Computers in Human Behavior 50 (2015) 9–24

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: www.elsevier.com/locate/comphumbeh

Exploring consumer perceived risk and trust for online payments: An

empirical study in China’s younger generation
Qing Yang a,1, Chuan Pang b, Liu Liu c, David C. Yen d,⇑, J. Michael Tarn e
a
Institute for Financial Studies, School of Economics, Fudan University, Shanghai 200433, PR China
b
School of Business, Macao University of Science and Technology, Macao, PR China
c
School of Economics, University of Zurich, Switzerland
d
School of Economics and Business, 226 Netzer Administration Bldg., SUNY College at Oneonta, Oneonta, NY 13820, USA
e
Department of Business Information Systems, Haworth College of Business, Western Michigan University, Kalamazoo, MI 49008, USA

a r t i c l e i n f o a b s t r a c t

Article history: The uncertainties of transaction handling and consumer perception toward risk have been identified as
some of the major problems causing consumers’ hesitance toward taking advantage of online payments.
With the ever-growing implementation of trusting mechanisms for online payments, consumer
Keywords: confidence has greatly increased. This research explores the elements of perceived risk and trust – the
Online payment two most vital factors influencing consumer behavior of online payment – in the relatively mature stage
Perceived risk of China’s online payment environment. It also analyzes and classifies perceived risks of different nature
Trust
into two categories: systematic perceived risk and transactional perceived risk according to their
Trust beliefs
different roles in affecting consumer trust. Specifically, we propose a conceptual model of trust and
specific risk facets as well as two new constructs of comparison and evaluation, based on the framework
of TRA, TPB, TAM and DTPB. The proposed model was empirically tested using data gathered from 870
respondents in Shanghai and Macao, China, most of whom belong to the younger generation. The results
show that in the current stage of China’s online payment, consumers have built up trust first as an
antecedent of their perceived risks. Moreover, perceived total risk is negatively related to trust while
perceived risks can be classified into two types: system dependent risk which is positively related to trust
and transactional risk which is negatively related to trust. This article concludes with implications for
academia and practitioners.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction effectiveness and security of online payments. However, the pro-

In its e-Commence and Development Report 2001, the United
Nations Conference on Trade and Development (UNCTAD) defines
online payment as ‘‘the process of finance or payment mainly
using the medium of the Internet’’ (UNCTAD, 2001). According to
iResearch’s 2010 Online Payments in China Development Report
(www.iresearch.cn), online payments in China reached 10.1 trillion
RMB (or 160.3 billion USD) in 2010, which was more than double
from 2009. Online payments in China were predicted to be more
than 20 trillion RMB (or 292 billion USD) by 2012. The future
success of e-commerce in China depends not only on the integra-
tion of e-business models and services delivery, but also on the
⇑ Corresponding author. Tel.: +1 607 436 3458 (Office); fax: +1 607 436 2543.
E-mail addresses: qyang@fudan.edu.cn (Q. Yang), cpang@must.edu.mo (C. Pang),
liuliuxiaoshu@hotmail.com (L. Liu), David.Yen@oneonta.edu (D.C. Yen), mike.
tarn@wmich.edu (J. Michael Tarn).
1
Tel.: +86 13817651131; fax: +86 21 65112913.
cess of growth for online payments may involve uncertainty and
the sensitive subjects of security, privacy, and dependency, which
challenge consumers’ behaviors in the developing e-commerce
environment. Therefore, a comprehensive model describing factors
that either drive or restrain consumers to accept online payments
will offer insights to both academics and practitioners.
In the context of e-commerce, many researchers have spent
considerable effort trying to advance theoretical specifications by
studying possible predictors of consumer behavior. Researchers
argued that factors like intentions, attitudes, trust, and perceived
risk are critical in influencing consumer decisions in online shop-
ping (Bélanger & Carter, 2008; Dinev & Hart, 2006; Featherman &
Pavlou, 2003; Kim & Benbasat, 2006; Lopez-Nicolas & Molina-
Castillo, 2008; McKnight & Chervany, 2002; Shen & Chiou, 2010;
Teo & Liu, 2007). It is natural to assume that these factors will
apply in the area of online payment as well, considering that online
payment is a key part of e-commerce. Although e-trust and

http://dx.doi.org/10.1016/j.chb.2015.03.058
0747-5632/Ó 2015 Elsevier Ltd. All rights reserved.
10 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
perceived risk in some studies are investigated separately
(Cunningham, Gerlach, & Harper, 2005), the elements of trust
and perceived risk are probably the most important components
in past research. According to Aladwani (2001), internet security
and customer trust are the two most important challenges of
online banking. Hoffman, Novak, and Peralta (1999) focused on
security and privacy as the key drivers of online trust and Pavlou
(2003) combined two main streams of literature on TAM – e-trust
and risk. With the exceptions of some studies on online banking
(Yousafzai, Pallister, & Foxall, 2003), literature in the area of trust
and online payments is minimal and more focused on issues of
e-commerce from the standpoint of banks rather than of con-
sumers (Corbitt, Thanasankit, & Yi, 2003; Crespo & Rodríguez,
2008; Teo & Liu, 2007). Thus, financial service sectors will need
more multidimensional and specific analyses on the key underly-
ing factors affecting consumer behavior in order to increase their
market shares, which is the prime objective of our research.
Ho and Ng (1994) investigated consumer risk perception of
electronic payment systems, which mainly involved the
Electronic Fund Transfer (EFT) systems such as Point-of-Sales
(POS) and Automatic Teller Machine (ATM). Pavlou (2003) and
Featherman and Pavlou (2003) discussed perceived risk facets in
e-services adoption by combining risk and TAM without involving
e-trust. Yousafzai et al. (2003) gave merely a conceptual model of
trust with two main antecedents – perceived security and
perceived privacy – that were deemed to influence customer trust
in e-banking. Nicolaou and McKnight (2006) examined only the
effect of perceived information quality on total perceived risk
and trusting beliefs in interorganizational data exchange.
Featherman, Valacich, and Wells (2006) manipulated the risk of
e-payments within a controlled laboratory experiment but did
not test the effect of e-trust. Bélanger and Carter (2008) studied
the impact of trust and risk perceptions on one’s willingness to
use e-government services and defined trust as an exogenous
variable to affect negatively perceived risk, although it may be an
endogenous variable.
This study attempts to build on previous literature by proposing
a model comprised of detailed factors to better understand the
mechanism of consumer reaction and behavior, which has the
following three objectives:
 To propose a model comprising the interactive and complex
relationship between consumer perceived risk, trust, intention,
comparison and evaluation behavior in the context of online
payments based on the basic models and theories of consumer
behavior such as TRA, TPB, TAM and DTPB.
 To find the key perceived risk facets and types influencing
consumer trusting beliefs and behavior in online payments.
 To test the conceptual model of online payments to support the
antecedents and consequences of consumer trust and perceived
risk in online payments with samples from Shanghai and
Macao, China, which consist mostly of the younger generation.
This article is organized as follows: Section 2 provides literature
review and theoretical foundation. The research model is proposed
and the hypotheses are developed in Section 3. The research
method and data collection are discussed in Section 4, and data
and results analyses are elaborated in Section 5. The last section
provides the conclusions and implications of the study as well as
the research limitations and future studies.
2. Literature review and theoretical foundation
The fundamental stream of literature on consumer behavior
consists of the Theory of Reasoned Action (TRA) Ajzen &
Fishbein, 1980; Fishbein & Ajzen, 1975 and Theory of Planned
Behavior (TPB) Ajzen, 1991. Both are based on human intentions,
which helps lay down a solid theoretical basis for the Technology
Acceptance Model (TAM) widely studied by researchers since
Davis, Bagozzi, and Warshaw (1989) in the technology-driven
context of e-commerce study areas. Thereafter, Taylor and Todd
(1995) suggested the Decomposed Theory of Planned Behavior
(DTPB) to specify a technology adoption model in e-commerce.
TRA assumes that consumers be rational and not influenced by
unconscious inducement, so that they can unquestionably control
their own behavior. As a general basic behavior model, TRA
proposes that a consumer’s Actual Behavior (AB) is decided by his
or her Behavior Intention (BI), both of which are influenced by
personal Attitude toward Behavior (AT) and Subjective Norm (SN)
Dinev & Hart, 2006. However, TRA does not involve objective con-
straint variables such as self-control and situational variables from
outside of the system, which is a logical necessity because con-
sumer behavior will be obliged by the outside environment. Thus,
Ajzen (1991) added the extrinsic variable of Perceived Behavioral
Control (PBC) – including the broken line – into TRA to represent
consumer perception of required resources and opportunities to
perform the behavior of interest. This also influences consumer
behavior intention and indirectly determines the individual behav-
ior, thus resulting in the Theory of Planned Behavior (TPB). The
causal relations among variables are depicted in Fig. 1.
Technology Acceptance Model (TAM) states that consumer
behavioral intention (BI) is determined by the attitude toward
using (AT) which is influenced by innovation characteristics such
as Perceived Usefulness (PU) and Perceived Ease of Use (PEOU)
together Davis, 1989. As shown in Fig. 2, Davis (1989) eliminated
the constructs of subjective norm, as well as beliefs and motiva-
tions of TRA, while keeping the constructs of perceived behavior
control in TPB. PBC is represented by PU from the construct of
perceived use-performance of Expectation theory (Schultz &
Slevin, 1975) and PEOU from the construct of self-performance in
Bandura’s (1986) self-performance theory. Meanwhile, the parallel
relation between attitude and perceived behavior control in TPB is
replaced by the serial relation between attitude and PU or PEOU.
Although TAM is widely used in the context of e-commerce,
Legris, Ingham, and Collerette (2003) reported that it only inter-
preted 40–60% of consumer behavior intention, with nearly half
of the relative factors not explained. The main reason is that
TAM does not consider the subjective norm of TRA, even though
it is obvious that consumers will be impacted by their surround-
ings when they accept a new means of payment. Aiming to solve
the problem, scholars developed various forms of extended TAM
(Sun, 2007; Venkatesh & Davis, 2000). For instance, Yi, Jackson,
Park, and Probst (2006) suggested improving TAM by considering
Innovation Diffusion Theory, exterior and control variables, and
evaluation in the mature stage of e-commerce. There was also a
recent trial by Crespo and Rodríguez (2008) involving the
innovation diffusion process.
Combining TAM and TPB, Lee (2009) synthesized five specific
risk facets drawing from perceived risk theory – financial, secur-
ity/privacy, performance, social, and time risk – with the perceived
benefit of online banking and concluded that the intention to use
online banking is adversely affected mainly by the security/privacy
and financial risks and positively affected mainly by perceived
benefit, attitude, and usefulness. In terms of DTPB, it combines
TAM and TPB for the specific technology adoption model (Taylor
& Todd, 1995) since the overall TPB, with perceived usefulness
and ease of use added into TAM, is integrated with the third factor
of Perceived Compatibility (PCOM), which stands as the antecedent
of attitude toward technology (Fig. 3).
Based on theories of consumer behavior such as TRA, TPB and
DTPB, SN and PBC, consumers’ self-identity and behavior are
 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
Fig. 1. basic structure of TRA (Theory of Reasoned Action) and TPB (Theory of
Planned Behavior).
controlled by outside factors. Some researchers consider trust as
representative of SN and perceived risk as representative of PBC
in TAM (Chircu, Davis, & Kauffman, 2000; Jarvenpaa & Tractinsky,
1999; Pavlou, 2001; Schepers & Wetzels, 2007). Much empirical
evidence has confirmed that perceived risk and trust are the two
most salient factors in e-shop-related contexts (Cho, 2010; Corbitt
et al., 2003; Dinev & Hart, 2006; Featherman & Pavlou, 2003;
Malhotra, Kim, & Agarwal, 2004; Nicolaou & McKnight, 2006;
Pavlou & Gefen, 2004; Teo & Liu, 2007), particularly in the context
of online payment (Yousafzai et al., 2003). Furthermore, trusting
beliefs and risks are stated to exert a significant effect on behavioral
intentions, which are incorporated as predictors of willingness to
use e-service (Lopez-Nicolas & Molina-Castillo, 2008).
Hsieh, Po-An, Rai, and Keil (2008) recomposed PEOU as a factor
impacting PBC in TPB and found that it worked in understanding
digital inequality. Venkatesh, Morris, Davis, and Davis (2003)
conceptualized facilitating conditions as a construct that reflects
an individual’s perception about the control over his behavior
and formulated it by integrating constructs from perceived
behavior control of TPB. This means that the degree of perceived
control on transactions and the risk involved in online payment
actually influence consumer’s adoption behavior. Therefore, in this
article, the model with perceived risks (PR) as constructs of PBC is
proposed to provide full insight into the perceived behavioral
control variables.
On the other hand, trust is a complex concept which is very dif-
ficult to measure objectively. McKnight and Chervany (2002) view
it as the expectation that the system will not take advantage of a
consumer’s weakness in e-transaction, as well as a sort of attitude
of the consumer. Further, Li, Hess, and Valacich (2008) recomposed
it from multi-facet constructs in an initial trust model: Trust Bases
(basic trust factors forming beliefs), Trusting Beliefs, Trusting
Attitude, and Trusting Intention. Here, we may regard Trust as
trusting beliefs and attitude representative of SN, whereas PEOU,
PU, PR are constructed as trust bases related to external variables
forming PBC and affecting beliefs of consumer trust, which then
affects consumer intention and eventually their evaluation of
employing online payments. Therefore, we consider the simple
relationship among those variables as shown in Fig. 4 under the
framework of TAM and DTPB.
Besides the main structure, as a study that aims to improve the
TAM, we introduce two new concepts connected with consumer
attitude toward online payments. The first concept is the variable
of ‘‘comparison,’’ which is the attitude consumers hold when they
are comparing online payment with traditional payment. This is a
new concept introduced by our model to test its explanatory
power in consumer’s acceptance of online payment. The previous
study did not concern the psychological factor of comparison in
consumer’s decision in choosing online payment. Much of our
knowledge about ourselves, and about the world in which we live,
is based on a process of comparison. The general social comparison
Fig. 2. Basic structure of TAM (Technology Acceptance Model).
11
Fig. 3. Basic structure of DTPB (Decomposed Theory of Planned Behavior) (the real
line part is TPB (Theory of Planned Behavior)).
theory proposed by Leon Festinger (Festinger, 1954) states that
people will compare with others’ opinions and attitudes, but they
will also compare similar experiences to help make evaluations.
By determining the similarities and differences between traditional
and online payment, consumers will identify the relative advan-
tage and importance of online payment. This procedure should
be taken into account in explaining consumer’s behavior. The other
variable, ‘‘evaluation,’’ is introduced as a complement to intention.
Here, the concept of ‘‘evaluation’’ is different from the evaluation
construct in TRA, because it is a composition of consumers’ atti-
tudes before they are familiar with a certain new technology. It
is more of a post-hoc attitude, which will be formed in the mature
stage of a technology when consumers have already developed a
general impression of the technology – a situation matching the
current state of online payment in China. In our survey, there are
answers which show that the participants are not intending to
use online payment yet they show a positive attitude toward the
development and future of this tool. In order to understand fully
the relationship between subjective intention and objective
evaluation, we introduced this new construct to better depict the
current stage of China’s online payment. In our opinion, intention
is the antecedent of evaluation because it was after stepping out
of the online payment environment as a consumer that a more
objective opinion of the development of this technology could be
reached (see Fig. 5).
3. Research model and hypotheses
3.1. Research on perceived risk
3.1.1. Definition of perceived risk
The concept of perceived risk was originally proposed by
Harvard’s scholar Bauer (1964, 1967) in consumer behavior
research. Formally, Bauer (1967) defined perceived risk as ‘‘a
combination of uncertainty plus seriousness of outcome involved.’’
Peter and Ryan (1976) regarded the perceived risks as ‘‘the expec-
tation of losses associated with purchases and acts as an inhibitor
to purchase behavior.’’ Due to the difficulty of capture and
measurement as an objective reality, the notion of perceived risk
has been predominantly addressed and defined as a relevant
situational and specific contextual parameter by researchers in
e-commerce study (Bélanger & Carter, 2008; Crespo & Rodríguez,
2008; Gemunden, 1985; Teo & Liu, 2007; Yousafzai et al., 2003).
In the context of online payment, we accept the definition of per-
ceived risk given by Featherman and Pavlou (2003), ‘‘the potential
for loss in the pursuit of a desired outcome of using an e-service.’’
3.1.2. Facets of perceived risk
Facets of perceived risk are traceable to the research of
Cunningham (1967), who first identified two categories of
Fig. 4. Simple structure of perceived risk and trust for online payment.
12 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
Economic Perceived
Risk Usefulness
Function
Risk
Security
Risk
H4(a)
Time Risk
H1&H2 Total Risk
Privacy
Risk
Social Risk
H4(b)
Service
Risk
Perceived
Psychologic Ease of Use
al Risk
Fig. 5. Conceptual model on perceived risk and trust for online payments.
performance and psychosocial risk and further typified it with
six dimensions: performance/functional, financial/economic,
opportunity/time, safety/physical, social, and psychological loss.
Thereafter, Jacoby and Kaplan (1972) inferred an overall measure
of perceived risk from Bauer’s seminal work (Bauer, 1967). Stone
and Gronhaug (2007) reported that the former five to six factors
could explain almost 88.8% of all perceived risk. Next, Bellman,
Lohse, and Johnson (1999) argued the importance of time con-
sideration in predicting online purchase behavior. Featherman
and Pavlou (2003) reported the importance of privacy risk, overall
risk, and Cunningham’s (1967) five risks excluding physical risk by
testing TAM in e-services. At the same time, Corbitt et al. (2003)
also studied consumer risk perception considering performance,
financial, social, psychological, and time risk, but regarded reliabil-
ity, security, and privacy concern as technology trustworthiness.
Yousafzai et al. (2003) discussed economic, personal, functional,
and privacy risk related to transaction specific uncertainty in the
concept model of e-trust for e-banking. Closely related to online
payments, Cunningham et al. (2005) tested the original six facets
of risk given by Cunningham (1967). Furthermore, the mentioned
six facets of perceived risk were used to measure the overall risk
between the UK and China (Veloutsou & Bian, 2008). Thus, the
accountability of facets of perceived risk depends on the purchase
behavior, situation, and even cultural differences, which may vary
depending on specific transactions and environments (Stone &
Gronhaug, 2007; Veloutsou & Bian, 2008).
In practice, online payments do not involve physical risk; hence,
physical risk will not be considered as a risk factor in the following
study. Lopez-Nicolas and Molina-Castillo (2008) suggested that
technical risk, service risk, and delivery risk should also be consid-
ered in e-commerce. However, technical risk and delivery risk
should be contained in security risk concerning money transfer
safety in online payments. Therefore, we add both of them to
security risk and keep service risk in order to measure the proba-
bility of a poor payment service by the third-party or e-banking.
In the early stage of online payment, reliability or function of
the network is the most important concern for consumers. When
financial data is transmitted over the internet, there are risks that
unauthorized parties could intercept the information and cause
economic loss, making the economic and security risk as very
important perceived risk factors in online payments. However,
with the development of technology, the most important facet con-
cerning consumer perceived risk has changed. A newly developed
H8(a)
Comparison
H7(a)
H6
H3 Trust H5 Intention H9 Evaluation
H7(b) H8(b) Control Variables:
Gender; District;
Age; Experience
risk facet regarding privacy concerns is particularly salient in
understanding consumer risk perception (Dinev & Hart, 2006;
Featherman & Pavlou, 2003). Cunningham et al. (2005) did not
consider privacy risk in e-banking in his study. Furthermore the
importance of security risk in the later stage of e-service would
be replaced by privacy risk and economic risk, which are subject
to relatively easier exposure given increased technology used to
ensure safety.
Besides, time-conscious consumers are likely to guard against
the possible loss of time and functional risk because they chose
online payment for its advantages of convenience and time-saving
(Bellman et al., 1999; Featherman & Pavlou, 2003). However, to the
group of people who spend most of their time on the Internet and
would like to choose different channels of online payments, per-
ception of functional risk is even more significant than time risk
because consumers care more about stability and reliability of
their online payment environment instead of time.
Psychological risk and social risk may still play a role in the con-
text of online payment, but their effect will decrease due to the
digressive marginal utility. Finally, considering that overall risk
represents a general measure of risk, we also accept the construct
of perceived total risk which is added in the list for predicting
e-service adoption by Featherman and Pavlou (2003). However,
unlike their study, based on the understanding that the definition
and design of each perceived risk are different from each other, our
study assumes that the total perceived risk be a function of each
perceived risk to represent a general measure, instead of assuming
the existence of a second-order perceived risk measure (definition
and scales of perceived risks are listed in Appendix A). Therefore,
we propose the following hypotheses:
H1. (a) Perceived total risk in online payment includes the
multiple facets of risks: economic, functional, security, time,
privacy, social, service, and psychological risk, (b) of which
consumers are more sensitive to economic, function, and privacy
risks than others in perceived total risk by online payment means.
3.1.3. Types of perceived risk
Considering the exogenous effect of perceived risk composed of
behavior and environmental uncertainty, there exits two groups of
risk: (1) system dependent uncertainty risk caused by functional
defects, deposit costs or security problems in online payments
 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
and (2) transaction specific uncertainty risk caused by the conduct
of different parties who are involved in the online transaction
(Bélanger & Carter, 2008; Pavlou, 2003; Yousafzai et al., 2003).
Basically, system dependent uncertainty risk is characterized as
an exogenous variable because it is related to environmental
uncertainty of the transaction and its unpredictable nature
associated with unpredictable online payment technology, such
as potential technological sources of errors, current limited func-
tions, necessary time input and security gaps, which are all beyond
the control of online payments servers or consumers (Bélanger &
Carter, 2008; Pavlou, 2003; Yousafzai et al., 2003). However, trans-
action specific uncertainty is regarded as an endogenous variable,
namely market uncertainty resulting from decisions of different
parties caused by asymmetric information and an opportunistic
manner among them by taking advantage of a distant and imper-
sonal internet environment (Bélanger & Carter, 2008; Yousafzai
et al., 2003).
In the context of online payments, examples of exogenous sys-
tem uncertainty risk, suggested by Yousafzai et al. (2003), include
theft of credit card information, breach of private information, and
stealing of personal information by hackers. Endogenous transac-
tional risk examples include payment misrepresentation, false
identity demonstration, warranty denunciation, private informa-
tion leak, misleading advertisement, and outright fraud. So, in
our research, functional risk, security risk, time risks, and social
risks are classified as system dependent risk groups (S), which
are connected closely with the current situation of a certain stage
of online payments and cannot be eliminated due to their
systematic characteristic specifically described in survey items.
The economic, privacy, service, and psychological risks are
assigned to transactional risks (T), which are endogenous risks
since they mainly represent uncertainties triggered by possible
misbehaviors of various parties involved in the transaction
according to their definition and detailed items (see Appendix A).
As known in the financial field, system risk related to the public
environment cannot be reduced by portfolios of financial assets,
while transactional risks can be eliminated by investing in
different business cycle assets, the idea of which may be used for
reference in online payment environment. It means perceived risks
relative to system uncertainty are not easily changeable to the
consumer in short run until basic infrastructure conditions are
improved in online payment procedures. Whereas perceived
transactional risks caused by a certain party misbehaving, such
as misuse of consumer information, high costs, low quality of
service and even fraud, may be solved by choosing multi-channels
or the competition among different servers.
The different nature of perceived risks leads to different
consumer reactions. On one hand, consumers are sensitive to the
transactional risks that are prone to be dealt with: the more money
involved in online financial transactions, the greater the risk con-
sumers perceive. In that case, if consumers enlarge their transfer
amounts on the internet, higher risks associated with risks of
opportunistic behavior or moral hazard will be perceived due to
higher exposure to transaction specific uncertainty. Consumers
will increase their perceived total risk and decrease their trust in
e-banking or the third-party payment platform.
On the other hand, system dependent uncertainty may be
strongly influenced by the effectiveness of the third-party, trust-
certification bodies, and the public key encryption infrastructure
for ensuring basic system security to enhance trust of consumers
in the early stage of online payment (Hoffman et al., 1999), which
may decrease due to the improvement of technology and environ-
ment in e-service. If consumers perceive a higher degree of those
risks and are still willing to bear those uncontrollable system risks
(i.e., being attacked by a hacker) to transact by online payment,
this means that despite the existence of high system risks,
13
consumers perceive a comparatively low perceived total risk
and strongly trust the third payment platform or e-banking.
Consequently, we hypothesize the relationship between types of
risk and consumer total perceived risk as follows:
H2. (a) risks relative to system dependent uncertainty may be
negatively related with perceived total risk, while (b) risks may
contribute to perceived total risk positively if it involves transac-
tional uncertainty in online payment systems.
3.2. Research on trust in online payment
3.2.1. Definition and construct of trust
Mayer, Davis, and Schoorman (1995) argued that ‘‘trust is the
party’s expectation of fulfilling its specific obligations and willing-
ness to accept possible damage the transaction, irrespective of its
ability to control the other party.’’ Similar concepts also involve
perceived trustworthiness which is defined as ‘‘trusting beliefs’’
or treated as consumer attitude by McKnight and Chervany
(2002) and further decomposed into trusting beliefs and trusting
attitudes (Li et al., 2008).
In our investigation, we follow the definition of Mayer et al.
(1995) and refer to consumer trust as representative of trusting
beliefs and attitudes in online payments with a specific definition
as follows: ‘‘a psychological state leading to the willingness of cus-
tomers to perform payment transaction over the internet and expect
the payment platform fulfilling its obligations, irrespective of cus-
tomer’s ability to monitor or control payment platform’s actions.’’
3.2.2. Relationship between trust and perceived risk
Mayer et al. (1995) argued that ‘‘one does not need to risk any-
thing in order to trust; however, one must take a risk in order to
engage in trusting action.’’ According to their research study on
the relationship between trust and risk, it is not certain whether
risk is an antecedent to trust, or is an outcome of trust since trust
is interactive with risk.
Some authors cite that the relationship between trust and risk is
parallel, and the two factors together influence willingness to buy
in e-shopping. Kim and Prabhakar (2000) argue that an insignifi-
cant cause-effect relationship may exist between perceived risk
and trust. To some extent, trust and risk factors exist parallel with
interactive relations (Das & Teng, 2004).
Other research studies state that the relation between trust and
perceived risk is not parallel but serial, and trust affects consumer
intention through perceived risk. For example, Pavlou (2003)
regarded trust as a function of perceived risk. Also, trusting beliefs
have been found to have a negative influence on risk beliefs signifi-
cantly in privacy concerns of e-commerce (Malhotra et al., 2004),
in the community of Sellers (Pavlou & Gefen, 2004) or even in
the different cultural environments in the US, Singapore and
China (Teo & Liu, 2007). Yousafzai et al. (2003) also suggested that
trust is a function of the degree of perceived risk for electronic
banking in his concept model. However, the relationship between
trust and perceived risk is complex, and initial trust beliefs may
change with time and consumer experiences (Corbitt et al., 2003;
McKnight & Chervany, 2002). Combining trust and perceived risk,
Kim, Ferrin, and Rao (2008a, 2008b) verified the impact of
Internet consumers’ trust and perceived risk on their purchasing
decisions. Furthermore, consumer disposition to trust, reputation,
privacy concerns, security concerns, the information quality of
the Website, and the company’s reputation, have strong effects
on Internet consumers’ trust.
Generally speaking, in the early stage of online payments,
consumers may not trust this new way of payment without the
protection of regulation and modern technology, such as online
14 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
identification and public key encryption. As a result, initial institu-
tion-based trusting belief mechanisms are needed to facilitate
service success, such as the third party payment platform or
Certificate Authority (CA). Consumers trust the servers if they visit
an unknown e-server at first, and then they take risk in the trans-
action. Thus, system dependent risks are the most important
factors forming initial trusting during early stage of e-commerce.
However, it may vary from building trust to confirming and main-
taining trust over time. Perceived risk may also impact consumer’s
trusting beliefs when they use online payments many times and
become familiar with the transactional environment in the mature
stage of service. Transactional risk, particularly, may become a
more important factor related to trust beliefs (Jarvenpaa,
Tractinsky, & Vitale, 2000). Research of Corbitt et al. (2003)
supported that web users’ experiences positively influenced
perceived trust and their perceived risk negatively influenced
perceived trust. This meant that perceived risk was the antecedent
of trust in the later stage of e-commerce. In practice, consumer’s
initial trust in such institutions may reduce their perceived total
risk and improve their trusting beliefs to use it continuously.
Therefore, formation of a trusting belief may be based on the
level of perceived risk related to the environment and transaction
itself. The conceptual model of e-trust for electronic banking
proposed by Yousafzai et al. (2003) looked at perceived compe-
tence, security and privacy as the antecedents of trust, which
may be regarded as the perceptions of the internet environment
(McKnight & Chervany, 2002) or the part of trust base in the
research of Li et al. (2008). Furthermore, Kim and Benbasat
(2006) regarded risks concerning privacy, security, product price,
and customer service as trust-related factors, the less of which
would promote more trust in e-shopping. Dinev and Hart (2006)
gave us the evidence that perceived privacy risk also had negative
impacts on e-trust in e-transactions. These factors are resources of
risk which may cause consumer loss.
As a general measure, total perceived risk is commonly thought
of as representative of the uncertainty regarding possible negative
consequences of using a service. If consumers can predict the
prospect or the degree of loss based on information regarding their
experience when they conduct transactions using online payment,
their worries and uncertainties about the platform may be reduced
and their trusting beliefs in e-servers may be enhanced. Therefore,
a lower level of perceived risk may be related to a higher level of
trust beliefs:
H3. Consumer perceived total risk may influence negatively the
beliefs of their trust in the developing stage of online payment
3.3. Proposed research model
3.3.1. Consequences of perceived risk and the related factors
As discussed, technology adoption intention can be explained
partly by the concretization of perceived behavior control in terms
of PEOU and PU in TPB (Fig. 4). In the context of online payments,
consumers have been found to prefer using credit for the payment
of oil because of its ease of use (PEOU), low costs Carow & Staten,
1999, convertibility of funds, security, and trust, all of which are
important to the attitude toward using ATM electronic payment
(Abrazhevich, 2001). Also, PEOU has been found to have a negative
influence on perceived risk in the e-server field (Featherman et al.,
2006), and perceived usefulness (PU) has a positive effect on
attitude toward e-service (Kim & Forsythe, 2008). Therefore, the
following hypothesis is proposed:
H4. Consumer PU and PEOU may influence negatively their
perception of risk in online payment
Perceived risk is proved to be the key antecedent of intention
and willingness to purchase over the Internet (Liu & Wei, 2003),
and it also negatively influences consumer attitudes or intentions
toward online transactions (Featherman et al., 2006; Forsythe &
Shi, 2003; Pavlou, 2001). Consumer perceptions of e-service risk
reduce consumer beliefs and intention of using it (Featherman &
Pavlou, 2003). Similarly, Park, Lennon, and Stoel (2005) supported
the existence of a negative relationship between perceived risk and
apparel purchase intentions during an on-line product pre-
sentation environment. Lopez-Nicolas and Molina-Castillo (2008)
showed that the customer perceived risk associated with different
CKM tools had negative influence on consumer intentions. In
accordance with the above research, we present:
H5. Consumer perceived risks have an indirect negative impact on
the intention to use online payment through trusting beliefs and
attitudes.
3.3.2. Antecedents and consequences of trust
Generally, trust is believed to be the key in its relationships
with other variables, which needs to be clarified in our study.
Because the developing stage of online payment in China is the
main subject of our research, we take it for granted that the
payment environment with both traditional and online payments
available prompts consumers to make comparisons between them.
With the development of online payment, consumers’ understand-
ing and comparison of the two payment means influence the
degree of their trust in certain payments via their gathered
experiences in e-commerce. Therefore, consumers gain more
experiences and form a clear comparison between online payment
and traditional payment, which helps to establish their trust in
new technology as long as the comparison shows more attractive-
ness to online payment. Thus, we suggest:
H6. The more consumers prefer to use online payment channels
compared to traditional channels, the higher level they trust in
online payment.
It has been argued that trust in online payment systems has a
positive relationship with perceived behavior control such as PU
and PEOU, though the direction is uncertain. Gefen and Straub
(2000) and Chircu et al. (2000) showed that trust increased the
perceived usefulness and ease of use because it helped reduce
the consumer’s effort to understand, monitor, and control the
transactional processes. PEOU was identified as the antecedents
of trusting belief in e-shopping (Malhotra et al., 2004;
Pennington, Wilcox, & Grover, 2003). Moreover, some research
regarded trust as the function of perceived risk (Bélanger &
Carter, 2008), usefulness, and ease of use (Teo & Liu, 2007;
Yousafzai et al., 2003). The definition that consumer perceived use-
fulness, risk, and ease of use may be looked at as the factors of
trusting bases resulting in the outcome of trust and attitude
(Li et al., 2008; Pavlou, 2003), leading to behavior intention of
online payment and finally evaluating those means after they
choose the best one among multi online payment channels in the
developing environment of e-commerce. Then, we propose:
H7. (a) Consumer perceived usefulness significantly increases
their trust in online payment and (b) Consumer perceived ease of
use will positively affect their trust beliefs in the online payment
environment.
Consumers may be more inclined to use online payments if
more trust was created. Previous empirical research has revealed
that trust is significantly related to attitude and signifies consumer
purchase intentions (Jarvenpaa & Tractinsky, 1999; Kim, Chung, &
Lee, 2011; Yousafzai et al., 2003) in different culture situations
 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
(Teo & Liu, 2007). Basically, PEOU, PU, and trust are all positively
related to online purchasing attitude and intention (Van der
Heijden, Vehagen, & Creemers, 2003). It is then natural to assume
that PU and PEOU will also influence consumer’s comparison
between online payment and traditional payment. Besides, trust-
ing beliefs in internet payment environment is positively asso-
ciated with consumer behavior intention which will lead to a
positive evaluation toward the development of online payment.
Therefore, it follows that,
H8. (a) Consumer perceived usefulness and (b) Consumer per-
ceived ease of use in online payment both have positive impacts on
comparison.
H9. Consumer trusting beliefs and intentions contribute positively
to a good evaluation of online payments with objective opinions on
the future of online payment.
3.3.3. Control variables
Consumer traits such as gender, age, or experience usually
influence their risk perceptions and trusting beliefs and are used
as control variables related to consumer behavioral expectations
in order to find out meaningful implications in detail with the
sub-group sample study (Malhotra et al., 2004; Venkatesh,
Brown, Maruping, & Bala, 2008). Corbitt et al. (2003) adopted
income and education as the control variables to test the trust
model of e-commerce. Besides, useful control variables, for exam-
ple age, culture, and gender of consumers, are all good complemen-
tary variables for similar research (Yi et al., 2006). In this sense the
similar control variables on online payment in this research will be
involved in the following model.
3.3.4. Conceptual model
Based on the above discussion and literature review, we pro-
pose a relationship between perceived total risk and trust in the
online-payment model, with perceived risk composed of 8 facets;
namely, economic, functional, security, time, privacy, social, ser-
vice and psychological risks, parallel with perceived usefulness,
perceived ease of use. Together with comparison, the trusting
bases influence consumer trust beliefs or attitudes, then increases
their intention or willingness, and finally decides the degree of
evaluation on online payments, which will be influenced by control
variables such as gender, age, district and experiences characteris-
tics of different consumer groups. Thus far, the development of our
model is entirely in terms of theoretical constructs discussed above
and the conceptual model as follows:
4. Research method and data collection
4.1. Survey procedure, questionnaire design and scale development
A survey instrument was designed and employed to collect
representative and quantitative data for the model and hypothesis
testing. To refine the investigation on the subject, a pilot study and
extended study were conducted. The student samples from
Shanghai (159 usable surveys) were involved in the pilot study.
The second study was extended to samples from both Shanghai
and Macao, which consisted mostly of the younger generation,
and had an adjusted questionnaire with perceived total risk and
consumer experiences in online payments added after the pilot
model was tested.
Before the formal survey began, several items were dropped
and word changes were made. For example, subtitles which may
interfere with the participants’ judgments were excluded.
15
Additional items such as perceived total risk and the quality of e-
service were added according to the test model and suggestion of
consumers in practice. Specifically, in the survey, items of various
constructs were developed by adopting existing measures to the
research context in reference to previous studies and practical
situations of online payments. Items of different facets of perceived
risk such as economic risk, functional risk, privacy risk, time risk,
and psychological risk were designed according to the concept
and pattern of the instruments developed by Featherman and
Pavlou (2003) and Veloutsou and Bian (2008). Items of security risk
are from Pennington et al. (2003) and those of service risk,
psychological risk and social risk benefit from the research of
Lopez-Nicolas and Molina-Castillo (2008). The constructs of trust
and intention are setup under the framework of McKnight and
Chervany (2002), Mayer et al. (1995), and Jarvenpaa et al. (2000).
As for the items of comparison and evaluation which are intro-
duced freshly in our investigation, they are mainly designed for
the entire practicability of the model and relevant accountability
of the constructs. The items of the model with their definitions
and sources are presented in Appendix A.
All items are scored on a five-point Likert scale ranging from (1)
Strongly Disagree to (5) Strongly Agree. The survey questionnaire
consists of three sections. In the first section, respondents are
asked to answer questions on their online payment experiences,
Multi online payment means or channels, what they choose online
payment for and general evaluation. The second section, as a main
part of the survey, requires respondents to indicate the extent to
which they agree or disagree based on the model items. The third
section contains basic demographic characteristics including age,
gender, nationality, education level, occupation, monthly income,
and internet experience. Respondents are required to provide a
valid email address and are also encouraged to fill in any sugges-
tions or comments at the end of the questionnaire.
4.2. Data collection
In the process of the survey, ePanel offered to account for the
online survey part. However, considering a comparatively large
number of items and high quality control, the rigorous way of
on-the-scene guidance and monitoring was adopted and all the
surveys were collected in groups and checked before data input.
We gathered all 951 surveys, of which 81 unusable questionnaires
were excluded. Thus, the quality of this survey is especially high
with a response rate of 91.5%.
This formal research drew the samples from the population of
teachers, MBA on-the-job postgraduates from high-tech companies
and service businesses, undergraduates and postgraduates of a
large and prominent university in Shanghai, as well as stable
groups of office-bearers working for either the government or the
army. 265 undergraduate students from a large university in
Macao were also included in the survey. The relatively homoge-
neous population of the samples with mainly younger generation
represents an important target market for the research context
and is therefore considered appropriate and suitable for hypothe-
ses testing and development. The final survey respondent profile
(sample size 870) is given in Table 1 of Appendix B.
5. Data and results analyses
5.1. Samples of those who have experienced online payments
Our main objective was to study China’s online payments in its
developing stage. Therefore, the samples were reduced to only
those with ample online payment experiences; i.e. 648 sub-sam-
ples. We first analyzed each item to verify whether they were
16 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
eligible in discriminating between different subsamples. Then, the
exploratory principal components factor analysis with direct
oblimin rotation on perceived risks and the validity and reliability
of model constructs (Structural Equation Modeling, SEM) were
conducted using SPSS 15.0 and a LISREL 8.7 path diagram.
5.1.1. Item analysis
Item analysis was conducted to test each construct in discrim-
ination and distraction in terms of critical ratio. The t-values of
other variables are significant except for social risk (social_risk1,
p = 0.962; and social_risk2, p = 0.336). Therefore, the original
perceived risk facets are reduced to 7 risks. The exclusion of social
risk reveals that consumers generally do not possess obvious
sensitivity to the negative remarks or opinions from outside when
they decide to use online payments in China today.
5.1.2. Principal components analysis
The exploratory Principal Components Analysis (PCA) helps find
problem items – those whose loadings on the intended construct
are lower than one or have more cross-loadings. The result is
shown in Table 2 of Appendix B, and both the KMO (0.83) and
Bartlett’s Test (6174.544) were significant. All facets of perceived
risks extract 75% of the variance with all items of each construct
load highly (greater than 0.6) on corresponding common factor
and no items have cross-loadings above the standard maximum
cutoff of 0.4 (Boudreau & Gefen, 2001; Hair, Anderson, Tatham, &
Black, 1998). All the factors’ Eigen values are above 1.0 except that
of private risk which, however, is very close to 1.0. The results
support the hypothesis that the items of perceived risks are
discriminate.
5.1.3. Construct validity and reliability
Commonly, there are two approaches to assessing the conver-
gent validity of the constructs: Firstly, individual items exhibit
factor loading coefficients above the minimum standard of 0.5
and a result more than 0.70 would be excellent. Secondly, the
average variance extracted (AVE) for each construct is higher than
the recommended minimum value of 0.5 (Gefen & Straub, 2000).
According to the psychometric properties of the items in Table 3
after the confirmatory factor analysis (CFA), these two require-
ments are both met, reflecting unidimensionality and convergent
validity.
As for discriminant validity, it is usually assessed by testing
whether the correlations between pairs of construct items are
significantly different from unity. According to Fornell and
Larcker (1981), constructs have adequate discriminant validity if
the square root of the AVE for a construct is higher than the vari-
ance shared between the construct and other constructs in the
model. As can be inferred from Table 3 of Appendix B, the square
root of the AVE for each construct is above 0.7, which is definitely
higher than the correlations between each pair of constructs
shown in Table 4 of Appendix B, supporting the discriminate valid-
ity of the constructs. Measure validation is also examined for relia-
bility analysis by three approaches. Cronbach’s alpha coefficients
indicate that each construct exhibits strong internal reliability
(all above 0.70). A more rigorous yet commonly used standard is
composite reliability with a tolerable reliability above 0.70 which
is also met. Finally, like CR, AVE is also additional measure of inter-
nal consistency, the estimates of which are all above 0.50 as shown
in Table 3 of Appendix B, providing further evidence of reliability.
5.1.4. Model fit
The fit indices of the measurement models and corresponding
good model standards are both listed for comparison in Table 5
of Appendix B. The CFA model shows a converged and reasonably
fit result with a low v2 per degree of freedom (2.15) and excellent
fit indices (with CFI 0.98 and RMSEA 0.042). In sum, factor loading
coefficients, squared multiple correlations, composite reliability,
AVE, and model fit indices all suggest that the measured variables
explain much of the variances of corresponding latent constructs,
supporting the validity of the measures.
5.1.5. Structural equation modeling and results
As shown in Table 4, overall fit indices and effect weights for
SEM model indicate a good fit to the data: all goodness of fit
statistics are in the acceptable range of a well-fit model, only
5.1% of the variance of the data is not explained. All the completely
standardized path coefficients, t-values, and R2 of the research
model are provided in Fig. 6 for testing hypotheses.
For the sub-model of perceived risks, not all perceived risks
have a significant effect on perceived total risk. Economic risk,
functional risk and privacy risk are significantly related to
perceived total risk. The relatively weak effects of security, time,
service and psychological risks suggest that consumers who have
online payment experience usually do not pay as much attention
to those four risks as they do on economic risk, functional risk
and privacy risk, which supports H1(b). Different from previous
studies (Featherman & Pavlou, 2003) which suggest that security
risk is a significant risk construct, here we find it insignificant.
However, correlation between privacy risk and security risk is
the highest among correlations of different perceived risk facets
and the correlation of economic risk and security risk is also rela-
tively high (see Appendix B), which suggests a further direction
of research because it may mean that in the developing stage of
online payment when the coverage of security risk is decomposed
into that of economic and privacy risk, the latter two may be con-
sidered a replacement of the transactional risk part of security risk
from the perspective of consumers. In sum, this result indicates
that in the developing stage of online payment, the significant risk
perception may only include specific risks which from the angle of
consumers may be more lasting and influential in their life
compared with the other risks.
Besides, functional and time risks have a negative contribution
to perceived total risk, different from the other risk constructs.
Based on our analysis in Section 3.1.3, this likely reveals that these
two risks are relative to system dependent uncertainty and are
negatively related with perceived overall risk and positively
related to trust. It is a reasonable inference since according to the
design of items of these two risks, the functional risk primarily
depends on the stability of the online payment environment and
instrument, while time risk refers to the input of time due to
online payment operation, both of which are associated with
unpredictable Internet technology beyond the control of online
payments servers or consumers, which satisfies the components
of system dependent uncertainty risk. However, the relationship
between security risk and perceived total risk is positive, which
is not in line with our presumption that could be in some way
attributed to the replacement discussed in the above paragraph.
It has to be noted that in our definition and item design of
perceived risk, the dominant characteristic (system dependent or
transaction specific) of the risk does not exclude the existence of
the other characteristics due to the complexity in understanding
of each perceived risk in different culture environment. The other
risks: financial, privacy, service and psychological more or less
result from decisions of economic factors and are usually caused
by asymmetric information and opportunistic manner among
transaction parties, all of which can be categorized into transaction
uncertainty risks, thus partly supporting H2.
For the main body of the research model, trust, PU, and PEOU all
have a significant negative relationship with perceived total risk,
path coefficient 0.42, 0.21 and 0.23, respectively, supporting
H3 and H4. Trust is positively related with intention, path
 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
Perceived
Economic
Usefulness
Risk
43%
Function
Risk
0.19
-0.15 (3.21) -0.21
Security (-2.86) (-3.46)
Risk
0.04
(0.61)
(13.33)
-0.07 Total Risk
Time Risk
(-1.38) 31%
0.19
(2.76)
Privacy
Risk -0.23
0.04
(-3.90)
(0.80) 0.06
(1.12)
Service
Risk
Psychologic Perceived
al Risk Ease of Use
Fig. 6. Research model with results (SEM model).
coefficient 0.74 (p < 0.001), which also means that consumer per-
ceived risks have a negative impact on the intention to use online
payment through trusting beliefs, supporting H5. The newly intro-
duced construct of comparison proves to be a significant predictor
of trust with the path coefficient of 0.48 (p < 0.001), supporting H6.
However, H7 is not fully validated because of an insignificant
relationship between PU and trust, which does not verify H7(a), a
result indicating that in the developing stage of online payment,
perceived usefulness is no longer influential in the formation of
consumer trust. H7(b) is confirmed by the positive effect PEOU
exerts on trust. Totally, 78% of trust is explained by the model.
The path coefficient of PU and PEOU in comparison is 0.33 and
0.22, both significant, supporting H8. Likewise, H9 is confirmed
with the newly introduced construct of evaluation significantly
and positively related to intention. Together with H6, the
confirmation of H9 testifies that the introduction of constructs of
comparison and evaluation better explain consumer behavior
and make the research model and theory even more complete.
5.2. Overall sample analysis with control variables
To test whether there is any significant difference among differ-
ent groups classified by control variables such as gender, district,
age and online payment experiences, analysis of independent
samples t-test was run based on the entire usable sample collected.
The basic result of significance is outlined in Table 6 of Appendix B.
For gender, the sample was divided into male and female; for
district, into those from Shanghai and from Macao; for age, into
under 25 and above 25; for experience, into those with and
without online payment experience.
Given the significance level of 0.01, no significant difference
exists between male and female, which is not in line with the
results emphasized in the research of Gefen and Straub (1997)
and can be attributed to the current developing stage of online
payments in China compared with the timing of their research.
For different age groups, only time risk is obviously differently felt,
which according to statistics reveals that youngsters under
25 years old don’t feel time risk so strong as those above 25 years
old, many of whom have more concern for time arrangement after
beginning a career. However, for those from different parts of the
country and with different online payment experiences, the
17
0.33
(5.26)
Comparison
0.06 25%
(1.22)
0.48
0.66 (11.72)
-0.42 Trust 0.74 Intention 0.76 Evaluation
(-10.50) 78% (16.03) 55% (15.36) 58%
0.23
0.22
(4.99)
(3.76)
difference is significant in almost all the variables, indicating that
cultural background and experiences do affect how people form
trust and perceive risk in online payment context. To be specific,
those from Shanghai show less perceived total risk and more trust
in online payment. Likewise, a reasonable result reflects that those
who lack online payment experiences sense perceived risk more
but trust online payment less.
6. Conclusions and implications
6.1. Contributions
This study explored the relationship between perceived risk and
trust in online payment area by assigning Perceived risk, PU and
PEOU as specific descriptive constructs of perceived behavioral
control (PBC). We put forth the structural model based on TPB,
TAM and DTPB which offer a sound theoretical foundation for
the research model. First, to our best knowledge, this research is
the first empirical study to classify and test eight facets (functional,
social, service, psychological, economic, time, privacy and security)
into two types of perceived risks (system dependent risk related to
infrastructure and transactional risk relative to transaction parties
and processes), which investigate their impacts on perceived total
risk and trust in the online payment environment in China. We
classified consumer perceived risk into system dependent uncer-
tainty risk and transactional uncertainty risk, and propose the
assumption that risks relative to system dependent uncertainty
may be negatively related with perceived total risk. The assump-
tion we empirically testified has significance in explaining the
great potential of online payment in China, because system depen-
dent uncertainty risk is more or less ignored by consumers pursu-
ing convenience and usefulness of online payment, which indicates
e-banking or third-party payment platform should focus more on
reducing transactional risk.
The conceptual model we propose explains well China’s current
stage online payment data, which shows that even based on tradi-
tional theories, in modern China, a modified model is valid and
powerful in studying new economy’s e-commerce development.
With its large population of online users and growing number of
consumers adopting online payment, this model suited especially
18 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
for the current stage of China’s online payment have its obvious
significance.
Secondly, unlike most other studies which do not include the
construct of trust or consider trust as an antecedent of other con-
structs in consumer behavior models (Cunningham et al., 2005;
Pavlou, 2003; Yousafzai et al., 2003), we consider perceived risk
to be the antecedent of trusting beliefs and attitudes, which
explains the significant negative relationship between them. This
idea is acquired through studying the current stage of China’s
online payment: unlike immature stages where only after perceiv-
ing and evaluating the online payment tools’ risks would con-
sumers build up their confidence, in the second stage of China’s
online payment, consumers who become familiar with this new
payment way already possess trust in it, but further consideration
of possible risks might interact with their trust which still affects
their attitude. Moreover, we introduce new constructs of compar-
ison and evaluation to better explain consumer behavior in online
payments, which are both supported by the empirical study.
Last but not least, given the characteristics of our samples, we
investigated whether control variables representing different cul-
tural backgrounds and experiences affect consumer behavior in
online payment and obtain corresponding results of confirmation.
Theory-guided empirical studies of consumer behavior in e-
shopping have been making progress in the past years. However,
studies on consumer behavior in online payment as a specific
and independent process are fewer. This research study fills in this
gap. Further, as one of the first studies, it collected and compared
the samples from different districts in China and analyzed
corresponding characteristics of the samples. The results, which
offer specific and new insights into this area, are in accordance
with the present situation of the development of online payment
in China as well as its fast growing number of online payment
users. In our opinion, this model for mature online payment stage
could also be adapted to further study in other countries with
mature e-commerce such as the USA and EU.
6.2. Limitations and future research
Although 78% of trust has been explained by the model, the
problem proposed by Legris et al. (2003) still exists in our model
which only interprets 55% of consumer behavior intention, with
45% not explained. The problem, though it may not have been
mentioned, also appears in previous research studies (Kim &
Forsythe, 2008; Nicolaou & McKnight, 2006; Pavlou & Gefen,
2004). They attempted to integrate trust and perceived risks into
consideration, which also suggests that the unrevealed factors
affect consumer behavior intention. Even though we collected
samples with good representation, a portion of the subjects were
recruited from college students. As listed in the descriptive fea-
tures of the sample, MBA students occupied nearly 20% of the
entire sample. Thus, the results seem to be derived from a homoge-
neous sample base. However, MBA students come from all walks of
life, so, in this case, homogeneity is not a concern. In addition,
Shanghai and Macao represent two districts in China with a similar
economic level. Future research should be expanded to include
broader differences in subject demographics and in different set-
tings. Also, items may need to be more precise and avoid overlap-
ping in order to clearly study their respective impacts in terms of
subjective factors and cultural background. Another limitation is
related to the development of trust. It has been noted that trust
develops gradually over time, which suggests that the consecutive
study on different stages of online payment would describe the
technology acceptance procedure more completely. We investi-
gated the developing stage of online payment but have not com-
pared the initial and mature stages of online payment. The above
provides a possible direction for further research.
6.3. Implications for academia
This study represents an important step toward understanding
the online payment environment in China and Chinese consumer
behavior, focusing on younger generation. The research model is
adapted from the framework of TRA, TPB, TAM, and DTPB to pro-
vide the basic logical sequence for perceived risk and trust beliefs.
It demonstrates that this extended model can explain a significant
amount of variance in the developing stage of online payment.
From a theoretical perspective, this study provides a comprehen-
sive analysis of consumer behavior in the context of developing
online payment environment, distinct from previous studies con-
cerning on early stage of a new technology. Furthermore, the com-
bined model bridges two important factors, trust and perceived
risks, with PU and PEOU. It also successfully introduces two new
constructs of comparison and evaluation to better explain the
subject of consumer behavior, which provides another direction
for future research in online payment.
In particular, the different risk perceptions which were identi-
fied as significant in this study point to a new direction for helping
to understand risk perceptions that tend to persist or disappear
with the development of new technology. Unlike Featherman
and Pavlou (2003), this study assumes that the total perceived risk
be a function of each perceived risk to represent a general measure
instead of the existence of a second-order perceived risk measure.
Furthermore, the different results of this study from previous
research on similar topics (Veloutsou & Bian, 2008) show that con-
sumers are more sensitive to economic, functional, and privacy
risks as well as functional, security and time risks which could
be categorized as the system dependent risk group in line with
both theory and practice. Therefore, in the long run, security risk
may be replaced by privacy risk and economic risk, which may
be a new characteristic of online payments in its developing stage.
However, our results are not fully in line with other previous
studies in that the lack of effect of PU on trust is somewhat
confusing. Our explanation is that in the developing stage of online
payment, only a small monetary amount per transaction is allowed
compared with traditional payment although consumers under-
stood the usefulness of this new technology. The advantage of
online payment attributes to more on its ease of use as obviously
different from traditional payment. Nevertheless, the insignificant
relationship between PU and trust requires further confirmation
and explanation to identify whether in the developing stage of
online payment the influence of some factors will decrease.
Finally, the analysis of control variables reveals that age, dis-
trict, and experience as background factors do affect consumer’s
perceived risk and trust in the context of e-service, suggesting
the need of more research of comparison to reveal other major fac-
tors that will play a role when samples are large enough to com-
prise different characteristics of consumers. The observed
between-group differences of this study warrant elaboration and
further examination as they represent promising directions for
more insight into differential behavioral patterns and their causes.
6.4. Implications for practitioners
The results from this study support the contention that e-ser-
vices of less risky categories will be trusted and intended to be
used more. From a pragmatic perspective, the discovery of eco-
nomic risk, functional risk, and privacy as important factors that
contribute to perceived total risk in China offers a hint to organiza-
tions involved in online payments to recognize the importance of
these factors of online payments in modern China. For business
corporations, improvement in maintaining good reputation by pro-
tecting consumers’ privacy and more input in stabilizing the online
payment environment will be the best approach to attracting
 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
potential consumers and satisfying the existing customers. In addi-
tion, the results of this study suggests that in comparison with
their traditional and online competitors, banking institutions and
third party payment platforms involved in online payments should
promote the relatively ease of use of their online payment proce-
dure and instrumental design to their customers.
The research results show that the development of online pay-
ment as a new way of payment method has been accepted and
Appendix A
The definition of perceived risk and trust model items (T: Transactional uncertainty risk; S: System dependency risk).
Risk type Risk definition
Overall risk A general measure of feelings of
(Total_risk1-2) uncertainty, discomfort and/or
anxiety, concern for online
payments aroused in
consumers
Economic risk (T) The risk because of potential
(Economic_risk1-3) monetary losses or financial
prices and fees overcharged by
service suppliers
Functional risk (S) The risk that the online
(Function_risk1-2) payments will not render the
benefits as consumers expect
due to limitations of the
instrument and system
Security risk (S) The risk that the system or
(Security_risk1-4) software of e-banks and third
parties is not stable or
vulnerable to network attacks
Privacy risk (T) The risk of disclosure of
(Private_risk1-3) consumer’s private information
due to carelessness or illegal
activities of banks, third parties
or others
Time risk (S) The risk of time wasting in
(Time_risk1-3) application, learning and
waiting for the completion of
the entire online payment
process
19
adopted by customers in a fast upward trend, especially for those
who are more experienced with online payment and show more
trust beliefs in online payment. For practitioners such as public pol-
icy makers and online payment servers, understanding this trend
and developing the more suitable environment for online payment
is critical. Furthermore, understanding what key control variables
affect consumer behavior across different groups is also necessary
for effectively planning and, thus, implementing online payment
control for better and more successful future development.
Constructs: (strongly disagree/agree, 1–5 scale) Sources
Generally, I do NOT think online payment is Jacoby and Kaplan
risky (1972)
It’s rare to meet unexpected accidents in online Featherman and Pavlou
payment (2003)
I worry: Featherman and Pavlou
(2003)
 that network cost and other cost will make Lopez-Nicolas and
total expenditure higher than expected (T) Molina-Castillo (2008)
 that if I request refunding, the third party Veloutsou and Bian
payment platform won’t coordinate (T) (2008)
 that monetary loss can be caused by impro-
Yousafzai et al. (2003)
per operations of third party platform (T)
Corbitt et al. (2003)
 that limited functions of online payment Yousafzai et al. (2003)
tools won’t satisfy my demands (S) Lopez-Nicolas and
 that I can’t pay online due to unstable net- Molina-Castillo (2008)
work (S) and Veloutsou and Bian
(2008)
 that the transaction information will be Pennington et al.
hacked (S) hackers or viruses caused by (2003)
security lapses of operation system or online
payment tools (S)
 that my account information will be illegally
used (T) that payment platform cannot work
properly (S)
 that my personal information can be stolen Featherman and Pavlou
in transaction because of carelessness or (2003)
illegal activities of others including banks, Veloutsou and Bian
third parties, etc. (T) (2008)
 that my personal information can be Yousafzai et al. (2003)
improperly collected and used by venders,
banks, or service providers (such as junk
mails) (T)
 that hackers can steal my personal informa-
tion from service providers’ database and
use it illegally (T)
 that it will cost me much time to apply for Featherman and Pavlou
an online payment account (S) (2003)
 that it will cost me much time to learn to use Veloutsou and Bian
online payment (S) (2008)
 that the online payment process will cost
Corbitt et al. (2003)
me much time in confirming ID and other
document (S)
(continued on next page)
20 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
Appendix A (continued)
Risk type Risk definition
Service risk (T) The risk of encountering low
(Service_risk1-2) quality service in online
payments
Psychological risk (T) The risk that the uncertainty of
(Psycho_risk1-3) above mentioned risks will
affect the consumers’ mood or
cause anxiety
Social risk (S) The risk of potential loss of
(Social_risk1-2) status in one’s social group
because of being untrendy or
failure in adopting the product
or service.
Perceived Usefulness The prospective user’s
(usefulness1-3) subjective probability that
using online payment means
will increase one’s performance
within daily-life context
Perceived Ease of Use The degree to which an
(PEOU) individual believes that
(easy_to_use1-5) performing the behavior of
interest would be free of effort
when they adopt online
payments
Trust belief (e_trust1-3) A psychological state leading to
the willingness of consumers to
perform payment transaction
over internet and expect the
payment platform fulfilling its
obligations, irrespective of
consumers’ ability to control
payment platform’s actions
Comparison The attitude consumers hold
(Comparison1-3) when they are comparing the
online payment with the
traditional payment means
Intention (Intention1-3) Consumer’s inclination to use
the online payment
Evaluation (Evaluation An objective evaluation on
1-3) online payments consumers
form when they are in the later
stage of online payment
Constructs: (strongly disagree/agree, 1–5 scale) Sources
 that if I apply for an online payment Lopez-Nicolas and
account, it’s difficult to communicate with Molina-Castillo (2008)
the bank or third party platform when I need Corbitt et al. (2003)
help (T)
 that if I use an online payment account, it’s
difficult to communicate with the bank or
third party platform when I need help (T)
 that I will feel pressure if there is loss during Featherman and Pavlou
online payment (T) (2003)
 that I will be anxious before the payment is Veloutsou and Bian
confirmed (T) (2008)
 that I will be anxious if the payment can’t be
Lopez-Nicolas and
transferred timely (T)
Molina-Castillo (2008)
Corbitt et al. (2003)
 that once be cheated in online payment, my Featherman and Pavlou
friends will have negative comments on me. (2003)
(S) Lopez-Nicolas and
 that if I don’t use online payment, my Molina-Castillo (2008)
friends will think I’m out. (S) Corbitt et al. (2003)
I believe: Davis et al. (1989)
 online payment can save much time and
speed up money transfer during online
purchasing.
 online payment can make online purchasing
process more efficient.
 online payment is more useful and conveni-
ent compared with other payment methods,
all things considered.
I think: Davis (1989)
 Online payment is very simple and Davis et al. (1989)
convenient
 Online payment is very easy to use
 Guideline of online payment is very clear
 Online payment won’t take much time and
effort
 Online payment operation process is easy to
learn
 Although it’s risky, I still willing to pay Mayer et al. (1995)
online
 I would like to encourage my friends to pay Yousafzai et al. (2003)
online
 I feel comfortable when pay online
 Compared with traditional payment system,
I am more interested in online payment
 Compared with traditional payment system,
I prefer to transact with online payment
 I choose online payment because it gradu-
ally replaces some outdated ways of tradi-
tional payment
 In general, I’m willing to use online payment Jarvenpaa and
 If possible, I will go on choosing to pay Tractinsky (1999)
online Jarvenpaa et al. (2000)
 Although there are some merits of online
payment, I do NOT want to use it
 Generally, I think online payment is positive
 I believe online payment have a brighter
future
 Objectively, online payment is a trend
 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24 21

Appendix B

See Tables 1–6

Table 1
Descriptive statistics of survey respondents (N = 870).

Gender Age Income

Male 394(45.3%) <18 years 2(0.2%) Unstable 397(45.6%)
Female 459(52.8%) 18–25 years 460(52.9%) ¥ 1000–¥ 3000 131(15.1%)
Missing 17 (2%) 26–30 years 211(24.3%) ¥ 3000–¥ 5000 110(12.7%)
Occupation 31–35 years 134(15.4%) ¥ 5000–¥ 10,000 86(9.9%)
Students 447(51.4%) 36–40 years 40(4.6%) >¥ 10,000 132(15.2%)
MBA (on the job)a 186(21.4%) 40–50 years 19(2.2%) Missing 14(1.6%)
Company employee 194(22.3%) >50 years & missing 4(0.4%) Experience for e–service
Work for government 75(8.6%) Experience for online payments By internet 683(78.5%)
Service business 30(3.4%) Experienced 648(74.6%) Nonexperienced 186(21.4%
Teachers 41(4.7%) Nonexperienced 221(25.5%) Missing 1(0.1%)
Others 77(8.9%)/6(0.7%) Missing 1(0.1%)
a
MBA(on the job) were double counted as both MBA students and employees.

Table 2
Principle Component Factor Analysis of the 7 perceived risks and perceived total risk.

Component
SEC TIM PSY TOT ECO SER FUN PRI
Eigen values 6.201 2.278 1.892 1.726 1.291 1.199 1.086 0.959
Cumulative% 28.185 38.538 47.139 54.984 60.854 66.304 71.241 75.601
Security_risk2 .898
Security_risk1 .880
Security_risk3 .862
Security_risk4 .627
Time_risk2 .854
Time_risk1 .842
Time_risk3 .777
Psycho_risk2 .905
Psycho_risk3 .884
Psycho_risk1 .764
Total_Risk2 .916
Total_Risk1 .912
Economics_risk1 .849
Economics_risk2 .841
Economics_risk3 .694
Service_Risk1 .935
Service_Risk2 .885
Function_Risk2 .878
Function_Risk1 .855
Private_risk2 .919
Private_risk1 .776
Private_risk3 .770

Table 3
Confirmatory factor analysis statistics.

Latent construct Measured variable Complete standardized loadings t-values R2 Cronbach’s a Composite reliability AVE
Comparison Comparison1 0.86 26.31 0.74 0.84 0.85 0.66
Comparison2 0.92 28.68 0.84
Comparison3 0.62 16.85 0.39
Trust e_trust1 0.74 20.92 0.55 0.79 0.78 0.56
e_trust2 0.74 20.71 0.54
e_trust3 0.76 21.42 0.57
Total risk Total_Risk1 0.89 25.00 0.80 0.82 0.83 0.70
Total_Risk2 0.78 21.37 0.61
Intention Intention1 0.90 28.08 0.81 0.81 0.83 0.63
Intention2 0.83 25.02 0.69
Intention3 0.62 16.87 0.39
Evaluation Evaluation1 0.69 19.00 0.48 0.81 0.82 0.60
Evaluation2 0.87 25.89 0.76
Evaluation3 0.76 21.50 0.58

(continued on next page)

22 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24

Table 3 (continued)

Latent construct Measured variable Complete standardized loadings t-values R2 Cronbach’s a Composite reliability AVE
Perceived usefulness usefulness1 0.80 22.31 0.64 0.76 0.78 0.55
usefulness2 0.81 22.68 0.65
usefulness3 0.60 15.63 0.36
Perceived ease of use easy_to_use1 0.76 22.14 0.58 0.90 0.90 0.64
easy_to_use2 0.84 25.48 0.70
easy_to_use3 0.79 23.39 0.62
easy_to_use4 0.79 23.35 0.62
easy_to_use5 0.83 24.98 0.68
Economic risk Economic_risk1 0.58 14.69 0.34 0.76 0.77 0.53
Economic_risk2 0.78 20.95 0.61
Economic_risk3 0.81 21.78 0.65
Function risk Function_Risk1 0.78 14.86 0.60 0.70 0.70 0.54
Function_Risk2 0.68 13.74 0.46
Security risk Security_risk1 0.86 26.74 0.74 0.86 0.87 0.63
Security_risk2 0.90 28.82 0.82
Security_risk3 0.81 24.40 0.66
Security_risk4 0.57 15.18 0.32
Time risk Time_risk1 0.70 18.35 0.50 0.78 0.79 0.55
Time_risk2 0.84 22.37 0.70
Time_risk3 0.68 17.61 0.46
privacy risk Privacy_risk1 0.78 22.02 0.60 0.84 0.83 0.63
Privacy_risk2 0.77 21.76 0.59
Privacy_risk3 0.83 24.12 0.69
Service risk Service_Risk1 0.82 21.96 0.68 0.87 0.88 0.78
Service_Risk2 0.94 25.39 0.88
Psychological risk Psycho_Risk1 0.69 18.56 0.47 0.83 0.83 0.62
Psycho_Risk2 0.81 22.77 0.66
Psycho_Risk3 0.86 24.52 0.74

Table 4
Correlation matrix of principle constructs.

Mean SD Correlation matrix

1 2 3 4 5 6 7 8 9 10 11 12 13 14
Comparison 3.32 0.81 0.81
Trust 3.32 0.73 0.62 0.75
Total Risk 2.82 0.89 0.43 0.58 0.83
Intention 3.78 0.66 0.51 0.59 0.41 0.79
Evaluation 3.97 0.59 0.46 0.50 0.35 0.64 0.77
PU 3.95 0.70 0.43 0.40 0.29 0.49 0.46 0.74
PEOU 3.81 0.74 0.39 0.49 0.31 0.46 0.38 0.58 0.80
Economic 3.64 0.82 0.08 0.20 0.24 0.11 0.02 0.02 0.04 0.73
Function 3.20 0.87 0.10 0.10 0.08 0.04 0.09 0.07 0.11 0.21 0.73
Security 4.07 0.70 0.00 0.05 0.16 0.17 0.23 0.17 0.13 0.42 0.18 0.79
Time 3.08 0.86 0.10 0.19 0.10 0.19 0.12 0.17 0.29 0.19 0.15 0.10 0.74
Privacy 4.04 0.76 0.01 0.15 0.21 0.12 0.13 0.15 0.06 0.40 0.12 0.56 0.23 0.79
Service 3.36 0.93 0.08 0.12 0.18 0.08 0.01 0.08 0.10 0.33 0.20 0.27 0.30 0.36 0.88
Psychological 3.39 0.85 0.07 0.16 0.17 0.10 0.01 0.09 0.06 0.35 0.27 0.32 0.17 0.27 0.33 0.79
⁄
The bold value on the diagonal is the square root of the Average Variance Extracted (AVE).

Table 5
Goodness of fit statistics for the measurement and structural model.

Goodness of Fit v2 (d.f.) v2/d.f. NFI CFI IFI RFI GFI AGFI RMSEA
Good model fit <3.00 >0.90 >0.90 >0.90 >0.90 0.90 >0.80 <0.080
CFA model 1561.96 (728) 2.15 0.96 0.98 0.98 0.95 0.90 0.87 0.042
SEM model 2068.46 (780) 2.65 0.95 0.97 0.97 0.94 0.87 0.85 0.051
 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24 23

Table 6 Fishbein, M., & Ajzen, I. (1975). Attitude, intention and behavior. Reading, MA:
Significance of difference between classified groups by control variables. Addison-Wesley Publishing Co..
Fornell, C., & Larcker, D. (1981). Structural equation models with
Gender District Age Experience unobservable variables and measurement error. Journal of Marketing Research,
18(1), 39–50.
Economic risk 0.109 0.855 0.305 0.008
Forsythe, S., & Shi, B. (2003). Consumer patronage and risk perceptions in Internet
Functional risk 0.486 0.000 0.519 0.044
shopping. Journal of Business Research, 56(11), 867–875.
Security risk 0.798 0.008 0.182 0.365 Gefen, D., & Straub, D. W. (1997). Gender differences in the perception and use of e-
Time risk 0.568 0.383 0.002 0.002 mail: An extension to the technology acceptance model. MIS Quarterly, 21(4),
Privacy risk 0.229 0.000 0.061 0.344 389–400.
Social risk 0.423 0.720 0.261 0.038 Gefen, D., & Straub, D. W. (2000). The relative importance of perceived ease of use in
Service risk 0.411 0.000 0.077 0.939 IS adoption: A study of e-commerce adoption. Journal of the Association for
Psychological risk 0.469 0.041 0.177 0.099 Information Systems, 1(8), 1–30.
Total risk 0.833 0.000 0.439 0.000 Gemunden, H. G. (1985). Perceived risk and information search: A systematic
PU 0.046 0.000 0.165 0.000 meta-analysis of the empirical evidence. International Journal of Research in
PEOU 0.040 0.000 0.967 0.000 Marketing, 2, 79–100.
Trust 0.479 0.000 0.399 0.000 Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (Eds.). (1998). Multivariate
Comparison 0.031 0.000 0.036 0.000 data analysis with readings (5th ed.. Englewood Cliffs, NJ: Prentice Hall.
Ho, S., & Ng, V. (1994). Customers’ risk perceptions of electronic payment systems.
Intention 0.089 0.000 0.430 0.000
International Journal of Bank Marketing, 2(8), 26–38.
Evaluation 0.079 0.000 0.914 0.000
Hoffman, D., Novak, T., & Peralta, M. (1999). Building consumer trust online.
Communications of the ACM, 42(4), 80–85.
Hsieh, J. J., Po-An Rai, A., & Keil, M. (2008). Understanding digital inequality:
Comparing continued use behavioral models of the socio-economically
References advantaged and disadvantaged. MIS Quarterly, 32(1), 97–126.
Jacoby, J., Kaplan, L. (1972). The components of perceived risk. In Proceedings of the
Abrazhevich, D. (2001). Electronic payment systems: Issues of user acceptance, 3rd annual conference for consumer research (pp. 382–393).
working paper. <http://www.ipo.tue.nl/homepages/dabrazhe/ps/Library/data/ Jarvenpaa, S., & Tractinsky, N. (1999). Consumer trust in an internet store: A
ebiz2001.PDF>. cross-cultural validation. Journal of Computer Mediated Communication, 5(2),
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human 1–35.
Decision Processes, 50(1991), 179–211. Jarvenpaa, S., Tractinsky, N., & Vitale, M. (2000). Consumer trust in an internet store.
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social Information Technology and Management, 1, 45–71.
behavior. Englewood Cliffs, NJ: Prentice-Hall. Kim, D., & Benbasat, I. (2006). The effects of trust-assuring arguments on consumer
Aladwani, A. M. (2001). Online banking: A field study of drivers, development trust in Internet store: Application of Toulmin’s model of argumentation.
challenges, and expectations. International Journal of Information Management, Information System Research, 17(3), 286–300.
21, 213–225. Kim, M. J., Chung, M., & Lee, C. K. (2011). The effect of perceived trust on electronic
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. commerce: Shopping online for tourism products and services in South Korea.
Englewood Cliffs, NJ: Prentice Hall. Tourism Management, 32, 256–265.
Bauer, R. A. (1964). Consumer behavior as risk taking. Dynamic Marketing for a Kim, D. J., Ferrin, D. L., & Rao, H. R. (2008a). A trust-based consumer decision-
Changing World, American Marketing Association Proceedings, 389–398. making model in electronic commerce: The role of trust, perceived risk, and
Bauer, R. A. (1967). Consumer behavior as risk taking. In D. Cox (Ed.), Risk Taking and their antecedents. Decision Support Systems, 44, 544–564.
Information Handling in Consumer Behavior. Cambridge, MA: Harvard University Kim, D. J., Ferrin, D. L., & Rao, H. R. (2008b). Trust and satisfaction, two stepping
Press. stones for successful e-commerce relationships: A longitudinal exploration.
Bélanger, F., & Carter, L. (2008). Trust and risk in e-government adoption. Journal of Information System Research, Articles in Advance, 1–21.
Strategic Information System, 17, 165–176. Kim, J., & Forsythe, S. (2008). Sensory enabling technology acceptance model (SE-
Bellman, S., Lohse, G. L., & Johnson, E. J. (1999). Predictors of online buying behavior. TAM): A multiple-group structural model comparison. Psychology and
Communication of the ACM, 42(12), 32–38. Marketing, 25(9), 901–922.
Boudreau, M. D., & Gefen, D. (2001). Validation in IS research: A state-of-the-art Kim, K., & Prabhakar, B. (2000). Initial trust, perceived risk, and the adoption of
assessment. MIS Quarterly, 25(1), 1–24. Internet banking. Proceedings of ICIS, Brisbane, Australia, 10–13.
Carow, K. A., & Staten, M. E. (1999). Debit, credit or cash: Survey evidence on Lee, M. C. (2009). Factors influencing the adoption of internet banking: An
gasoline purchases. Journal of Economics and Business, 51(5), 409–421. integration of TAM and TPB with perceived risk and perceived benefit.
Chircu, A. M., Davis, G. B., & Kauffman, R. J. (2000). Trust, expertise and ecommerce Electronic Commerce Research and Applications, 8, 130–141.
intermediary adoption. In J. DeGross (Ed.), Proceedings of the sixth Americas Legris, P., Ingham, J., & Collerette, P. (2003). Why do people use information
conference on information systems (pp. 710–716). New York: ACM. technology? A critical review of the technology acceptance model. Information
Cho, S. E. (2010). Perceived risks and customer needs of geographical accessibility in and Management, 40(3), 191–204.
electronic commerce. Electronic Commerce Research and Applications, 9, Li, X., Hess, T. J., & Valacich, J. S. (2008). Why do we trust new technology? initial
495–506. trust formation with organization information systems. Journal of Strategic
Corbitt, B. J., Thanasankit, T., & Yi, H. (2003). Trust and e-commerce: A study of Information Systems, 17, 39–71.
consumer perceptions. Electronic Commerce Research and Applications, 2, Liu, X., & Wei, K. K. (2003). An empirical study of product differences in consumers’
203–215. e-commerce adoption behavior. Electronic Commerce Research and Applications,
Crespo, Á. H., & Rodríguez, I. (2008). Explaining B2C e-commerce acceptance: An 2, 229–239.
integrative model based on the framework by Gatignon and Robertson. Lopez-Nicolas, C., & Molina-Castillo, F. J. (2008). Customer knowledge management
Interacting with Computers, 20, 212–224. and e-commerce: The role of customer perceived risk. International of
Cunningham, S. M. (1967). The major dimensions of perceived risk. In D. Cox (Ed.), Information Management, 28, 102–113.
Risk taking and information handling in consumer behavior. Cambridge, MA: Malhotra, N. K., Kim, S. S., & Agarwal, J. (2004). Internet uers’ information privacy
Harvard University Press. concerns (IUIPC): The construct, the scale, and a causal model. Information
Cunningham, L. F., Gerlach, J., & Harper, M. D. (2005). Perceived risk and e-banking Systems Research, 15(4), 336–355.
services: An analysis from the perspective of the consumer. Journal of Financial Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of
Services Marketing, 10(2), 165–178. organizational trust. Academy of Management Review, 20, 709–734.
Das, T. K., & Teng, B. S. (2004). The risk-based view of trust: A conceptual McKnight, D. H., & Chervany, N. (2002). What trust means in e-commerce customer
framework. Journal of Business and Psychology, 19(1), 85–116. relationships: An interdisciplinary conceptual typology. International Journal of
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance Electronic Commerce, 6(2), 35–59.
of information technology. MIS Quarterly, 13(3), 319–340. Nicolaou, A., & McKnight, D. H. (2006). Perceived information quality in data
Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer exchanges: Effects on risk, trust, and intention to use. Information Systems
technology: A comparison of two theoretical models. Management Science, Research, 7(4), 332–351.
35(8), 982–1003. Park, J., Lennon, S. J., & Stoel, L. (2005). On-line product presentation: Effects on
Dinev, T., & Hart, P. (2006). An extended privacy calculus model for e-commerce mood, perceived risk, and purchase intention. Psychology & Marking, 22(9),
transactions. Information Systems Research, 17(1), 61–80. 695–719.
Featherman, M., & Pavlou, P. (2003). Predicting e-services adoption: A perceived Pavlou, P. A. (2001). Integrating trust in electronic commerce with the technology
risk facets perspective. International Journal of Human-Computer Studies, 59, acceptance model: Model development and validation. Boston, MA: AMCIS
451–474. Proceedings.
Featherman, M. S., Valacich, J. S., & Wells, H. D. (2006). Is that authentic or artificial? Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: Integrating trust
Understanding consumer perceptions of risk in e-service encounters. and risk with the technology acceptance model. International Journal of
Information Systems Journal, 16, 107–134. Electronic Commerce, 7(3), 101–134.
Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, Pavlou, P. A., & Gefen, D. (2004). Building effective online marketplaces with
117–140. institution-based trust. Information Systems Research, 15(1), 37–59.
24 Q. Yang et al. / Computers in Human Behavior 50 (2015) 9–24
Pennington, R., Wilcox, H. D., & Grover, V. (2003). The role of system trust in
business-to-consumer transactions. Journal of Management Information Systems,
20(3), 197–226.
Peter, J. P., & Ryan, M. J. (1976). An investigation of perceived risk at the brand level.
Journal of Marketing Research, 13(5), 184–188.
Schepers, J., & Wetzels, M. (2007). A meta-analysis of the technology acceptance
model: Investigating subjective norm and moderation effects. Information &
Management, 44(1), 90–103.
Schultz, R. L., & Slevin, D. P. (1975). Implementation and organizational validity: An
empirical investigation. Implementing Operations Research/Management Science.
New York: American Elsevier.
Shen, C. C., & Chiou, J. S. (2010). The impact of perceived ease of use on Internet
service adoption: The moderating effects of temporal distance and perceived
risk. Computers in Human Behavior, 26, 42–50.
Stone, R. N., & Gronhaug, K. (2007). Perceived risk: Further considerations for the
marketing discipline. European Journal of Marketing, 27(3), 39–50.
Sun, J. J. (2007). Research on the united of TAM, TRA and TPB. Journal of Modern
Technology on Books and Intelligence, 8, 40–43 (in Chinese).
Taylor, S., & Todd, P. (1995). Understanding information technology usage: A test of
competing models. Information Systems Research, 6(2), 144–176.
Teo, T. S. H., & Liu, J. (2007). Consumer trust in e-commerce in the United States,
Singapore and China. Omega, 35(1), 22–38.
United Nations Conference on Trade and Development (UNCTAD), (2001). e-
Commerce and development report. <http://r0.unctad.org/ecommerce/docs/
edr01_en/edr01_en.pdf>.
Van der Heijden, H., Vehagen, T., & Creemers, M. (2003). Understanding online
purchase intentions: Contribution from technology and trust perspectives.
European Journal of Information Systems, 12, 41–48.
Veloutsou, C., & Bian, X. M. (2008). A cross-national examination of consumer
perceived risk in the context of non-deceptive counterfeit brands. Journal of
Consumer Behavior, 7, 3–20.
Venkatesh, V., Brown, S. A., Maruping, L. M., & Bala, H. (2008). Predicting different
conceptualization of system use: The competing roles of behavioral intention,
facilitating conditions, and behavior expectation. MIS Quarterly, 32(3), 483–502.
Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology
acceptance model: Four longitudinal field studies. Management Science, 46(2),
186–204.
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of
information technology: Towards a unified view. MIS Quarterly, 27(3), 425–478.
Yi, M. Y., Jackson, J. D., Park, J. S., & Probst, J. C. (2006). Understanding information
technology acceptance by individual professionals: Toward an integrative view.
Information and Management, 43(3), 350–363.
Yousafzai, S., Pallister, J., & Foxall, G. (2003). A proposed model of e-trust for
electronic banking. Technovation, 23(11), 847–860.