Reuse Intention of Third-Party Online Payments

15 Pages·2016·1.65 MB·English

sustainability
Article
Reuse Intention of Third-Party Online Payments:
A Focus on the Sustainable Factors of Alipay
YongrokChoi*andLiliSun
DepartmentofInternationalTrade,InhaUniversity,Inharo100,Nam-gu,Incheon402-751,Korea;
[email protected]
* Correspondence:[email protected];Tel.:[phone];Fax:[phone]
AcademicEditor:GiuseppeIoppolo
Received:24December2015;Accepted:29January2016;Published:4February2016
Abstract: An anonymous transaction environment and the advantage of virtual property have
resultedintrustplayinganimportantroleintherapidgrowthofonlineshoppinginChina. Tosatisfy
thistrustissue,Alibaba(China)Co.,Ltd. (Hangzhou,China)inventedAlipay,thelargestthird-party
onlinepaymentservice. Usingastructuralequationmodel(SEM),thispaperattemptstodetermine
whetherAlipay’sservicequalityfactorsaretrulysustainable. Theresultsindicatethatonlytwoof
fivefactors—convenienceandsecurity—aresignificantlymediatedbythesustainableperformance
of customer satisfaction as a mediator. The other three factors—usefulness, responsiveness and
economy—wererejectedfortheroleofcustomersatisfaction,eveniftheyareacceptedregardingthe
directeffectonreuseintention. ThisresultimpliesthatChinesewebcompaniesneedtomakegreater
effortsnottoensureinitialsuccess,butinsteadtoensuresustainableperformance.
Keywords:thirdpartypaymentservice;structuralequationmodeling;reuseintention;China;Alipay
1. Introduction
Inrecentyears,Chinahasexperiencedsignificantachievementsinallindustriesandparticularly
e-commerce. Alibaba, the leader of the Chinese e-marketers, invented Kwangkunjie (Single’s Day),
one-dayonlinesaleonNovember11thatissimilartoblackFridayafterThanksgivingintheUnitedStates.
Onthisdoubleelevenday,Alibaba’stotalturnoverfromallofitsplatformshasbeen91.2billionYuan
(US$13.68 billion), which involved 232 countries and 467 million logistics orders. These figures
representtheworldrecordforsingle-dayonlinetransactionvolume[1]. Alibaba’ssignificantsuccess
onKwangkunjiemayhaveresultedfromthemostsuccessfulintermediationofAlipay,thethird-party
online payment system that it launched in China in 2004. Alipay charges no transaction fees and
providesanescrowservicethatenablesconsumerstoverifytheirsatisfactionwiththegoodsthey
purchased before releasing money to the seller. This service is critical because it gives consumers
guaranteedintermediationbyanauthorizedthirdparty,anecessitygivenChina’sweakconsumer
protectionlawsandregulations. Itspro-activenessmakesAlipayauniquepayment-relatedservice;
for example, relative to U.S.-based Paypal’s simple third-party payment service, Alipay provides
guarantees for transactions. Alipay also provides enhanced online payment functions through
integrateddiversefunctions,suchasformobilegames,moviesandothers. Incontrast,SamsungPay
inKoreasimplyprovidesflexiblepaymentservices. Alipayismuchmoreattractivetocustomersgiven
itsintegratedone-stopmulti-user-friendlyservices.
China’sthird-partyonlinepaymentvolumeincreasedtoapproximatelyeighttrillionYuan(US$1.2trillion)
in2014, withgrowthof50.3%comparedto2013. Alipayhasledthisrapidlyincreasinggrowthin
third-partyonlinepayments,witha49.6%marketshareor3.87trillionYuan,asshowninFigure1[2].
Morethan300millionpeopleintheworlduseAlipayforpaymentservicesatTaobaoandTmall,as
Sustainability2016,8,147;doi:10.3390/su8020147 www.mdpi.com/journal/sustainability
Sustainability2016,8,147 2of15
Sustainability 2016, 8, 147 2 of 15
wellaTsaaotbmaoa annydo Tfflmianlel, satso wreesll. aAs laitp maya’nsys otrfoflninge esstocrreosw. Asluippapyo’sr stteronnagb leesdcroAwli bsuapbpaotrot ernapabidleldy Ainlicbraebaas eits
onlinetom raarpkiedtlys hianrcereianseC hitisn ao,nalinnde omthaerkrefto lslhoawree risne fCfehcintiav,e laynden oatbhleerd fcoollnoswuemrse resffiencttihvieslyl aergnaebcloedu ntry
tointecgornastuemoenrsli nine tshhiso plaprgineg cowuinthtryo ntoli ninetepgaryatme eonntlsinaen sdholopgpiinstgi cwsitthhr oonulginhe opnaey-msoeunrtsc ea,nmd uloltgii-scthicasn nel
through one-source, multi-channel transactions. In 2013, Alipay launched a financial product
transactions. In2013,AlipaylaunchedafinancialproductplatformcalledYu’ebaothatenablesAlipay
platform called Yu’ebao that enables Alipay users to deposit any remaining money in an account and
userstodepositanyremainingmoneyinanaccountandearninterest. Thisinterest-yieldingnew
earn interest. This interest-yielding new invention by the third-party payer made Alipay the leader
inventionbythethird-partypayermadeAlipaytheleaderinthise-businessnewfrontier.
in this e-business new frontier.
Currently,Alipayprovidesboththird-partyescrowguaranteesononlinetransactionsandintereston
Currently, Alipay provides both third-party escrow guarantees on online transactions and
theunusedportionofanAlipayaccount.Inadditiontothesedirectsupportservicesforonlineshopping,
interest on the unused portion of an Alipay account. In addition to these direct support services for
manyomnolirnee asphpolpicpaintigo,n msaenxyis tmfoorreA alpippaliycautsioenrss wexhisot ffroerq Auelinptalyy uvsiseirtsd wivheor sferegqaumenetlayn vdisoitth deirveernstee rgtaaminem ent
sitesaanndd footrhuers eernsteerntagiangmeednitn siitnefso armnda tfiovre ucsoemrs menugnaigtyeda cinti vinitfioersm.Aatilvipe acyombemguannitays aacntivointileisn. eAslhipoapyp ing
suppobretegranb yasb aenin ognalinnee sschroopwpipngro svuipdpeorratenrd bbye bceaimnge aanm eusclrtoi-wch parnonveidlethr iarndd-p baerctaymoen lai nmeubltain-ckhianngnpelo rtal
site. Tthhierdc-opnatritny uoinnlginseu bcacneksisngo fpAorltiapla syitec.e rTthaei ncloynatirniusiensgf rsoumccetshs eopf rAolaipcatiyv eceortraginalnyi zaaritsieosn farlocmu ltthuer eof
proactive organizational culture of Alibaba (China) Co., Ltd., which is akin to reading users’ minds
Alibaba(China)Co.,Ltd.,whichisakintoreadingusers’mindsandpullingtheirdemandinstead
and pulling their demand instead of engaging in pushing-oriented forceful promotional marketing.
ofengaginginpushing-orientedforcefulpromotionalmarketing. Alipay’slong-termperformance
Alipay’s long-term performance could be sustainable given its user-friendly, performance-oriented
could be sustainable given its user-friendly, performance-oriented marketing efforts, because the
marketing efforts, because the value creation based on network management with diverse
valuecreationbasedonnetworkmanagementwithdiverseparticipantsofacertainproductiskeyto
participants of a certain product is key to sustainable e-business management [3].
sustainablee-businessmanagement[3].
Yeepay, 3.2%
IPS, 2.7%
Others, 1.6%
ChinaPnR, 5.2%
Alipay, 49.6%
99Bill, 6.8%
Chinaums,
11.4%
Jaibutong,
19.5%

Figure1.Third-partyonlinepaymentmarketinChina(2014).Source:China’sthird-partyonlinepayment
markeFtiginur2e0 114. T[4h]i.rd-party online payment market in China (2014). Source: China’s third-party online
payment market in 2014 [4].
However, as an online payment service, Alipay may require more sustainable operations for
However, as an online payment service, Alipay may require more sustainable operations for
successful performance, because it is in the initial stages of the market, and the rapidly changing
successful performance, because it is in the initial stages of the market, and the rapidly changing
environmentinChinacannotguaranteethatinitialluckwillcarryforwardtothelongterm.Inparticular,
environment in China cannot guarantee that initial luck will carry forward to the long term. In
webmpaarrtkiceutilnarg, wfoerba mnaornkelitnineg pfoary mane onntlisneer vpiacyemisenatl ssoerbviacsee ids aolnsos ubassteadin oanb lseusrtealiantaibolnes rheilpatsiownsithhipus sers.
Accorwdiinthg utsoerBsh. aAtctcaocrhdeirnjge eto( 2B0h0a1t)ta,cthheersjeues (t2a0i0n1a)b, ltehep seurfsotarimnaabnlec epeorffowrmebansceer voifc wesesbh soeurvldicenso sthboeulbda sed
onsimnoptl eb,ei nbastsaedn toann esoimupslein, tinensttainotnasneinoutsh ientsehnotirotntse rinm t,hbe usthoornt treerums,e biuntt eonnt iroeunssef ionrtetnhteiolnosn fgort ethrme [5].
Chinelsoengc utesrtmom [5e]r.s Cahrieneqsue iccuksltyomaderosp atrien gqu“imckulyl taicdhoapntinngel ”“mshulotipcphainngnebl”e hshaovpioprintgh baethiasvciorer atthiantg isb oth
creating both opportunities and challenges for social e-commerce brands/companies, retailers and
opportunities and challenges for social e-commerce brands/companies, retailers and companies.
companies. Taobao and its payment system Alipay certainly provide the most appropriate marketing
TaobaoanditspaymentsystemAlipaycertainlyprovidethemostappropriatemarketingplatform
platform in China [6]. As previously mentioned, the U.S.-based PayPal does not provide escrow
in China [6]. As previously mentioned, the U.S.-based PayPal does not provide escrow services,
services, and Samsung Pay in Korea provides very flexible access points for payments, but no other
andSamsungPayinKoreaprovidesveryflexibleaccesspointsforpayments,butnootherservices.
services. Compared to these third-party online payment services, Alipay provides diverse, yet
Compinarteegdrattoedt hseersveictehsi rfdor- pmarutlyti-ochnalninneel puasyerms.e Hntowseervveirc, ems,oArel iips anyotp arlowvaiydse sbedttievre, rasned, ya eptaiynmteegnrta ted
servicesformulti-channelusers. However,moreisnotalwaysbetter,andapaymentservice’score
compe tenceshouldmaintainanappropriate“reach”toenablereusebyusers[7]. Thepurposeofthis
researchistodetermineAlipay’ssustainablefactorsasathird-partyonlinepaymentserviceprovider.
Toevaluatethesustainablefeasibilityoftheonlinepaymentservice,thispaperempiricallyteststhe
Sustainability2016,8,147 3of15
relationshipbetweenthewebmarketingmixofAlipayandthereuseintentionofusers. However,
alinkisalwaysmissingbetweenwebmarketingmixanditsperformance[8]. Becauseofthismissing
link,previousstudiesconcludethatinitialtrustinthewebmarketingmixisnoteffectiveenough[6],
whereasothersinsistthatthewebmarketingmixissignificant[8]. Therefore,thisresearchempirically
teststheroleofconsumersatisfactionasamediatorbetweenthewebmarketingmixandthereuse
intention of consumers on third-party online payment systems, with an emphasis on the unique
connectionalcharacteristicsofAlipay.
2. TheoryandHypotheses
2.1. CharacteristicsofThird-PartyOnlinePayments
Third-partyonlinepaymentservicesarewellknownasescrowservices,becauseanescrowgenerally
refers to money held by a third party on behalf of the transacting parties. These services are based
onacontractualarrangementinwhichathirdpartyreceivesanddisbursesmoneyordocumentsfor
theprimarytransactingparties, withthedisbursementdependingonconditionsagreedtobythese
parties. AstheAlipayserviceindicates,thisprocesssometimesmeansthatanaccountisestablished
byabrokertoholdfundsonbehalfofthebroker’sprincipaluntiltheconsummationorterminationof
atransaction.Ingeneral,mostonlinepaymentservicesprovideanobjectivetransferofmoneyalongthe
linesofacommercialtransaction. Third-partyonlinepaymentsworkbyplacingmoneyinthecontrol
ofanindependentandlicensedthirdpartytoprotectbothbuyerandsellerinatransactionoverthe
Internet.Whenbothpartiesverifythatthetransactionhasbeencompletedperthesetterms,themoney
isreleased. Ifatanypointadisputeoccursbetweenthepartiesinthetransaction,theprocessmovesto
disputeresolution. Theoutcomeofthedisputeresolutionprocessdetermineswhathappenstothe
moneyinescrow.
AsthemostrepresentativeonlinepaymentserviceinChina,Alipayhassuccessfullyoperated
athird-partyonlinepaymentserviceforonlineshoppingatTaobao,itsbrotherwebsite. Becauseonline
paymentsarewebservices,Alipay’ssuccessfuloperationmayresultfromtheappropriateutilization
of the web marketing mix. Thus, the core competence of web services, such as online payments,
shouldbebasedonthecapabilitytocreateandmaintainrelationalmarketing,whichisdifferentfrom
traditionalmarketing[9].
Similartootherservices,anonlinepaymentservicecouldbeevaluatedthroughitsservicequality.
Ingeneral,servicequalityisdefinedastheintegratedcapacityfortheconsumertobesatisfiedwiththe
expectedservicepackagelevelanditsrelatedfunctions[10]. Parasuramanetal. (2005)definee-service
qualityastheextenttowhichawebsitefacilitatesefficientandeffectiveshopping,purchasingand
delivery[11]. WolfinbargerandGilly(2003)definee-servicequalityasthebeginningtotheendof
atransaction,includinginformationsearch,websitenavigation,order,customerserviceinteractions,
deliveryandsatisfactionwiththeorderedproduct[12]. Inthesedefinitions,ane-servicehasamuch
broader spectrum than traditional or general service quality. In particular, Wilfinbarger and Gilly
(2013)arguethate-servicequalityshouldbeself-sufficientinfulfillingtheproceduralsetofanonline
transaction. Becausetheonlinepaymentserviceispartofanonlinetransaction, theevaluationof
e-servicesinthisresearchfocusesontheextenttowhichwebsitescreateandmaintainrelationships
throughusersatisfaction. Therefore,theextentofthecapacityforwebsitestocreateandmaintain
relationshipswillbeexaminedingreaterdetail.
2.2. OnlinePaymentServiceQualityHypotheses
Ingeneral,servicequalityrepresentsthecomparativedegreebetweenexpectationsandperformance.
Generally, customers tend to compare the service they experienced with the service they expected.
Unfortunately,theexpectedserviceandtheperceivedservicearesometimesnotequal,thuscreating
agapthatmayresultinacomparativelevelofservicequality.Parasuramanetal.(1988)suggestedthatthe
fivedimensionsofthisgaparesummarizedintheso-calledRATERmodelasfollows:Responsiveness,
Sustainability2016,8,147 4of15
Assurance(security),Tangibles,Empathyandservice,Reliability(convenience)[13]. However,these
determinantsareofflinebased.Thus,Blutetal.(2015)conductedameta-analysisofstudiesone-service
qualitypublishedduring2000–2014andfoundfourdeterminantsofe-servicequality: websitedesign,
fulfillment(usefulness),customerservice(convenience)andsecurity/privacy[14]. LiaoandCheung
(2002)empiricallytestede-bankingservicesusingsevendeterminants: accuracy,security,transaction
speed,userfriendliness,userexperience(usefulness),userinvolvementandconvenience. Theyfound
thatmostofthesedeterminantsarestatisticallysignificant,whereassomedeterminantsdependon
otherfactors,implyingastepwiseorinter-relationalstructureofthedeterminants[15]. Consideringall
oftheseargumentsfromtheperspectiveofathird-partyonlinepaymentservice,wecouldusethe
commonfactorsofusefulness,convenience,securityandresponsiveness. Moreover,Alipayprovides
interestontheremainingmoneyinauser’saccount,aneconomythatmaybeanimportantsustainable
factor. Altogether, as a third-party online payment service provider, Alipay provides five factors
relatedtoobtainingsustainableperformancethroughusers’reuseintention;usefulness,convenience,
security,responsivenessandeconomy.
Unliketraditionalmarketing,thewebmarketingmixemphasizestheinvisiblevalueofrepurchase
orreuseintentiontomeasure“sustainable”performance. Inthiscontext, reuseintentionrefersto
the likelihood that customers prefer to use a particular product service again. Reuse intention is
positivelyrelatedtothefinalpurchase(oruse)probability,whichleadstodirecteconomicbenefitsfor
companies[16–18]. Reuseintentionrepresentsconsumers’subjectivepreferencesforusingaspecific
serviceandrecommendingittotheirfamilyandfriends. EngelandBlackwell(1982)definereuse
intention as the probability of an individual’s desire for a certain action to be implemented [19].
NewmanandWeber(1973)empiricallyprovethatifauserisunsatisfiedwithaservicebecausethe
servicequalityislowerthanhisexpectation,thentheprobabilityofreusemaybemuchlower[20].
Thus,thewebmarketingmixaimsforwebservicequalitytodrivereuseintentiononthebasisof
relationalmarketingontheInternet.Toincreaseconsumers’reuseintention,websitesshouldprovide
easierandmorereliableaccess(convenience)thatisveryinformative,amazingandsuitableforweb
surfing(usefulness);further,theyshouldensurelong-termprofits,suchasinterest(economy),andshould
respondpromptlytousers’requests(responsiveness),alongwithstrongerprotectionforthemoney
transfer(security). Thesefiveservicequalityfactorsshouldbeintegratedintowebsitestocreateand
maintainvalue,becausetheyareinterrelatedwitheachotherasasetofservicequality.Onthebasisof
thesearguments,weproposethefollowinghypothesesforthird-partyonlinepaymentservicequality.
‚ H1: The degree of usefulness of a third-party online payment service has a positive effect on
reuseintention.
‚ H2: Thedegreeofconvenienceofathird-partyonlinepaymentservicehasapositiveeffecton
reuseintention.
‚ H3: The degree of security of a third-party online payment service has a positive effect on
reuseintention.
‚ H4: Thedegreeofresponsivenessofathird-partyonlinepaymentservicehasapositiveeffecton
reuseintention.
‚ H5: The degree of economy of a third-party online payment service has a positive effect on
reuseintention.
2.3. IntermediationRoleofUserSatisfactiononReuseIntention
Even given similar web content, different actual reuse intentions could result from a customer’s
satisfaction.Severalpriorstudiesshowthepositiveeffectsofcustomersatisfactiononreuseintention[21–23].
Oliver[21]reportsthatcustomersatisfactionisapositiveantecedentofauser’sattitudeandakeyfactor
thatinfluenceshisorherreuseintention. However,MorgesonandPetrescu[22]pointoutthatproduct
(service)qualitymaydifferdependingonperceivedperformanceandmaynotresultinpurchase(orreuse)
intention. LiaoandCheung[15]alsofindthatthedeterminantsofservicequalitymayhaveastepwise
Sustainability2016,8,147 5of15
orinter-relationalstructuretoeachother;thus,sometimes,ahigherservicequalitymaynotresult
instrongerreuseintention. Dowetal.[22]analyzetheeffectofcustomersatisfactionthroughe-mail
Sustainability 2016, 8, 147 5 of 15
marketingandfindthateveniftheempiricalresultsgenerallysupportthenotionthatthedeterminants
canbeaunsde dmtaoy enffoetc trievsuellty imn opduerclhea-sme a(oilr urseeurses)a tinistfeanctitoionn. ,Ltihaoe yanfidn dChneousntga t[i1s5t]i caallsloy -fsiingdn itfihacta nthtee vidence
thatspdaemteramffiencatnstso ovfe srearlvlicues qeuraslaittyi smfaacyt ihoanvei na satnepew-misea oilr sinytsetre-rmel.ational structure to each other; thus,
Alsloomfetthimeesse, aa rhgiguhmere snetrsviscue pqupaolirttyt mheayp noosts riebslueltm ini ssstrionnggelirn rkeubsee tiwnteenetniown. eDboswe retv ailc. e[2q2u] aanliatlyyzaen dreuse
the effect of customer satisfaction through e-mail marketing and find that even if the empirical results
intention,andcustomersatisfactioncouldfillthismissinglink. Therefore,webservicequalitymight
generally support the notion that the determinants can be used to effectively model e-mail user
beeffectiveonlywhencustomersarefully(orpartially)mediatedbyperceivedsatisfaction. Onthe
satisfaction, they find no statistically-significant evidence that spam affects overall user satisfaction
basisofthesearguments,weproposethefollowinghypothesesfortheroleofamediatorbetweenweb
in an e-mail system.
servicequaAliltly ofa tnhdesea awrgeubmsietnet’ss spueprpfoorrt mthae npcoessfioblre rmeuissseinign ltinenk tbieotnw.een web service quality and reuse
intention, and customer satisfaction could fill this missing link. Therefore, web service quality might
‚ H6: The degree of usefulness of a third-party online payment service has a positive effect on
be effective only when customers are fully (or partially) mediated by perceived satisfaction. On the
cubsatsoims oefr thsaesteis afargcutimonen.ts, we propose the following hypotheses for the role of a mediator between
‚ H7w:eTb hseervdieceg qreuealiotyf caondn va ewneibesnitcee’s opferafotrhmiradn-cpe aforrt yreounsel iinnetenptaioynm. entservicehasapositiveeffecton
cus tomHe6r: Tsahtei sdfeagcrteieo no.f usefulness of a third-party online payment service has a positive effect on
‚ H8: Tchuestodmegerr esaetiosffacsteiocnu. rity of a third-party online payment service has a positive effect on
 H7: The degree of convenience of a third-party online payment service has a positive effect on
customersatisfaction.
customer satisfaction.
‚ H9: Thedegreeofresponsivenessofathird-partyonlinepaymentservicehasapositiveeffecton
 H8: The degree of security of a third-party online payment service has a positive effect on
customcuesrtosmateirs fsaactitsifoanct.ion.
‚ H1 0: THh9e: Tdheeg dreegereoef oef croesnpoomnsyiveonfeass tohf iard th-piradr-ptyarotyn olinnlienep paayymmeenntt sseervrviciec ehahs aas paospitoivsei teifvfeecte ffect on
customone rcussattoimsfearc stiaotinsf.action.
 H10: The degree of economy of a third-party online payment service has a positive effect on
‚ H11: Thehigherthecustomersatisfactionfromathird-partyonlinepaymentservice,thehigher
customer satisfaction.
thereuseintentionthatresultsfromthepaymentservice.
 H11: The higher the customer satisfaction from a third-party online payment service, the higher
the reuse intention that results from the payment service.
Using the proposed hypotheses, the conceptual model of this study is presented in Figure 2.
Figure2indUisciantge sthteh patrocpuossteodm heyrpsoathtiessfeasc, ttihoen caosnctehpetuinalt emrmodeedl ioaft othrisc ostuulddy aifsf epcrtestehnetesdt eipn wFiigsueraep 2p. roachto
Figure 2 indicates that customer satisfaction as the intermediator could affect the stepwise approach
driving“sustainable”reuseintention. Therefore,weexaminethesehypothesesusingthedirectmodel
to driving “sustainable” reuse intention. Therefore, we examine these hypotheses using the direct
(reuseintentiononly)and/ortheindirectmodel(customersatisfactionasamediator).
model (reuse intention only) and/or the indirect model (customer satisfaction as a mediator).
Service Quality
Usefulness H6 Customer
H1 H7 Satisfaction
Convenience
H8
H2 H9 H11
Security
H3
H10
Reuse
Responsiveness H4 Intention
H5
Economy

Figure 2. Research Framework.
Figure2.Researchframework.
3. Methodology
3. Methodology
3.1. Data
3.1. Data

Fortheempiricalresearch,aquestionnairesurveywasusedtocollectthedata.First,apilotstudy
wasconducted.ThreeexpertswithrichexperienceinAlipaypaymentserviceswereinvitedtoevaluate
Sustainability2016,8,147 6of15
thequestionnaires.Inaddition,10Alipayuserswereinvitedtocompletethequestionnaireasapilottest.
Thequestionnairewasmodifiedonthebasisofthefeedbackreceivedinthepilotstage.Then,afull
questionnairesurveywasconductedfrom15October–12November2015. Atotalof410questionnaires
weresentrandomlytousersofChinesesociale-commercewebsitesbye-mailorthroughmessengers
usinginformationretrievedfromtheYellowPagesforvarioussectorsandprovinces.Excludingunusable
or inappropriate responses and responses from those with no experience with Alipay resulted in
280uniqueandusableresponses(foraresponserateof68.29%). Toanalyzethedescriptivestatisticsof
thedata,theresponsesgiventothequestionswerestoredinanExcelfile. SPSSandAMOSwereused
toevaluatethestrengthoftherelationshipbetweenvariablesandtoachievetheresearchobjectives.
Thesurveysamplerepresentsvarioussectors: collegegraduatesaccountedfor49.3%ofthesample,
whereas respondents who were graduates and higher constituted 28.6% of the sample, implying
thatmostcustomers(78%)familiarwithe-paymentshavearelativelygoodeducationalbackground.
Table1presentsthedemographiccharacteristicsoftherespondents. Allofthesurveyrespondents
reportedtohaveusedane-paymentsystemmorethanonce.
Table1.Demographiccharacteristics.
Characteristic Items Percentage
Male 55.7%
Gender
Female 44.3%
0–20 12.9%
20–25 35.0%
AgeRange 26–30 30.4%
31–40 17.1%
41andolder 4.6%
Highschoolorlowereducation 6.4%
College 15.7%
EducationLevel
Undergraduate 49.3%
Graduate/masterorhigher 28.6%
Lessthanayear 8.9%
1–2year 34.4%
UserExperience
3–4year 41.4%
Morethan5years 15.3%
To address the problem of non-response bias, we compared early and late respondents, as
recommendedbyArmstrongandOverton[24]. The280questionnaireswereclassifiedintotwogroups
onthebasisoftheirfinishedtime. Anindependentt-testwasusedtoidentifyanydifferencesbetween
theanswersoftheearlyandthelategroups. Theresultsindicatednosignificantdifferencesinthe
variousitemsofthesetwogroups,suggestingthatoursampleisfreefromnon-responsebias.
3.2. Measurement
Tomeasureallobservedvariablesinthisstudy,weadoptedwell-establishedmultiple-itemfive-point
Likert scales. These scales ranged from 1 (“strongly disagree”) to 5 (“strongly agree”). The detailed
measurementitemsarepresentedintheAppendix.
3.2.1. OnlinePaymentServiceQualityVariables
Onthebasisofpriorstudies[6,8,9],thefollowinghypothesesweretestedempirically: online
paymentservicequalityenablesconvenientaccessforusers;usefulcontentretainsinterestonsites;
onlineproceduresmakeusersfeelsafeandsecure; andwaitingtimeforaserviceisadequateand
indicatesgoodresponsiveness. Subsequently,ausershouldgaineconomicbenefits,suchasinterest
on the account in the long term, to develop sustainable reuse intention. Thus, online payment
service quality variables (the inputs in the analysis) include five factors: usefulness (three items),
Sustainability2016,8,147 7of15
convenience(fouritems),security(fouritems),responsiveness(threeitems)andeconomy(threeitems).
Thesefivedimensionsweremeasuredusingafive-pointLikertscalerangingfrom“notconsidering
itatall”(1)to“perfectlyagreed”(5)inresponsetoquestionsonthedegreeofsubjectiveperception
regardingthird-partyonlinepaymentservices.
3.2.2. ReuseIntention
Bhattacherjee[25]arguedthatthesuccessfulperformanceoftheservicesystemshouldnotbe
basedonsimple,instantaneoususe,butonsustainablereuseintention. Reuseintentionisdefinedas
thelevelofconsumers’subjectivepreferencesforusingtheservicesagainandforrecommendingthis
servicetofamilyandfriends. Beforecustomersusetheseservices,theyhavespecificrequirementsin
mind. Onthebasisoftheseneedsandafterbrowsingtherelatedinformationgiveninitialexperiences,
customersfilter,evaluateandcomparetheinformation/servicesandsubsequentlymakearational
decisiontoselectthesameonlineserviceagain. Therefore,inthisresearch,weconsiderreuseintention
astheoutputofthemodelasmeasuredbyfouritems.
3.2.3. CustomerSatisfactionasaMediator
Reuseintentionmaydependonthemediatoryroleofcustomersatisfaction.Oliver[21]empirically
shows that an online payment service could affect perceived customer satisfaction, and customer
satisfactionmayaffectusers’attitudes. Finally,users’attitudesdeterminethedegreeofreuseintention.
Therefore,thisstudyhandlesthemediatorofcustomersatisfactionusingfouritems.
Structuralequationmodeling(SEM)isusedtotesttheresearchhypothesesrelatedtothelatent
variables.TheSEMtechniqueinvolvesamultipleregressionanalysis,apathanalysisandaconfirmatory
factortest[25]. UsingSEMtotestthemediationcouldovercomethemeasurementerrorproblems
inthemediatorvariablescore, whichcouldresultindifficultiesrelatedtomodelingcausalityand
possiblereversecausalityinahierarchicalregression[6].
4. EmpiricalResults
4.1. ReliabilityTest
Thegoodness-of-fitofastatisticalmodelrepresentshowwellthemodelfitsasetofvariables.
Goodness-of-fitmeasurestypicallydescribethediscrepancybetweentheobservedvaluesandthe
expectedvaluesofthemodelinquestion. Inthevarianceanalysis,oneofthecomponentsintowhich
thevariancesarepartitionedmaybealack-of-fitsumofsquares. Totestthemodelfit,wealsoused
anumberofgoodness-of-fitindicesrecommendedbymanyresearchers,including(1)anormalized
χ2
chi-square[ ];(2)thegoodness-of-fit(GFI);(3)therootmeansquareerrorofapproximation(RMSEA);
df
(4)thecomparativefitindex(CFI)and(5)theincrementalfitindex(IFI)[26]. AsTable2indicates,allof
thediversegoodness-of-fitindicesarehigherthantherecommendedcriteria. Thus,thegoodness-of-fit
ofthemeasurementmodelisacceptable.
Table2.Goodness-of-fit(GFI)testresults.
Indices Recommended DirectModel MediationModel
χ2
<5.0(Hairetal.2010[27]) 3.089 2.921
df
GFI >0.8(ChoiandJin[6]) 0.833 0.817
RMSEA <0.10(ChoiandGao[8]) 0.089 0.086
CFI 0.924 0.922
>0.9(ChoiandJin[6])
IFI 0.924 0.922
RMSEA,rootmeansquareerrorofapproximation;CFI,comparativefitindex;IFI,incrementalfitindex.
Sustainability2016,8,147 8of15
Ourproposedmodelmustbestatisticallyreliableandvalid.Thereliabilitytestevaluatestheoverall
consistencyofameasure;theproposedmethodshouldreflectsimilarresultswhenretestedunderthe
sameconditions. Forthereliabilitytest,weusethewidely-knownCronbach’salphacoefficient,the
correcteditem-totalcorrelationcoefficientandtheconstructreliabilitycoefficient.Thesecriteriaarethe
mostcommonlyusedtomeasurereliability[26].AsTable3shows,eachmeasureexceedsthesuggested
thresholds of 0.7, 0.5 and 0.8, respectively. Composite reliability or reliability rho tests if it may be
assumed that a single common factor underlies a set of variables and thus crucial for the SEM [28].
However,theacceptablecutoffforcompositereliabilitywouldbethesameastheresearchersetsfor
Cronbachalpha,sincebothattempttomeasuretruereliability. Moreover,theresponsivenessconstruct
reliabilityseemstobehigh,anditmayindicatethatsomeitemsareredundant;therefore,theconstruct
exhibitsquestionablevalidity[28]. However,asChoandKim[28]argued,thedeletionofthevariable
alsomaynotgivebetterconsistency,andthus,weshallacceptit. Thus,themeasuresareconsidered
appropriateforconfirmingasatisfactorylevelofreliabilityinthisstudy.
Table3.Reliabilitytestofthelatentvariables.
CorrectedItem:Total Construct
LatentVariables No.ofItems Cronbach’sα(>0.7)
Correlation(>0.5) Reliability(>0.8)
Usefulness 3 0.929 0.757–0.818 0.929
Convenience 4 0.882 0.748–0.781 0.883
Security 4 0.875 0.647–0.746 0.875
Responsiveness 3 0.978 0.773–0.834 0.979
Economy 3 0.903 0.712–0.785 0.905
ConsumerSatisfaction 4 0.944 0.759–0.826 0.944
ReuseIntention 4 0.926 0.746–0.801 0.929
4.2. ValidityTest
Thevalidityofanassessmentisthedegreetowhichitmeasureswhatitissupposedtomeasure(true
information).Validityanalysisincludesbothcontentandconstructvaliditycomponents.Contentvalidityanalysis
teststherepresentativenessoftheitemsinthequestionnaire. Noreportsexistofmisunderstandings
duringthepilottest. Theintervieweesstatedthattheitemsareeasilyunderstood,indicatinggood
contentvalidity[26].
Confirmatory factor analysis (CFA) is one of the most effective tools to test construct validity.
AccordingtoCampbellandFiske[29],constructvaliditytypicallyteststheextenttowhichthedata
providethefollowing:(1)convergentvalidity,theextenttowhichdifferentassessmentmethodsindicate
similarmeasurementsofthesametrait(i.e.,ideally,theconstructvaluesshouldbemoderatelyhigh);
and(2)discriminantvalidity,theextenttowhichindependentassessmentmethodsindicatedivergent
measurementsofdifferenttraits(ideally,thesevaluesshoulddemonstrateminimalconvergence).
AssummarizedbyChoiandYu[26],convergentvalidityoccurswhen(1)allfactorloadingsare
significantlyhigherthanthe0.5cutoffpointand(2)theaveragevarianceextracted(AVE)fromthe
itemsbytheirrespectiveconstructsisgreaterthan0.5. Table4indicatestheresultsoftheconvergent
validitymeasuredusingCFA.Themeasurementscaleindicatesastrongconvergentvaliditybecause
allofthefactorloadingsaresignificantandhigherthan0.5,andtheAVEofalloftheitemsisgreater
than0.5.
Sustainability2016,8,147 9of15
Table4.Convergentvaliditytestonthemeasurementmodel.
PotentialVariable Items SFLa C.R.b AVEc
Usefulness1 0.985 27.903
Usefulness Usefulness2 0.802 18.66 0.821
Usefulness3 0.922 -
Convenience1 0.795 16.143
Convenience2 0.797 16.586
Convenience 0.668
Convenience3 0.746 14.98
Convenience4 0.921 -
Security1 0.814 -
Security2 0.836 13.929
Security 0.641
Security3 0.807 13.553
Security4 0.743 13.39
Responsiveness1 0.996 -
Responsiveness Responsiveness2 0.921 35.704 0.938
Responsiveness3 0.987 81.505
Economy1 0.841 16.784
Economy Economy2 0.914 18.46 0.759
Economy3 0.857 -
CustomerSatisfaction1 0.823 -
CustomerSatisfaction2 0.852 15.939
CustomerSatisfaction 0.721
CustomerSatisfaction3 0.846 18.71
CustomerSatisfaction4 0.874 16.343
ReuseIntention1 0.783 -
ReuseIntention2 0.915 19.758
ReuseIntention 0.790
ReuseIntention3 0.862 18.226
ReuseIntention4 0.983 21.738
aSFLisanestimateofstandardizedfactorloading;bC.R.isthecriticalratio;cAVEistheaveragevarianceextracted.
AsFornellandLarker(1981)suggest[30],theestimateofthestandardizedfactorloadingsshould
besignificantlyhigherthanthe0.5cutoffpoint. Weevaluatedtheconvergentvalidityoftheconstructs
usingAVE.AsTable3shows,theAVEsofallsevenconstructsaremuchhigherthanthecriticalvalue
of0.5,indicatingthatthesesevenconstructscapturemorethan50%ofthevarianceintheirobservable
measures. Therefore,westatethattheconvergentreliabilityoftheconstructsissatisfactory.
DiscriminantvalidityisachievedwhenthesquarerootoftheAVEfortheconstructsisgreater
thananyofthecorrespondinginter-constructcorrelations. Table5indicatesthatthesquarerootofthe
AVEofallofthevariablesisgreaterthantheirinter-correlations,whichsupportsdiscriminantvalidity.
Table5.Discriminantvaliditytestofthemeasurementmodel.
ConstructItems USE CON SAF REA ECO CS REU
Usefulness(USE) 0.906a
Convenience(CON) 0.011 0.817a
Security(SAF) 0.042 0.131 0.801a
Responsiveness(REA) 0.006 0.018 0.049 0.969a
Economy(ECO) 0.024 0.022 0.097 0.005 0.871a
Satisfaction(SAF) 0.068 0.245 0.301 0.117 0.133 0.849
ReuseIntention(REU) 0.045 0.173 0.300 0.080 0.099 0.385 0.889a
aThesquarerootoftheAVEasacriterionofthecuttingpointforcorrelation.
Sustainability2016,8,147 10of15
Sustainability 2016, 8, 147 10 of 15
4.3. HypothesesTests
4.3. Hypotheses Tests
After confirming the reliability and validity of the questionnaire items, we test the research
After confirming the reliability and validity of the questionnaire items, we test the research
hypotheses. ThemaximumlikelihoodestimationmethodisusedforhypothesestestingwiththeSEM.
hypotheses. The maximum likelihood estimation method is used for hypotheses testing with the
Theempiricalfindingsofthefirststageofthedirectmodel(withoutconsideringcustomersatisfaction
SEM. The empirical findings of the first stage of the direct model (without considering customer
assthateismfaecdtiioanto ars) tahree msheodwiantoirn) aFrieg ushroew3.n in Figure 3.
InItnh ethdei dreircetcmt modoedleiln invvoolvlvininggt htheefi fivvee sseerrvviiccee qquuaalliittyy ffaaccttoorrss (u(usesfeufulnlensess, sc,ocnovnevneienniecne,c see,csuerciutyr,i ty,
resrpeosnpsoinvseinveesnsesasn danedc oencoomnoym)ayn) danredu rseeuisnet einnttieonnt,iotnh,e trhees urletssuinltds iicnadteictahtae ttthhaet atlhleo fatllh eofs etrhvei cseerqvuicaeli ty
facqtouraslihtya vfaecatoprso shiativvee ai mpopsaitcitvoe nimcupastcot mone rcsu’sptoumrcehrass’ epuinrtcehnatsieo nin.tTenhtuiosn,H. T1h–uHs,5 Har1e–Hsu5p aproe rstuepdp(opr<ted0. 1).
Ev(epn <i f0H.1)1. Eisvmena irfg Hin1a lisly maacrcgeipntaeldly aatctchepetseidg naitfi thcaen scigenliefvicealnocfe 0le.1v,etlh oifs 0r.e1s,u thltisim repsluielts imthpaltieast tlheaats tatt he
inilteiaalsts uthcec einsistioafl sAulcicpeasys oref sAullitpsafyr oremsubltesi nfrgomw eblel-ionrgg wanelilz-eodrg,acnuiszteodm, ceur-sotormienert-eodrioerntceuds otor mcuizsteodmsiezrevdi ce
quaselirtvyi.cAe mquoanligtyt.h Aemcooenffig ctiheen tcso,etfhfieciceonntsv,e tnhiee ncocnevfaenctioenrscein fdacictoartse itnhdeiclaartge etshtei mlarpgaecstt oimnpreaucts eonin rteeunsteio n.
Thuinst,etnhteiopnr. iTmhaursy, tphuer pproismeaorfya ptuhriprdo-spe aorft ya othnilridn-epparatyym oennlitnsee prvayicmeemnat ysebrevitcoe amdadyr ebses ttoh eadudrgreesnst tnheee d
forucrognenvte nnieeendt ffoarc icloitnavtieonnieonft ofancliilniteatsihoonp opf ionngl.ine shopping.

FigFuigreur3e. 3R. eRseuslutsltos fotfh tehed idriercetcts tsrturucctuturraall mmooddeell.. **,, **** aanndd ****** ddeennootete aalll lsisginginfiicfiacnatn ctoceoffeifcfiiecnietsn atst a0.t10, .1,
0.005.0a5n adn0d. 001.0,1re, srepsepceticvtievleyl.y.
To test the mediating effect of customer satisfaction, the relationships among the variables
Totestthemediatingeffectofcustomersatisfaction,therelationshipsamongthevariablesshould
should satisfy all of the conditions suggested by Baron and Kenny [31]: (1) the independent variable
satisfyalloftheconditionssuggestedbyBaronandKenny[31]: (1)theindependentvariableshould
should significantly influence the dependent variable; (2) the independent variable should
significantly influence the dependent variable; (2) the independent variable should significantly
significantly influence the mediator; (3) the mediator must significantly influence the dependent
influence the mediator; (3) the mediator must significantly influence the dependent variable; and
variable; and (4) the impact of the independent variable on the dependent variable must diminish
(4)theimpactoftheindependentvariableonthedependentvariablemustdiminishaftercontrolling
after controlling for the effects of the mediator. If any of these conditions are not satisfied, no
fortheeffectsofthemediator. Ifanyoftheseconditionsarenotsatisfied,nomediationoccurs. Ifallof
mediation occurs. If all of these conditions are satisfied, and the influence of the independent variable
theseconditionsaresatisfied,andtheinfluenceoftheindependentvariablebecomesnon-significant
becomes non-significant in the presence of the mediator, the effects of the independent variable are said
inttoh ebep r“ecsoemnpcleetoeflyt”h eorm “efudlilayt”o rm,ethdeiaetefdfe cbtys tohfe tmheediniadtoerp. eInf daleln otf vtharei acobnlediatrioenssa aidret soabtiesfi“ecdo manpdl ethteel y”
ori“nffululleyn”cem oef dthiae tienddebpyenthdeenmt veadriiaatboler. reImf aalilnso fsitghneificcoanndt iinti othnes parreesesnacteis ofife tdhea mndedtihaetoirn, tflhue eenffceectos fotfh e
indthepe einnddeepnetnvdaerniat bvlaerriaebmlea ianrse ssiagidn itfioc baen t“pinartthiaellpyr”e smeendceiaotefdt h[3e2m]. ediator,theeffectsoftheindependent
variablearesaidtobe“partially”mediated[32].