
beproved.Thoughthenewreliabilitytestwasdoneincluding6factorsoutof8andthevalueofChronebach’salpha(0.882)againpresentedastronginternalconsistencyofdata(Table4-5).
Table4-5:Reliability Statistics (step 2)
Cronbach'sAlphaCronbach'sAlphaBasedon
StandardizedItems
NofItems
.882 .884 28
4.3. Discussion
4.3.1. Factoranalysis
RotatedComponentMatrix squarepresents “factor loading” forall variables. The correlation
between factors and variables is called factor loading (Hair JR,Wolfinbarger Celsi, H.Money,
Samouel, & J.page, 2011). The highest factor loading for each variable is chosen. Since we
selectedsixfactors,relatedvariableswithhighestfactorloadingaregroupedineachfactor.
Table4-6showsafinalsummaryoutputoffactoranalysisinthreecolumns:Factornamesthat
showsthenameofagroupofitems,thesecondcolumnisItems(variables)andtheirnumber
andfinallythelastcolumnwhichpresentsthefactorloading.Asitispresentedinthirdcolumn,
allitems’factorloadingareabove0.3whichisconventionallyacceptedbyresearchers(HairJR,
WolfinbargerCelsi,H.Money,Samouel,&J.page,2011).
Six factorsare included in table4-6 thataccording to their items’ factor loadings their relied
hypotheses couldnot be rejected and theother variables couldnot be correlated andmake
anotherfactorssincetheirfactorloadingscouldnotbeaccumulatedinonespecificfactor.
35
Table4-6:FactorAnalysistable
Herethedetailsoffactorsarediscussed.
FactorName Items FactorLoadingsF1:Quality(11.07%oftotalvariance)
• X1:Oneofthemaincharacteristicsofluxuryproductsisquality. .710• X6:Luxuryproductsarelonglasting. .809• X18:Themainreasonforbuyingluxurygoodsistheirquality. .844• X30:Youalwaysshouldpaymoreforthebestquality. .412• X33:Luxuryitemsareexpectedtoobservehighstandards. .323• X38:Ithinkthatthehighpriceofluxuryisforitsquality. .831
F2:Uniqueness(10.86%oftotalvariance)
• X5:Ienjoyhavingthingsthatothersdonothave. .325• X11:Icanspendlotsofmoneyforrarethings. .522• X17:Fewpeopleownatrueluxuryproduct. .730• X32:Inmyopinion,productsthatallcanaffordarelessvaluable. .482• X39:Luxuryproductscannotbemassproduced. .760• X45:Idonotliketobuyproductsthatthemajorityofpeoplecan
affordtobuy..435
• X46:Aluxuryproductshouldnotbesoldinnormalstores. .763F3:Usability(10.14%oftotalVariance)
• • X3:Itendtouseluxuryitems. .477• X15:Inmyopinion,havingluxurygoodsisnotnecessary. 0.751• X23:Usingluxuryproductsincreasesthequalityoflife. 0.624
F4:conspicuousness(10.08%oftotalVariance)
• X27:Itendtobuytheluxuryitemthatisanewstyle. 0.315• X37:IlikepeopletoknowthepriceofthegoodsIbought. 0.714• X44:Havingluxuryproducts,IcanshowpeoplethatIamrich. 0.513• X47:ItendtoshowofftheluxuryitemIbought. 0.692• X52:Others’opinionsaboutmylookareanimportantissueforme. 0.607
F5:Hedonic(7.99%oftotalvariance)
• X19:Luxuryproductsshouldmakemehappy. 0.471• X31:Luxuryproductsshouldmakemefeelbetteraboutmyself. 0.428• X34:Luxuryconsumptionshouldbringmeself-satisfaction. 0.805
F6:Materialistic(7.04%oftotalvariance)
• X14:Ithinkthatmylifewouldbebetterbybuyingluxuryproducts. 0.315• X26:Thecriterionforassessingpeopleistheamountofmoneythey
have.0.670
• X36:Ibelievethatmoneybringshappiness. 0.593• X48:Therateofpeople’sassetsinsocietyshowsthedegreeof
successinlife.0.733
36
Factor1:QualityValue
This value talks about the quality and performance. The highest loading belongs to variable“X18: Themain reason for buying luxury goods is their quality”with 0.844 and followed by“X38:Ithinkthatthehighpriceofluxuryisforitsquality.”withloading0.831.Wecaninterpretthat people expect the luxury items to be expensive if they are qualified enough. Themostimportantpointisthatqualityandlonglastingarethereasons,peoplebuyluxuryitemsandliketospendhighpricesfor.
Factor2:UniquenessValue
Peoplebelievethatluxuryitemsarenotproducedinmassnumbers.Rarityanduniquenessarevalues that people pay attention to when buying luxury items. In this factor the highestloading,belongto”X46:Aluxuryproductshouldnotbesoldinnormalstores.”Sotheserviceandatmospherepeopleexpecttobuyluxuryproductsareimportantaswell.
The second place belongs to X39 in which people believe “luxury items cannot be mass-produced”.
Factor3:UsabilityValue
The biggest factor loading in this factor belongs to variable number 23,which says that thequalityoflifewillincreasebyusingluxuryitems.However,variableX15showsworkingwomenbelieve“havingluxuryitemsisnotnecessary”butanywaytheydesiretohaveit.
Factor4:Conspicuousnessvalue
TheitemthatcanbestdescribethisfactorisX37,whichhasthehighestloadinginthisfactor.Itsaysthatworkingwomenlikeotherstoknowthepriceoftheluxuryitemstheybuy.Inaddition,X47,showingoffisthesecondstrongestvariablethatcandescribethisfactor.TheinterestingpointisthatinWiedmanresearchinvaluebasedsegmentationofluxuryconsumptionbehavior(Wiedmann,Hennigs,& Siebels, 2009), it seems that conspicuousness valuewasnotproved.Nevertheless,inIran,mostofthewomenliketoshowoffwiththeirluxuryitemsanditisapartoftheculture.TheideathatwasprovedbyIranianstudyofTeimourpourandHeidarzadeh,intheirstudytheconspicuousnessvaluehasthehighesttotalvarianceinFactoranalysisandthehypothesiswasproved.See(HeidarzadehHanzaee,Teimourpour,&Teimourpour,SegmentingConsumersBasedonLuxuryValuePerceptions,2012)
37
Factor5:HedonicValue
X34,Self-satisfactionisthemostimportantvalueinthisfactorbyloadingnumberof0.805.
Itindicatesthatworkingwomensaysluxuryitemsshouldgivethemself-satisfactionandmakethemhappyandshouldmakethemfeelmuchbetterintheirlife.
Factor6:Materialisticvalue
This factor shows that Iranianworkingwomenbelieve the rateof people’s assets shows thedegreeofsuccessinlife.Themoresomeonehas,themoresuccessheorshehasgained.Thisvaluehasthehighestfactorloading,0.733,whichisfollowedbyX23,describingthatpeopleareassessedbytheamountofmoneytheyhave.Thisfactorshowsthatpeoplebelievelifewouldbebetterbyhavingmoneyandmoneybringshappiness.
Herewetakealookatourhypothesisagain:
Hypothesis1TheperceivedlevelofusabilitywithregardtosuperiorfunctionalvalueisasuitablecriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis2TheperceivedlevelofqualitywithregardtosuperiorfunctionalvalueisasuitablecriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis3TheperceivedlevelofuniquenesswithregardtosuperiorfunctionalvalueisanappropriatecriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis4Theperceivedlevelofself-identityintermsofsuperiorindividualvalueisasuitablecriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis5 TheperceivedlevelofhedonismtowardaluxurybranditemwithregardtosuperiorindividualvalueisarightfulcriterionforvaluebasedsegmentationofIranianworkingwomen.
Hypothesis6Thelevelofrequirementtomaterialsintermsofsuperiorindividualvalue,isarightfulcriterionforvaluebasedsegmentationofIranianemployedwomen.
38
Hypothesis7TheperceivedvalueofconspicuousnessofaluxurybranditemintermsofsuperiorlevelofsocialvalueisanappropriatebasisofvaluebasedsegmentationofIranianworkingwomen.
Hypothesis8TheperceivedlevelofprestigeofaluxurybranditemintermsofsuperiorlevelofsocialvalueisanappropriatebasisofvaluebasedsegmentationofIranianworkingwomen.
Aswecansee,theresultoffactoranalysisissixfactorssothehypothesisnumber1,2,3,5,6and7cannotberejectedandhypothesis4and8arerejected.
39
4.3.2. ClusteringClusteranalysisisamultivariateapproachbywhichobjectsareunitedintogroupsinawaythatobjectsofeachgrouparesimilartoeachotherandmeanwhiletheydifferfromobjectsofothergroups.Clusteringcanbeperformedinhierarchicalandnonhierarchicalprocedures(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011).
Inorder toachieverelevantsegmentationof theresulted factor in this study,hierarchical,k-mean and two-Step clustering process has been conducted based on the factors’ accepteditems.Thenumberofnot-accepteditemswhichwereremovedfromclusteringprocesswas11andalsothe14itemsbelongtoself-identityandprestigevalueswerecompletelyeliminatedasthey could not be proved as an effective perceived value in factor analysis process for theirfactor loading was less than 0.3 (Hair JR, Wolfinbarger Celsi, H.Money, Samouel, & J.page,2011).Thereforethetotalnumberofvaluesbywhichclusteringwasperformedis28outof53items.
4.3.3. NumberofClustersBefore partitioning the data using k-mean clustering method, a hierarchical clustering withwardmethodwasperformedinordertoidentifyingthenumberofclustersandmaximizingtheheterogeneityaswellashomogeneitywithinandbetween them(Hair JR,WolfinbargerCelsi,H.Money, Samouel,& J.page, 2011). Thenumberof clusterswas estimated3 and this resultwaseitherconfirmedbytwo-stepclusteringmethod.
Thenak-meanclusteringhasbeenconductedin2stepsinwhichthenumberofsegmentswassetas3and2respectively.Despite in3clustersstagethepopulationofoneclustercontainsonly1percentofthesample;itwasdecidedtokeepitandtoaccept3clusters.
According to the result of final k-mean clustering (Table 4.1-1), the most important valueswhichplayedmainrolesincharacterizingtheclustersare“Conspicuousness”and“Uniqueness”withinwhichthehighestFratiosexist(X37,X44,X47),(X5,X17,X46).
Fratioindicatestheresultofcomparing“Variancebetweengroups(VB)”and“Variancewithingroups(VW)”anditisusedinANOVAtestwhichisusedasapartofclusteringprocedure.Thehigher F ratio, themore significant differences between the groups andmost probably nullhypothesiswillberejected(HairJR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011).
InasimplewordtheabovedefinitionforFratiocanbeshownasbelow:
F=Variancebetweengroups(VB)/Variancewithingroups(VW)
Table4.1-1also reveals thedetailed informationaboutcreatedclusters, their relatedmeansandFvalues.Theaverageforeachfactor’sgroupofmeansiscalculatedineachclusterinordertosimplifyreadingthewholetable.Theresultsinthistablearegoingtobemoredescribedandinterpretedinsection4.4(introducingtheclusters).
40
4-1:ClustersStatistics
Value ItemsUsedinClustering
Cluster1
(n=392)Mean
Cluster2
(n=11)Mean
Cluster3
(n=569)Mean
FactorLoading F Sig.
Conspicuousness
X27:Itendtobuytheluxuryitemthatisanewstyle. 3.48 3.64 2.98 0.315 26.431 .000
X37:IlikepeopletoknowthepriceofthegoodsIbought. 2.85 4.36 1.66 0.714 270.863 .000
X44:Havingluxuryproducts,IcanshowpeoplethatIamrich.
3.18 4.09 1.92 0.513 232.685 .000
X47:ItendtoshowofftheluxuryitemIbought. 3.08 3.09 2.00 0.692 178.797 .000
X52:Others’opinionsaboutmylookareanimportantissueforme.
3.70 4.55 3.26 0.607 29.349 .000
3.26 3.28 1.97
Hedonic
X19:Luxuryproductsshouldmakemehappy. 4.13 4.27 3.22 0.471 113.057 .000
X31:Luxuryproductsshouldmakemefeelbetteraboutmyself.
3.99 3.91 2.95 0.428 137.666 .000
X34:Luxuryconsumptionshouldbringmeself-satisfaction.
4.47 3.64 3.94 0.805 43.901 .000
4.19 3.94 3.37
Materialistic
X14:Ithinkthatmylifewouldbebetterbybuyingluxuryproducts.
2.73 2.73 1.75 0.315 164.427 .000
X26:Thecriterionforassessingpeopleistheamountofmoneytheyhave.
3.00 3.36 2.46 0.670 29.211 .000
X36:Ibelievethatmoneybringshappiness. 3.77 4.27 2.90 0.593 74.906 .000
X48:Therateofpeople’sassetsinsocietyshowsthedegreeofsuccessinlife.
3.13 3.73 2.42 0.733 44.918 .000
3.16 3.52 2.38
Quality
X1:Oneofthemaincharacteristicsofluxuryproductsisquality.
3.95 2.09 3.55 0.710 35.170 .000
X6:Luxuryproductsarelonglasting. 3.47 4.00 3.19 0.809 11.568 .000 X18:Themainreasonforbuyingluxurygoodsistheirquality.
3.49 2.64 2.95 0.844 25.291 .000
X30:Youalwaysshouldpaymoreforthebestquality. 3.96 2.36 3.38 0.412 44.830 .000 X33:Luxuryitemsareexpectedtoobservehighstandards. 4.58 3.91 4.36 0.323 14.412 .000 X38:Ithinkthatthehighpriceofluxuryisforitsquality. 3.37 2.64 2.86 0.831 26.101 .000
3.80 2.94 3.38 Uniqueness
X5:Ienjoyhavingthingsthatothersdonothave. 3.58 2.55 2.09 0.325 207.344 .000 X11:Icanspendlotsofmoneyforrarethings. 3.00 4.00 2.13 0.522 98.315 .000 X17:Fewpeopleownatrueluxuryproduct. 3.23 1.36 1.99 0.730 171.147 .000 X32:Inmyopinion,productsthatallcanaffordarelessvaluable.
3.19 1.09 2.10 0.482 168.085 .000
X39:Luxuryproductscannotbemassproduced. 3.50 1.91 2.63 0.760 96.465 .000 X45:Idonotliketobuyproductsthatthemajorityofpeoplecanaffordtobuy.
3.08 2.64 2.03 0.435 172.897 .000
X46:Aluxuryproductshouldnotbesoldinnormalstores. 3.93 2.00 2.59 0.736 219.680 .000
3.21 2.22 2.22
Usability
X3:Itendtouseluxuryitems. 4.15 1.55 3.45 0.477 125.186 .000
X15:Inmyopinion,havingluxurygoodsisnotnecessary. 2.80 2.00 1.82 0.751 147.045 .000
X23:Usingluxuryproductsincreasesthequalityoflife. 3.32 4.00 2.40 0.624 93.563 .000
3.42 2.51 2.56
41
Discriminantanalysisisdefinedasastatisticalprocedurebywhichanindividual’smembershipto group can be predicted by using two or more metric independent variables by which itexamines difference in groups; or it finds discriminant functions (linear combination ofindependent variables) that determines the differences between group means (Hair JR,WolfinbargerCelsi,H.Money,Samouel,&J.page,2011).Discriminantanalysiswasdoneinthisresearch in order to verify cluster grouping results; as one of the several purposes of thisanalysis is classifying cases to groups and statistical significance tests confirm how well thefunctionsseparatedthegroups.
Toperformdiscriminantanalysis in this study, thenumberofdetectedclusters calculatedbyclustering procedure, used as grouping variable which was 3 (dependent variable indiscriminantanalysis).Thusthedefinedrangeforgroupingvariablewasset to1asminimumand 3 asmaximum. The number of discriminant functions is always equal to the number ofdependent variables minus one; so with regard to the number of dependent variables, thenumberofdiscriminantfunctionsis2inthisstudyasitisstatedinthetable4.1-2either.
Likeothersimilarstatisticalanalyses,thefirstjobtobedoneindiscriminantanalysisistocheckhow much significant is the statistical relationship between dependent (clusters) andindependentvariables.Table4.1-2istheanswertothisquestionforitpresentsasummaryofcanonicaldiscriminantfunctions.
Table 4.1-2 shows a summary output of discriminant analysis in this study. According to thenumberofintroducedgroups(3),twodiscriminantfunctionsarepresented.BasedondefinitiontwovalueofEigenvalueindicatedhowwellafunctiondifferentiatesthegroups,thehigherthisvalue, the better the identification is (BӦKEOĞLU& BUYKOWTURK, 2008). Eigenvalues tableshowsthatthefirst functionaccountsfor84.2%ofdiscriminatingabilityofthediscriminatingvariablesandthesecondfunctionaccountsfor15.8.Thisalsocanbestatisticallycalculatedandre-confirmed by achieved eigenvalues shown in the table. Canonical correlation shows thecorrelationbetweengroupsandfunctions.Thesigvalue(.000)ofWilk’sLambdathatevaluatesthestatisticaldifferenceorsignificancebetweengroups(HairJR,WolfinbargerCelsi,H.Money,Samouel, & J.page, 2011) shows that Wilk’s lambda is highly significant and therefore theresultscanbe trusted for furtherdiscussions.Thevalueof0.179 inWilks’Lambda’s first rowindicatedthat17.9%of totalvariability isnotexplainedwhich isacceptable for this research.Thevalueofchi-squarechecksifthevalueofcanonicalcorrelationofeachfunctioniszeroorittests the null hypothesis for each function that assumes each function has no discriminantability (idre, 2015), which is rejected in this study according to the values of canonicalcorrelationandChi-square.
42
Table4-2:SummaryofCanonicalDiscriminantFunctions
Eigenvalues
Function Eigenvalue %ofVariance Cumulative% Canonical
Correlation
1 2.700a 84.2 84.2 .854
2 .508a 15.8 100.0 .580
a.First2canonicaldiscriminantfunctionswereusedintheanalysis.
Wilks'Lambda
Testof
Function(s)
Wilks'Lambda Chi-square df Sig.
1through2 .179 1642.357 56 .000
2 .663 392.315 27 .000
Table4.1-3containsthecentroid(mean)ofeachgroup’sdiscriminantscores.Accordingtothecentroidsforfunction1,cluster1and2havepositivevalueswhilecluster3hasnegativevaluewhichmeansfunction1separatedcluster3fromclusters1and2;butforfunction2,cluster1and3havepositivevaluesandcluster2hasnegativevalueand itmeans that function2hasseparatedcluster2fromtheothers.
Table4.1-3:FunctionsatGroupCentroid
ClusterNumberFunction
1 2
1 1.962 .159
2 1.474 -6.619
3 -1.380 .019
Table 4.1-4 presents correlation between all the independent variables and discriminantfunctions.Bytheprovidedinformationinthistableitcanbesaidthatwhichvariable(s)ismoreimportantforeachfunction.Inthefollowingtablethestareditems(rowsfromX37toX6)areimportantforfirstdiscriminantfunctionwhilevariablesfromX3toX33areconcernedmorebytheseconddiscriminantfunction.
43
Table4-4:StructureMatrix
Items Function1 Function2X37:IlikepeopletoknowthepriceofthegoodsIbought. .436* -.299X44:Havingluxuryproducts,IcanshowpeoplethatIamrich. .415* -.176X5:Ienjoyhavingthingsthatothersdonothave. .395* .109X46:Aluxuryproductshouldnotbesoldinnormalstores. .394* .259X47:ItendtoshowofftheluxuryitemIbought. .370* -.028X45:Idonotliketobuyproductsthatthemajorityofpeoplecanaffordtobuy.
.363* .050
X14:Ithinkthatmylifewouldbebetterbybuyingluxuryproducts. .354* -.025X17:Fewpeopleownatrueluxuryproduct. .346* .240X32:Inmyopinion,productsthatallcanaffordarelessvaluable. .333* .304X15:Inmyopinion,havingluxurygoodsisnotnecessary. .332* .112X31:Luxuryproductsshouldmakemefeelbetteraboutmyself. .324* -.011X19:Luxuryproductsshouldmakemehappy. .293* -.044X11:Icanspendlotsofmoneyforrarethings. .265* -.166X23:Usingluxuryproductsincreasesthequalityoflife. .263* -.112X39:Luxuryproductscannotbemassproduced. .254* .220X36:Ibelievethatmoneybringshappiness. .236* -.085X48:Therateofpeople’sassetsinsocietyshowsthedegreeofsuccessinlife.
.181* -.086
X34:Luxuryconsumptionshouldbringmeself-satisfaction. .175* .128X26:Thecriterionforassessingpeopleistheamountofmoneytheyhave.
.147* -.058
X27:Itendtobuytheluxuryitemthatisanewstyle. .142* -.031X52:Others’opinionsaboutmylookareanimportantissueforme. .138* -.135X38:Ithinkthatthehighpriceofluxuryisforitsquality. .136* .089X18:Themainreasonforbuyingluxurygoodsistheirquality. .132* .098X6:Luxuryproductsarelonglasting. .087* -.082X3:Itendtouseluxuryitems. .239 .453*X1:Oneofthemaincharacteristicsofluxuryproductsisquality. .112 .275*X30:Youalwaysshouldpaymoreforthebestquality. .159 .218*X33:Luxuryitemsareexpectedtoobservehighstandards. .087 .137*
Table 4.1-5 presents a summary of classification procedure results including number andpercentofitemsthatareclassifiedcorrectlyindiscriminantanalysisprocess.Thisoutputshowsthat96.4%originalandcrossvalidatedgroupedcasesclassifiedcorrectly.
44
Table4.1-5:ClassificationResults
ClassificationResultsa,c
ClusterNumberofCase PredictedGroupMembership Total
1 2 3
Original Count 1 371 0 21 392
2 0 11 0 11
3 14 0 555 569
% 1 94.6 .0 5.4 100.0
2 .0 100.0 .0 100.0
3 2.5 .0 97.5 100.0
Cross-validatedb Count 1 371 0 21 392
2 0 11 0 11
3 14 0 555 569
% 1 94.6 .0 5.4 100.0
2 .0 100.0 .0 100.0
3 2.5 .0 97.5 100.0
a.96.4%oforiginalgroupedcasescorrectlyclassified.
b.Crossvalidationisdoneonlyforthosecasesintheanalysis.Incrossvalidation,eachcaseis
classifiedbythefunctionsderivedfromallcasesotherthanthatcase.
c.96.4%ofcross-validatedgroupedcasescorrectlyclassified.
45
4.4. IntroducingtheClustersAccording to theoutput of k-mean clustering anddiscriminant analysis, the resulted clustersseem close regarding the nature and effective parameters; however by takingmore preciselooktotheresults,theclusterscanbedefinedasfollows.
4.4.1. Cluster1-RationalHedonistsThisclustercontains392peopleor40.3%ofthewholesampleinwhichthemeanageis31.28andthemajorityofwomenaresingleandtheincomeisnormallyhigherthanRLS10,000,000whichisconsideredasaveragerevenue(Table4.1-6).
4-6:Cluster'sDemographicInformation
Variable N % Cluster1 Cluster2 Cluster3Age 14-65 972(Mean=32.10) 100 39.3%
(Mean=31.28)1%(Mean=28.09)
59.7%(Mean=32.74)
MaritalStatus Single 418 44.8 21.6% 0.1% 23.1%Married 554 55.2 18.2% 0.7% 36.3%Education UnderDiploma 8 0.7 0.2% 0 0.5%Diploma 92 8.4 4.1% 0.1% 4.2%BachelorDegree 514 52.7 20% 0.4% 32.3%MasterDegreeorabove
358 38.2 15.4% 0.3% 22.4%
Income(Rials) >=3,000,000 29 3.0 0.5% 0 2.5%=6,000,000
31 3.1 1.0% 0.1% 2.0%
=10,000,000
96 9.8 5.0% - 4.7%
=20,000,000
212 20.8 8.3% - 12.6%
=50,000,000
325 31.8 13.1% 0.4% 18.3%
46
With regard todiscriminant aswell asdescriptive information, thehighest averagemeans inthisclusterbelongtoHedonic,QualityandUsabilityvaluesrespectively.Thisresult identifiesthatbeside self-pleasure, the functionalityof luxuriousgoods is important to theconsumers.Accordingtotheitems,highestmeansineachvaluesuchas“Luxuryconsumptionshouldbringme self-satisfaction” and “Luxury items are expected to observe high standards” it can beinferred that thequality (or functionality) of luxury good should satisfy the consumer in thisclustertogiveherthefeelingofenjoyandhappinessofluxuryconsumption.
4.4.2. Cluster2-OstentatiousIndividualistsThisclusteristhesmallestoneregardingthesizewhichcomprises11peopleoronly1percentofthewholesampleandregardingthedemographicinformationincludesyoungerwomen.Themeanageinthiscluster is28andthemajorityofwomenaremarried.Theaverageincomeismorethan20,000,000Rialswhichisabsolutelygoodrevenue.
Regarding the discriminant and descriptive information, “Hedonic”, “Materialistic” and“Conspicuousness” values are themost important perceived values in this cluster. Similar tofirstcluster,hedonicvaluehasahighvalueinthiscluster,speciallytheitem“Luxuryproductsshould make me happy” carries a high mean, but the means more than 3 and 4 in“Conspicuous” and “Materialistic” values aremore considerable: “I like people to know thepriceofthegoodsIbought”,“Others’opinionsaboutmylookareanimportantissueforme”and“Having luxuryproducts, I canshowpeople that Iamrich”, “Ibelieve thatmoneybringshappiness” and “The rate of people’s assets in society shows the degree of success in life”.From all above items it can be inferred that consumers in this segment who are rich andyounger women care more about show-off in their social life and beside that luxuryconsumption brings them self-pleasure and happiness. The high means in materialisticdimensionalsoemphasizeonstrengthofindividualdimensionwhichindicatesthatwomencarelessaboutthequalityorfunctionalityofluxurygoodsastheyprefertoenjoyandshow-offbyspendingmoneytohaveluxurygoods.
4.4.3. Cluster3-QualitySeekers
This clusterwhich is the biggest one regarding the size, includes 569 (59.7%