Professional Documents
Culture Documents
1 s2.0 S0957417416301816 Main
1 s2.0 S0957417416301816 Main
a r t i c l e i n f o a b s t r a c t
Article history: Smartphone credit card (SCC) is an emerging payment method using NFC-enabled smartphones. The
Received 30 January 2016 proximity payment allows consumers to pay their products and services by waving their smartphones
Revised 14 April 2016
with a NFC reader. While there are advantageous adopting SCC, the adoption rate has not been encour-
Accepted 15 April 2016
aging. Interestingly, existing research work on past information technology and system models have so
Available online 18 April 2016
far focused primarily on organizational context and adopted specifically for work. Furthermore, past an-
Keywords: tecedents were mainly constructed using electronic commerce literatures which do not reflect the ac-
Mobile commerce tual mobile environment. In contrast SCC is mainly adopted voluntarily by mobile users and for personal
Near field communication (NFC) purposes. Thus this leads to the difficulty in drawing meaningful conclusion. The study addresses these
Mobile technology acceptance model limitations by proposing a new mobile technology acceptance model (MTAM) which consists of mobile
(MTAM) usefulness (MU) and mobile ease of use (MEU) to determine SCC adoption. In anticipating on the com-
Mobile payment
plexity which exists in the mobile environment, additional mobile constructs namely mobile perceived
Partial least squares-structural equation
modelling-artificial neural network
security risk (MPSR), mobile perceived trust (MPT), mobile perceived compatibility (MPC) and mobile per-
(PLS-SEM-ANN) ceived financial resources (MPFR) were incorporated into the parsimonious MTAM. The integrated model
was applied to 459 mobile users through a questionnaire approach and tested using partial least square-
structural equation modelling-artificial neural network (PLS-SEM-ANN) has provided a new impact and
a possible new research methodology paradigm as it is able to capture both linear and non-linear rela-
tionships. While the model confirms the role of MU in MTAM, MEU needs for more attention in practice.
The results from the extended model showed that only three of the proposed hypotheses were non-
significant in this study and thus warrant further investigation. The study contributes to academia by
proposing new mobile constructs that brings together MTAM to assess the likelihood of mobile users to
adopt SCC. The study also offers several important managerial implications which can be generalized to
the mobile studies of other transportation, hotel, banking, and tourism industries.
© 2016 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.eswa.2016.04.015
0957-4174/© 2016 Elsevier Ltd. All rights reserved.
34 K.-B. Ooi, G.W.-H. Tan / Expert Systems With Applications 59 (2016) 33–46
world and the real world of physical objects (Teh, Ahmed, Cheong, organizational settings. The cost is mainly absorbed by individuals
& Yap, 2014). The emergence of smartphone credit card (SCC) over- and therefore the adoption decision is not solely based for utilitar-
comes all of the limitation of traditional m-payment solutions. ian reasons. Moreover, Tsai and LaRose (2015) stressed that users
SCC provides simple communication setup, ease of use, mobiquity must have the sufficient knowledge and skills to use new innova-
and requires only little power consumption (Egger, 2013; Madl- tion when compared to employees in organization where training
mayr, Langer, Kantner, & Scharinger, 2008; Teh et al., 2014). SCC and technical support are provided. While TAM has been widely
can be defined as a payment of goods and services that is con- validated in many mobile studies and able to explain many adop-
ducted using smartphone with the aid of Near Field Communi- tion studies, they do not draw meaningful conclusion in view of
cation (NFC) technology. In broad terminology, NFC works using the differences. UTAUT was advocated by Venkatesh, Morris, Davis,
short range wireless technology between two devices of 10 cm or and Davis (2003) and put forward four core constructs that are di-
less. By waving the NFC-enabled smartphone over a sales termi- rect determinants intention and subsequently usage behaviour. The
nal, checkout process can be completed faster using minimum ef- constructs namely performance expectancy (PE), effort expectancy
fort. This is because signature is not required during the payment (EE), social influence and facilitating conditions (FC) are also mod-
process (Zmijewska, 2005; Leong, Hew, Tan, & Ooi, 2013). The cur- erated by gender, age and experience in the acceptance of IT.
rent SCC technology includes Apple Pay, Google Wallet, Visa Pay- Venkatesh, Thong, and Xu (2012) however criticized that the model
Wave, Simply Tapp and BellID. With benefits that will overcome is again based on organizational contexts and developed primarily
the limitation of traditional m-payment solutions, SCC is foreseen based on the acceptance of technology among employees. Technol-
as a vital catalyst to the existing m-payment solutions and will ogy adoption outside the workplace differs on a number of dimen-
potentially alter consumers’ lifestyles and change the m-payment sions such as differences in types of tasks and complexity of inter-
industry as a whole. SCC is even considered as a fourth wave of actions (Brown, Venkatesh, & Bala, 2006). Thus UTAUT has limita-
income for mobile operators after data, voice and text plan. While tion in describing mobile studies adoption such as SCC because po-
there are tremendous benefits adopting SCC as an alternative m- tential adopters are actual mobile consumers and not simply tech-
payment solution, the adoption rate is far from full utilization. Teo, nology users with decisions mainly centralized from the personal
Tan, Ooi, Hew, and Yew (2015a) explained that SCC has yet to context. In view that many past IT/IS models was adopted based
reach mass market which leads to frustration among m-payments on the workplace environment, Brown and Venkatesh (2005) pro-
providers. Feijóo, Gómez-Barroso, and Ramos (2016) explain that posed Model of Adoption of Technology in Households (MATH).
market analysts are still sceptical on the adoption. This is despite The model which is based on Theory of Planned Behaviour (Ajzen,
the availability of SCC infrastructures since 2010 (Tan et al., 2014a). 1991) framework, focused on the adoption of technology in the
Leong et al. (2013) elaborated on the variety of merchants adopt- household using three constructs namely attitudinal beliefs, nor-
ing to SCC ranging from hypermarket, retailer, specialty store, con- mative beliefs and control beliefs. While the model can explain 50
venience store, petrol station, cinema to public transportation. percent of the variance, the model however is limited to the ap-
As the success of SCC is dependent on the consumers’ up- plicability of adoption of personal computers (PC) for home use
take, without a critical mass of consumers, merchants will be dis- (Brown et al., 2006). Teo, Tan, Cheah, Ooi, and Yew (2012) argued
couraged from investing on the infrastructure. According to Teo that there are differences in the adoption of new technology be-
et al. (2015a), by 2017, global m-payment transaction is expected tween desktop and mobile. Finally, Roger (1983) introduces DOI
to reach $1.3 trillion. Feijóo et al. (2016) further added that m- which consists of observability, relative advantage (RA), trialabil-
payments are expected to increase twofold from 2012 to 2016 with ity, complexity (CO) and compatibility. The model explains that
NFC technologies playing an increasing role. The introduction of prior to the adoption of the new innovation; consumers will go
Apple Pay in 2014 which runs on NFC technology on iPhone is also through a four stage process which includes knowledge, persua-
expected to represent huge business opportunities for merchants sion, decision and confirmation. Balachandran and Tan (2015) high-
and retailers in the coming years (Apple, 2015). Idate even esti- lighted that trialability and observability in their original form are
mated that the volume of NFC transactions will reach €53.8 billion not useful when predicting a new IT/IS adoption. Lancelot Miltgen,
Euros in 2018 from €4.6 billion in 2014 (Idate.org, 2014). In ad- Popovič, and Oliveira (2013) explained that since the new technol-
dition, NFC-enabled phone penetration is expected to reach 1907 ogy has yet to reach mass wide, the majority of consumers would
million by 2018. With such huge potential, businesses are increas- not be able to test the technology and thus both models are not
ing investments on the NFC infrastructures. Therefore, there is a well-suited. As most of the conventional frameworks has also lim-
need to understand critical prerequisites that affect consumers’ in- ited ability to explain the adoption of new technologies, many re-
tention to adopt SCC. SCC is in essence an information technol- searchers have extended and modified the model with new con-
ogy (IT)/information system (IS). Hence, to understand the accep- structs (e.g., Sim, Tan, Wong, Ooi, & Hew, 2014; Tan, Sim, Ooi, &
tance process of SCC there is a need to comprehend IT/IS adoption Kongkiti, 2012; Teo, Tan, Ooi, & Lin, 2015b). The approach helps
models. Among prominent and common IT/IS models adopted to to increase the variance explained and making the argument for
understand the adoption and utilization of new technological in- model sufficiency. Although the constructs are rich in studies on
novations includes “Technology Acceptance Model (TAM)”, “Diffu- the factors of technology adoption, the innovation being studied is
sion of Innovation (DOI)”, “Unified Theory of Acceptance and Use mainly borrowed from electronic commerce literatures. A majority
of Technology (UTAUT)” and “Model of Adoption of Technology in of consumers however behave and react differently in an electronic
Households (MATH)”. The models however are not without limi- environment as compared to mobile. Mobile users perception of
tations. TAM asserted that the actual use is related to individual’s ease of use behaviour for example could be heavily influenced by
attitude towards using technology and the model identified two the access speed, screen size, text input facilities, storage and bat-
salient determinants of system use behaviour namely perceived tery life span as compared to desktop computers (Tan, Ooi, Leong,
usefulness (PU) and perceived ease of use (PEOU) (Davis, 1989). & Lin, 2014b). Hence again they are not suited to be adopted in
Originally, TAM was constructed to explain individual’s adoption of mobile studies.
electronic mail system in an organizational settings. The adoption In order to close the research gap, the study proposed a
was based on work purposes with the individual employee of the new base model termed as ’mobile technology acceptance model’
organization. As such the cost is normally borne by organization (MTAM) which consists of mobile usefulness (MU) and mobile ease
with the adoption mandatory. In contrast, SCC is adopted volun- of use (MEU) based on the most relevant mobile technology liter-
tarily and mainly for personal purpose by individuals in a non- atures. Both constructs were given priority based on the similarity
K.-B. Ooi, G.W.-H. Tan / Expert Systems With Applications 59 (2016) 33–46 35
2. Literature reviews
intention to adopt NFC-enabled mobile credit card. Their findings (BusinessNewsDaily.com, 2015). The adoption of SCC can also come
which adopted 262 consumers using neural network approach re- with annual credit card fees and charges (Hsbc.com.hk, 2015).
mains mix which warrant further investigations. In another study, Given the high financial considerations to adopt SCC, the technol-
Teo et al. (2015a) employed convenience sampling on 194 con- ogy may be out of reach for consumers with limited and constraint
sumers to gauge on their opinion on the effects of perceived trans- financial resources. The practical benefits and easiness of use on
action convenience (PTC) and perceived transaction speed (PTS) to the adoption would therefore diminish. Based on 262 valid ques-
adopt m-payment. Based on the combination model of UTAUT, they tionnaires, Leong et al. (2013) provide evidences that MPFR is not
discovered that EE, FC and PTS influences adoption rate. All the a significant factor with MU and MEOU in their study on NFC-
above studies provide early possibilities on the barriers in accept- enabled mobile credit card acceptance in Malaysia. Finally accord-
ing SCC solutions. ing to Balachandran and Tan (2015), the adoption of SCC includes
conversion, maintenance and accessing costs. They found that the
3. Hypotheses formulation costs prevent consumers from the IU to adopt NFC mobile payment
in Malaysia and suggested to reduce the cost through creative pro-
3.1. Mobile perceived compatibility (MPC) motion and pricing techniques. Thus the following hypotheses are
being constructed:
MPC refers to the extent that the mobile innovation is rec-
oncilable with potential adopters past experiences, needs, value H2: MPFR has a negative and significant relationship with MEU.
and behavioural pattern. The adoption of SCC is not something H8: MPFR has a negative and significant relationship with IU.
unusual. Consumers are already adopting to the traditional m- H10: MPFR has a negative and significant relationship with MU.
payment methods using technologies such as General Packet Radio
Service, Unstructured Supplementary Service Data, Wireless Appli- 3.3. Mobile ease of use (MEU)
cation Protocol and short messaging services in everyday life. As
SCC is based on m-devices technology, consumers are more recep- MEU which has the same meaning as ease of use and complex-
tive to adopt if they have experience using m-devices. Wong, Tan, ity of using a particular IT/IS (Venkatesh et al., 2003). Therefore
Ooi, and Lin (2015a) commented that consumers in Malaysia have MEU combines the constructs mentioned above and refers to the
adopted smartphone to purchase products and services (Wong et perception of complexity to learn and use for potential adopters
al., 2015a). In addition, consumers also using smartphone devices when adopting to m-devices. The adoption of SCC requires min-
to access television (Wong, Tan, Loke, & Ooi, 2014), banking (Teo imum effort to use and learn when compared to the old fash-
et al., 2012) and listening to music (Sim et al., 2014). NFC tech- ion m-payment methods. Thus this leads to favourable impressions
nology has also been adopted by consumers to exchange large for potential adopters. By selecting Apple Pay and placing the in-
information of data, play games, share information by bumping dividual’s finger on the ‘Touch ID’, payment could be conducted
phones (Nearfieldcommunication.org, 2015). Balachandran and Tan in a breeze (Apple, 2015). Consumers also do not need to require
(2015) in Malaysia further explicated that consumers are adopting much effort in registering SCC prior to usage. Consumers can regis-
m-devices to check in at locations and to share data. Thus past ter their credit card details such as name card number and expiry
experience of using smartphone devices can be regarded as an in- date using the camera icon in their smartphone (Appleinsider.com,
dicator of mobile compatibility. Yang, Lu, Gupta, Cao, and Zhang 2014). When consumers feel that SCC is easy to use and learn; they
(2012) provided evidences that MPC influences the adoption of m- will more likely to adopt this new innovation. Dutot (2015) ex-
payment in China. Similarly MPC was also found to have influence plained that when consumers start to adopt a new system, they
on the adoption of mobile shopping in Malaysia whereby the im- tend to stress on the possibility of learning success and in this
portance of consumers’ past belief and life style was emphasized case directly affect their IU to use a particular technology. MEU
(Wong et al., 2015a). Based on a research conducted by Schierz, was found to influence the IU to adopt contactless m-payment and
Schilke, and Wirtz (2010), MPC was also found in influencing use- m-payment in Malaysia (Teo et al., 2015a, 2015b). In another study
fulness of m-payment system. Finally, as SCC runs on ’Touch ID’ in France using 320 smartphone users based on PLS-SEM analysis
and proximity payment (Apple, 2015) consumers might consider on NFC mobile payment, MEU was found to be significant with IU
the concept as new and hence experience a struggle to adopt. Thus (Dutot, 2015). These studies stressed on the friendly interface, sys-
if SCC is designed to be compatible with the adopters’ past expe- tem operation and function in ensuring higher users’ acceptance.
riences, this would lead to the perception that SCC is easy to use. Hence the next hypothesis has been constructed:
In view of these findings, the following hypotheses have been con-
structed: H3: MEU has a positive and significant relationship with IU.
H1: MPC has a positive and significant relationship with MU. 3.4. Mobile usefulness (MU)
H7: MPC has a positive and significant relationship with IU.
H9: MPC has a positive and significant relationship with MEU. MU is similar with sub-constructs of usefulness such as rel-
ative advantage and extrinsic motivation and explains how they
3.2. Mobile perceived financial resources (MPFR) lead to improvement in job performance for the particular individ-
ual (Venkatesh et al., 2003). Likewise in this research, MU refers
MPFR refers to the extend individuals belies that the he or she to the perception of usefulness enhancement rendered for poten-
has the financial resources to use SCC (e.g., cost of device, transac- tial adopters when adopting to m-devices. Compared with tradi-
tion fees and bank specific charges). Wong et al. (2014) commented tional m-payment methods, SCC which is based on contactless pay-
that financial resources usually hamper the adoption of new in- ment technology results to effectiveness in payment such as con-
novation among consumers. SCC devices are normally pricier than venient and quicker checkout. Itbusinessedge.com (2015) explained
traditional phones. This is due to the higher cost in production that payment can be made by placing the individual’s iPhone near
which requires a redesigning of traditional models and therefore the NFC reader. Thus no card swiping is involved. In addition, mul-
expensive to produce. Some of the m-devices could come with tiple credit cards can be loaded in SCC. This allows consumers the
bundle contract such as 24-months contract (Tan et al., 2014a). flexibility to choose their credit card for each specific transaction
Consumers would also need to pay for transactional costs for usage (Nearfieldcommunication.org, 2015). The ability to store multiple
K.-B. Ooi, G.W.-H. Tan / Expert Systems With Applications 59 (2016) 33–46 37
cards in SCC reduces the risk of losing credit cards. A study by identity theft and mobile fraud. The following hypothesis has been
Tan et al. (2014a) in Malaysia on NFC m-devices found that MU constructed:
leads to higher intention to adopt for m-payment. Similar result
was also illustrated by Teo et al. (2015b) whereby the adoption H6: MPT has a positive and significant relationship with IU.
of m-payment brings time saving and portability. Interestingly MU
was found to be non-significant according to the study by Dutot
4. Research method
(2015) and Balachandran and Tan (2015) in France and Malaysia
respectively despite the advantages rendered by NFC m-payment.
4.1. Sampling procedure and data collection
In view of these contradicting findings, the following hypothesis
has been constructed:
A paper-based survey was conducted between October and
H4: MU has a positive and significant relationship with IU. November 2015 at five major shopping malls at Klang Valley,
Malaysia to validate the research model. Klang Valley was cho-
sen since area has high population density with about 1/5 of
3.5. Mobile perceived security risk (MPSR)
Malaysian’s total population (Pan, Chew, Cheah, Wong, & Tan,
2015). The sampling location is also a major hub for economic,
MPSR is a vital construct in the research of wireless environ-
financial and culture in the county with a fair balance of ma-
ment as it determine the decision of consumers whether to adopt
jor ethnics in Malaysia such as Malay, Chinese and Indians
a new innovation. MPSR refers to the perception of protection
(Wordpopulationreview.com, 2016). Thus this would give a fair
against risk associated with mobile transactions. This could be risk
characteristics presentational of the gender, age and ethnicity of
of losing confidential data which leads to financial losses. Mail
the general population in Malaysia which are relevant to this con-
Online (2015), reported on a new breed of digital pick pocket-
text of study. The five major shopping malls were chosen due to
ing known as digital skimming. By using an app and standing 6
the high number of potential mobile users available which will ex-
inches away, data can be stolen during any NFC transactions within
pedite our process in data collection. The questionnaire distribu-
the store. Thus, the NFC infrastructure needs to be configured cor-
tion was assisted by four research assistants and was based on
rectly by merchants and retailers to avoid any misuse or attacks
systematic sampling method whereby, every second participants
(Hoy, 2013). Zupanovic (2015) stressed that NFC technology per-
who set foot in the shopping malls were approached and inquired
mits consumers to experience safer interactions due to its bet-
to voluntarily complete the questionnaires. Prior to distribution
ter encryption of transacted data. In another study on NFC smart-
the participants were brief on the purpose of study and guaran-
phone access control system based on 40 participants, Cheong et
teed anonymity. In addition, to avoid a respond bias in the sur-
al. (2014) found that MPSR has significant relationship with IU.
vey, names were excluded. Upon agreement, a short video clip
While MPSR is considered to be important, interestingly findings
on smartphone credit card (e.g., https://www.youtube.com/watch?
by Tan et al. (2014a) found that perceived risk does not influence
v=4I9MbIrlEUw) was presented. The approach has been widely
the IU of NFC-enabled m-devices in Malaysia. Thus, the construct
adopted in recent years by international researchers (Tan et al.,
requires further investigation. The following hypothesis has been
2014b). The samples were subsequently briefed on the terminol-
constructed:
ogy and definitions adopted in the questionnaires. While SCC in-
H5: MPSR has a negative and significant relationship with IU. cludes different technologies and can be applied in various sectors
we explained that SCC refers to the overall concepts and was mea-
sured from a general perspective. The respondents were also en-
3.6. Mobile perceived trust (MPT)
couraged to ask the research assistants on the questionnaire items
that they didn’t understand prior to answering the questionnaire.
MPT refers to a psychological expectation that others will be
This would help to eliminate the possibility of respondents’ filling
sincere in keeping promises and will not behave opportunistically
out the questions without exactly understanding the content of the
in expectation of a promised service derived from mobile transac-
questionnaire. On a whole, 501 questionnaires were handed out.
tions. If the retailers show benevolence, integrity and competence,
The questionnaires were scrutinized and only 459 valid and usable
consumers are more likely to engage with SCC transactions (Leong
samples were adopted, yielding a valid response rate of 91.61 per-
et al., 2013). According to Yang and Lin (2014), it is more diffi-
cent among volunteered participants.
cult for a retailer and consumer to build online trust compared
to an offline trust. This is caused by the absent of any face-to-
face interactions between both parties during any online transac- 4.2. Instrument development
tions. Hence MPT is an important factor in the online environment
such as in any mobile transactions. Otherwise relationships cannot The items in the survey instrument were based on our review
be established over the long-term due the uncertainty. Addition- of existing and past literatures relevant to the research model.
ally online environment promotes betrayal, lack of accountability Some of the wordings were modified and reworded to fit the con-
and opportunity for deceit due to its association with risk char- text of SCC. The research were anchored on a seven-point Likert-
acteristics (Immonen & Sintonen, 2015). Sztompka (1999) pointed type scale measurement varying from “1 (strongly disagree)” to “7
out that trustworthiness is based on the trustee’s reputation and (strongly agree)”. A pre-test was also performed with three mobile
performance. Tamimi and Sebastianelli (2015) remarks that privacy commerce scholars with a background on mobile payment to en-
seals, third party assurances, security policies, provide feedbacks sure that the questionnaires have no semantic problems. Some of
help to increase the consumer’s perception on online trustwor- the clarity of questions, length of instruments, the completeness
thiness. Qualitative interviews by Mallat (2007) provide evidences of content and structure which were found to be ambiguous and
that trust in reliable providers reduced risks in m-payment ser- vogue were further amended based on the provided feedback. The
vices. Based on past researchers, MPT has been proven to increase instruments were then further pilot-tested with 20 mobile pay-
intention for online payments in China (Yang, Pang, Liu, Yen, & ment users. The participant group has knowledge on the mobile
Michael Tarn, 2015). Similar Teo et al. (2015b) also found support- payment systems in Malaysia. Slight modifications were made to
ing evidences on the influential effect of MPT and IU. They con- the wording resulted from the pre-test. The multiple phases of in-
cluded that consumer would only transact if the IT/IS is free of strument development helped to establish the appropriate level
38 K.-B. Ooi, G.W.-H. Tan / Expert Systems With Applications 59 (2016) 33–46
Table 1
Questionnaire source and number of items.
Mobile Usefulness (MU) 5 Davis et al. (1989) and Tan et al. (2014a)
Mobile Ease of Use (MEU) 5 Davis et al. (1989) and Tan et al. (2014a)
Mobile Perceived Compatibility (MPC) 5 Rogers (1983) and Wong et al. (2015a)
Mobile Perceived Trust (MPT) 5 Zhou and Lu (2011) and Leong et al. (2013)
Mobile Perceived Security Risk (MPSR) 3 Tan et al. (2014a)
Mobile Perceived Financial Resources (MPFR) 4 Leong et al. (2013)
Intention to use (IU) 5 Tan et al. (2014a)
relationships (Leong et al., 2015), however due to the ‘black box’ Table 3
Item loadings, AVE, CR, and Cronbach’s α .
operation of the ANN algorithm, it is not suitable for parametric
hypothesis testing (Sim et al., 2014). Hence, in order to take ad- Constructs AVE CR Cronbach’s α Item loadings
vantage of the both, we have used a combined PLS-ANN approach
IU 0.8787 0.9731 0.9654 IU1 0.9435
as this approach will enable us to test the hypothesis and on the IU2 0.9536
same time detect both linear and non-linear relationships that ex- IU3 0.9460
ist in the research model. IU4 0.9058
Outer measurement model was firstly assessed for both reliabil- IU5 0.9372
MEU 0.7503 0.9375 0.9166 MEU1 0.8227
ity and validity; while inner structural model was later on exam-
MEU2 0.8794
ined. This is known as two-step approach, and it is consistent with MEU3 0.8957
Wong, Tan, Hew, and Ooi (2016). Moreover, mediating test was MEU4 0.8482
also conducted and reported afterward. There are seven latent vari- MEU5 0.8829
ables and 32 observed variables in the proposed model. Through MPC 0.8190 0.9577 0.9447 MPC1 0.8870
MPC2 0.9208
the calculator by Soper (2015), which is suggested by Westland
MPC3 0.9015
(2010,2012), the recommended sample size is 88. The following MPC4 0.9231
settings were made in the calculator: MPC5 0.8920
MPFR 0.7594 0.9264 0.9031 MPFR1 0.8829
• anticipated effect size - 0.3 (Dutot, 2015) MPFR2 0.9444
MPFR3 0.8464
• desired statistical power level - 0.8 (Cohen, 1988)
MPFR4 0.8062
• significance probability level - 0.05 (Fisher, 1925) MPSR 0.9104 0.9682 0.9511 MPSR1 0.9424
MPSR2 0.9586
In addition, Wong (2013) posited that under the most extreme MPSR3 0.9614
situation, wherein there are 10 arrows pointing to a construct, 91 MPT 0.7786 0.9462 0.9293 MPT1 0.8860
samples would be required. In view of these, the sample size of MPT2 0.8944
MPT3 0.8517
this study, 459, is considered sufficient. MPT4 0.8910
MPT5 0.8882
MU 0.7894 0.9493 0.9332 MU1 0.8829
5.2.1. Assessing outer measurement model MU2 0.8937
Reliability and validity of the outer measurement model were MU3 0.8576
assessed by item reliability, convergent validity, and discriminant MU4 0.9142
MU5 0.8932
validity. Following the measures suggested by both Kim and Park
(2013), and Chen and Shen (2015), Cronbach’s α and composite re-
liability (CR) are the measures to ascertain reliability; average vari-
Table 4
ance extracted (AVE) and item loadings are used to evaluate con-
Square root of the AVE and construct correlations.
vergent validity; while square root of the AVE was assessed to eval-
uate discriminant validity. Fransen et al. (2014), p.1376) referred IU MEU MPC MPFR MPSR MPT MU
Table 6
Mediation test and Sobel’s test.
IV -> DV MV -> DV
Table 7
RMSE values of the ANN models.
Table 8
Normalized importance of predictors in ANN models.
Relative importance
Output: MU Output: MEU Output: IU
The finding explains that when SCC fits well with the consumers’
lifestyle, they will find that SCC beneficial. For example, if con-
sumers are already adopting to the traditional payment method
such as mobile peer-to peer, a shift to a newer payment method
such as proximity payment will lead to the perception that SCC
is more efficient during payment. This is because consumers are
able to compare the usefulness rendered by both m-payment sys-
tems. The finding also confirms the study by Balachandran and
Tan (2015) whereby MPC was found to have significant influence
with IU. This could be due to the experience of consumers in us-
ing m-devices for m-commerce purposes (Wong et al., 2015a). Fur-
ther, the study also revealed that MPC has significant relationship
with MEU. Thus, it implies that consumers will find SCC easy to
use when the innovation is reconcilable with the potential users’
current habits, needs, value and past experiences.
Previous studies in Malaysia and Taiwan on NFC-enabled mobile
credit card and mobile services, indicated that MPFR has no signifi-
cant relationship with MEU (Leong et al., 2013; Wang, Lin, & Luarn,
2006). The study coincides with our findings whereby MPFR was
found to be non-significant with MEU. This means that the per-
ception of financial resources will not reduce the operational com-
plexity of SCC. On the contrary our results confirmed that MPFR
is significant with MU. The result found empirical support from
Sim et al. (2014) and Wang et al. (2006) on the study on mo-
bile technologies. This leads us to believe that when SCC is value
for money, potential adopters will have higher expectation on the
benefits of usefulness in their daily lives. Interestingly MPFR is a
non-significant determinant of IU. Our findings validate the view
by Tan et al. (2014a) on NFC mobile credit card in Malaysia. Teo
et al. (2015b) accredited the reason with the affordable prices of
smartphones and telecommunication plans in Malaysia. Similarly
Wong et al. (2015a) mentioned that the cost of m-devices has re-
duced significantly over the last few years due to the intense com-
petition among manufactures and telecommunication providers.
Regarding MEU, our study highlight that the construct is not a
significant predictor for IU. The findings contradicts with all the
past studies conducted on NFC m-devices where consumers would
accept the new innovation if the system is effortless (Balachandran
& Tan, 2015; Dutot, 2015; Leong et al., 2013; Tan et al., 2014a). The
finding is also in disagreement with past findings conducted on
mobile studies for shopping and learning (Wong, Lee, Lim, Chua, &
Tan, 2012; Wong, Tan, Loke, & Ooi, 2015c). One possible explana-
tion is that the number of steps needed to perform a SCC trans-
action is lesser than other m-payment systems. Therefore majority
of the consumers perceived that SCC is user friendly to use and
probably requires lesser effort to learn.
Fig. 3. The ANN models in the study. MU on the other hand was found to be significant with
IU. The findings coincides with the past study performed by
42 K.-B. Ooi, G.W.-H. Tan / Expert Systems With Applications 59 (2016) 33–46
Tan et al. (2014a) which highlighted that the more benefits per- 7.2. Managerial implications
ceived by mobile payment adopters, the more favourable towards
consumers’ IU. MU has also been proven to be a predictor of IU in The cost of developing an infrastructure for NFC is not cheap.
other mobile studies such mobile learning (Tan et al., 2012), mobile Thus careful planning is essential before implementing. Based on
advertising (Wong, Tan, Tan, & Ooi, 2015b), mobile banking (Teo et traditional m-payment methods, SCC offers advantages such as
al., 2012), mobile marketing (Pan et al., 2015), mobile healthcare quicker checkout which translate to time saving and convenience
(Tan, Siah, Ooi, Hew, & Chong, 2015) and mobile television (Wong for consumers. In addition, SCC promotes safety as consumers do
et al., 2016). In this context, consumers place high importance on not need to carry cash when shopping. Since MU is significant with
time saving, convenience and safety features rendered by SCC as IU, in order to improve MU, merchants and retailers should pro-
contrast to offline payment. mote the advantages of SCC in their mobile marketing campaigns.
On the contrary, MPSR is not significant with IU which the find- Strategic alliances should also be created between mobile compa-
ings coincides with the results conducted by Tan et al. (2014a) and nies and merchants. Apple Pay for example has teamed up with
Cheong et al. (2014).One possible explanation is when compared over 220,0 0 0 stores from paying gas, amusement park, to vending
to other traditional mobile payment methods; SCC is portrayed machines in the United States (Digital Trends, 2015). This creates a
to have better encryption (Zupanovic, 2015). Thus consumers have favourable impression that the adoption will bring usefulness and
confidence that their personal information will not be intercepted productivity for consumers in their job performance. Apart from
during any transaction and hence do not view MPSR as important. m-payment, the NFC technology can also be adopted for various
Finally it was found that MPT is significant with IU to adopt functions. According to Leong et al. (2013). p. 5605), “NFC-enabled
SCC. Similar results were found by mobile payment scholar such mobile phones may also read NFC tags that can host a diversity
as Teo et al. (2015b) whereby MPT was found to be the most influ- of content such as vouchers, videos, social media pages, compe-
ential construct on IU to adopt m-payment. This implies that con- titions, travel information and etc.”. Thus, another strategy is to
sumers will be more likely to adopt SCC if they believe that the integrate NFC functions in every life situations of consumers. The
banks and retailers are honest, sincere in keeping promises and strategy will lead to conversion of non-users to potential users in
take responsibility in the eventuality of any identity theft, fraud the long run when they find that NFC is useful. Next when de-
or security issues. veloping NFC apps, businesses need to consider consumer’s expec-
tations on the functions and usability of SCC. Businesses should
7. Implications improve on functions based on suggestions from consumers such
as MEU interface so that consumers can easily operate and trans-
7.1. Theoretical implications act with minimum effort even though they are non-significant in
this study. Apple Pay’s credit card picture in m-devices for ex-
The main contribution of the research paper is the development ample can be adjusted; tap, drag and move forward or backward
of MTAM which comprises of MEU and MU. MTAM is more com- to improve view (Appleinsider.com, 2014). The registration pro-
prehensive model in understanding the adoption of SCC in com- cess should also be simple without much procedure. According to
parison with traditional IT/IS models as it examines the adoption Appleinsider.com (2014), the registration of credit card using Ap-
based on actual mobile users in a non-organizational settings. Tra- ple Pay includes camera icon whereby consumers can scan the
ditional IT/IS model may not be suitable as they are based on or- name, card number, expiration date. Supports should also be ex-
ganizational settings and related to work purposes which often is tended not only through offline but also online. Automated phone
compulsory. The basic framework of MTAM would allow other con- call, blogs, online FAQ, video instructions, competent personal are
structs to be incorporated easily to better understand the adop- some of the several aspects which can help to increase the per-
tion of mobile technologies from the perspective of mobile con- ception of EE among consumers. MPSR should also not been ne-
sumers. Further, the general applicability of the model can be ap- glected although the study indicates otherwise. When compared
plied across different mobile technologies and users. Second as with e-commerce, transactions that build upon wireless environ-
most past constructs were developed from e-commerce theories ment are more vulnerable for security attacks and interceptions.
and adopted for usage in mobile studies, the research may not One way to reduce the perception of MPSR is to adopt advanced
reflect the actual mobile environment. This is because SCC offers technology infrastructure. This includes adopting wireless encryp-
unique characteristics which are different from the electronic com- tion technologies, sign-on security systems and other reliable secu-
merce environment. As such consumers would behave differently rity measures. Hsbc.com.hk (2015) for example reported that their
in view of the differences. The study contributes by introducing SCC includes pin before payment for security reasons and the con-
four new mobile constructs namely MPSR, MPT, MPC and MPFR. tactless payment features automatically turn off after 30 s when
They are introduced since consumers’ choice of payment depends not in use. Businesses should also adopt third-party legal certifica-
on their trust, characteristics and financial resources. The new mo- tions and continuously improvement mechanism to increase pay-
bile constructs provides a more precise explanation on the adop- ment security. Strict security policies should also be adhered to
tion and investigation of SCC. Future researcher should be able to guarantee security during payment. Retailers should make efforts
readily relate the new constructs with other mobile technology to verify the identity of consumers prior to transaction. In addition,
studies. Third, research on m-payment is mostly focused on tra- safe shopping guarantee and other protection mechanism should
ditional technologies. SCC is currently an emerging research topic be properly communicated to all consumers. The guarantee and
among academicians and scholars. We contribute to the existing protection mechanism should contain description that are not too
literatures on m-payment by providing information on the barriers lengthy and should avoid the use of technical jargons which makes
that prevents consumers to adopt. Lastly, one of the vital theoret- consumers comprehension difficult. Lee, Au, and Law (2012) even
ical contributions of this study is the use of expert systems and suggested communicating security policies in video format as part
artificial intelligent systems via the PLS-ANN approach in compari- of the communication with consumers. Consumers should also
son to other related papers which used pure ANN approach as the be encouraged to adopt specific security measures to secure and
ANN approach has a "black box" operation that renders it not suit- safeguard their private information. Based on the suggestions by
able for hypothesis testing. However with the new PLS-SEM-ANN Nearfieldcommunication.org (2015), they includes only (1) con-
approach proposed in this study, we are able to address this limi- ducting transactions with merchants that adopts secure channels,
tation. (2) m-devices should be in active-passive pairing and (3) adopting
K.-B. Ooi, G.W.-H. Tan / Expert Systems With Applications 59 (2016) 33–46 43
password or any keypad locks when SCC device is turned on. When at an infancy stage of implementation. The assumption is based
consumers can integrate SCC in their daily activities, this will lead that IU is closely related to actual behaviour (Venkatesh & Davis,
to a favourable perception of MPC that they are easy to use and 20 0 0) with the method closely shared by many studies conducted
has many advantages and thus influence the adoption intention. on mobile adoptions. However the actual adoption behaviour may
Therefore, mobile service providers should develop and advertise not necessary be reflected by the IU with Wang et al. (2006). p.
SCC in a manner that matches consumer’s prior experience, need 175) for example reported weak links in their study. Hence fu-
and behavioural pattern. Among the advantages of adopting SCC ture studies could be improved by employing actual behaviour in
is usefulness and convenience. Since SCC render faster checkout the study to closely observe the later stage of SCC implementa-
during payment process (Tan et al., 2014a), the innovation should tion. Third, SCC technologies range from Apple Pay, Google Wal-
be marketed as a lifestyle products. Thus this will suit busy con- let, Visa PayWave, SimplyTapp to BellID. Moreover SCC can also
sumers who are always ‘time starved’ and constantly looking to be applied in various sectors such as transportation, hotel, bank-
complete their tasks with minimum time. Businesses should also ing, tourism and etc. The study however did not focus on any
focus by highlighting on the values of SCC as compared to tradi- specific SCC technology, no on any specified sector and was mea-
tional m-payment services. By analysing consumer’s values, a more sured from a general perspective. Based on different scenarios and
personalized marketing campaign can be emphasized. Consumers circumstances, the perception for the same service could be dif-
perceived that using SCC is more expensive than traditional credit ferent. While this may not necessary affect the results since the
card. This is due to the perception on of high financial cost asso- study’s intention is to understand the motivation of consumers to
ciated with using SCC. SCC-enabled m-devices are not cheap (e.g., adopt SCC, the study provides some direction for future research.
Samsung Galaxy, iPhone). In addition to the cost of device, con- Fourth, although the survey was conducted at five major shopping
sumers need to pay annual fees of credit card. There are also addi- malls and covers the main financial and economic hub in Malaysia,
tional transaction costs involved (BusinessNewsDaily.com, 2015). In the study may suffer from biases derived from the geographical
order to ease consumers concern on MPFR, appropriate promotion context of the current study. This is because the respondents of
and pricing strategies should be created to reduce the financial the study are from Malaysia. While the findings is applicable to
barrier facing potential consumers. Banks for example could lower other South East Asia countries which shares the similar character-
their credit card annual fees and commission for the first few istics and demographics, the findings may not necessary be appli-
years. Once consumers discover the benefits and get used to the al- cable other emerging countries that lagged behind ICT infrastruc-
ternative payment method, they will be inclined to continue usage. ture or with cross-culture (e.g., individualism, collectivism) differ-
Retailers are also encouraged to absorb some of the transaction ences. Hence future studies should conduct cross-national research
costs rather than passing the cost to consumers. Telecommuni- such as by concentrating on consumers from different countries to
cation providers could offer subsidized NFC-enabled m-devices to strengthen and validate the research. Fifth, the study is restricted
spur adoption rate. Finally in order to increase MPT, businesses for to only six attributes to understand on the consumers’ intention to
example should encouraged their clients to post up their indepen- adopt SCC. There are obviously other important factors that have
dent reviews on the discussion board, social network sites, blogs been excluded which might influence IU. Research for example
and etc. since the construct is significant in this study. The commu- shown that influences system characteristics, government supports,
nication should be properly managed with prompt responds. This past adoption behaviour or other psychological behaviours influ-
is relevant for increasing consumer’s trust as some consumer based ences adoption. In addition different demographic variables such as
their decisions on the online reviews. Banks should also allow the income, education qualification or individual innovativeness may
possibility of free trial of SCC prior to adoption. Once tried, this di- also influence adoption (Sim, Tan, Ooi, & Lee, 2011). Future stud-
minishes the uncertainty faced by consumers and will lead to an ies should therefore integrate or extend the current theoretical
increase of use. Banks, telecommunication providers are also sug- model to discuss any discrepancies in terms of results rendered by
gested to be transparent regarding the issue of privacy. Customers the unrevealed constructs. Finally the study focuses on SCC which
need to be informed on what basis their private information is be- is at the early adoption stage and is an on-going process. This
ing shared. The information should be provided on the websites cross-sectional study did not portray how consumers’ attitude can
for consumers’ viewing. Apple for example has uploaded their pri- change over time (Tan et al., 2012). Davis et al. (1989) for example
vacy overview on their website (Apple.com, 2015). All of this can revealed that ease of use of technology adoption was significant
help to build trust with consumers which will lead to increase of after fourteen weeks of implementation contradicting to their ear-
reputation in the long run and higher intention to adopt SCC. lier findings. Kim, Mirusmonov, and Lee (2010) echoed the similar
sentiments whereby they concluded that the usefulness has no ef-
8. Limitations and future directions fect on m-payment systems for early adopters since useful features
cannot be expected from the technology at the initial stage. Thus
Our research has several weaknesses that future studies should future studies can employ a longitudinal study at the later stage of
address. First the majority of the respondents in our demographic SCC implementation as it would show the changes in consumer’s
information are slightly biased towards younger group with 54.1% behaviour.
of them below the age of 30.Although the official data from
Malaysia Communication and Multimedia Commission (MCMC) 9. Conclusion
(2014) indicates that majority of m-devices in Malaysia are young
users, we cannot claim that it is representative and therefore can- The research conducted on mobile studies for decades have
not be generalize to other areas or the elder. Many studies have been mainly centred on electronic commerce constructs. At the
found that older respondents are reluctant to adopt new innova- same time findings are also inconclusive due to the inconsistency
tion due to their risk adverse and lack of familiarity with technol- of theoretical models applied across different mobile studies. Fur-
ogy (Venkatesh et al., 2003). Additionally young consumers have thermore consumers react differently in an electronic environment
generally tended to be early adopters of new technology (Tan, as oppose to mobile. These lead to the difficulty in comparing the
Chong, Ooi, & Chong, 2010). Future research may explore different research findings and drawing meaningful conclusion. Recognizing
age groups to see if there are possible differences in IU. Second our the needs for a specific mobile constructs and model the study
research model measure IU as the dependent variable rather than draws on theories grounded from mobile technology literatures
actual behaviour given that the nature of SCC services which is still and proposed a new MTAM framework. The proposed framework
44 K.-B. Ooi, G.W.-H. Tan / Expert Systems With Applications 59 (2016) 33–46
was integrated with additional mobile constructs such as MPC, Apple.com. (2015). Apple Pay security and privacy overview – Apple Support. Avail-
MPSR, MPT and MPFT to increase predictive power. The frame- able at: https://support.apple.com/en-us/HT203027 (Retrieved 7 March 2015) .
Appleinsider.com. (2014). How to set up Apple Pay on an iPhone 6 or 6 Plus running
work was also applied to mobile users as approach to many past iOS.1, and where you can use it right now. Available at: http://appleinsider.
models which adopted respondents from organizations. The frame- com/articles/14/10/20/how- to- set- up- apple- pay- on- an- iphone- 6- or- 6- plus-
work which is exclusively tailored from the mobile perspective running- ios- 81- and- where- you- can- use- it- right- now (Retrieved 5 March
2015).
provides insights and valuable information for businesses, govern- Azhari, F. (2014). Quick detection of NFC vulnerability. Information Management &
ment agencies, retailers, merchants, banks in designing their mo- Computer Security, 22(2), 134–140.
bile marketing strategies and the development and provisions of Bagozzi, R., & Yi, Y. (1988). On the evaluation of structural equation models. Journal
of the Academy of Marketing Science, 16(1), 74–94.
mobile services with effective technological designs.
Balachandran, D., & Tan, G. W. H. (2015). Regression modelling of predicting NFC
mobile payment adoption in Malaysia. International Journal of Modelling in Op-
erations Management, 5(2), 100–116.
Baron, R., & Kenny, D. (1986). The moderator-mediator variable distinction in so-
Acknowledgement cial psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51(6), 1173–1182.
This is a revised version of a paper presented at the 6th In- Brown, S., Venkatesh, V., & Bala, H. (2006). Household technology use: Integrating
household life cycle and the model of adoption of technology in households.
ternational Research Symposium in Service Management (IRSSM- The Information Society, 22(4), 205–218.
6), UiTM, Kuching, Malaysia, 11-15 August 2015 by Garry Wei-Han Brown, S. A., & Venkatesh, V. (2005). Model of adoption of technology in the house-
Tan and Keng-Boon Ooi and was awarded the ‘Emerald Literati hold: A baseline model test and extension incorporating household life cycle.
MIS Quarterly, 29, 399–426.
2015 Highly Commended Paper Award’ by Journal of Service Man- BusinessNewsDaily.com (2015). 27 Ways to accept mobile payments. Avail-
agement, Emerald (UK). In addition, this research is supported by able at: http://www.businessnewsdaily.com/4457- mobile- payment- solutions.
the Universiti Tunku Abdul Rahman Research Publication Scheme htm (Retrieved 6 March 2015).
Chen, J., & Shen, X. (2015). Consumers’ decisions in social commerce context: An
(UTARRPS), under the project number 6251/G03. empirical investigation. Decision Support Systems, 79, 55–64.
Chen, K., & Chang, M. (2013). User acceptance of ‘near field communication’ mobile
phone service: An investigation based on the ‘unified theory of acceptance and
use of technology’ model. The Service Industries Journal, 33(6), 609–623.
Appendix A Chen, Y., Wu, J., Peng, L., & Yeh, R. (2015). Consumer benefit creation in online group
buying: The social capital and platform synergy effect and the mediating role
of participation. Electronic Commerce Research and Applications, 14(6), 499–513.
Cheong, S., Ling, H., & Teh, P. L. (2014). Secure encrypted steganography graphical
Acronyms password scheme for near field communication smartphone access control sys-
ANN Artificial Neural Network tem. Expert Systems with Applications, 41(7), 3561–3568.
AVE Average Variance Extracted Chiang, W. Y. K., Zhang, D., & Zhou, L. (2006). Predicting and explaining patronage
CO Complexity behavior toward web and traditional stores using neural networks: A compara-
CR Composite Reliability tive analysis with logistic regression. Decision Support Systems, 41(2), 514–531.
DOI Diffusion of Innovation Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, N.J.:
L. Erlbaum Associates.
EE Effort Expectancy
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance
IS Information System
of information technology. MIS Quarterly, 13(3), 319–339.
IT Information Technology Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer
IU Behavioural Intention technology: A comparison of two theoretical models. Management Science, 35(8),
MATH Model of Adoption of Technology in Households 982–1003.
M-device Mobile Device Digital Trends (2015). Visa boosts mobile payment security in Europe, ready for
MEU Mobile Ease Of Use eventual Apple Pay launch. Available at: http://www.digitaltrends.com/mobile/
MLP Multi-layer Perceptron apple- pay- partners- news/ (Retrieved 5 March 2015).
M-payment Mobile Payment Dutot, V. (2015). Factors influencing near field communication (NFC) adoption: An
MPC Mobile Perceived Compatibility extended TAM approach. The Journal of High Technology Management Research,
MPFR Mobile Perceived Financial Resources 26(1), 45–57.
MPSR Mobile Perceived Security Risk Egger, R. (2013). The impact of near field communication on tourism. Journal of Hos-
MPT Mobile Perceived Trust pitality and Tourism Technology, 4(2), 119–133.
Feijóo, C., Gómez-Barroso, J., & Ramos, S. (2016). Techno-economic implications of
MTAM Mobile Technology Acceptance Model
the mass-market uptake of mobile data services: Requirements for next gener-
MU Mobile Usefulness
ation mobile networks. Telematics and Informatics, 33(2), 600–612.
NFC Near Field Communication
Fisher, R. (1925). Statistical methods for research workers. Edinburgh, UK: Oliver &
PE Performance Expectancy Boyd.
PEOU Perceived Ease Of Use Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unob-
PLS-SEM-ANN Partial Least Square-Structural Equation servable variables and measurement error. Journal of Marketing Research, 18(1),
Modelling-Artificial Neural Network 39–50.
PTC Perceived Transaction Convenience Fransen, J., D’Hondt, E., Bourgois, J., Vaeyens, R., Philippaerts, R. M., &
PTS Perceived Transaction Speed Lenoir, M. (2014). Motor competence assessment in children: Convergent and
PU Perceived Usefulness discriminant validity between the BOT-2 Short Form and KTK testing batteries.
RA Relative Advantage Research in Developmental Disabilities, 35(6), 1375–1383.
RMSE Root Mean Square of Error Garner, P., Edwards, R., & Coulton, P. (2006). XEPS – Enabling card-based payment
SCC Smartphone Credit Card for mobile terminals. 2006 IEEE International Symposium on Consumer Electronics.
doi:10.1109/isce.2006.1689509.
TAM Technology Acceptance Model
Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2014). A primer on partial least
UTAUT Unified Theory of Acceptance and Use of
squares structural equation modeling (PLS-SEM). Los Angeles, CA: Sage Publica-
Technology
tions.
Haykin, S. (2004). Neural networks: A comprehensive foundation (2nd ed.). Upper Sad-
References dle. River, NJ: Prentice Hall.
Hew, J. J., Lee, V. H., Ooi, K. B., & Wei, J. (2015). What catalyses mobile apps usage
Aibinu, A., & Al-Lawati, A. (2010). Using PLS-SEM technique to model construction intention: An empirical analysis. Industrial Management & Data Systems, 115(7),
organizations’ willingness to participate in e-bidding. Automation in Construc- 1269–1291.
tion, 19(6), 714–724. Hoy, M. (2013). Near field communication: Getting in touch with mobile users. Med-
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human ical Reference Services Quarterly, 32(3), 351–357.
Decision Processes, 50, 179–211. Hsbc.com.hk (2015). HSBC Mobile Payments. Available at: https://www.hsbc.com.hk/
Ang, M., Ramayah, T., & Amin, H. (2015). A theory of planned behavior perspective 1/2/hk/cards/mobilepayments (Retrieved 6 March 2015).
on hiring Malaysians with disabilities. Equality, Diversity and Inclusion: An Inter- Hulland, J. (1999). Use of partial least squares (PLS) in strategic management re-
national Journal, 34(3), 186–200. search: A review of four recent studies. Strategic Management Journal, 20(2),
Apple (2015). Apple – Apple Pay. Available at: https://www.apple.com/apple-pay/ 195–204.
(Retrieved 4 March 2015).
K.-B. Ooi, G.W.-H. Tan / Expert Systems With Applications 59 (2016) 33–46 45
Zhou, T., & Lu, Y. (2011). The effects of personality traits on user acceptance of Zupanovic, D. (2015). Implementation model for near field communication in croa-
mobile commerce. International Journal of Human-Computer Interaction, 27(6), tian ferry ticketing system. Procedia Engineering, 100, 1396–1404.
545–561.
Zmijewska, A. (2005). Evaluating wireless technologies in mobile payments – A cus-
tomer centric approach. In Proceedings of the fourth International Conference on
Mobile Business (ICMB).