ArticlePDF Available

An Overview of Online Trust: Concepts, Elements, and Implications

Authors:

Abstract and Figures

Lack of trust has been repeatedly identified as one of the most formidable barriers to people for engaging in e-commerce, involving transactions in which financial and personal information is submitted to merchants via the Internet. The future of e-commerce is tenuous without a general climate of online trust. Building consumer trust on the Internet presents a challenge for online merchants and is a research topic of increasing interest and importance. This paper provides an overview of the nature and concepts of trust from multi-disciplinary perspectives, and it reviews relevant studies that investigate the elements of online trust. Also, a framework of trust-inducing interface design features articulated from the existing literature is presented. The design features were classified into four dimensions, namely (1) graphic design, (2) structure design, (3) content design, and (4) social-cue design. By applying the design features identified within this framework to e-commerce web site interfaces, online merchants might then anticipate fostering optimal levels of trust in their customers. http://www.sciencedirect.com/science/article/pii/S0747563203001092 http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.90.2184&rep=rep1&type=pdf
Content may be subject to copyright.
Computers in Human Behavior 21 (2005) 105- -- 125
Final Version
An overview of online trust:
Concepts,
elements, and
implications
Ye Diana Wang
*
,
Henry H.
Emurian
Information Systems Department, College of Engineering and Information Technology, UMBC,
1000 Hilltop Circle, Baltimore, MD 21250, USA
Abstract
Lack of trust has been repeatedly identified as one of the most formidable barriers to
people
for engaging in e-commerce, involving
transactions
in which financial and personal infor-
mation is submitted to merchants via the Internet. The future of e-commerce is
tenuous
without a general climate of online trust. Building consumer trust on the Internet presents
a
challenge for online merchants and is a research topic of increasing interest and
importance.
This paper provides an overview of the nature and concepts of trust from
multi-disciplinary
perspectives, and it reviews relevant studies that investigate the elements of online trust.
Also,
a framework of trust-inducing interface design features articulated from the existing
literature
is presented. The design features were classified into four dimensions, namely (1)
graphic
design, (2) structure design, (3) content design, and (4) social-cue design. By applying the
design features identified within this framework to e-commerce web site interfaces,
online
merchants might then anticipate fostering optimal levels of trust in their
customers.
Keywords: Trust; Online trust; E-commerce; Interface
design
1. Introduction and backgrou
nd
The phenomenal growth in the number of Internet users and the
enormous
potential of electronic commerce (e-commerce) via the Internet have
attracte
d
*
Corresponding author. Tel.: +1-410-455-3206; fax: +1-410-455-1073.
E-mail addresses:
ywang8@umbc.edu
(Y.D. Wang),
emurian@umbc.edu
(H.H.
Emurian
).
Y.D. Wang, H.H. Emurian
merchants to conduct their businesses online. The Census Bureau of the
Department
of Commerce reported that the estimate of US retail e-commerce sales for the
thir
d
quarter of 2002 was $11.061 billion, an increase of 34.3% from the third quarter
of
2001. 1 However, while Internet technologies and
infrastructures
to support e-
commerce are now in place and
obviously
being used by consumers, several
recent
surveys show that the majority of users of the Internet in the US are still
hesitant
about shopping online.
2
In one national survey of 1017 Internet users conducted by the Pew Internet
and
American Life Project (PIP), 68% expressed concern about revealing nancial in-
formation, but 48% had actually made a purchase online with a credit card;
almost
3% reported being cheated by an online merchant or having had a credit
card
number stolen (Fox, 2000). The survey also reported that online consumer
concerns
were little different from those expressed about other merchants. In a more
recent
survey of 1677 Internet users conducted in October 2002, it was estimated that
only
3% of 64 million Internet users in the US purchased a product online on an
average
day (PIP, 2002). In yet another national survey of 1500 Internet users conducted
by
Princeton Survey Research Associates (2002) for Consumer WebWatch, a project
of
Consumers Union, 64% of users expressed
reservations
about trusting e-com
merce
sites.
The most recent survey data on online purchasing were reported by PIP for the
month of December 2002. Based on a national survey of 1220 Internet users,
only
28% purchased holiday gifts online, and this projects to
approximately
30 milli
on
purchasers (Rainie & Horrigan, 2002). Taken together, the results from these
severa
l
surveys suggest that consumer distrust regarding
transactions
on the Internet
may
hinder the future
valuation
of even the leading
‘‘dot.coms’’
(The Economist, 2001),
and online merchants presently face a difficult challenge - -- to build consumer trust
to
make purchases on the
Internet.
Trust with all of its
connotations
has been studied in numerous disciplinary fie
lds,
and it presents important research
opportunities
and
applications
in an
online
context. If online trust can be understood and enhanced by reputable online mer-
chants, then the number of people who engage in e-commerce should
increa
se
substantially.
More
importantly,
a general climate of trust online will be created in
which buyers will feel more at ease in disclosing sensitive information, sellers will
feel
confident to conduct business online, and there will be intensive
inter
actions,
transactions,
and
associations
to benefit consumers and merchants alike. Put
simp
ly,
the future of e-commerce depends on trust, and at least one training
org
anization,
m-eTrust, 3 now offers a certificate program purporting to teach the skills
and
competencies considered vital to implement and manage
trustworthy
e-com
merce
web
sites.
1 US Department of Commerce News: http://www.ce
nsus.gov/mrts/www/current.html.
2 Internet is used here to include the World Wide Web because that is now a common way to refer to the
medium for electronic
communications
and information
exchanges.
3 http://www
.m-etrust.co
m/
.
Y.D. Wang, H.H. Emurian
To build online trust, however, is a formidable task. Because trust is a
complex
and abstract concept, it is difficult to define trust and to identify the elements
that
construct it. Moreover, due to the nature of e-commerce, online merchants
mainly
depend on their web sites to represent themselves to consumers. As stated by
Ang
and Lee (2000), ‘‘If the web site does not lead the consumer to believe that the
merchant is
trustworthy,
no purchase decision will result’’ (p. 3). Hence, there is
a
need to focus on web interface implications that online merchants can consider
and
adopt to enhance their perceived
trustworthiness
to potential
customers.
Against that
background,
this paper has three objectives: (1) to provide
an
overview of various
approaches
in
conceptualizing
trust in multiple elds; (2)
to
examine the state of trust studies that investigate the elements of trust in
online
contexts; and (3) to articulate the existing web interface design implications,
most
ly
based on
human- -- computer
interaction (HCI) knowledge, and propose a
framew
ork
of design features that may be helpful to induce greater online
trust.
The paper addresses the following questions. How is trust defined in both
offline
and online contexts? What are the
characteristics
of trust in both offline and
online
contexts? What are the elements, antecedents, underlying dimensions,
determ
inants
or principles of online trust? How can online merchants build consumer trust by
web
interface design? Online trust in this paper refers to consumer trust directed
towards
e-commerce web sites or merchants on the Internet and only applies to
Business
-to-
Consumer (B2C) e-commerce
transactions.
Finally, consumers' concerns about the
ability of the Internet
infrastructure
in a technical sense and their perceptions
of
trustworthiness
of an online merchant should be recognized as two distinct
issues,
and the latter issue is the focus of this
paper.
The rest of the paper is organized as follows. First, a general overview of
trust
from the disciplines of
philosophy, psychology,
management, and marketing
is
provided to illustrate the nature and concept of trust. Based on this overview,
diverse
perspectives are integrated into a summary of
characteristics
of trust, and
these
characteristics
are further
elaborated
in an online context. Following this,
relevan
t
studies are reviewed in an attempt to identify the elements that are pertinent to the
formation of online trust. Next, a framework of trust-inducing features is
proposed
that is based on existing literature on enhancing online trust by web interface
design.
The paper concludes with suggestions for further research on online trust.
Unless
noted otherwise, the paper is based upon data collected from English speaking
users
of the Internet located within the
US.
2. The nature and concept of
trust
2.1. The difficulty of defining
trust
Trust, according to the Oxford English Dictionary (1971), is dened as
‘‘confi-
dence in or reliance on some quality or attribute of a person or thing, or the truth
o
f
a
statement’’
(p. 3423). Trust has existed as long as the history of human beings
and
the existence of human social interactions. Almost every aspect of a person's life
is
Y.D. Wang, H.H. Emurian
based on some form of trust.
Undoubtedly,
trust is positive and vital to
human
ity
since it is part of love and friendship, and meaningful
relationships
depend upon
it.
Presently, however, researchers have difficulty in
operationalizing
what exactly
trust
is, and they disagree even on basic denitions (e.g., Husted, 1998). This difficulty in
defining trust is further evidenced by numerous researchers. For example, Shnei-
derman (2000) quoted definitions of trust published by a former US State
Depart-
ment Analyst (i.e.,
Fukuyama,
1995) and by computer scientists (i.e., Fogg & Tseng,
1999) and also offered his own definition of trust.
Additionally,
Sirdeshmukh, Singh,
and Sabol (2002) quoted two prior studies' definitions of trust (i.e.,
Moorman,
Zaltman, & Deshpande, 1992, p. 315; Morgan & Hunt, 1994, p. 23) and also offe
red
a definition of their own. Despite the Oxford
Dictionary
'
s
definition, then, trust
is
often
conceptualized
by
researchers
according to the features of a
particular co
ntext.
The presence of multiple definitions of trust in the literature is likely due to
two
reasons. First, trust is an abstract concept and is often used
interchangeably with
related concepts such as credibility, reliability, or confidence. Thus, to define the
term and to delineate the distinction between trust and its related concepts
have
proven challenging for researchers. Second, trust is a multi-faceted concept
that
incorporates
cognitive, emotional, and
behavioral
dimensions (Lewis & Weiger
t,
1985). Trust has been widely studied in many disciplines, but each discipline has
its
own
understanding
of the concept and different ways to
operationalize
it. Writing
from the perspective of a critical analysis of
trustworthiness
as security,
Nis
senbaum
(2001) stated, ‘‘Trust is an
extraordinarily
rich concept, covering a variety of
rela-
tionships, conjoining a variety of objects. One can trust (or distrust) persons, insti-
tutions, governments, information, deities, physical things, systems, and more’’
(p.
104). There is, then, a lack of consistent principles by which to dene trust and all
of
its
manifestations
as the term may be used to characterize
consumer
- -- merc
hant
relationships.
2.2. A general overview of
trust
For obvious reasons, trust has been studied long before the emergence of the
Internet or e-commerce and has been
conceptualized
within different
disciplines.
These trust studies mainly focus on trust in general (i.e., ‘‘offline trust’’) in
attempts
to provide a fundamental
understanding
of the concept and, therefore, they
are
important in grounding us to examine trust in the online world. To provide a
general
overview of trust and to illustrate the various
approaches
to the denition of
trust,
several examples are presented from the extensive literature in the disciplines
of
philosophy,
psychology,
management, and
marketing.
2.2.1. Philosophy
In the discipline of philosophy,
speculations
on trust can be traced back to the
ancient Greeks. Most ancient philosophers studied trust in an attempt to draw
a
picture of human nature (Bailey, 2002). The Greeks believed that people trust
others
only if people are confident that the others fear detection and punishment
sufficien
tly
to dissuade them from harming or stealing. In addition, they identified love
and
Y.D. Wang, H.H. Emurian
sympathy to be the elements that would ensure more trust and, therefore,
more
cooperation
and peace, in a state of
nature.
Modern philosophers, pioneered by Annette Baier, focused on
interpersonal trust
and the morality of trust
relationships.
Baier (1994) dened trust as the
‘‘accepted
vulnerability
to another's possible but not expected ill will toward one’’ (p. 99).
Accepting trust as a three-place predicate (A trusts B with valued thing C),
she
suggested that trustor A needed good judgment to know how much discretion
to
give. Because trustee B would have had
discretionary
powers over the e
ntrusted
valued thing C, the trustor had to take the risk if the trustee abused the
granted
power.
Political philosophers have explored trust in terms of social values and
b
e
nefits.
For instance, Francis
Fukuyama,
in his politically oriented book entitled Trust, the
Social Virtues and the Creation of Prosperity argued that
‘‘social
capital is a
cap
a-
bility that arises from the prevalence of trust in a society or in certain parts of
it
(Fukuyama,
1995, p. 26).
Spontaneous sociability
engendered by trust is a mech
a-
nism by which values are shared and conrmed among group
member
s.
2.2.2.
Psychol
ogy
The majority of the
psychology
literature on trust has focused on
interperson
al
trust, but different
approaches
have been taken to study trust. Psychologists
general
ly
agreed that trust, especially
interpersonal
trust, was an important concept in
ps
y-
chology and vital to
personality
development (Erikson, 1963),
cooperation institution
(Deutsch, 1962), and social life (Rotter, 1980). Unlike
philosophers, psychologists
concentrated on individual differences in trust as
personality characteristics
and the
consequences of such dierences in early life and human
relationships.
For example,
psychoanalyst
Erik Erikson dened trust as therst stage of
his
model of human development. He suggested that such
‘‘basic
trust
’’
was a
centra
l
ingredient in the
‘‘healthy personality’’
and had a major impact on individual
traits
(Erikson, 1963, p. 7). Trust also plays an important role in establishing
cooperative
relationships.
According to Deutsch, ‘‘the initiation of
cooperation
requires
trust
whenever the individual, by his choice to
cooperate,
places his fate partly in the
hands of others’’ (Deutsch, 1962, p. 302).
A frequently cited definition of
interpersonal
trust was given by Rotter (1967),
who defined trust as ‘‘an expectancy held by individuals or groups that the
word,
promise, verbal, or written statement of another can be relied on’’ (p. 651). The
researcher pointed out that individuals differed in a generalized expectancy
and
developed a Likert-type scale to measure
interpersonal
trust. By conducting
exp
er-
iments, the researcher also conrmed the positive consequences of trust to
peop
le
and overall
societ
y.
2.2.3. Managem
ent
In the discipline of management, trust has been studied intensively in
or
ganiza-
tional contexts. Scholars have investigated the roles and importance of trust in or-
ganizations (Kramer, 1999). Driscoll (1978) considered trust and
participation as
predictors of
satisfaction
in
organizational
decision making.
Organizational
trust
was
Y.D. Wang, H.H. Emurian
defined as ‘‘the belief that the decision makers will produce outcomes
favorable to
the person's interests without any influence by the
person
(Driscoll, 1978, p. 44).
As
workforce
composition increased in diversity and
organizational
structure
changed,
trust was also identified as a control mechanism for enabling employees to
work
together more
productively
and effectively (Mayer, Davis, & Schoorman, 1995).
Another role of trust that had been recognized was to reduce the cost of
both
intra- and
inter-organizational transactions
(Uzzi, 1997). Management
resear
chers
generally believe that trust can enhance business
performance.
In addition, there
has
been interest in the link between
inter-organizational
trust and
governance. For
instance, Sako (1998) regarded trust as an informal governance structure,
‘‘whic
h
enhances the effectiveness of
transactions
whether they take place in markets
o
r
within a
hierarchy’’
(p. 91). Finally, the relationship of employee trust in
leader
ship
with job
satisfaction
and
organizational
commitment has been reported in a
recent
meta-analysis
of several decades of research (Dirks & Ferrin, 2002).
2.2.4. M
arketin
g
In the discipline of marketing, research on trust has been conducted within the
context of distribution channels (e.g.,
manufacturer- -- retailer)
and buyer- -- seller
rela-
tionships. One
particular
area of research focused on how to maintain long-term
relationships
in distribution channel
arrangements
given the fact that the
switching
costs were relatively high. Kumar (1996) and other researchers agreed that
trust,
rather than power, helped
manufacturers
or retailers receive more tangible
benefits
and realize their full potential, thus facilitating long-term
co
mmitme
nts.
Much attention on trust was drawn from relationship marketing
perspectives.
Relationship marketing focused not on discrete
transactions,
but rather on
relational
exchange, which was ‘‘longer in duration, reflecting an ongoing
process
’’
(Dwyer
,
Schurr, & Oh, 1987, p. 23). Trust, defined as
‘‘a
willingness to rely on an
exc
hange
partner in whom one has
confidence’’
(Moorman, Deshpande, & Zaltman, 1993,
p.
90), had assumed an essential role in establishing and maintaining a long-term re-
lationship between sellers and
customers.
Scholars in this discipline primarily focused on two targets of trust: supplier
rms
and their sales people. A dominant view was given by Doney and Canon (1997):
‘‘customers
can trust the supplier rm, its
salesperson,
or
both
’’
(p. 36). These in-
vestigators
had further verified that some
characteristics
of
salespersons,
such
as
expertise, likeability, and similarity to customers, played a signicant role in
buildi
ng
trust and strengthening the link between the customer and the supplierrm.
Finally,
Sirdeshmukh et al. (2002) found that consumers' reports of the behavior of
front
-line
employees and the practices and policies of management, both evaluated in terms
of
competency, benevolence, and problem solving, were related to reports of trust in the
businesses under
consideration.
2.3. Characteristics of
trust
While
acknowledging
the diverse opinions of scholars and researchers in
de
ning
trust, this paper does not attempt to provide an explicit definition of trust.
In
stead,
Y.D. Wang, H.H. Emurian
we attempt to integrate several emergent themes related to how trust was
viewed
from the selected studies in the previous section,
particularly
those revealing the
nature of trust. This
approach
is intended to facilitate a better
understanding
of
trust
in an online context. Against that
background,
four
characteristics
of trust
are
generally observed and accepted by researchers studying
trust:
1. Trustor and trustee. There must exist two specific parties in any trusting
relat
ion-
ship: a trusting party (trustor) and a party to be trusted (trustee). The two
parti
es,
comprised of persons,
organizations,
and/or products, include an
evaluation of
the actions of each party. The development of trust is based on the ability
of
the trustee to act in the best interest of the trustor and the degree of trust
that
the trustor places on the
trust
ee.
2. Vulnerability. Trust involves
vulnerability.
Trust is only needed, and
actually
ourishes, in an environment that is uncertain and risky. Trustors must be
willing
to make themselves
vulnerable
for trust to be
operational
by taking the risk of los-
ing something important to them and relying on the trustees not to exploit the
vul-
nerability.
3. Produced actions. Trust leads to actions, mostly risk-taking
behaviors.
The
form
of the action depends on the situation, and the action may concern something ei-
ther tangible or intangible. For instance, a person lends his or her money to
a
friend because the friend is trusted to pay back the money later; a couple gets mar-
ried because the parties trust each other to be loyal in the
relat
ionshi
p.
4. Subjective matter. Trust is a subjective matter. It is directly related to and
affected
by individual differences and
situational
factors. Different people view the role
of
trust differently in different scenarios and have different magnitudes of trust to-
wards different
trustees.
3. Online
trus
t
3.1. Characteristics of online
trust
Online trust shares similar
characteristics
to those of offline trust, but there
are
some important distinctions that are unique in an online environment. These dis-
tinctions can serve as starting points for seeking a deeper
understanding
of the
na
-
ture of trust in an online context. The
characteristics
of online trust can be
described
as
foll
ows:
1. Trustor and trustee. The two parties, trustor and trustee, are still vital for
estab-
lishing a trusting relationship in the online world, but they imply specific
entities
.
In the offline situation, the trustor and trustee positions can be lled by many dif-
ferent entities. In online trust, however, the trustor is typically a consumer who
is
browsing an e-commerce web site, and the trustee is the e-commerce web site,
or
more specifically, the merchant that the web site represents. Sometimes, the tech-
nology (mainly the Internet) itself is an object of trust (Marcella, 1999).
2. Vulnerability. Because of the high complexity and anonymity associated with e-
commerce, merchants can behave in an
unpredictable
manner on the
Internet.
Y.D. Wang, H.H. Emurian
Consumers are often uncertain about the risks at present and their full conse-
quences when transacting online. As stated by Gefen (2002), ‘‘Even when
online
consumers only examine a web site without purchasing from it, data may be
au-
tomatically
collected about their activities and later misused or distributed with-
out their consent or
knowledge’’
(p. 40). All these reasons contribute to
a
n
insecure environment where trust is needed and can potentially ourish,
posing
unique challenges to building and maintaining trust. Finally, in online
commer
cial
transactions,
consumers are
vulnerable
to specific trust
violations:
loss of
money
and loss of privacy (Friedman, Howe, & Kahn, 2000).
3. Produced actions. Consumer trust in online merchants generates two
specific
forms of action from the consumer: (1) making a purchase online from the mer-
chant,
presumably
including providing credit card and personal information in
the
transaction,
and (2)
‘‘window-shopping’’
at the merchant's web site. Either
of these actions brings positive outcomes to online merchants, such as actual
or
potential sales. To engage in such activities, consumers must be condent
that
they have more to gain than to
lose.
4. Subjective matter. Like offline trust that is
associated
with individual difference
s
and situational factors, online trust is inherently a subjective matter
(Grabner-
Kraeeter, 2002). The level of trust considered sufficient to make
transactio
ns
online is different for each individual. People also hold different attitudes
tow
ard
machines and
techno
logy.
3.2. Elements of online
trust
Without attempting to identify the elements that are pertinent to the formation
of
online trust, it is difficult to derive effective and reliable design principles or
impl
ica-
tions on enhancing consumer trust in e-commerce. This section reviews relevant
studi
es
on the subject of these trust elements, which are often referred to
interchangeably
as
antecedents, underlying dimensions, determinants, or principles of online trust.
In
general, these terms all refer to factors that can produce a sense of
trustworthiness
or
even determine whether consumers will trust an online merchant's web
site.
Gefen (2002) examined trust from a multi-dimensional perspective. According
to
the researcher, the specific beliefs of integrity, ability, and benevolence were seen
as
antecedents to overall trust. In the case of e-commerce, integrity was the belief
that
the online merchant adhered to stated rules or kept promises. Ability was the
belief
about the skills and competence of the online merchant to provide good
qua
lity
products and services. Benevolence was the belief that the online merchant,
aside
from wanting to make legitimate profits, wanted to do good to the customer
without
regard to making a
sale.
While holding a similar view, Ang, Dubelaar, and Lee (2001) proposed that
three
dimensions of trust were important for enhancing the perception of trust on the
Internet. These three dimensions were the ability of the online merchant to deliver
a
product or service that performs as promised, the willingness of the online
mercha
nt
to rectify should the purchase not meet the customer's
satisfaction,
and the
presence
of a privacy policy or statement on the web
site.
Y.D. Wang, H.H. Emurian
Based on the literature from multi-disciplines, Kim, Song,
Braynov,
and
Rao
(2001) investigated the determinants of online trust and divided the
determ
inants
into six dimensions, namely information content, product,
transaction,
techno
logy,
institutional, and
consumer-behavioral
dimensions. These dimensions, which
were
further broken down into many sub-dimensions or properties, formed a
theoret
ical
framework of online trust, covering the different stages that a consumer
went
through to complete an online
transaction.
In differing from most researchers,
Kim
et al. (2001) proposed that the consumer could perceive trust before, during, or
after
the online
transaction.
The
researchers
further concluded that different
determ
inants
of trust were associated with different stages of the
transac
tion.
Lee and Turban (2001) posited that four main antecedents influenced
consumer
trust in Internet shopping. The antecedents were the
trustworthiness
of the
Intern
et
merchant,
trustworthiness
of the Internet as a shopping medium,
infrastructual
(contextual) factors (e.g., third-party certification), and other factors (e.g.,
company
size). Given the fact that most
researchers
tended to ignore the role of the
computer
systems or the Internet through which
transactions
were executed and only focus
on
building trust between consumer and merchants, it was
remarkable
that Lee
and
Turban (2001) considered the perception of the shopping medium as a critical
factor
of consumer trust in Internet shopping. The researchers also pointed out that the
individual
'
s
trust propensity (‘‘trust
baseline’’)
could moderate the
relationships
between the antecedents of trust and trust
itself.
Two other areas of research have been repeatedly cited and considered
found
a-
tional, including the work by Egger (2001) and the Cheskin/Sapient Report (1999).
The
researchers
both took a similar
approach
to formulate the factors that
were
likely to affect consumer trust or to communicate
trustworthiness
to consumers in e
-
commerce by constructing a model of trust. Egger's model is called the Model
of
Trust for Electronic Commerce (MoTEC). The model consists of four
compo
nents:
the
pre-interactional
lters taking place before any online interaction, the
interface
properties of the web site, the information content of the web site, and
relationship
management. The researcher stated that
‘‘relationship
management reects the fa-
cilitating effect of timely, relevant, and personalized
merchant- -- buyer
interactions
on
trust development and
maintenance’’
(Egger, 2001, p. 6). The strength of this
mod
el
is that it covers the entire buyer- -- seller interaction process, and it also places em-
phasis on the effects of customer relationship management. On the other hand, the
Cheskin/Sapient Report (1999) focused on web site interface cues and presented
a
model of six building blocks of online trust. These six building blocks were seals
of
approval,
brand,
navigation,
fulfillment,
presentation,
and technology. The
buildi
ng
blocks could be further divided into a total of 28 components to establish
percei
ved
trustworthiness.
More recently, Hemphill (2002) conceptualized the foundation of online trust in
terms ofve fair information practice principles. Three principles are relevant to the
design of a web site. It was argued that an online merchant should post the
busines
s
'
s
policies on disclosure of personal information, provide options for how a
consumer
'
s
personal data might be used in other contexts, and allow consumers to access
and
view personal data.
Importantly,
the author argued that
‘‘without
an en
forcemen
t
Y.D. Wang, H.H. Emurian
and redress mechanism, a fair information practice code is merely a suggested set
of
guidelines rather than a prescriptive mechanism, and does not ensure
compliance
with the fair information practice
principles’’
(Hemphill, 2002, p. 2).
Hemphill (2002) was one of the few scholars in this area who considered and dis-
cussed the necessity of legislation to create civil remedies for consumers in the event
of
untrustworthy
interactions with an online merchant. Consistent with that
approach,
the author noted the Online Privacy Act of 1998 and the
Gramm- -- Leach- -- Bliley
Act
of
1999, together with recent Federal Trade Commission suits against several
online
merchants (e.g.,
Toysmart.com)
for committing privacy violations. Even with
such
regulations, however, online consumers still require protection from deceptive
web
sites, and effective design might be anticipated to contribute to that
outc
ome.
A final study to be considered is that reported by Ba and Pavlov (2002). Using 95
experienced eBay buyers, self-reports of trust in a seller's credibility were found to
b
e
sensitive to experimental
manipulations
of positive and negative feedback about the
seller. Positive feedback was associated with greater trust in a seller than was neg-
ative feedback, presented to the buyer subjects as
experimenter-controlled
reputation
profiles. Although the effects on trust were demonstrated by profile differences,
at
issue was the buyer's trust in the
accuracy
of the profiles themselves. Against the
background
of guidelines, surveys, and models, this was a rare experimental
eval-
uation of a trust-inducing
mechanism.
These preceding studies provide important insights into the elements of
online
trust. However, because the study of trust in an online context is relatively new,
there
are several consistent issues that exist across most of the studies. As
mentioned
earlier, the terms element, antecedent, dimension, determinant, and principle
are
sometimes used
interchangeably
due to the lack of agreement among researchers in
theeld on a clear definition for each term. A similar
observation
was made
b
y
Shankar, Urban, and Sultan (2002) regarding trust studies in the information sys-
tems and e-business literature. In addition, researchers sometimes propose a list
of
trust-inducing features in their studies without providing empirical evidence for the
relative importance of each feature to the establishment of
trust
.
There are exceptions, however, to include the study of over 4500 people by
Fogg
(2002) and his
associates,
who offer 10 specific
research-based
guidelines for
fostering
credibility on a web site, e.g., ‘‘Make it easy to verify the
accuracy
of the
information
on your
site.
Although this
investigator
and his colleagues make a distinction be-
tween trust and credibility (e.g., Fogg & Tseng, 1999),
trustworthiness
is viewed as
a
dimension of credibility, and the analyses contribute to the knowledge about
trust.
Furthermore,
the survey by Princeton Survey Research Associates (2002) also re-
ported several specific types of posted information (e.g., a privacy policy) that
were
said to be valued by Internet
users.
Despite these surveys and the experimental work of Ba and Pavlov (2002), the
guidelines proposed for
trustworthy
e-commerce web sites largely reflect the
opinions
of the researchers, which are often in response to survey data about the
report
ed
concerns of e-commerce consumers rather than to the purchasing actions of con-
sumers. This is the current body of work from which any potential implementation
is
to be
derived.
Y.D. Wang, H.H. Emurian
4. Implications for web interface design
4.1. Build online trust by web interface design
Based on the
accumulated
knowledge and
understanding
of the elements of
online
trust in the previous section, online merchants can take many managerial
measures
to build and enhance
merchant- -- customer relationships
before, during, and/or
after
any online
transaction,
which can increase consumers' trust. While those
manager
ial
measures might be useful in retaining loyal customers, online merchants need
more
efficient ways to convey their
trustworthiness
to rst-time visitors and to
transform
them into consumers. Moreover, the physical
appearances
of business buildings
and
facilities, together with the direct contact with company employees, which have
been
found to have effects on consumer trust in a
conventional
business context,
are
missing in the online environment. Online merchants depend primarily on
their
electronic storefront to attract potential customers and to communicate with them.
Therefore, applying trust-inducing features to the web sites of online merchants
is
the most effective method of enhancing online trust, given the current state
o
f
knowl
edge.
In a sense, a web site with trust-inducing features functions as a skillful
sales
-
person for the company
(Jarvenpaa,
Tractinsky, & Saarinen, 1999) and,
therefore,
moderates the
disadvantages
of an impersonal web site.
Karvonen
(2000) also re-
ported that Internet users admitted to making intuitive, and rather emotional, on-
the-spot decisions based on their perceptions of an online merchant's web site
when
shopping online. Finally, as discussed by Kubilus (2002), the implementation of
a
trustworthy
e-commerce interface will share many of the general design features
for
effective interface usability (e.g., Neilsen, 2000) when applied to e-commerce
web
sites (Scheffelmaier & Vinsonhaler, 2002).
4.2. A framework of trust-inducing
feat
ures
In this section, a framework of trust-inducing features is proposed in an effort
to
synthesize existing literature on enhancing online trust by web interface design. This
growing body of the literature, mainly in the eld of HCI, focuses on how to
design
web sites that are perceived by consumers as
trustworthy,
rather than using tech-
nological mechanisms to build strong
infrastructures.
In other words, the main
go
al
for related HCI researchers was to explore web interface design implications
to
maximize consumer trust or, more precisely, trust perception. The implications to
be
presented are to be understood in terms of optimally representing the
information
associated with many of the trust-inducing elements discussed
previously.
The
framework classifies these trust-inducing features into four broad dimensions
as
shown in Table 1. The four dimensions are discussed ne
xt.
4.2.1. Graphic design dimension
This dimension refers to the graphical design features on the web site that
nor
-
mally give consumers their rst impression. In an experimental investigation
of
Y.D. Wang, H.H. Emurian
Table 1
Framework
of trust-inducing
features
Dimension Explanation Features Literature
sources
Graphic
design
Structure
design
Content
design
Social-cu
e
design
Refers to the
graphical
design factors on the
web
site that normally
give
consumers a
rst
impression
Defines the
overall
organization and
accessibility
of
displayed
information
on the web
site
Refers to the
informational
components that
can
be included on the
web site,
either
textual
or
graphical
Relates to
embeddin
g
social cues, such
as
face-to-face
interaction
and social
presence,
into web interface
via
different
communic
ation
media
Use of three-dimen
sional,
dynamic, and
half-screen
size
clipart
Symmetric use of
moderate
pastel color of low
brightness
and cool
tone
Use of well-chosen,
good-shot
photographs
Implementation
of
easy-to-use
navigation
(simplicity,
consistency)
Use of accessible
information
(e.g., no broken links
and
missing
pictures)
Use of navigation
reinforce-
ment (e.g., guides,
tutorials,
instructions)
Application of page
design
techniques (e.g., white
space
and margin, strict
grouping,
visual
density)
Display of
brand-promotin
g
information (e.g.,
prominent
company logo or
slogan,
main selling
point)
Up-front disclosure of
all
aspects of the
customer
relationship (e.g.,
company
competence,
security
,
privacy,
nancial, or
legal
concerns)
Display of seals of
approval
or third-party
certificate
Use of
comprehens
ive,
correct, and
current
product
information
Use of a relevant
domain
name
Inclusion of
representative
photograph
or video
clip
Use of
synchronous
communication
media
(e.g.,
instant messaging, chat
lines,
video
telephony)
Karvonen and
Parkkinen (2001);
Kim
and Moon (1998)
Cheskin/Sapient
Report
(1999);
Karvonen and
Parkkinen (2001);
Neilsen (1998);
Zhang,
von Dran, Small,
and
Barcellos (1999)
Belanger, Hiller,
and
Smith (2002); Cheskin/
Sapient Report (1999);
Egger (2001); Hu, Lin,
and Zhang (2001);
Neilsen (1999);
Shneiderman (2000)
Basso,
Goldberg,
Greenspan,
and
Weimer (2001);
Riegelsberger and
Sasse
(2001);
Steinbruck,
Schaumburg, Duda,
and Kruger (2002)
alternative
cyber-banking
system interfaces conducted in Korea, Kim and
Mo
on
(1998) reported that the main clipart and overall color layout affected the
trustworthiness
of the web site. By
operationalizing trustworthiness
in terms of
users
'
Y.D. Wang, H.H. Emurian
self-reports of the reliability and
dependability
of alternative interface designs, the
researchers found that using three-dimensional, dynamic clipart that covered at
least
half of the total screen size could enhance the
trustworthiness
of a web
inter
face.
They also concluded that some attributes of interface color conveyed
trustworthi
-
ness. For example, the tone of the interface color should be cool, and its main
color
should be a moderate pastel color.
Additionally,
the colors should be of
low
brightness and
symm
etrical.
The use of real
photographs,
rather than
cartoons,
was recommended by
Kar-
vonen and Parkkinen (2001). Generally, high-quality
photographs
of products
and
well-chosen images generate consumer condence that can be
transferred
to
other
aspects of the web site. As Basso et al. (2001) indicated, ‘‘web retailers use eye-
catching graphics not only to grab a user's attention but also to convey
competence
or
professionalism’’
(p. 138). However, overuse of graphics will not help, but
may
damage the
professional
impression of a web site. As Lightner (2003)
reported,
mature and affluent users of e-commerce web sites tended to favor the absence
of
gratuitous sensory
impac
t.
Finally, an analysis of the comments of 2440 Internet users regarding factors
that
promote the credibility of a web site, concluded that
‘‘nearly
half of all consumers
(or
46.1%) in the [Stanford Persuasive Technology Lab] SPT study assessed the credi-
bility of sites based in part on the appeal of the overall visual design of a site, in-
cluding layout,
typography,
font size, and color
schemes
’’
(Stanford
Per
suasive
Technology Lab, 2002).
4.2.2. Structure design dimension
This dimension defines the overall
organization
and accessibility of
displ
ayed
information on the web site. Ease of
navigation
was frequently mentioned as a key
to
promote online trust (e.g., Cheskin/Sapient Report, 1999; Neilsen, 1998). In
other
words, users can easily locate the information they seek on the web site. This ease-
of-use reflects two
characteristics
of a
trustworthy
web site: simplicity and
consis
-
tency. Consumers appreciate simplicity or a clear design of e-commerce web
sites
because it reduces the perceived risks of wasting time, deception, and
frustration
.
Consumers could get annoyed when they see the design and format of
interface
elements varying from page to page within a site. When the structure and design
of
the web site are consistent, ‘‘users feel more condent using the site because they
can
transfer their learning from one sub-site to the next rather than having to
learn
everything over again for each new
page
’’
(Neilsen, 1998, p. 107).
Accessibility of the information on a web site is also instrumental to the
estab-
lishment of online trust. For example, broken links, meaningless images, and
similar
‘‘hygiene
factors
may relate to users'
dissatisfaction
with a web site interface (Zhang
et al., 1999).
Additionally,
the Cheskin/Sapient Report (1999) indicated that
navi-
gation reinforcement, such as prompts, guides, tutorials, and instructions, could
aid
and inform users to seek information or to perform
transactions
on the web site
and
,
hence, promoted consumer trust. And the
application
of page design techniques
that
can increase
readability,
such as
appropriate
amount of white space and
margins,
strict grouping, and visual density, also help to increase the overall
trustworthiness
Y.D. Wang, H.H. Emurian
of the web site
(Karvonen
& Parkkinen, 2001). Online consumers may report the
importance of such factors as privacy policies, but their intention to purchase
and
their buying behavior suggest that the design of a web site has a major influence
on
the decision to use it (Belanger et al., 2002).
4.2.3. Content design dimension
This dimension refers to the
informational
components that can be included on the
web site, either textual or graphical. Several researchers stress the importance
of
‘‘branding’’
in e-commerce, which is to promote the brand reputation of a
c
o
mpany
online. Egger (2001) suggested two ways to do so: (1) displaying a prominent logo
and
slogan to facilitate the easy identification of the company, and (2) presenting the
company
'
s
main selling point to arouse people's curiosity. The up-front disclosure
of
all aspects of the customer relationship was also considered important by
numerous
researchers
(e.g., Cheskin/Sapient Report, 1999; Egger, 2001; Neilsen, 1999; Shnei-
derman, 2000) to build consumer trust in an online environment. These aspects of the
customer relationship can be classified as company competence, security,
privacy,
nancial, and legal
responsibilities.
To verify its competence, an online merchant
can
provide company
background
and contact details or disclose patterns of past per-
formance and references from past and current customers on the web site. To
address
the remaining customer relationship aspects through the web site, an online
mercha
nt
can provide links to security and privacy policies, offer tracking information, state the
return policy, reveal shipping charges, and explain
contractual
term
s.
The
incorporation
of seals of
approval
or third-party certicates into
content
design has been widely accepted as a strategy to assure consumers that the web
sites
are
trustworthy.
These
assurance
symbols, whether in the form of a logo or state-
ment, have been proven to be useful in establishing online consumer trust
through
trusted third parties (TTPs), such as BBBOnLine, 4 TRUSTe, 5 and VeriSign. 6 The
study of Hu et al. (2001) classified the trusted seals intove types based on
their
functionality,
namely protecting privacy, providing security, demonstrating con-
sumer
satisfaction,
providing reliability, and providing
assurance
or guarantee. They
also confirmed that displaying a trusted seal could promote online trust and influ-
ence consumers' willingness to shop online. The other trust-inducing features in
this
dimension include providing
comprehensive,
correct, and current product
informa-
tion and using a domain name consistent with the brand or company name (Egger,
2001; Neilsen, 1999).
4.2.4. Social-cue design dimension
This dimension relates to embedding social cues, such as face-to-face
interaction
and social presence, into web site interfaces via different
communication
media. Be-
cause a lack of the
‘‘human
touch
’’
or presence may constitute a barrier for at
least
4 http://www
.bbbonline.o
rg/
.
5 http://www
.truste.org/.
6 http://www
.verisign.com
/
.
Y.D. Wang, H.H. Emurian
some consumers to trust online merchants (e.g., Riegelsberger & Sasse, 2002),
severa
l
researchers have investigated the possibility of bringing consumer online
exp
eriences
closer to
interpersonal,
face-to-face level interactions. This
approach
has been
referred
to as
‘‘virtual
re-emb
edding
(Riegelsberger & Sasse, 2002; Steinbruck et al., 2002).
Using
walkthroughs
by 15 users as a method for evaluating reactions to social
cues
(i.e.,
photographs
and names of customer service agents, chat and call-back
oppor
-
tunities,
photographs
of the company, and
photographs
of a customer receiving
an
item) displayed on mock-up e-commerce interfaces, Riegelsberger and Sasse (2001)
reported that medium experienced users responded
favorably
to
photographs
as
trust
inducing factors. Social cues that were functional, such as chat and call-back
oppor
-
tunities, were valued highest. Riegelsberger and Sasse (2002)
subsequently
reported
a
protocol analysis of 15 users' impressions of four mock-ups of the German
online
shopping site Amazon.de, all of which displayed a
photograph,
usually of a
person.
Photographs on the web site were
evaluated
positively for experienced shoppers if the
photograph was
appropriate
to the brand. However, non-shoppers with
pre-existing
low levels of general trust perceived
photographs
of people as a
manipulative
plo
y.
In an experiment in Germany that studied 45 users' responses on a trust
que
s-
tionnaire to three mock-up versions of an online-bank web site, Steinbruck et
al.
(2002) found that the web sites displaying a
photograph
of the
company
'
s repre-
sentative produced higher trust reports than did the web site lacking a
photograph
.
Including a photograph of a person, then, might be a simple but effective
techni
que
to increase the
trustworthiness
of an online
mercha
nt.
In addition, the interactivity of
communication
media was found to
influence
consumer trust (Basso et al., 2001).
Synchronous communication
media, such
a
s
instant messaging or video telephony, which
‘‘provides opportunities
for real-time
feedback and thus minimizing
misunderstandings
and demonstrating
atten
tiveness
(Basso et al., 2001, p. 138), increases judgments of the
trustworthiness
of the
sales
-
person and the online
mercha
nt.
Although the ndings from these studies do favor the use of
photographs
overal
l,
there are complex interactions between the display of real images on an e-commerce
web site and the prior experiences and the current objectives of the users. It should
be
emphasized, then, that the studies on applying social cues, especially
photographs, to
web site design are still in a preliminary stage. However, the researchers do present
a
potentially effective
approach
to enhance online trust by adding a surrogate
human
presence and actual contact
opportunities
to the otherwise impersonal e-commerce
interface.
4.2.5.
Exampl
e
A single e-commerce web site was chosen to illustrate some of design
features
discussed in the above
framework.
Fig. 1 presents the web site for Saks Fifth Ave-
nue, 7 represented in gray tones. The interface is presented in muted colors on the
web.
7
http://www.sak
sfifthavenue.
com/
.
Y.D. Wang, H.H. Emurian
Fig. 1. Saks.com e-commerce interface.
Reproduced
with permission of Saks Fifth
Avenue.
In terms of the graphic dimension, this interface has graphics on roughly
ha
lf
(48%) of the page. There are
photographs
displayed of products as well as people.
In
terms of the structure design dimension, the interface appears to be uncluttered
and
concise, thereby exhibiting simplicity, although this judgment is subjective and
not
based upon information density or complexity metrics studied elsewhere (e.g., Em-
urian, 1994; Emurian & Gonce-Winder, 1994; Emurian & Seborg, 1990). Text
and
image links are arranged in an orderly vertical or horizontal fashion, an
arrange
ment
that minimizes information complexity, and the text describing products is
large
r
than the text for other features of the site, such as ‘‘EMAIL UPDATES’’
and
‘‘Catalogs.
’’
Deeper in the site, the consistency of the interface is evidenced by the
common links at the top and bottom of each page. Product
photographs
and
links
are consistently positioned within the center
sectio
n.
In terms of the content design dimension, the Saks brand logo is clearly visible in
the upper left corner of each page. All aspects of the
consumer- -- merchant relation-
ship are presented at the bottom of each page. Included there are links to
statemen
ts
about privacy and return policies, among other types of information.
However,
there is no seal of
approval
concerning privacy or security from an
independent
Y.D. Wang, H.H. Emurian
certication of the web site. In terms of the social-cue design dimension, there is
a
link on the web site to a chat-line with a sales
representative.
Also presented are
an
email address and an 800 telephone
number.
The features represented in the Saks web site, then, illustrate many of the fac-
tors considered instrumental in a merchant's design of a
trustworthy e-commerce
interface.
5. Conclusion and future direction
s
Lack of trust has been repeatedly identified as one of the most formidable
barri
ers
to people for engaging in e-commerce, involving the disclosure of nancial
and
personal information to online merchants (Hoffman, Novak, & Peralta, 1999). The
future of e-commerce is tenuous without a general climate of online trust.
Hen
ce,
building and sustaining consumer trust on the Internet present a
con
siderab
le
challenge for online merchants and is a research topic of increasing interest
and
importance.
This paper provided an overview of the nature and concepts of trust from
multiple
disciplinary
perspectives, and it reviewed the relevant studies that investigate the
elements of online trust. Also, a framework of trust-inducing features
articu
lated
from existing literature was proposed. The design features have been synthesized
and
classified into four dimensions, namely graphic design, structure design,
con
tent
design, and social-cue design. This
framework,
though still preliminary and in
need
of refinement and
evaluation,
suggests how to implement trust-inducing features in
furtherance of enhancing online trust by web interface design. Finally, other re-
searchers are also developing
frameworks
from their
interpretation
of the
literat
ure,
and, for example, Nah and Davis (2002) proposed a model based on web site
con
-
tent, design, and external certications and
refer
ences.
Other
approaches
to enhancing trust between consumers and online
merchants,
such as the development and use of ‘‘Internet Trust
Brokers’’
discussed by
Fukuy-
ama (2001), were not considered within the current
framework.
An example of a
type
of such a broker is ePublicEye.com, which collects data from online consumers
and
posts
evaluations
of vendors offering various categories of goods and services
online.
Online trust has become a prominent research topic, and it offers promising op-
portunities for further research. Based upon the present overview, ve
potenti
al
areas of suggested research include: (1) the effects of culture on online trust; (2) the
effects of domain (.com, .edu, .org) on online trust; (3) the reasons for losing
online
trust and the ways to repair it; (4) the importance of civil remedies for consumers in
case of
violations
of privacy laws; and (5) the
transferability
of online trust from the
Internet to other activities. A more extensive, but somewhat different, set of
future
directions for research was suggested by Shankar et al. (2002) who proposed eight
avenues of potential investigation based on their analysis of
multi-stakeholder fac-
tors in online trust. In addition, there is obvious need for more
comprehen
sive
frameworks
to model online trust and more empirical
evaluations,
rather
than
survey-based
and hypothetical
approaches,
to include the present
formulation.
The
Y.D. Wang, H.H. Emurian
present
framework
is not exhaustive in the sense that it does not intend to
capture
every trust-inducing feature that web designers can apply. It only suggests a
starting
point for experimental
evaluations
and
validations
of its components. Finally, it
should be pointed out that while well-crafted web interfaces can induce those
who
intend to purchase online, online merchants should also pay attention to
other
methods, such as customer relationship management (CRM) and offline
market
ing
strategies, to obtain consumer trust and nurture strong business
relationships (Tan,
Yen, & Fang, 2002).
Even when an interface is optimized to induce trust, the safety and benets of e-
commerce will still require an educated consumer who is informed about the
risks
and protections that are present in such
transactions.
The fact that few
consumer
s
have reported problems with online shopping and that known violations of con-
sumer privacy have resulted in publicized actions taken against
untrustw
orthy
merchants bodes well for the future of e-commerce. Since a person's degree of
trust,
commonly called individual trust propensity, is a dynamic function