ethics research 1
TRANSCRIPT
-
8/3/2019 Ethics Research 1
1/29
THEORY AND REVIEW
PRODUCT-RELATED DECEPTION IN E-COMMERCE:
ATHEORETICAL PERSPECTIVE1
Bo Xiao
Department of Computer Science (Computing and Information Systems), Faculty of Science,
Hong Kong Baptist University, Kowloon, Hong Kong, CHINA {[email protected]}
Izak Benbasat
Sauder School of Business, University of British Columbia, 2053 Main Mall,
Vancouver, BC V6T 1Z2, CANADA {[email protected]}
With the advent of e-commerce, the potential of new Internet technologies to mislead or deceive consumers has
increased considerably. This paper extends prior classifications of deception and presents a typology of
product-related deceptive information practices that illustrates the various ways in which online merchants
can deceive consumers via e-commerce product websites. The typology can be readily used as educational
material to promote consumer awareness of deception in e-commerce and as input to establish benchmarks
for good business practices for online companies. In addition, the paper develops an integrative model and
a set of theory-based propositions addressing why consumers are deceived by the various types of deceptive
information practices and what factors contribute to consumer success (or failure) in detecting such deceptions.
The model not only enhances our conceptual understanding of the phenomenon of product-based deception and
its outcomes in e-commerce but also serves as a foundation for further theoretical and empirical investigations.
Moreover, a better understanding of the factors contributing to or inhibiting deception detection can also helpgovernment agencies and consumer organizations design more effective solutions to fight online deception.
Keywords: Product-based information practices, electronic commerce, typology, stimulusorganismresponse
framework, model of deception detection
Introduction1
The rapid growth of electronic commerce (e-commerce) has
created fertile ground for online fraud and deception (Federal
Trade Commission 2001; Grazioli and Jarvenpaa 2000). The
Federal Trade Commission (FTC), the U.S. Securities and
Exchange Commission (SEC), the Federal Bureau of Inves-
tigation (FBI), as well as consumer protection agencies, such
as the National Consumers League (NCL), have all voiced
concerns over Internet consumer fraud and have initiated
specialized programs targeted at detecting and prosecuting
such practices (Grazioli and Jarvenpaa 2003b). According to
the annual report released by the Internet Crime Complaint
Center (IC3)a partnership between the FBI, the National
White Collar Crime Center (NW3C), and the Bureau of
Justice Assistance (BJA)in 2008, the IC3 received a total of
275,284 complaints from consumers claiming to have been
defrauded online, which represents a 33.1 percent increase
over the previous year; the total dollar loss linked to online
fraud was U.S. $265 million, about $25 million more than in
2007; and the average individual loss amounted to $931
(Internet Crime Complaint Center 2008). In addition to finan-1M. Lynne Markus was the accepting senior editor for this paper. Choon
Ling Sia served as the associate editor.
MIS Quarterly Vol. 35 No. 1 pp. 169-195/March 2011 169
-
8/3/2019 Ethics Research 1
2/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
cial damage, victims of Internet deception often suffer from
the psychological damage of being victimized, the loss of
time for filing complaints and refund requests, and the loss of
private information (Grazioli and Jarvenpaa 2000). The
increases in online fraud can also adversely affect honest
online businesses. When consumers are reluctant to make
purchases online for fear of deception, online businesses as a
whole suffer from loss in sales and reputation.
Consumer deception is clearly an important concern for
e-commerce. Prior studies in offline and online deception
(Boyle 2003; Davies and Parasuraman 1982; Gilovich 1991;
Johnson et al. 1993; Johnson et al. 2001; Klein et al. 1997)
have focused on investigating factors contributing to indi-
viduals detection (or nondetection) of another partys
deliberate attempts to deceive. However, they have not
investigated the rather obvious question ofwhy individuals
may actually be deceived by deliberately deceptive practices.
In a world where deceptive practices are perpetrated viatelephone, in print, on radio, or on TV, deception theorists
may not have asked this question because of the commonly
held assumption that people are deceived because they do not
pick up the nonverbal behavioral cues leaked that can serve
as telltale signs of deception (i.e., changes in deceivers
behaviorsuch as pupil dilation, higher vocal pitch,
fidgeting, blinkingas a result of the physiological arousal,
emotional reactions, cognitive effort, and attempted control
associated with deception) (Ekman and Friesen 1969;
Zuckerman et al. 1981, 1986). Media richness theory (Daft
and Lengel 1986) and social presence theory (Short et al.
1976), which are often used as the theoretical foundations for
empirically examining deception detection across different
media, suggest that people should be more accurate in
detecting deception in richer media (e.g., face-to-face)
where more nonverbal cues are available for judgment. For
instance, Short et al. (1976) found that people are more
responsive to fraud attempts via phone than via face-to-face
communication because the lack of visual cues in the phone
medium makes people more vulnerable to strategic
manipulation.
However, with the advent of e-commerce, the potential of
new Internet technologies to mislead or deceive consumers is
considerable. As noted by Heckman and Wobbrock (2000),every new technology applicable to commerce (e.g., tele-
graph, telephone, radio, or television) helps the unscrupulous
to swindle the unwary. E-commerce is no exception. Al-
though many deceptive information practices in e-commerce
settings are variations of well-known deception types already
used in the traditional physical shopping contextsuch as the
misrepresentation of merchants, products, and return/refund
policiesthe advent of e-commerce has not only made
deception more likely and the perpetration of deceptive acts
easier, but has also introduced new avenues for deception
(Grazioli and Jarvenpaa 2001, 2003a).
First, the unique characteristics of the Internet, such as digital
environment, low entry barriers, spatial/temporal separation,
and anonymity, have made it a fertile ground for deception.
The Internet is a digital environment, which lowers the effort
for online companies to create and change information
content as well as to manipulate the presentation and produc-
tion of such information content in order to achieve deception.
For instance, web pages can be constructed to attract/distract
attention, encourage/discourage cross-comparisons, force
choices, and create pressure to buy immediately (Aditya
2001). The Internet also lowers the resources needed to set up
a genuine-looking online storefront, thus making the mer-
chants identity easy to falsify and difficult to authenticate.
In addition, the physical distance between the web merchant
and its customers, as well as the temporal separation ofpayment and product delivery, makes it harder for customers
to verify the truthfulness of the website or its claims. More-
over, since anonymity provides people with a low threat
setting, it may breed disinhibited antisocial behavior (Suler
2004, 2005), including deception.
Second, the various innovative technologies supporting
e-commerce have also given rise to novel forms of deceptive
practices. Prior research (e.g., Benassi 1999; Glover and
Benbasat 2006; Grazioli and Jarvenpaa 2000) has revealed a
number of information technology (IT) mechanisms that can
be used by online companies to increase consumers trust ine-commerce websites and/or mitigate their risk perceptions
associated with online shopping. However, as illustrated in
Appendix A, both trust-building mechanisms and risk-
reducing tools can be exploited by dishonest companies to
deceive consumers (Grazioli 2004; Grazioli and Jarvenpaa
2000). For instance, an online company may design a product
recommendation agent (PRA) that provides biased product
recommendations to serve the interests of the company. It
may provide a virtual product experience (VPE) (Jiang and
Benbasat 2005, 2007a) that does not represent consumers
real experiences with a product (Bloom et al. 1994). It may
also employ multimedia technologies (e.g., Flash, animations)
to excite consumers in a manner similar to what fast-talking
salespersons can do in physical shopping settings, thus
leading consumers to make impulse purchases they may regret
later (Bloom et al. 1994). In electronic marketplaces such as
eBay, a seller can collude with a group of buyers in order to
receive unfairly high ratings, thus allowing that seller to
receive more orders from buyers and at a higher price than
deserved (Dellarocas 2000).
170 MIS Quarterly Vol. 35 No. 1/March 2011
-
8/3/2019 Ethics Research 1
3/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
Concerned with the identity crisis in the IS research com-
munity, Benbasat and Zmud (2003) urged IS researchers to
pay close attention to the information technology artifact and
to incorporate it in IS theories and design recommendations.
In the case of e-commerce, there is significant potential for IT
(i.e., the Internet and the technologies supporting
e-commerce) to contribute to consumer deception. This paper
makes the following contributions to research and practice:
It extends prior classifications of deception and presents
a typology of product-related deceptive information
practices that illustrates the various ways online mer-
chants can deceive consumers via e-commerce product
websites. The typology can be readily used as educa-
tional material to promote consumer awareness of
deception in e-commerce and as input to establish bench-
marks for good business practices for online companies.
It develops an integrative model and a set of theory-basedpropositions addressing why consumers are deceived by
the various types of deceptive information practices and
what factors contribute to consumer success (or failure)
in detecting deceptions. It enhances our conceptual
understanding of the phenomenon of product-based
deception and its outcomes in e-commerce and serves as
a foundation for further theoretical and empirical investi-
gation. Moreover, a better understanding of factors
contributing to or inhibiting deception detection can also
help government agencies and consumer organizations
design more effective solutions to fight online deception.
This paper focuses onproduct-relateddeceptive information
practices. A 1999 survey by Indiana University and the
professional service organization KPMG revealed that
product information is the most important concern for online
customers over the age of 25 (Williams and Larson 2000).
The importance of product information to consumers has
motivated online merchants to perform deceptive manipu-
lations on such information so as to influence consumers
judgment and decision making in e-commerce settings.
According to Pavlou and Gefen (2005), product mis-
representation is one of the most common forms of Internet
fraud reported. The NCL has also consistently ranked product
misrepresentation among the top two Internet scams (e.g., NCLs Fraud Center 2007). Many consumers enjoy the
convenience and low prices offered by online shopping, yet
these benefits may be countermanded by the increased risk
associated with the products purchased online.
In the next section, we present a typology of product-related
deception information practices in e-commerce. Following
that, we introduce the conceptual model of why consumers ay
be deceived by deceptive practices embodied in websites, and
then go on to define the constructs that make up the model.
Next, we present propositions concerning the relationships
among the constructs. In the final section, we discuss
contributions to research and practice and suggest directions
for future work.
A Typology of Product-Related DeceptiveInformation Practices in E-Commerce
Deceptive practices revealed in prior research include:
(1) insufficient information disclosure of refund policies,
warranty information, and cancellation terms; (2) contract
default; (3) product-related deceptive information practices,
such as product misrepresentation; (4) nondelivery or late
delivery of product; and (5) misuse of personal and financial
information (Pavlou and Gefen 2005; Singsangob 2005).
To obtain an understanding of deceptive information practices
in e-commerce, we first identified the definitions of deception
suggested by psychologists and communication researchers.
They include
A communicators deliberate attempt to foster in others
a belief or understanding which the communicator
considers to be untrue (DePaulo and DePaulo 1989, p.
1553).
Message distortion resulting from deliberate falsifica-
tion or omission of information by a communicator with
the intent of stimulating in another, or others, a belief that
the communicator himself or herself does not believe
(Miller 1983, p. 92).
The deliberate attempt, whether successful or not, to
conceal, fabricate, and/or manipulate in any other way
factual and/or emotional information, by verbal and/or
nonverbal means, in order to create or maintain in
another or in others a belief that the communicator him-
self or herself considers false (Masip et al. 2004, p.
148).
Three characteristics of deception appear to be consistent
across the above definitions.
1. Deception is an intentional or deliberate act. The ele-
ment of intentionality is what distinguishes deception
(i.e., intentional distortion of messages) from misinfor-
mation (i.e., unintentional distortion of messages) (Masip
et al. 2004, p. 148).
MIS Quarterly Vol. 35 No. 1/March 2011 171
-
8/3/2019 Ethics Research 1
4/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
2. Deception is accomplished by manipulating information
in some way.
3. Deception has an instrumental end goalthat is, to create
or maintain a belief in another that the communicator
herself believes to be false.
In accordance with these characteristics, this paper defines
product-relatede-commerce deceptive information practices
as the deliberatemanipulation of product-related information perpetrated by online merchants to mislead consumers in
order to induce desired attitudinal and behavioral changes in
consumerschanges that are detrimental to consumers and
beneficial to the merchants.
Prior research has suggested many different ways of classi-
fying deception. Summarizing the multitude of previous
categorizations of deception, Buller, Burgoon, and theircolleagues (Buller et al. 1994; Burgoon et al. 1994) dis-
tinguished among three relatively distinct types of deception:
(1) concealment to withhold, omit, or disguise relevant
information; (2) equivocation to present information
vaguely and/or ambiguously; and (3)falsificationto present
false or exaggerated information. Extending Buller and
Burgoons concealmentequivocationfalsification classifica-
tion, which is generic enough to be applied to different
deception cases, we include a new dimension detailing the
specific operational-level deception techniques in which the
three types of deception are carried out in the e-commerce
context. This results in a 3 3 typology of deceptive infor-
mation practices that describes how deception works within
e-commerce. The typology has two dimensions: one repre-
senting the three deception types already examined in prior
deception research and the other representing the following
three specific implementation techniques:
1. The manipulation ofinformation content, which refers to
the direct alteration of the content of product information
provided at an e-commerce website.
2. The manipulation ofinformation presentation, which
refers to the manipulation of the design of how product
information is presented to consumers at an e-commerce
website.
3. The manipulation of information generation, which
refers to the manipulation of the dynamic production of
product information at an e-commerce website, based on
consumer interests, needs, and/or preferences obtained
explicitly or implicitly.
Since the dimension of these three different deception tech-
niques is the main contribution of this paper, in the remainder
of this section, we will discuss each in greater detail.2
It should be noted many similarities exist between deception
in e-commerce product websites and deception in contexts(e.g., advertising, personal selling, close relationship, employ-
ment), particularly in advertising. While an exact demarca-
tion between honest advertising and deceptive advertising is
often difficult to make, we consider a piece of advertising
deceptive when it employs any of the deceptive manipulations
specified in our 3 3 typology of deceptive information
practices in order to mislead the consumer acting reasonably
in the circumstances, to the consumers detriment (Ford and
Calfee 1986, p. 86).
Manipulations Performed on
Information Content
Prior studies in interpersonal deception have focused
predominantly on the nonverbal cues accompanying decep-
tion, such as pitches and tones, body gestures, and facial
expressions, rather than on the deceptive messages commu-
nicated by the deceivers (Buller and Burgoon 1996).
Research investigating deceptive messages has primarily
examined manipulations that can be performed on the content
rather than thepresentation of these messages. For instance,
McCornack (1992) and Buller and Burgoon (1996) propose
that individuals can manipulate the content of information
simultaneously along several different dimensions such ascompleteness, clarity, and veridicality, which correspond to
the three deception types concealment, equivocation, and
falsification, respectively.
For product-related information provided at an e-commerce
website, content can be concealed, equivocated, and/or falsi-
fied by online companies. For instance, an online company
can withhold negative information (e.g., a known safety
problem) about a product (i.e., concealment); provide vague
information about the total cost (e.g., selling price, tax,
shipping and handling fee) of a product (i.e., equivocation);
give ambiguous information concerning product return andrefund policies (i.e., equivocation); automatically filter out
negative consumer reviews (i.e., concealment); pose as con-
2While we will focus onpersonalizedinformation (i.e., information tailored
to the specific needs of the individual customer) when discussing the
manipulation ofinformation generation, our discussion ofinformation pre-
sentation and information contentmanipulation will center on nonperson-
alizedinformation.
172 MIS Quarterly Vol. 35 No. 1/March 2011
-
8/3/2019 Ethics Research 1
5/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
sumers to write positive reviews about products and services
received from the company (i.e.,falsification); or even sell a
nonexistent product (i.e.,falsification).
In sum, by altering the availability and quality of information,
online companies can manipulate the content of productinformation at an e-commerce website so as to enhance
consumers evaluation of those particular products.
Manipulations Performed onInformation Presentation
In addition to the direct alteration of the content of product
information, deceptive manipulations can also be performed
on the presentation of product information at an e-commerce
website. Kleinmuntz and Schkade (1993) note that the
presentation of information can be designed to encourage
effective decision making. In the same vein, the design of
information presentation and delivery can be manipulated to
lead to biased decision making. Since the number of potential
ways for presentation and delivery of certain information
content is vast, we focus on two important characteristics that
apply to a broad range of contexts:presentation media and
information organization.
Presentation Media
Deceptive manipulations can be performed on the presen-
tation of product information at an e-commerce website viathe manipulation ofpresentation media . Information content
can be presented using a variety of media in the e-commerce
context, such as text, graphics, audio, video, and animations,
so that users can make better sense of the information
available (Lim and Benbasat 2000). Heller and Martin (1995)
have categorized four different types of presentation media
with increasing complexity: text,graphics,sound, and motion.
We add to this classification a new media type, virtual experi-
ence (e.g., virtual reality), which involves active consumer
interaction with the media (Jiang and Benbasat 2005, 2007a,
2007b).
Online companies can manipulate presentation media in three
ways to achieve their deceptive ends. First, an online com-
pany can alter the individual features of a medium to either
inhibit correct product understanding or foster incorrect
product understanding; for instance, an online company may
use images (still images or videos) of small size and low
fidelity to present a product that has some exterior problems
(i.e., concealment). It can also manipulate the response rate
of an online product demonstration to mislead consumers
about a particular product feature; for example, it may
demonstrate a shorter delay between shots of a digital camera
(i.e.,falsification).
Second,by manipulating the level ofvividness of a presen-tationthat is, the extent to which the presentation is
emotionally interesting, imagery provoking, and inherently
appealing (Sundar and Kalyanaraman 2004)deceptive
online companies can direct consumers attention toward
irrelevant information, distract their attention from relevant
information, and shape their overall attitude toward and
judgment of certain products. For instance, an online com-
pany may choose a text-only presentation for a product with
desirable functionalities but unappealing appearance (i.e.,
concealment). The same company may use flashy animations
as a decoy to distract consumers from processing non-vivid
yet more useful and informative textual descriptions (i.e.,
concealment). In addition to attracting consumers attention,
the sensory stimuli supplied by a vivid presentation may also
trigger intense emotional responses (e.g., pleasure and
arousal) that can overwhelm consumers self-observation
during online shopping, leading to unregulated buying
behavior (e.g., impulse buy) (LaRose 2001).
Finally, a common deceptive practice utilized by online
companies is to present conflicting information via different
media. For instance, a digital camera retailer may state truth-
fully in the textual description that the interchangeable lens of
a single lens reflex (SLR) camera is not included in the
package; however, it may display an image of the camera withlens attached, with no annotation to indicate that the lens is
not included (i.e., concealmentand equivocation). Since
information conveyed in images is given greater weight in
consumer judgment than that conveyed in text, particularly
when the two types of information are in conflict (Argyle et
al. 1971; Bone and France 2001), consumers are likely to be
misled into believing that the lens featured in the image is
actually part of the package and may potentially make a
purchase decision to their own detriment.
Information Organization
The manipulation of information presentation can also be
achieved via the manipulation ofinformation organization.
Information content in an e-commerce website can be
organized meaningfully into groups (Jarvenpaa 1989;
Kleinmuntz and Schkade 1993), hierarchies, and/or sequences
(Kleinmuntz and Schkade 1993). Information organization
provides a cognitive incentive system for decision makers by
MIS Quarterly Vol. 35 No. 1/March 2011 173
-
8/3/2019 Ethics Research 1
6/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
influencing the effort and accuracy associated with infor-
mation processing strategies and, therefore, inducing the use
of different strategies (Kleinmuntz and Schkade 1993).
Unscrupulous online companies can manipulate the way
information is organized in their web pages to encourage or
discourage the use of certain information processing strategies
by consumers. For instance, to encourage consumers to com-
pare different products based on a certain attribute (e.g., the
only attribute on which the promoted products have advantage
over other products), online companies may provide the
functionality to sort by that attribute alone (i.e., concealment).
Likewise, since information at a deep level of navigation
requires more effort to access (Chau et al. 2000), online com-
panies can hide negative product information and consumer
reviews such that consumers have to traverse many levels of
navigation to locate such information (i.e., concealment).
The sequence in which information appears may influence
consumers judgment as to the relevance and/or importance ofthe individual pieces of information (Kleinmuntz and Schkade
1993; Schkade and Kleinmuntz 1994), particularly when
consumers hold certain expectations as to the order of infor-
mation presentation; for instance, when they expect that
products are arranged in order of decreasing popularity.
Deceptive online companies can thus influence consumers
decision making by manipulating the sequence of presented
information in accordance with their expectations. For
example, online companies may present a promoted product
at the top of their bestselling list in order to increase the
chance that consumers will choose that particular product
(i.e.,falsification).
In summary, the manipulation of information presentation can
be achieved in two ways. First, by a combination of the
different means of manipulating presentation media, online
companies can influence consumers product understanding
and attention as well as emotional responses conducive to
deception. Second, by manipulating the way product infor-
mation is organized at an e-commerce website, deceptive
online merchants can influence consumers information pro-
cessing strategies as well as the accessibility and the per-
ceived relevance of information.
Manipulations Performed onInformation Generation
The third type of deceptive manipulation is performed on
information generation, which refers to the dynamic produc-
tion of product-related information at an e-commerce website,
based on consumer interests, needs, and/or preferences
obtained explicitly or implicitly. Examples of e-commerce
technology supporting dynamic production of information
include search engines, product catalogs, and online product
recommendation agents (PRAs). Our discussion of the
manipulation of information generation focuses on PRAs
because they are likely to be the first technological artifacts
with which a consumer interacts at an e-commerce website
that has a vast number of products or choices.
PRAs are software artifacts that take as input individual
consumers product-related interests or preferences, obtained
either explicitly or implicitly, and subsequently provide
recommendations for products that match the consumers
expressed interests or preferences (Xiao and Benbasat 2007).
Appropriately designed PRAs can enable consumers to make
informed purchase decisions by reducing their information
overload and search complexity, while improving their
decision quality. However, the degree to which PRAs
actually empower consumers depends upon the veracity and
objectivity of the PRAs (Hill et al. 1996; King and Hill 1994).
Users of automated decision aids have been found to place
undue trust in such technologies, leading to the abusive use of
such systems and biased decision-making processes, par-
ticularly when the competence of the systems far exceeds that
of their users (e.g., Mosier et al. 1998; Skitka et al. 1999).
Two classes of errors that often emerge in highly automated
decision-making environments are omission errors (which
occur when people fail to respond to system irregularities/
events because the automated decision aid fails to detect or
indicate them) and commission errors (which occur when
people incorrectly follow the directive or recommendation of
the automated decision aid, despite contraindications from
other sources of information) (Mosier et al. 1998; Skitka et al.
1999). Likewise, consumers (particularly those with little
experience in the intended product category) may over-rely on
the PRAs to make decisions for them, rather than using the
PRAs as one component of a thorough monitoring and
decision-making process (Skitka et al. 2000). Moreover,
people have a tendency to trust experts or specialists. Just as
we trust advice from human experts rather than that offered
by nonexpert friends and relatives, we also trust IT artifacts
that claim specialty. People who lack expertise in the subject
matter are likely to perceive a technology labeled as
specialist (e.g., PRAs acting as online sales advisors) to becredible (Reeves and Nass 1996; Tseng and Fogg 1999).
Unscrupulous online companies can prey upon consumers
double vulnerabilities, that is, theirtrust inautomation and
trust in specialist, by designing deceptive PRAs that provide
recommendations biased toward their own interest. For
instance, a PRA can focus consumers attention only on
criteria on which the promoted products have a distinctive
174 MIS Quarterly Vol. 35 No. 1/March 2011
-
8/3/2019 Ethics Research 1
7/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
competitive advantage (Wagner et al. 2001) (i.e., conceal-
ment); provide false decision guidance to influence con-
sumers decision criteria (i.e.,falsification); give priority to
promoted products by manipulating the underlying algorithm
for generating recommendations (Aksoy and Bloom 2001)
(i.e.,falsification); mix promoted products (which do not go
through the PRAs filtering process) with those that actually
fit consumers preferences in an unordered set of recom-
mendations (i.e., equivocation); exclude products that best fit
consumers preferences from its recommendation list (i.e.,
concealment); and provide value or overly general explana-
tions on how the recommendations are generated (i.e.,
equivocation). There have already been some reported
incidents of companies generating false recommendations to
consumers. For instance, Amazon.com has admitted to using
faux recommendations to drive business to its new clothing
store partners. The false recommendations were positioned
right next to the legitimate recommendations for books,
music, and other items that were generated based oncustomers purchase histories (Wingfield and Pereira 2002).
In summary, by manipulating how personalized information
is generated at an e-commerce website (e.g., through product
recommendations), online merchants can influence con-
sumers product evaluation and subsequently their decision
making.
Summary
In this section, we have proposed a two-dimensional typology
of deceptive information practices in e-commerce. The threemajor types of deceptionconcealment, equivocation, and
falsificationexamined in prior deception research comprise
one dimension, while the other dimension consists of the three
operational-level deception techniquesthe manipulation of
information content, the manipulation of informationpresen-
tation, and the manipulation of information generationin
the e-commerce context, with the latter dimension being our
focus of discussion. Please see Table 1 for examples of
deceptive information practices in e-commerce based on our
new typology.
A Theory of Product-Related DeceptiveInformation Practices in E-Commerce
In the previous section, we presented a 3 3 framework of
the various ways by which online merchants can exploit the
capabilities of IT to deceive consumers with e-commerce
websites. In this section, we present our theoretical model of
product-related deceptive information practices in e-
commerce. Drawing primarily from thestimulusorganism
response framework in environmental psychology (Mehrabian
and Russell 1974) and the model of deception detection
(Johnson et al. 1993; Johnson et al. 2001), our model theo-
rizes (1) how the various types of deceptive information
practices exploiting IT capabilities (based on the typology that
we developed and presented in the previous section) con-
tribute to consumer deception, and (2) what factors contribute
to the success (or failure) of consumer deception detection at
an e-commerce website embodying deceptive information
practices.
Thestimulusorganismresponse framework posits that the
variousstimuliwithin a shopping environment together affect
a consumers cognitive and/or affective processes (organism),
which in turn determine the consumers behavioralresponses,
expressed as either an approach behavior or an avoidance
behavior. Whereasstimuli are cues external to the customer
that rouse or incite him consciously or subconsciously intoaction (Belk 1975), organism refers to the intervening internal
processes (e.g., the perceptual, feeling, and thinking activities)
between the stimuli and the reaction of the consumer (Bagozzi
1986). The concept ofapproachavoidance is defined in a
broad sense to include physical movement toward, or away
from, an environment or stimulusdegree of attention,
exploration, favorable attitudes such as verbally or non-
verbally expressed preference or liking, [and] approach to a
task (Mehrabian and Russell 1974, p. 96). Approach
behaviors are positive attitudes/actions toward a product/
brand, service, or shopping environment (such as staying in
the store, purchasing products, or returning to the store), while
avoidancebehaviors are the opposite (Parboteeah 2005). The
stimulusorganismresponse framework has been extensively
elucidated (Bagozzi 1986) and widely adopted in past
research, with promising results, to model the impact of
environmental stimuli on consumer responses in both offline
and online shopping contexts (e.g., Baker et al. 1994; Eroglu
et al. 2001; Fiore and Kim 2007; Sherman et al. 1997). In an
e-commerce context, the various types of deception tactics
can be deliberately employed by online merchants to create
the stimuli that can activate consumers internal cognitive/
affective processes, which, in turn, promote an approach
behavior toward the e-commerce website. As such, the
stimulusorganismresponseframework serves as an appro-priate overarching framework for our theoretical model.
Johnson and his colleagues (Johnson et al. 1993; Johnson et
al. 2001) developed a model of deception detection that
describes four subprocesses by which individuals, based on
their domain knowledge and the available information cues,
determine whether information provided by a sender is
deceptive (Grazioli 2004):
MIS Quarterly Vol. 35 No. 1/March 2011 175
-
8/3/2019 Ethics Research 1
8/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
Table 1. Examples of Product-Related Deceptive Information Practices in B2C E-Commerce
Deception
Types
Manifestations of Deception Types
Manipulation of Info
Content
Manipulation of Info Presentation
(Media and Organization) Manipulation of Info Generation
Concealment
The website withholds
negative information
(e.g., a known safety
problem) about a
product.
The website
automatically filters out
negative consumer
reviews.
The websites use flashy animations as a
decoy, that is, to distract consumers
from processing non-vivid yet more use-
ful and informative textual descriptions
( Presentation media).
To encourage consumers to compare
different products on a certain attribute
(e.g., the only attribute on which the
promoted products have advantage over
other products), the website provides the
functionality to sort by that attribute alone
( Information organization)
The PRA focuses consumers
attention only on criteria on which
the promoted products have a
distinctive competitive advantage.
The PRA excludes products that
best fit consumers preferences
from the recommendation list.
Equivocation
The website provides
vague information about
the total cost (e.g.,
selling price, tax,
shipping and handling
fee) of a product.
The website gives
ambiguous information
concerning product
return and refund.
The website shows conflicting informa-
tion about what is included in the pack-
age with different media. For instance,
while displaying a vivid image of an SLR
camera with lens attached (with no
annotation to indicate that the lens is not
included), the website states (truthfully)
in the textual description that the lens of
the camera is sold separately (
Presentation media).
The PRA presents an unordered
set of recommendations, mixing
promoted products (which do not
go through the PRAs filtering
process) with those that actually fit
consumers preferences.
The PRA provides vague or overly
general explanations as to how the
recommendations are generated.
Falsification
The website sells a
nonexistent product.
Staff of the online com-
pany pose as prior
consumers and write
positive reviews about
products and services
received from the
company.
The website manipulates the response
rate of an online product demonstration
to mislead consumers about a particularproduct feature (e.g., demonstrate a
shorter delay between shots of a digital
camera) ( Presentation media).
The website presents promoted products
at the top of its bestselling list (
Information organization).
The PRA gives priority to promoted
products by manipulating the
underlying algorithm for generatingrecommendations.
The PRA provides false decisional
guidance (i.e., guidance as to how
to choose a certain product) to
influence consumers decision
criteria.
1. The activation subprocess consists of allocating attention
to cues, based on the presence of discrepancies between
what is observed and what is expected.
2. The hypothesis generation subprocess is where indi-
viduals generate interpretive hypotheses to explain the
anomalies detected during the activation process.
3. The hypothesis evaluation subprocess is where pre-
viously generated deception hypotheses are evaluated (by
comparison with some criterion) to determine their
acceptability.
4. Theglobal assessmentsubprocess consists of combining
the accepted hypotheses into one synthetic assessment of
deceptiveness.
According to Johnson and his colleagues, individuals are
deceived because they cannot detect deception due to their
inability to identify anomalies/inconsistencies in their inter-
actional environment, generate deception hypotheses to
explain the detected anomalies/inconsistencies, evaluate
already generated deception hypotheses, or arrive at an appro-
priate global assessment of the other partys deceptiveness.
176 MIS Quarterly Vol. 35 No. 1/March 2011
-
8/3/2019 Ethics Research 1
9/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
There is some empirical work in applying the model of
deception detection in the online shopping context. Inte-
grating the process-oriented model of deception detection
(Johnson et al. 1993; Johnson et al. 2001) with the broader
deception, trust, and risk model of Internet consumer behavior
developed by Grazioli and his colleagues (Grazioli 2004;
Grazioli and Jarvenpaa 2000; Grazioli and Wang 2001)
empirically investigated the determinants of deception detec-
tion success by comparing the information processing
behavior of successful and unsuccessful detectors during the
four processes of activation, hypothesis generation, hypoth-
esis evaluation, and global assessment. They found that
peoples competence at evaluating the hypothesis of decep-
tion was a strong differentiator between successful and
unsuccessful detectors. In addition, successful and unsuc-
cessful detectors relied on different types of cues when
evaluating the deception hypothesis. Whereas successful
detectors relied on assurance cues (e.g., third-party seals,
warranties, news clips, and physical location of the onlinestore) and discounted trust cues (e.g., seller reputation,
customer testimonials, and seller size), unsuccessful detectors
did the opposite.
In a similar vein, researchers in psychological contract
breach3 in organizational settings (Morrison and Robinson
1997; Robinson 1996; Robinson and Morrison 2000) have
specified a two-stage model for detecting psychological
contract breach, in which individualsperceive psychological
contract breach before attributing it to either purposeful
reneging, reneging due to inability, or misunderstanding. In
this model, the hypothesis generation and hypothesis evalua-
tion subprocesses of the model of deception detection are
combined into a single stage and the global assessment
subprocess is omitted.
Informed by the model of deception detection and in line with
the more parsimonious model for detecting psychological
contract breach, our model focuses on two sequential sub-
processes of individuals deception detection process, namely,
the noticing of anomaly and the attribution of anomaly.
Individuals competence in noticing anomalies and in making
proper attributions of such anomalies will likely help them
perform better in deception detection.
In the remainder of this section, we will introduce our
theoretical model and define the constructs in the model.
Then, we will present propositions concerning the relation-
ships among the constructs in the model.
Theoretical Model
Our theoretical model is presented in Figure 1. As shown in
Figure 1, the key constructs of the model (from right to left)
are consumers approach behavior toward target product(s);
consumers perception of website deceptiveness; affective
mechanisms; cognitive mechanisms; use of deceptive informa-
tion practices at e-commerce website; individual, product,
and situational characteristics; and deception detection
support mechanisms .
In the remainder of this section, whenever we mention the
terms anomaly, deception, and deceptiveness, we referspecifically to product-related anomaly, deception, and
deceptiveness.
Adopting the definition of approach behavior from
Mehrabian and Russell (1974), we define approach behavior
toward target product(s) as the positive attitude or action
toward the target product(s) (i.e., the product offering(s) on
which deceptive information practices have been performed).
It is the potential outcome, or the instrumental end goal, of
deceptive information practices. An approach behavior
toward target product(s) is exhibited when
A consumer generates an attitude toward the target
product(s), an attitude that she would not otherwise have
generated in the absence of the deceptive information
practices performed by the online merchants, or
A consumer makes a purchase of the target product(s)
that she would not otherwise have made in the absence of
the deceptive information practices performed by online
merchants, a purchase that may be to the detriment of the
consumer.
A consumer is deceived when she exhibits an approach
behavior toward the target product(s).
Perceived product-related deceptiveness in e-commerce
website is the extent to which a consumer believes that an
e-commerce website is deceptive in the content, presentation,
and generation of product-related information. It indicates the
consumers awareness of the online merchants intent to
deceive. Perceived deceptiveness is triggered by a negative-
valenced violation of consumers preconceived expectations,
3Research in psychological contract breach is closely related to deception
research in that deception can be considered as purposeful reneging of
psychological contract whereas incompetence can be considered as reneging
due to inability.
MIS Quarterly Vol. 35 No. 1/March 2011 177
-
8/3/2019 Ethics Research 1
10/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
Figure 1. A Theoretical Model of Product-Related Deceptive Information Practices in E-Commerce
often as a result of the deceptive information practices per-formed by online merchants, or the recognition of cues
suggesting deceptive information practices. It is the outcome
of the consumers deception detection process.
Cognitive/affective mechanisms refer to the cognitive and/or
affective mediating processes through which deceptive infor-
mation practices influence consumer behavior. Cognitive
mechanisms consist of beliefs, thoughts, or perceptions about
products, brands, and e-commerce websites, formed either
through direct interaction with product offerings or through
the processing of secondary source information. This model
includes one cognitive mechanism promoting approach
behaviornamely, perceived productvalue, referring to a
consumers overall assessment of the utility of a product
based on an assessment of what is received and what is given
(e.g., the trade-off between quality and price) (Sweeney and
Soutar 2001; Zeithaml 1988). Various deceptive information
practices are employed to either enhance consumers positive
assessment or reduce/mitigate their negative assessment of the
value of the target product(s).
The model also includes a cognitive mechanism potentiallyleading to an avoidance behavior (i.e., a negative attitude/
action)namely, deception detection process at an
e-commerce website. Websites that manipulate product-
related information with deceptive information practices often
exhibit anomalies that violate consumers preconceived
expectations, thus triggering a two-step deception detection
process: the noticing of product-related anomalies in the
e-commerce website, followed by the attribution of product-
related anomalies noticedin the e-commerce website. The
deception detection process may result in consumers belief
that the e-commerce website is deceptive and subsequently
lead to an avoidance behavior on the part of the consumers.
Affective mechanisms refer to emotional reactions activated by
the stimuli in the shopping environment. This paper examines
two such emotional reactions:pleasureand arousal. Whereas
pleasure refers to a subjective feeling state in which a person
feels good, joyful, or happy about the target product(s),
arousalis a subjective feeling state in which a person feels
excited, stimulated, alert, or active about the target product(s).
P4, P6
Affective Mechanisms
Cognitive Mechanisms
Use of Deceptive
Information Practices at
E-Commerce Website
Types of Deception Concealment
Equivocation
Falsification
Types of Deception
Techniques
Manipulation ofinformation content
Manipulation of
information presentation
Manipulation of
information generation
Approach
Behavior
Toward Target
Product(s)
Individual, Product, and
Situational Characteristics
STIMULUS ORGANISM RESPONSE
Perceived Product Value
Pleasure
Arousal
Perceived
Product-Related
Deceptiveness in
E-Commerce
Website
Dominance
Deception Detection
Process
Noticing of Product-
Related Anomaly at
E-Commerce Website
Attribution of Product-
Related Anomaly
Noticed at E-Commerce
WebsiteDeception Detection
Support Mechanisms
P4-5
P18-19
P25-26
P13-17
P20-24
P27-28
P8-11
P7
P1b
P1a
P12
P2
P3a
P3b
Causal link
Process arrow
178 MIS Quarterly Vol. 35 No. 1/March 2011
-
8/3/2019 Ethics Research 1
11/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
Deceptive information practices are sometimes performed to
induce pleasure and arousal in a consumer, which in turn
may result in an approach behavior on the part of the con-
sumer. Dominance was originally conceptualized as a third
emotional reaction to environmental stimuli (Mehrabian and
Russell 1974), defined as the extent to which a person feels in
control of or free to act in a situation. However, Russell and
Pratt (1980) proposed a modification of the Mehrabian
Russell theory that deletes the dominance dimension. They
argued that dominance required a cognitive interpretation by
the person and is therefore not purely applicable in situations
calling for affective responses. Moreover, prior empirical
studies have found the dominance factor to be of little
predictive value (Donovan and Rossiter 1982; Russell and
Pratt 1980). Hence, although dominance is included in our
theoretical model for the reason of completeness, we will not
state any propositions related to this factor.
Use of deceptive information practices at e-commercewebsite refers to the application of one or more types of
deceptive manipulations at an e-commerce website, based on
our newly developed typology of product-related deceptive
information practices presented in the previous section.
Individual, product, and situational characteristics refer to
the individual, product, and situational factors that moderate
whether or not the use of deceptive information practices
succeeds in deceiving customers. The individual factors
explored in this paper include consumers motivation (i.e.,
involvement with product and involvement with purchase
decision), experience with online/offline shopping in general,
product expertise,prior interaction(s) with the e-commerce
website,prior trust toward the e-commerce website, truth bias
(i.e., the predisposition to assume that others communication
is truthful) (Carlson et al. 2004), and the receipt of third-party
information about the e-commerce website. The product-
related factor isproduct type (i.e., whether the product is a
search product that can be assessed based on the values
attached to its attributes or an experience product charac-
terized by attributes that need to be experienced prior to
purchase). The situational factors include types of consumer
purchase (i.e., whether the purchase is a plannedone that
occurs when consumers have a specific purchase in mind or
an impulse purchase made with no pre-shopping intentions)(Beatty and Ferrell 1998; Moe 2003), task complexity, and
consultation of alternative resources (i.e., whether alternative
information resources, such as another website, are consulted
by consumers during online shopping).
Deception detection support mechanisms refer to support
mechanisms (IT-based or non-IT-based) that can help con-
sumers carry out the processes of deception detection,
namely, noticing or identifying anomalies (resulting from
deceptive information practices) in the e-commerce website
and attributing such anomalies to deception by online
merchants.
Next, we present propositions concerning the relationships
among the constructs in the theoretical model.
Propositions
Effects of Deceptive Information Practices
In this section, we theorize about the positive (i.e.,pleasure,
arousal, and perceived product value) and negative ante-
cedents (i.e.,perceived deceptiveness of the e-commerce web-
site) of consumers approach behavior toward target pro-
duct(s). We also examine the effects of deceptive informationpractices on these antecedents and the moderating effects of
situational (i.e., type of consumer purchases and task
complexity) and product (i.e.,product type) characteristics.
In accordance with the stimulus-organism-response model,
the environmental cues in e-commerce websites that embody
various deceptive information practices (as delineated in the
typology of e-commerce deceptive information practices)
create the stimuli that invoke different cognitive/affective
mechanisms by consumers. The cognitive mechanism ex-
amined in this paper is consumersperceived product value .
When consumers perceive higher value associated with a
particular product offered at an e-commerce website as aresult of the deceptive manipulation performed on product-
related information, they will be more likely to exhibit
positive attitude/action toward this product (an approach
behavior) compared to when deception does not exist.
Similarly, heightenedpleasure and arousal(the two affective
processes examined in this paper) triggered by sensory stimuli
created by deceptive information practices undermine con-
sumers self-control (LaRose 2001) and reduce their cognitive
attention (Cenfetelli 2006), thus leading to positive attitude/
action toward the productan approach behaviorwithout
active information processing (Forgas 1995). It is thus
proposed that
P1a: Perceived value of a target productis positively
associated with the likelihood that consumers will
exhibit approach behavior toward the target
product.
P1b: Affective reactions (i.e., pleasure and arousal)
toward a target product are positively associated
MIS Quarterly Vol. 35 No. 1/March 2011 179
-
8/3/2019 Ethics Research 1
12/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
with the likelihood that consumers will exhibit
approach behavior toward target product(s).
However, deceptive information practices will not succeed in
inducing desired consumer responses if consumers, during
their interaction with the e-commerce website, are able to
perceive the online merchants intent to deceive and thus
make appropriate and adequate corrections. When product-
related anomalies in the e-commerce website are noticed by
consumers and subsequently attributed to deception by online
merchants, an approach behavior toward the target pro-
duct(s) is less likely to take place, even when consumers have
enhanced perception of the value of and/or heightened affec-
tive reaction (i.e., pleasure and arousal) toward the target
product (effected by deception). It is thus proposed that
P2: Perceived product-related deceptiveness in e-commerce
websites is negatively associated with the likelihood that
consumers will exhibit approach behavior toward targetproduct(s).
P3: The effect ofperceived product value (P3a) as well as
that ofpleasure and arousal (P3b) on consumers
approach behavior toward target product(s) will be
attenuated when consumers perceive product-related
deceptivenessin an e-commerce website.
Individuals often lack the requisite cognitive resources to
form well-defined preferences that are stable over time and
invariant to the context in which decisions are made (Bettman
et al. 1998). Instead, they tend to construct their preferences
on the spot when they must make a choice (Bettman et al.
1998; Payne et al. 1992). Since these preferences are con-
structed, rather than absolute, they are sensitive to the charac-
teristics of the decision environment. E-commerce deceptive
information practices are performed to influence consumer
preference construction, potentially to their detriment.
All three types of deception techniquesmanipulation of
information content, manipulation of information presen-
tation, and manipulation of information generationwork by
enhancing consumers perception of the valueassociated with
target products and/or triggering feelings ofpleasure and
arousaltoward those products. However, each type exerts adifferential effect on consumers cognitive/affective pro-
cesses. The manipulation of information contentoperates
directly on the content of product information and is thus
expected to exert greater influence on consumers cognitive
evaluation of the value of a product. The same is true with
the manipulation ofinformation generation, which influences
the relative attractiveness of a particular product offering via
biased product recommendations. The manipulation of
information presentation, however, often achieves deception
by activating affective (i.e.,pleasure and arousal) processes
that may accentuate the perceived value associated with
particular product offerings. For instance, online companies
can use vivid presentations to trigger intense emotional
responses in consumers that can overwhelm their self-
observation during online shopping, potentially to their
detriment. It is thus proposed that
P4: Websites that exhibit deceptive information practices are
more likely to enhance consumers perceived value
associated with a target product and trigger strong
feelings ofpleasure and arousalin consumers toward a
target product.
P5: Websites that employ the manipulation of information
content or informationgeneration are more likely to
enhance consumersperceived value associated with a
target product than websites that employ the manipu-lation ofinformationpresentation.
P6: Websites that employ the manipulation of information
presentation are more likely to trigger strong feelings of
pleasure and arousal in consumers toward a target
product than websites that employ the manipulation of
informationcontentor the manipulation ofinformation
generation.
Type of Purchase Situation as Moderator
The type of purchase situation interacts with different decep-tive information practices to influence consumers purchase
decision making. Prior research has dichotomized consumer
purchase intoplanned purchase and impulse purchase (e.g.,
Beatty and Ferrell 1998; Rook and Fisher 1995). Aplanned
purchase occurs when consumers have a specific purchase in
mind (Moe 2003), whereas an impulse purchase is a sudden
and immediate purchase with no pre-shopping intentions
(Beatty and Ferrell 1998). Research in impulse purchase has
shown that sensory stimuli in a retail environment may make
consumers less attentive to their purchasing behavior by
generating a sense of pleasure and arousal about the product
that overwhelms their self-control (LaRose 2001). Shoppingenvironments can be carefully crafted to facilitate impulse
purchases by undermining consumers self-control mech-
anisms (LaRose 2001). Moreover, positive emotions trig-
gered by sensory stimuli in the shopping environment may be
relied upon by consumers as a heuristic and thus lead to
behavior without active information processing (Forgas
1995). Positive emotions signal that the environment is favor-
able and, therefore, little cognitive attention is required
180 MIS Quarterly Vol. 35 No. 1/March 2011
-
8/3/2019 Ethics Research 1
13/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
(Cenfetelli 2006). Insomuch as the manipulation of informa-
tion presentation exploits media characteristics that empha-
size attractive product stimuli (e.g., enlarged photos, anima-
tions, and interactive graphics), such tactics may trigger
feelings of pleasure and arousal in consumers, which can
reduce their self-control and information processing, thus
leading to impulse purchases (LaRose 2001). In contrast, the
instrumental nature of the planned purchase suggests that
consumers are likely to engage in focused, deliberate,
thoughtful search and evaluation during the shopping process.
They will scrutinize the content of product information and
assess the merit of the products, thus attending less to
peripheral cues (e.g., image vividness, relative position in a
list of recommendation by PRAs) (Petty and Cacioppo 1986).
Since the manipulation of information content operates
directly on the content of product information, it is expected
to exert greater influence on a consumers cognitive
evaluation of the value of a product than the manipulations of
information presentation orinformation generation. It is thusproposed that
P7: Forimpulse purchases, websites that employ the manipu-
lation of information presentation are more likely to
trigger strong feelings of pleasure and arousal in
consumers toward a target product than websites that
employ the manipulation of information contentor the
manipulation of information generation.
P8: Forplanned purchase s, websites that employ the manip-
ulation of information contentare more likely to enhance
consumers perceived value associated with a target
product than websites that employ the manipulation of
information presentation or the manipulation of infor-
mation generation.
Product Type as Moderator
Product type can also influence a consumers vulnerability to
deception at an e-commerce website. Products are generally
categorized into search products and experience products.
Whereas search products (e.g., cameras, calculators, or
books) are characterized by attributes (e.g., color, size, price,
and components) that can be assessed based on the valuesattached to them without necessitating the need to experience
them directly, e xperience products (e.g., movie, music,
clothing, or cosmetics) are characterized by attributes (e.g.,
taste, smell, softness, and fit) that need to be experienced
prior to purchase. Because of the inherent difficulty asso-
ciated with the evaluation of experience products prior to
purchasing online, consumers tend to feel uncertain as to
whether the products would meet their expectations (Spieker-
mann 2001). When faced with difficulties in making choices,
consumers may resort to heuristics, resulting in decision
biases (Bettman et al. 1998; Payne et al. 1993). Thus, the
manipulation of information presentation and the manipu-
lation informationgeneration, which increase the visual
salience as well as the perceived relevance/importance of the
target products, will likely exert greater effect on consumers
cognitive evaluation of experience products than on their
evaluation ofsearch products.
In addition, King and Balasubramanian (1994) found that
product type had a significant impact on consumers reliance
on a particular decision-making process. Consumers eval-
uating asearch productwere more likely to use own-based
decision-making processes (i.e., to rely on themselves for
product search, evaluation, and purchase); in contrast, those
evaluating an experience producttended to use other-based
decision-making processes (i.e., to subcontract either part or
all of their decision-making process). Since prior research hasshown that consumers who relied on others for decision
making are likely to make purchasing decisions in keeping
with salespersons recommendations (Formisano et al. 1982),
Xiao and Benbasat (2007) conclude that consumers eval-
uating experience products are more likely to rely on
salespersons assistance and adopt their recommendations.
As the role of product recommendation agents in online
shopping is similar to that of salespersons in a traditional
shopping environment, consumers shopping forexperience
products are likely to be influenced by the recommendations
provided by PRAs and thus render themselves vulnerable to
the manipulation of information generation performed by
online merchants via the design of biased PRAs. We thus
propose
P9: When the target product is an experience product,
websites that employ manipulation of information
generation are more likely to enhance consumers per-
ceived value associated with the target product than those
employing manipulation of information presentation or
manipulation of information content.
Task Complexity as Moderator
An additional situational characteristic that moderates the link
from deceptive information practice to the affective and
cognitive mechanisms is task complexity, typically deter-
mined by the number of attributes on which each alternative
is compared (Jacoby 1977). As consumers tend to evaluate
products on several dimensions (e.g., attributes or features)
before making a choice, the greater the number of alternatives
(or dimensions), the more complex the decision task will be
MIS Quarterly Vol. 35 No. 1/March 2011 181
-
8/3/2019 Ethics Research 1
14/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
due to the increasing volume of relevant information that cus-
tomers must assess. When task complexity is low, consumers
are able to evaluate all the relevant information in detail prior
to making their final choice. Under such circumstances, the
manipulation of information contentwhich alters the con-
tent of product information to make the target products more
attractivewill likely exert greater influence on consumers
assessment of product value. In contrast, when task com-
plexity is high, due to the limitations in information pro-
cessing capacity, it is not feasible for consumers to evaluate
all relevant product information to determine the value of
each alternative. Instead, consumers are likely to attend to
and be influenced by peripheral cues in their decision-making
process (Petty and Cacioppo 1986). For instance, consumers
may choose a target product simply because it is presented at
the top of a websites bestselling list or is recommended as
the best-fitting product by a PRA at the website, without
processing available information thoroughly or searching for
additional information. As such, the manipulations of infor-mation presentation and information generationwhich
increase the visual salience and the perceived relevance/
importance of the target productswill have greater effect on
consumers assessment of the value of the target products and
hence their purchase decision making when the task com-
plexity is high. It is thus proposed that
P10: Websites that employ the manipulation of informa-
tion contentare more likely to enhance consumers
perceived value associated with a target product
when task complexity is low.
P11: Websites that employ the manipulation of informa-
tion presentation or the manipulation of information
generation are more likely to enhance consumers
perceived value associated with a target product
when task complexity is high.
In summary, while cognitive/affective mechanisms such as
perceived product value, pleasure, and arousalcontribute
positively to consumers approach behavior toward target
product(s), perceived deceptiveness of the e-commerce
websitehas a negative impact. Different deception techniques
exert differential impacts on the cognitive/affective mech-
anisms, with the manipulation of information content or
manipulation ofinformationgeneration having greater impact
on consumersperceived product value, and the manipulation
of informationpresentation on feelings ofpleasure and
arousal. Type of consumer purchase,product type, and task
complexity also interact with type of deception technique to
influence the cognitive/affective mechanisms.
Consumer Deception Detection
This section examines the effects of deceptive information
practices as well as individual (i.e., motivation, experience
with online/offline shopping in general, product expertise,
prior interaction(s) with the e-commerce website,prior trust
toward the e-commerce website,truth bias, and the receiving
of third-party information about the e-commerce website),
product (i.e., product type), and situational (i.e., type of
consumer purchase and consultation of alternative resources)
characteristics on the deception detection process and its
outcomes (i.e., hits ormisses).
Signal detection theory (Davies and Parasuraman 1982)
differentiates between two classes of eventsnoise (i.e., the
background) andsignal (i.e., stimulus that deviates from
background noise and thus may be detected)and explains
the performance of individuals who strive to determine the
presence of a signal. Signal detection theory specifies fourpossible outcomes in error detection tasks: (1) hits, (2) misses,
(3)false alarms, and (4) correct rejections. In the context of
deception detection, information that is free from manip-
ulation might be considered noise, whereas information that
is tampered by manipulation of its content, presentation,
and/or generation would provide a signal (Biros 1998). As
such, there are also four possible outcomes in a deception
detection task, which determine an individuals deception
detection performance:
1. A hitoccurs when consumers detect deception when it
does exist.
2. A miss occurs when consumers fail to detect deception
when it does exist.
3. Afalsealarm occurs when consumers report deception
when it does not exist.
4. A correctrejectionoccurs when consumers do not report
deception when it does not exist.
This paper focus on hits and misses as indicators ofdeception
detection success and deception detection failure, respec-
tively, since our theoretic model assumes the existence of
deception in the e-commerce website. Hits occur whenconsumers perceive the e-commerce website to be deceptive;
misses occur when consumers perceive the website to be free
from deceptive manipulations.
Prior research in deception detection (e.g., Johnson et al.
1993; Johnson et al. 2001; Morrison and Robinson 1997;
Robinson 1996; Robinson and Morrison 2000) suggests that
182 MIS Quarterly Vol. 35 No. 1/March 2011
-
8/3/2019 Ethics Research 1
15/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
individuals follow two sequential subprocesses in detecting
deception, namely, the noticing of anomaly and the attribution
of anomaly. To be successful in detecting deception in an
e-commerce website, consumers must be able to (1) notice the
anomalies as a result of the deceptive information practices,
and (2) attribute the noticed anomaly to deception by online
merchants. Thus, we propose
P12: Hits are more likely to occur when consumers are
able to notice product-related anomalies in the
e-commerce website and attributesuch anomalies to
deception.
The extent to which a consumer is able to notice and correctly
attribute anomalies (as a result of deceptive information
practices) in an e-commerce website is influenced by a variety
factors and processes, including the amount and type of
deception deployed in the website, the availability of decep-
tion detection support mechanisms, and various individualand situational characteristics. In the remainder of this sec-
tion, we will first examine factors influencing the likelihood
of consumers to notice and attribute anomalies resulting from
deceptive information practices. We will then explore the
effects of deception support mechanisms on deception detec-
tion performance.
The Noticing of Product-Related Anomaly: Individual
Characteristics, Amount/Type of Deception, and
Situational Characteristics
People hold preexisting expectations regarding the behaviorof those with whom they are interacting (Burgoon and
Walther 1990). These expectations may stem from social
norms (Burgoon et al. 1995; Jones 1986), prior interactions
(Burgoon et al. 1995; Honeycutt 1991; Jones 1986), informa-
tion from a third party (Levine et al. 2000), or stereotypes
(Burgoon and Le Poire 1993; Jones 1986). According to the
model of deception detection (Johnson et al. 1993; Johnson et
al. 2001), an anomaly is an inconsistency between what is
observed and what is expected. When an interactants
observed behavior sufficiently deviates from what is ex-
pected, an expectation violation is said to occur (Levine et al.
2000), raising suspicion, creating doubt, and ultimatelyleading to judgments of deceptiveness.
However, for the occurrence of expectation violation to be
noted, people must have situation-relevant expectations
regarding an interactants behavior in the first place. For
instance, in an online shopping context, consumers may have
expectations regarding the selection and prices of products
available at the e-commerce website; pre-sale and post-sale
services provided by the online company; website func-
tionalities facilitating product search, evaluation, choice, and
purchase; or other relevant issues. As discussed above, such
expectations may derive from a variety of bases, such as the
consumers general experience in shopping online and/or
offline, their prior interaction with this particular website, and
information about the website from their friends and relatives,
all of which help consumers create expectations that serve as
criteria against which the behavior of the e-commerce website
is evaluated and inconsistencies, if any, are noted. Thus, we
propose
P13
P15:
Consumers who are experienced in online/offline
shopping in general (P13), who have had prior
interaction(s) with the e-commerce website (P14),
and/or who have received third-party information
about the e-commerce website (P15) are more likely
to generate expectations that enable them to notice
product-related anomalies in the e-commercewebsite (thus resulting in hits).
As posited by the signal detection theory, increased attention
and effort as a result of increased motivation enable indi-
viduals to better distinguishsignalfrom noise in a deception
detection task (Boyle 2003; Gilovich 1991; Klein et al. 1997).
Similarly, individuals with greaterexpertise in a task domain
are better equipped to detect a wider set of signals from
background noise. In an online shopping context, consumers
motivation is manifest in their involvement with a particular
product category and/or their involvement with a particular
purchase decision. Consumers who are highly involved either
with a product or with a purchase decision are likely to
engage in thorough information search and comprehensive
information processing (Clarke and Belk 1978; Zaichkowsky
1985), and thus are more likely to identify deception cues
when they are present. Likewise, the moreproduct expertise
the consumers possess, the more likely for them to notice
anomalies in their interaction with the e-commerce website
during the online shopping process (e.g., Biros 1998; Johnson
et al. 2001) and to distinguishsignals from noise. It is thus
proposed
P16: Motivated consumers are more likely to notice
product-related anomalies in the e-commerce web-site (thus resulting in hits) than unmotivated con-
sumers.
P17: Consumers with a high level ofproduct expertise are
more likely to notice product-related anomalies in
the e-commerce website (thus resulting in hits) than
consumers with a low level ofproduct expertise.
MIS Quarterly Vol. 35 No. 1/March 2011 183
-
8/3/2019 Ethics Research 1
16/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
The interactants observed behavior must sufficiently deviate
from what is expected in order for the inconsistency to be
noticeable. This is in line with the signal detection theory,
which states that strong signals are more likely to be detected
than weak signals (Davies and Parasuraman 1982). There-
fore, in an e-commerce context, the more deceptive infor-
mation practices are employed by the online merchant, the
more likely for consumers to take notice of the anomalous
behavior. Also, although empirical research in detection
performance associated with different types of deception is
scarce, there exists some evidence suggesting the differential
detectability of different deception tactics. Burgoon and
Buller (1994; see also Burgoon et al. 1996) found that of the
three types of deception (i.e., equivocation, concealment, and
falsification), equivocation produced the greatest degree of
detection accuracy. The vague and indirect nature of equiv-
ocal messages makes it more noticeable than the other two
types of deception. Fatt (2001) also suggests that those who
wish to deceive others should engage in falsification ratherthan equivocation, since the latter is more likely to result in
detection. While Grazioli and Jarvenpaa (2003a) argue that
falsification is hard for online consumers to identify because
detection requires assessing the content of an offer to transact
via the Internet, we believe that the same applies to conceal-
ment. It is thus proposed that
P18: Consumers are more likely to notice product-related
anomalies in an e-commerce website that employs
more deceptive information practices than in one
that employs fewer.
P19: Consumers are more likely to notice product-related
anomalies as a result of equivocation than those
resulting fromfalsification orconcealment.
The type of purchase situation also influences consumers
performance in noticing anomalies. Aplanned purchase is
characterized by focused, deliberate, thoughtful search and
evaluation (Moe 2003); in contrast, an impulse purchase
occurs after experiencing a compelling urge to buy (Beatty
and Ferrell 1998). As noted by Kacen and Lee (2002), the
rapidity of an impulse purchase decision precludes thoughtful,
deliberate consideration of all information and choice alter-
natives. This lack of reflection often renders consumersvulnerable to unwanted strategic influences exerted by online
merchants via the implementation of various deceptive
information practices. It is thus proposed that
P20: Product-related anomalies in the e-commerce web-
site are less likely to be noticedby consumers (thus
resulting in misses) when they are engaged in
impulse purchases than inplanned purchase.
Whether a shopper consults alternative information sources
during online shopping (a situational factor) also influences
the likelihood that he will notice anomalies. Since all three
types of deceptive information practices work (albeit to
various extent) by influencing the availability and quality of
information presented to consumers, when alternative externalinformation resources (e.g., product manufacturers websites,
third-party expert review or consumer review websites such
as CNET.com and ConsumerReports.org, or independent
product recommendation agents, such as those designed by
Active Decisions) are available to consumers and when
consumers actually make use of such resources to verify the
information presented at an e-commerce website, they are
more likely to uncover inconsistencies and anomalies in the
website. It is thus proposed that
P21: When shopping at websites that exhibit deceptive
information practices, consumers who consult alter-
native information sources are more likely to noticeproduct-related anomalies in the e-commerce web-
site (thus resulting in hits) than consumers who do
not.
The Attribution of Product-Related Anomaly:
Individual Characteristics, Type of Deception,
Type of Deception Techniques
Individuals frequently try to make sense of the world around
them by making attributions about the cause of events/
behaviors (Tomlinson and Mayer 2009), particularly when theevents/behaviors are negative in valence. According to
causal attribution theory (Weiner 1986), the perception of a
negative outcome (that has occurred when the trustor per-
ceives disconfirming evidence regarding her expectations)
leads the individual to identify the causes of the outcome.
Similarly, the model of deception detection specifies that,
upon noticing an anomaly, individuals will often make
attributions of the anomaly by generating potential hypotheses
to explain it (Johnson et al. 2001; Koonce 1993). Since
human behavior is frequently amenable to different inter-
pretations (e.g., Borkenau 1986; Higgins 1989), the same act
is quite often considered from multiple perspectives
(Wojciszke 2005). As such, several alternatives (e.g., inno-
cent mistakes, incompetence) are often available to explain
the observed anomaly, with deception being one of the
alternatives. In an e-commerce context, factors such as
individual characteristics (e.g., trustand truth bias) as well as
type of deception technique influence the likelihood for
consumers to attribute the observed anomalies to deception by
online merchants.
184 MIS Quarterly Vol. 35 No. 1/March 2011
-
8/3/2019 Ethics Research 1
17/29
Xiao & Benbasat/Product-Related Deception in E-Commerce
Robinson (1996) defines trust as ones expectations, assump-
tions, or beliefs about the likelihood that anothers future
actions will be beneficial, favorable, or at least not detri-
mental to ones interests (p. 576). As a general positive
attitude toward another social entity, trust plays a significant
role in influencing ones interpretation of social behaviors
within a relationship (Robinson 1996). Two types of trust
relevant to the e-commerce deception context areprior trust
(a kind of knowledge-based trust) and calculative-based trust.
Individuals tend to maintain cognitive consistency by inter-
preting information in ways that reinforce their prior beliefs
and attitudes, while avoiding or ignoring interpretation that
disconfirms their prior beliefs (Fiske and Taylor 1984); this is
referred to asselective interpretation bias in judgment and
decision-making literature. In an online shopping context,
consumers with highprior trusttoward a website (probably
resulting from positive prior interaction with the website) will
expect the e-commerce website not to behave in a waydetrimental to their interests. Therefore, upon noticing an
anomaly, they will be less likely to attribute the anomaly to
deception and more apt to attribute it to extenuating circum-
stances (e.g., incompetence, honest oversight) (Robinson
1996). In contrast, consumers with lowprior trusttoward a
website will be more likely to interpret the identified
anomalies as cues of deception. This is done to reinforce and
maintain their prior theories and hypotheses about the
e-commerce website. It is thus proposed that
P22: Consumers with a high level ofprior trustare less
likely to attribute product-related anomalies noticed
in the e-commerce website to deception (thus
resulting in misses) than consumers with a low level
ofprior trust.
Calculative-based trustis shaped by the rational assessment
of t