Decision Support Systems 79 (2015) 171–183

Contents lists available at ScienceDirect

Decision Support Systems

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

A customer value, satisfaction, and loyalty perspective of mobile

application recommendations
Chenyan Xu a,⁎, Daniel Peak b,1, Victor Prybutok b,2
a
Computer Science and Information Systems Program, School of Business, Stockton University, 101 Vera King Farris Dr, Galloway, NJ 08205, United States
b
Department of Information Technology & Decision Sciences, College of Business, University of North Texas, 1155 Union Circle #311160, Denton, TX 76203-5017, United States

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

Article history: Although both anecdotal and academic sources have suggested that interpersonal recommendation is an impor-
Received 17 February 2013 tant means to promote mobile applications (apps), little to no research has investigated this topic. To fill the void,
Received in revised form 16 July 2015 we develop a research model based on the Customer Value, Satisfaction, and Loyalty Framework (VSL). While
Accepted 21 August 2015
prior VSL studies treated customer value as a black box and examined its effects as a whole, this research breaks
Available online 1 September 2015
down customer value to utilitarian benefits (consisting of app utility and app quality), hedonic benefits
Keywords:
(consisting of app aesthetics and enjoyment), monetary sacrifices and non-monetary sacrifices (consisting of
Mobile application technicality, knowledge of alternative quality and privacy risk). This research extends VSL to the mobile app con-
Customer value text and discerns factors that influence app recommendation. Based on the data analysis of 347 app users, we
Satisfaction found that intention to recommend is a strong predictor of app recommendation behavior. Satisfaction with
Recommendation apps, users' app continuance intention, and hedonic benefits obtained from using apps are the direct antecedents
of intention to recommend. Moreover, non-monetary sacrifices and utilitarian benefits have indirect effects on
intention to recommend through satisfaction and app continuance intention. Contrary to our hypotheses, we
did not find the significant effects of monetary sacrifices in the refined VSL. We also provide a detailed discussion
about these findings, and the theoretical and practical implications of the research.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction The importance of recommendation is well-established in commu-

The past decade has witnessed a phenomenal growth of smartphones,
as well as its killer application—mobile applications (apps). Buoyed by an
enormous user-base, mobile apps have captured the attention of business
[58]. In order for organizations to reap the benefits of information
technology (IT), they must understand user behaviors. Presumably, the
forgoing argument also applies to mobile apps, an emerging IT artifact.
Organizations' initiatives in apps cannot succeed without a deep under-
standing of individuals' app usage behaviors. To date, only a handful of in-
formation systems (IS) research directly address the topic of mobile app
usage [7,79]. Among them, the main focus is on examining users' inten-
tion regarding the initial acceptance and the continuance of apps. Given
that IT usage exists in multiple forms beyond simple adoption [10], we
believe that the current literature provides a limited explanation of app
users' behaviors. Thus, the first objective of this research is to examine
users' recommendation of mobile apps, a type of user behavior in the
post-adoption stage of IT that is not well addressed in prior research.
⁎ Corresponding author. Tel.: +1 609 626 5572.
E-mail addresses: chenyan.xu@stockton.edu (C. Xu), daniel.peak@unt.edu (D. Peak),
victor.prytutok@unt.edu (V. Prybutok).
1
Tel.: +1 940 369 7210.
2
Tel.: +1 940 565 4767.
nication, social psychology, consumer, and IS literature. According to
the Diffusion of Innovation Theory [62], most people rely on recommen-
dation as a reliable, credible information source for assurance to adopt
an innovation. In the social psychological discipline, numerous studies
showed that interpersonal influence plays an important role in various
contexts, such as consumption [37]. Consumer research has confirmed
that recommendation impacts customer purchase decisions [87] and
that recommendation is a critical success factor in the contexts where
the information giver and receiver are powerful customers [20]. With
the rise of the Internet, customers continue to use recommendation
from others to make purchase decisions in the online environment [9].
In the IS field, the Theory of Reasoned Action, the Theory of Planned Be-
haviors, the Model of PC Utilization, the Unified Theory of Acceptance
and Use of Technology (UTAUT) and UTAUT2 [e.g., 77,78] indicate that
individuals' IT adoption decisions are largely dependent on social
influence, the degree to which an individual perceives that others be-
lieve s/he should use the new technology. In the context of mobile
apps, recommendation is of particular importance. Apps are a form of
“experiential computing” technology [85], and people are more likely
to rely on recommendation when products/services are intangible and
experiential [57].
On the other hand, voices from the practical world also consolidate
the importance of recommendation in the mobile app context. Mobile

http://dx.doi.org/10.1016/j.dss.2015.08.008
0167-9236/© 2015 Elsevier B.V. All rights reserved.
172 C. Xu et al. / Decision Support Systems 79 (2015) 171–183
business consultants have viewed referral and recommendation as the
most effective way of promoting mobile apps [6,22,42]. Industrial re-
ports have provided further support for this proposition, revealing
that recommendation plays an important role in mobile app down-
loads: personal recommendation is the most important factor for 43%
of the surveyed people to make download decisions [56].
Thus, motivated by the gap between a dearth of research on mobile
app recommendation and the aforementioned theoretical and practical
justification regarding its importance, we believe it necessary, relevant
and timely to examine factors that influence app users' recommenda-
tion behaviors as a topic to organizations' benefits. To this end, we
draw upon the Customer Value, Satisfaction and Loyalty Framework
(VSL) [45,84]. According to VSL, customers who perceive they obtain
greater value from using a product/service tend to be more satisfied
with the product/service in question; both customer value and satisfac-
tion are significant antecedents of customer loyalty. Considering that
recommendation is an important facet of customer loyalty [45,84,86],
we believe that VSL is an appropriate theoretical lens to examine why
users make recommendation of apps.
Furthermore, we refine the original VSL using auxiliary theories
and empirical studies from prior IT adoption research. Specifically,
this work: 1) acknowledges the predominant conceptualization
that customer value is the customers' overall assessment of what is
received (benefit) versus what is given (sacrifice), 2) breaks down
the “black box” view of customer value used in prior VSL studies,
and 3) explores specific benefits and sacrifices associated with
using mobile apps. Therefore, the second and third objectives of
this research are to adapt VSL to the mobile app context and to inves-
tigate how the decomposed value components influence satisfaction
and recommendation.
By accomplishing the foregoing three research objectives, this re-
search makes several theoretical contributions. First, it fills the gap cre-
ated by the importance of mobile app recommendation and a scarcity of
research on this topic. Second, given that recommendation (or more
broadly speaking, word-of-mouth (WOM)) is essentially a type of infor-
mation exchange where messages flow from senders to receivers
through channels, by focusing on app users who will potentially initiate
recommendation, this research fills the gap that a handful of IS studies
investigated the “sender” elements in the WOM process. Third, given
that recommendation resides along the positive end of the WOM va-
lence continuum, this research fills an IS research gap identified by de
Matos and Rossi [20] and Harrison-Walker [34] that only a few WOM
studies examined WOM valence and this dimension of WOM was only
measured indirectly. Fourth, it echoes several research calls. By examin-
ing the recommendation topic, this research responds to Chin and
Marcolin's [10] call that IS research shift its attention to complex
“deep usage” of IT. By examining the mobile app context, this research
responds to Yoo's [85] call that it is time for IS researchers to focus on
experiential computing. Fifth, the refined VSL we developed to examine
recommendation in the mobile app context also bears valuable theoret-
ical implications. Identifying the specific benefits/sacrifices that shape
customer value and testing their effects within VSL, this research en-
hances VSL by decomposing the value construct that was unanimously
treated as a “black box” by past works. In addition, by considering sacri-
fices factors such as knowledge of alternative quality and privacy risks,
the refined VSL framework here offers insights into why satisfied
users are not loyal (e.g., not recommending apps), an increasingly criti-
cal issue in the marketing and IS domains.
The remainder of the research is organized as follows. The literature
review section details VSL and integrates it with prior IT adoption re-
search to propose a model of the app recommendation. The hypothesis
development section develops the hypotheses within the refined VSL
model. The methodology section describes the research method used
to empirically test the research model. The data analysis and results sec-
tion presents the results of data analysis. The discussion and implica-
tions section discusses the theoretical and practical implications of the
results in greater detail, lists the limitations of this study, and suggests
directions for future research.
2. Literature review
In this section, we first present an overview of VSL and explain the
key constructs (i.e., value and satisfaction). Then, we introduce the use
of VSL in the IS field and summarize three major issues underlying
prior research that employed VSL. Next, we discuss the key constructs
in our research model, which expands and builds upon the original VSL.
2.1. The customer value, satisfaction and loyalty framework
Originating in the marketing discipline, VSL has been developed and
validated by a plethora of literature in the strategic marketing, service
management, and consumer behavior areas that examined the interre-
lationships among customer value, satisfaction and customer loyalty in
both business-to-business and business-to-consumer contexts [e.g., 45].
VSL holds that customer value and satisfaction are two primary ante-
cedents of customer loyalty while customer value positively influences
satisfaction.
Customer value is defined as “consumer's overall assessment of the
utility of a product based on perceptions of what is received and what
is given” [89]. Therefore, the formation of customer value involves a
trade-off process where customers evaluate the benefits received and
the sacrifices given from using a product/service. Early marketing stud-
ies focused on product quality as the primary “get” component and
product price as the main “give” component [e.g., 72]. Later, Sinha and
DeSarbo [64] pointed out that customer value is a complex and multi-
dimensional construct and called for exploring other factors that
shape customer value. That is to say, there are various benefits/sacrifices
beyond product quality/price from using a product, for example, shop-
ping experiences and product risks [8]. Another important characteristic
of customer value is that its benefit/sacrifice components vary with new
consumption contexts [8]. Therefore, prior IS literature that took the
value perspective to examine IT adoption was busy with identifying
new compositions of customer value and exploring their effects in dif-
ferent IT contexts.
Another important factor in VSL is satisfaction—the key to cultivate
and retain loyal customers [2]. Satisfaction has taken a central,
predominating position in marketing theory and practice because, as a
major outcome of marketing activities, it transforms the initial con-
sumption and purchase to post-purchase phenomena such as attitude
change, re-patronage and brand loyalty [15]. The American Customer
Satisfaction Index is an important customer-based measurement sys-
tem that assesses performance of firms, industries and national econo-
mies [15].
2.2. The customer value, satisfaction and loyalty framework in the
information systems field
IS researchers have drawn upon VSL to examine user behavior in the
post-adoption stage of a variety of IT contexts such as SNS services [52],
China's telecommunication market [80], organizational IT services [38],
mobile value-added services [43], and mobile-commerce [48]. This re-
search supplements this stream of research by addressing three issues
related to employing VSL.
First, despite its multi-dimensionality, VSL research treated per-
ceived value mainly as a black box: little to no VSL research identifies
specific benefits/sacrifices that shape customer value and tests the influ-
ences on satisfaction and customer loyalty. This is surprising because IS
researchers commonly decompose customer value in using this con-
struct, leading to a series of value-based technology adoption models,
for example, the Value-based Adoption Model of Technology [e.g., 40].
Customer value is context-based, so a new environment, like mobile
apps, may change what constitutes the value construct. Therefore, this
 C. Xu et al. / Decision Support Systems 79 (2015) 171–183
research not only builds on VSL, but also: 1) decomposes customer
value, 2) explores specific benefits and sacrifices related to app usage,
and 3) examines how the benefits and sacrifices influence other key fac-
tors in our refined VSL.
Second, even for the limited VSL and IS research [e.g., 52,73,74] that
decomposed customer value, most inquiries focus on the benefit compo-
nent of value, leaving the sacrifice component less examined. In fact, the
ignorance of sacrifice is a known historical issue in VSL studies [45].
Echoing Lam et al. [45], this research refines VSL by identifying primary
sacrifice factors and examining the effects of the sacrifice component in
VSL.
Third, to enhance its explanatory power, prior research supplement-
ed VSL with addendum variables (e.g., service quality and trust). Al-
though this research provided sound justifications for adding the
addendum variables to VSL, one may still question not adding other var-
iables such as information quality and cognitive absorption. Referring to
several versions of UTAUT that streamlined factors influencing technol-
ogy acceptance [77,78], we believe that it is important to introduce a
systematic approach that justifies categorizing similar addendum vari-
ables into the same group so that we can investigate the effect of each
group in VSL as Venkatesh et al. [77,78] did in developing UTAUT and
UTAUT 2. Motivated by the research [14,40] that holds a value-based
view of IT adoption, this research borrows the utilitarian-and-hedonic
classification of IT benefits and monetary and non-monetary classifica-
tion of IT sacrifices to synthesize the addendum variables of VSL.
2.3. Constructs in the refined customer value, satisfaction and loyalty
framework
Building upon early discussion regarding the necessity of refining
VSL, here, we present an overview of the constructs that were used in
this research to further VSL and discuss the details of these constructs.
2.3.1. Utilitarian and hedonic benefits of perceived value in mobile apps
IS researchers have long explored the utilitarian and hedonic mo-
tivations (also called extrinsic and intrinsic motivations) to use IT,
and compared their relative strengths across different types of IS
[19,78]. This paper defines utilitarian benefits of mobile apps as
functional benefits delivered by the performance of the apps that is
instrumental in achieving valued outcomes distinct from the
usage itself, and defines hedonic benefits of mobile apps as non-
functional benefits delivered by the performance of the apps for no
apparent reinforcement other than the process of using the apps to
obtain fun and self-fulfillment. Similar to prior works [49], our iden-
tification of utilitarian/hedonic factors is based on literature review
and interviews with mobile app users. The identified utilitarian
benefits are app utility and app quality, and the hedonic benefits are
enjoyment and app aesthetics.
2.3.1.1. App utility. App utility is the extent to which a person believes
that using an app will deliver desired, superior outcomes; we derive
app utility from system utility. Reflecting the utilitarian benefits of
using IS, system utility has been represented by perceived usefulness
in the Technology Acceptance Model (TAM) [18] and performance ex-
pectancy in UTAUT and UTAUT2 [77,78]. With the extensive application
of TAM, UTAUT and UTAUT2 to examine IT adoption in numerous vol-
untary and non-voluntary settings, research shows system utility to be
a salient predictor of users' behavioral intention. Furthermore, it is the
sole factor that describes utilitarian benefits in the Motivation Model
of IT Adoption [19] and the Value-based Adoption Model of Technology
[40]. Hence, we believe app utility is a representative factor of the utili-
tarian benefits derived from using apps and enjoys the potential to ex-
plain user behavior in most apps. In the Mobile Internet context,
utility is an important value component [40]. The more useful a person
perceives the Mobile Internet, the greater value he/she receives. While
people employ utilitarian mobile apps for utility such as information-
173
seeking, file-synchronization and music-identification, our focus group
discussion revealed that utility is also the major concern of most hedon-
ic app users. This is consistent with previous findings that perceived
usefulness also plays a critical role in users' intention to play online
games and mobile games [32,35]. Specifically speaking, people play
gaming apps for entertainment, relaxation, escape from reality, and
other reasons; a gaming app becomes a favorite when it serves their
needs best compared to other alternatives, which addresses the utility
of hedonic apps.
2.3.1.2. App quality. Consistent with Wixom and Todd's [82] concep-
tualization of system quality, in this research, app quality reflects
the quality of app features such as reliability, responsiveness and in-
tegration. Wixom and Todd [82] incorporated system quality into
their IT adoption model as a supplement to system utility (equiva-
lent to perceived usefulness advocated in TAM) because system
quality is an object-based belief. While system utility reflects a
behavior-based belief about the consequences of using a system,
object-based beliefs such as system quality that describe the system
itself also matters to user behaviors. Given the extensive use of the
Wixom and Todd's Model in various IT contexts, we believe the in-
clusion of app quality as another dimension of utilitarian benefits
can further predict user behavior in mobile apps. Previous research
found app quality to be an important factor influencing users' post-
adoption intention of mobile apps [7]. Per our focus-group inter-
views, we observed individuals are concerned with app quality is-
sues such as whether an app can run smoothly on the smartphone
platform, whether it quickly responds to users' inputs, and whether
it can deliver its utility in a reliable manner.
2.3.1.3. Enjoyment. Enjoyment is the extent to which using an app is en-
joyable in its own right, apart from any anticipated performance conse-
quences. In various settings, IS research indicates that enjoyment
influences technology acceptance and use intention directly [e.g., 76],
including the use of utilitarian systems [28]. It was treated as the most
salient factor to represent hedonic benefits in UTAUT 2 [78] and the Mo-
tivation Model of IT Adoption [19]. Given that enjoyment has a direct ef-
fect on user attitudes toward mobile games [32], we believe its effects
will be similar for hedonic apps. Prior studies [7] revealed that people
experience enjoyment when they use such utilitarian apps that deliver
information in a timely fashion. Thus, we also believe that enjoyment
is a critical value component, despite the utilitarian/hedonic nature of
apps. Our focus-group discussion provides insights. For example, one
member mentioned that she was put into a good mood by using
Evernote (a utilitarian app that helps users take notes with online syn-
chronization and backup provided) because it reminded her just how
technologically-savvy she is.
2.3.1.4. App aesthetics. Aesthetics is an emerging guideline in IS de-
sign [16]. Akin to system quality, aesthetics is an object-based belief
about the visual appearance of a system. Recent IS research shows
that system aesthetics plays a determinant role in influencing IT
adoption in many settings [16,50]. The effect of system aesthetics
on IT user behavior can be explained by many theories such as the
Signaling Theory [81], the Visual Rhetoric Theory [16] and the “beau-
ty is good” stereotype [71]. While prior research recognizes the influ-
ence of aesthetics on conventional IT products [e.g., 71], aesthetics
also affects user appraisals of mobile games and websites displayed
on mobile phones [17,32]. Anecdotal sources [e.g., 68] viewed app
aesthetics as the key in helping a mobile app stand out in a crowded
market and determining its success; our focus group interview sup-
ports this perspective. As one respondent commented, how the
snapshot of an app looks really matters, because it determines
whether or not she will download it, especially when she has little
knowledge of the app. Moreover, others agreed that an appealing in-
terface brings about a wonderful user experience.
174 C. Xu et al. / Decision Support Systems 79 (2015) 171–183
2.3.2. Monetary and non-monetary sacrifices of perceived value in mobile
apps
Prior research suggested that sacrifices that decrease customers'
value perceptions are both monetary and non-monetary [49]. The
term monetary sacrifices refers to what one has to give up financially
to obtain a product/service [14] whereas non-monetary sacrifices usual-
ly include time, physical effort, or mental effort needed to acquire and
use a product [40]. We find both monetary and non-monetary sacrifices
relevant to this research because in the mobile context, individuals
bear the dual role of consumer and technology user. As a consumer, mo-
bile app users pay at iTunes or Google Pay to download paid apps
(e.g., Monopoly). The effects of non-monetary sacrifices are also pro-
nounced. For example, as technology users, individuals need to work
at learning to gain familiarity with the interface and functions of apps
in a manner similar to using other IT products. Therefore, we incorpo-
rated both monetary and non-monetary sacrifices in the refined VSL.
We used perceived price to measure monetary sacrifices, a construct
that was unanimously adopted in previous research [14,40] to
operationalize monetary sacrifices.3 Using the same procedures to
obtain utilitarian and hedonic benefits, we identified three non-
monetary sacrifices, namely privacy risk, knowledge of alternative
quality and technicality, and added them to the research model.
2.3.2.1. Perceived price. Price is an important financial cost required to
obtain and use a product. Perceived price is the result of encoding a
product's price by customers in a way that is meaningful to them [89].
A large body of marketing literature shows that perceived price influ-
ences perceived value. If a price is perceived as unacceptable, customers
will then assess the product with little or no value [92]. The recent de-
cade has witnessed the rise and popularity of IT products that require
payment and that are used voluntarily (known as consumer technolo-
gies). Against this backdrop, an increased number of IS research
[e.g., 78] posited that in the consumer technology context, IT users are
also consumers because they bear the monetary costs associated with
using the technologies. This stream of research [14,49] examined the
role of perceived price and found it significant in predicting IT adoption
in various settings. Particularly, perceived price (or its variant) was
shown to be a critical construct in the Value-based Adoption Model of
Technology [40] and UTAUT 2 [78] that compliments conventional fac-
tors such as perceived usefulness and perceived ease of use in under-
standing IS success. Also, the price of using paid mobile apps includes
the download fee and charges associated with value-added features.
Therefore, we believe it is necessary to incorporate perceived price
into the refined VSL.
2.3.2.2. Privacy risk. Privacy risk refers to the app developers' malicious
behavior concerning the collection and use of an app user's individual
information. The inclusion of privacy risk as a sacrifice component has
its roots in conventional consumer literature [70] that conceptualized
product price and product-related risk as two common sacrifice compo-
nents that decrease customer value: while the effect of product price oc-
curs at the time of purchase, product-related risk has a longer-term
impact during the ownership of a product. Thus, the consideration of
product-related risk is advocated in assessing customer value at the
post-consumption stage [70]. The use of technologies is not free of
risks. In fact, privacy risk as IT users' largest concern is of a growing in-
terest to IS researchers that have spawned numerous studies that ex-
plored its antecedents and outcomes [66]. This is because online
companies generally tend to behave opportunistically with users' per-
sonal information to realize inappropriate gains [67].
While privacy risk has long been a critical issue in mobile commerce
[29], the advent of mobile apps offers seemingly unbounded options for
collecting, processing, distributing, and using private information of app
3
The remainder of the paper will use the terms “perceived price” and “monetary sacri-
fices” alternatively.
users, which renders privacy risk a sacrifice component that merits our
immediate attention. In mobile app usage, one problem area is the gen-
eralized use of Facebook verification. For example, a mobile app may re-
quire users to provide verification to register for an offered service, but if
users sign in through their Facebook accounts, these apps gain access to
users' Facebook information. Apps that offer more personalized services
may require more detailed private information. Apps may track individ-
uals' usage behaviors, and with GPS smartphones, record usage and lo-
cation data. In an attempt to mitigate potential privacy risks, California's
Online Privacy Protection Act has extended its footprint to the mobile
app context [63]. Specifically, the Act requires app developers to include
privacy policies in their apps so that users will be informed about the
data that apps will access, use, and share before they download the
apps. Despite these precautions there is simply no guarantee that app
developers will not invade users' privacy. All these facts underscore
that privacy risk continues to be an important source of sacrifices asso-
ciated with app usage.
2.3.2.3. Knowledge of alternative quality. This research defines knowledge
of alternative quality as the extent to which a person is aware that an
alternative app, when compared to the primary app, is able to fulfill
his/her needs. The concept is grounded in the relationship perspective
of marketing [55] and the Commitment and Trust Model of IT Continu-
ance [46]. Both theories posit that intention of relationship continuance
with a company, post-consumption of a product, or IT continuance is
also dependent on alternative quality. In the IT context, Li et al. [46] in-
dicated that although users may be satisfied with a particular website,
they may incline to switch to an alternative site as long as they perceive
that an alternative website works better. The role of alternative quality
is also salient in the experiential computing context. Under the para-
digm of the value perspective, Lin et al. [49] conceptualized knowledge
of alternatives as a vital sacrifice that reduces customer value of using
experiential technology such as IPTV. Due to the competitive app mar-
ket where the core function of an app can be easily achieved by others,
consistent with Lin et al.'s work, we included knowledge of alternative
quality to capture another dimension of sacrifice when people use mo-
bile apps (another form of experiential technology).
2.3.2.4. Technicality. This study defines technicality as the extent to
which a mobile app is perceived as being technically-difficult to use. Ac-
cording to the Value-based Adoption Model of Technology [40], techni-
cality is a primary non-monetary sacrifice in using mobile-related
services. In the same vein, we conceptualized technicality as a non-
monetary sacrifice. In the context of mobile apps, high technicality can
be attributable to a wide range of factors. For example, mobile app
users rely on touch interfaces (fingers) rather than standard interfaces
(mouse) to type or operate mobile apps; they are required to change
their interfacing habits to adapt to touch screen design. This new IT con-
text will foreseeably produce significant psychological discomfort
among users such as frustration, anxiety and mental fatigue [40]. Thus,
technicality represents a significant factor in non-monetary sacrifices.
2.3.3. Intention to recommend, app continuance intention and
recommendation
This research draws a clear distinction between recommendation
and WOM activities. WOM has two dimensions [20]. The first dimension
is termed as “WOM activities” which addresses the frequency of the
WOM communications, the number of people influenced, and the infor-
mation supplied by the WOM initiators. The second dimension is
termed as “WOM praise” (or “WOM valence”) which reflects the va-
lence of the WOM communications that could be positive, neutral and
negative. Thus, recommendation differs from WOM activities in that it
represents the positive valence of WOM whereas WOM activities
might involve disseminating negative information of a product. Consis-
tent with prior VSL and WOM works [e.g., 45,84], this research
 C. Xu et al. / Decision Support Systems 79 (2015) 171–183
conceptualized recommendation as the extent to which one says posi-
tive things about an app and encourages others to use an app.
This research examined both intention to recommend and the actual
recommendation behavior in the refined VSL. Although the intention-
behavior relationship is strong and robust in the IT context, we believe
it necessary and valuable to examine the actual actions of mobile app
users for two reasons. First, recommendation behavior is easy to mea-
sure in the context of this research. A review of the IS research that
employed use intention as a proxy for usage behavior shows that
most authors recognized the limitation and they did so because actual
usage is difficult to assess [e.g., 93]. However, the assessment issue
does not pose a challenge in this research because the participants in
this study have already adopted the apps which allows us to measure
the recommendation behaviors that occurred. Second, usage behavior
lacks scrutiny in the IS field [94] and particularly, recommendation as
a form of “deep usage” behaviors of IT receives little attention [10];
therefore, our approach could expand the discipline's understanding
in this regard.
This research included app continuance intention in the refined VSL
to be another antecedent of intention to recommend. This construct re-
flects individuals' willingness to continue to use an app. Past marketing
research found that WOM is closely associated with customer loyalty
[20] which is characterized by repatronage [65,88]. On the other hand,
IS research [4] posited that IS continuance (intention) is in the same
spirit of repatronage and it is a sign of loyalty in the IT context. Taken to-
gether, it is reasonable to expect that app continuance intention might
affect intention to recommend. This is particularly true when recent IS
literature found that intentions to adopt biometric technologies in-
creases users' intention to recommend the technologies to their friends
[54]. The relationship is important to explore to provide better under-
standing of the outcome variable.
3. Hypothesis development
Fig. 1 summarizes our research model. In this research, we identified
value components that are specific to the mobile app context and
Note:
1. PR (Privacy Risk), Knowledge of Alternative Quality (KAQ), App Continuance Intention (ACI), Intention to Recommend (IR)
Fig. 1. Proposed research model.
175
incorporated app continuance intention. In addition, we refined the
original VSL in two respects. First, we conceptualize utilitarian benefits,
hedonic benefits, and non-monetary sacrifices as second-order con-
structs. We pursue this methodology for three reasons.
1) One of the primary interests of this research is to break down the
“black box” view of customer value used in previous VSL studies
and examine the individual effects of the value components
(i.e., utilitarian benefits, hedonic benefits and non-monetary sacri-
fices) on satisfaction, app continuance intention and intention to
recommend. Accordingly, conceptualizing various dimensions of
utilitarian benefits, hedonic benefits and non-monetary sacrifices
as a whole respectively fits the purpose of this research. If we draw
inferences at higher levels based on the analysis of a decomposed
model, it will result in atomistic fallacy [61]. For example, in a
decomposed model, while privacy risk impacts intention to recom-
mend and knowledge of alternative quality does not, it would be ar-
bitrary to say non-monetary sacrifices matters (or does not matter)
to intention to recommend. As Petter et al. [61] posited, “Relation-
ships that hold, or do not hold at lower units of measurement and
aggregation, may not hold at the higher levels”.
2) For these three value components, the decomposed dimensions
(e.g., app utility and app quality) have the same relationships with
other constructs. Under this circumstance, it is suggested to develop
a second-order construct with first-order factors as its formative in-
dicators and then to formulate research hypotheses at the higher-
order factor level rather than at the individual sub-construct level
[36]. The advantage of this approach is to produce a parsimonious
model that provides abstractions that generate insightful explana-
tions about complex phenomena [61]. If we were to decompose util-
itarian benefits, hedonic benefits and non-monetary sacrifices and
then test the effects of each dimension, we would have 12 extra
paths. While it is not the intent of this research to know the effects
of the decomposed value components of mobile apps at the cost of
creating a complex model, the higher-order approach offers a viable
solution to this paradox.
176 C. Xu et al. / Decision Support Systems 79 (2015) 171–183
3) We recognize that multiple approaches exist to approximate
second-order constructs [11]. For example, the repeated indicator
approach measures second-order constructs directly with items of
all its first-order constructs [51]. But this approach is most effective
when all the first-order factors have equal number of indicators.
The second approach models the paths between first-order and
second-order constructs [23]. This research takes the latter approach
because the two dimensions of utilitarian benefits, two dimensions
of hedonic benefits and three dimensions of non-monetary sacrifices
have unequal number of indicators.
Second, we propose that user satisfaction, app continuance inten-
tion, and intention to recommend are directly influenced by utilitarian
benefits, hedonic benefits, and non-monetary sacrifice without testing
the mediation of customer value. According to Liang and Xue [47], the
exclusion of a mediator can precisely delineate the direct effects of the
mediated factors which allows for a rich understanding of the mediated
factors. Decomposing the black box of customer value and investigating
the individual effects of utilitarian benefits, hedonic benefits, and non-
monetary sacrifices shed light on how app developers can better man-
age user satisfaction, maintain continued use and foster users to recom-
mend apps. Next, this section presents empirical evidence that supports
the relationships in our research model.
3.1. Satisfaction to app continuance intention and intention to recommend
According to the Expectation Confirmation Model [4], satisfaction
positively influences IS continuance. Such effect was confirmed in a
wide range of IT environments [12,90,91]. On the other hand, VSL
studies found that satisfaction positively influences customers/users'
intention to recommend products/services [24,45,84]. Thus, we hypoth-
esize that:
H1. Satisfaction positively influences users' app continuance intention.
H2. Satisfaction positively influences users' intention to recommend
mobile apps.
3.2. Utilitarian benefits, hedonic benefits, perceived price and non-
monetary sacrifices to app continuance intention and intention to
recommend
Lin et al. [49] found that customer value is a function of benefits and
sacrifices (including both monetary and non-monetary sacrifices) per-
ceived by users and customer value is a predictor of continuance inten-
tion of using IPTV. Similarly, Kim et al. [40] found that benefits such as
usefulness and enjoyment and sacrifices such as technicality and per-
ceived price are determinants of customer value, which in turn affects
individuals' intention to adopt mobile Internet. In addition, VSL works
in the IT context [48,52] indicated that customer value influences
repatronage intention/continued use intention and intention to recom-
mend. Given the definition that customer value is consumers' overall as-
sessment of the utility of a product based on the received benefits and
given sacrifices [89], perceived benefits and sacrifices capture different
facets of customer value and each should affect app continuance inten-
tion and intention to recommend. Thus, we hypothesize:
H3. Non-monetary sacrifices negatively influence users' app continu-
ance intention.
H4. Non-monetary sacrifices negatively influence users' intention to
recommend mobile apps.
H5. Perceived price negatively influence users' app continuance
intention.
H6. Perceived price negatively influence users' app intention to recom-
mend mobile apps.
H7. Utilitarian benefits positively influence users' app continuance
intention.
H8. Utilitarian benefits positively influence users' intention to
recommend mobile apps.
H9. Hedonic benefits positively influence users' app continuance
intention.
H10. Hedonic benefits positively influence users' intention to
recommend mobile apps.
Per VSL, customer value impacts satisfaction, an association that has
received substantial empirical support [e.g., 12]. Based on the forgoing
argument that utilitarian benefits, hedonic benefits, perceived price
and non-monetary sacrifices are significant components of customer
value, we hypothesize:
H11. Non-monetary sacrifices negatively influence satisfaction.
H12. Perceived price negatively influences satisfaction.
H13. Utilitarian benefits positively influence satisfaction.
H14. Hedonic benefits positively influence satisfaction.
As noted earlier, motivated by de Matos and Rossi's [20] meta-
analysis of prior WOM literature, this research explores the effect of
app continuance intention on intention to recommend. This approach
is consistent with recent IT adoption literature that began to use
intention to recommend as outcome variable and examined how
users' intention to adopt an IT product influences their intention to
recommend the technology [e.g., 54]. Thus, we hypothesize:
H15. App continuance intention positively influences intention to
recommend.
The Theory of Reasoned Action and the Theory of Planned Behavior
support the contention that behavioral intention is highly correlated
with and is an immediate and strong predictor of actual behavior [1,3,
25]. The intention-behavior relationship was confirmed in numerous
IS research studies [e.g., 77]. Thus, we hypothesize:
H16. Intention to recommend positively influences app users'
recommendation behaviors.
4. Methodology
4.1. Measurement development
We developed all measurements based on their theoretical under-
pinning and previous relevant literature. The two utilitarian benefits
dimensions–app utility and app quality–were contextualized from
Davis's [18] and Wixom and Todd's [82] works. On the other hand,
app aesthetics was developed based on previous research [17,41,75]
that examined the influence of visual aesthetics in contexts relevant to
mobile apps; the questions for enjoyment, the other dimension of he-
donic benefits, were drawn from van der Heijden's [76] and Chen
et al.'s [11] works. We measured perceived price and technicality with
a contextualized version of the questions used by Kim et al.'s [40] and
Lin et al.'s [49] studies. For the other two dimensions of non-monetary
sacrifices, we derived measurement for knowledge of alternative quali-
ty from Li et al.'s work [46], and privacy risk from Dinev and Hart's [21]
and Kim et al.'s [39] works. We measured satisfaction with the items
adapted from Bhattacherjee's [4] research, app continuance intention
from Chen et al.'s [7] research, intention to recommend from Turel
et al.'s [73] research and recommendation from Harrison-Walker's
[34] research. We evaluated all the questions by a 5-point Likert scale.
To further assess the content validity of all items, we invited experts
 C. Xu et al. / Decision Support Systems 79 (2015) 171–183 177

in relevant areas to examine the items. We revised the questionnaire
based on the comments received.
4.2. Data collection
A total of 347 respondents participated in the study and successfully
completed the questionnaires. The respondents were undergraduate
and graduate students at a major public state university in the south-
western U.S. We administered the questionnaires in 10 business classes
with the permission of the instructors. Participation in this research was
voluntary. Extra credit was given as an incentive.
5. Data analysis and results
We used Partial Least Squares (PLS) with WarpPLS 5.0 to validate the
measurements and tested the proposed hypotheses.
5.1. Measurement validation
Construct reliability was measured by Cronbach's alpha and com-
posite reliability. According to Table 1, the Cronbach's alpha values of
all constructs ranged from 0.81 to 0.95, higher than the minimum cutoff
score of 0.70 suggested by Hair et al. [33]; the lowest composite reliabil-
ity value is 0.88, which satisfies the benchmark of 0.70 [27]. In addition,
all of the average variance extracted (AVE) values are higher than the
threshold of 0.50. According to Table 2, the square roots of all AVEs
are greater than the off-diagonal elements, which demonstrate the dis-
criminant validity of all constructs.
To test for common method variance (CMV), we employed
Harman's one-factor test, which revealed ten factors with the first factor
accounting for 30.46% of the variance. We thus concluded that CMV was
not a concern.
To test the appropriateness of conceptualizing a construct as second-
order, researchers [49,60] have suggested that moderate rather than

Table 1
Descriptive statistics of reliability indices and factor loadings for constructs.

Item Factor analysis Cronbach's alpha Composite reliability AVE

U AE PE KAQ PP T PR Sat RI R ACI

U1 0.77 0.34 0.26 −0.10 0.15 −0.15 −0.01 0.39 0.22 0.22 0.38 0.87 0.90 0.61
U2 0.81 0.21 0.10 −0.14 0.13 −0.08 −0.11 0.30 0.22 0.17 0.36
U3 0.71 0.10 −0.01 −0.02 0.09 0.02 −0.11 0.14 0.13 0.19 0.13
U4 0.85 0.18 0.05 −0.08 0.16 −0.05 −0.05 0.23 0.20 0.24 0.24
U5 0.83 0.29 0.19 −0.04 0.20 −0.03 −0.02 0.33 0.23 0.25 0.33
U6 0.72 0.36 0.28 −0.21 0.11 −0.21 −0.14 0.42 0.25 0.23 0.35
AE1 0.31 0.89 0.48 −0.16 0.20 −0.21 −0.05 0.46 0.27 0.24 0.40 0.93 0.95 0.79
AE2 0.28 0.91 0.48 −0.14 0.26 −0.16 −0.02 0.44 0.33 0.34 0.37
AE3 0.28 0.91 0.48 −0.14 0.24 −0.19 −0.08 0.42 0.34 0.33 0.37
AE4 0.29 0.91 0.46 −0.15 0.23 −0.16 −0.07 0.47 0.34 0.30 0.39
AE5 0.25 0.83 0.41 −0.07 0.14 −0.15 −0.06 0.42 0.33 0.37 0.33
PE1 0.17 0.46 0.85 −0.17 0.20 −0.30 −0.17 0.59 0.42 0.33 0.57 0.89 0.93 0.76
PE2 0.23 0.50 0.89 −0.13 0.18 −0.26 −0.21 0.64 0.44 0.40 0.50
PE4 0.05 0.39 0.85 −0.15 0.11 −0.20 −0.18 0.56 0.25 0.32 0.38
PE5 0.18 0.46 0.89 −0.12 0.15 −0.25 −0.17 0.65 0.38 0.37 0.52
KAQ1 −0.08 −0.02 −0.03 0.79 −0.06 0.30 0.18 −0.02 −0.01 0.03 −0.07 0.91 0.93 0.73
KAQ2 −0.10 −0.16 −0.18 0.87 −0.16 0.49 0.26 −0.20 −0.11 −0.06 −0.23
KAQ3 −0.10 −0.15 −0.17 0.88 −0.18 0.41 0.21 −0.13 −0.08 −0.02 −0.16
KAQ4 −0.14 −0.17 −0.14 0.89 −0.17 0.43 0.19 −0.18 −0.11 −0.03 −0.21
KAQ5 −0.11 −0.13 −0.16 0.84 −0.07 0.41 0.22 −0.16 −0.14 −0.05 −0.23
PP1 0.17 0.20 0.16 −0.13 0.95 −0.02 −0.05 0.19 0.16 0.12 0.17 0.94 0.96 0.89
PP2 0.18 0.24 0.18 −0.14 0.97 −0.04 −0.05 0.17 0.19 0.15 0.18
PP4 0.15 0.24 0.18 −0.16 0.91 −0.09 −0.06 0.24 0.20 0.17 0.25
T1 −0.09 −0.19 −0.26 0.46 −0.03 0.91 0.33 −0.24 −0.13 −0.01 −0.29 0.92 0.94 0.81
T2 −0.13 −0.23 −0.31 0.42 −0.11 0.89 0.32 −0.29 −0.14 −0.04 −0.29
T3 −0.08 −0.11 −0.21 0.39 −0.01 0.88 0.34 −0.21 −0.08 0.10 −0.25
T4 −0.08 −0.16 −0.28 0.46 −0.03 0.91 0.35 −0.25 −0.14 0.04 −0.29
PR1 −0.08 −0.04 −0.17 0.23 −0.03 0.31 0.87 −0.15 −0.16 −0.11 −0.17 0.94 0.96 0.78
PR2 −0.14 −0.06 −0.19 0.22 −0.08 0.32 0.87 −0.19 −0.21 −0.13 −0.20
PR3 −0.04 0.01 −0.11 0.17 −0.01 0.28 0.84 −0.10 −0.13 −0.08 −0.10
PR4 −0.07 −0.08 −0.19 0.24 −0.06 0.36 0.90 −0.18 −0.19 −0.09 −0.19
PR5 −0.06 −0.08 −0.22 0.23 −0.06 0.35 0.92 −0.20 −0.25 −0.14 −0.22
PR6 −0.10 −0.09 −0.22 0.23 −0.04 0.35 0.90 −0.22 −0.27 −0.13 −0.25
S1 0.37 0.46 0.64 −0.19 0.20 −0.27 −0.20 0.92 0.48 0.40 0.70 0.91 0.93 0.78
S2 0.36 0.48 0.65 −0.18 0.22 −0.27 −0.19 0.93 0.48 0.41 0.67
S3 0.30 0.33 0.53 −0.06 0.12 −0.18 −0.16 0.79 0.32 0.35 0.47
S4 0.33 0.48 0.65 −0.15 0.20 −0.25 −0.15 0.89 0.47 0.45 0.56
R1 0.28 0.37 0.44 −0.12 0.23 −0.18 −0.21 0.55 0.90 0.60 0.62 0.92 0.95 0.82
R2 0.24 0.32 0.36 −0.10 0.18 −0.12 −0.18 0.43 0.94 0.64 0.50
R3 0.22 0.32 0.37 −0.08 0.17 −0.12 −0.17 0.44 0.93 0.61 0.50
R4 0.22 0.30 0.36 −0.08 0.11 −0.08 −0.28 0.39 0.85 0.53 0.45
IR1 0.27 0.39 0.49 −0.11 0.19 −0.10 −0.15 0.51 0.68 0.83 0.46 0.81 0.88 0.64
IR2 0.26 0.39 0.48 −0.10 0.22 −0.11 −0.13 0.51 0.59 0.79 0.41
IR3 0.14 0.13 0.17 0.06 0.04 0.15 −0.05 0.21 0.41 0.79 0.12
IR4 0.22 0.22 0.15 0.06 0.03 0.14 −0.07 0.22 0.42 0.79 0.13
ACI1 0.34 0.37 0.55 −0.18 0.22 −0.33 −0.22 0.65 0.52 0.28 0.90 0.95 0.96 0.87
ACI2 0.37 0.42 0.55 −0.23 0.21 −0.28 −0.18 0.64 0.58 0.36 0.92
ACI3 0.38 0.38 0.51 −0.20 0.19 −0.29 −0.21 0.63 0.53 0.34 0.96
ACI4 0.34 0.38 0.49 −0.17 0.17 −0.25 −0.19 0.61 0.52 0.34 0.94

Note: U (App Utility), AE (App Aesthetics), PE (Perceived Enjoyment), KAQ (Knowledge of Alternative Quality), PP (Perceived Price), T (Technicality), PR (Privacy Risk), S (Satisfaction), R
(Recommendation), IR (Intention to Recommend), ACI (App Continuance Intention).
We highlight the factor loadings of the measurement items on the constructs that they belong to.
178 C. Xu et al. / Decision Support Systems 79 (2015) 171–183

Table 2
Correlation of latent variables.

U AE PE KAQ PP T PR Sat IR R ACI

U 0.78
AE 0.32 0.89
PE 0.18 0.52 0.87
KAQ −0.13 −0.15 −0.16 0.85
PP 0.18 0.24 0.18 −0.15 0.94
T −0.11 −0.19 −0.29 0.48 −0.05 0.90
PR −0.09 −0.06 −0.21 0.25 −0.05 0.37 0.88
Sat 0.38 0.50 0.70 −0.16 0.21 −0.28 −0.20 0.88
IR 0.27 0.36 0.43 −0.11 0.20 −0.14 −0.23 0.50 0.90
R 0.28 0.36 0.41 −0.03 0.15 0.03 −0.13 0.46 0.66 0.80
ACI 0.38 0.42 0.56 −0.21 0.21 −0.31 −0.22 0.68 0.57 0.35 0.93

Note: U (App Utility), AE (App Aesthetics), PE (Perceived Enjoyment), KAQ (Knowledge of Alternative Quality), PP (Perceived Price), T (Technicality), PR (Privacy Risk), S (Satisfaction), R
(Recommendation), IR (Intention to Recommend), ACI (App Continuance Intention).
The diagonal elements in bold are the square root of the average variance extracted.

high correlations among the first-order constructs and low collinearity
among the first-order constructs should be expected. For the hedonic
benefits construct, the impacts of enjoyment and aesthetics on hedonic
benefits are significant (see Fig. 2); the correlation between them is low
at 0.52 (see Table 2), suggesting that a reflective model is less likely to
be present [60]; VIF values are also low (app aesthetics = 1.69,
enjoyment = 2.36). For the utilitarian benefits construct, the impacts
of app utility and app quality on utilitarian benefits are significant (see
Fig. 2); the correlation between app utility and app quality is 0.41 (see
Table 2); low VIF values (app utility = 1.38, app quality = 1.93)
shows that collinearity is less likely to be a problem. For the non-
monetary sacrifices construct, technicality, privacy risk and knowledge
of alternative quality have significant impacts on non-monetary sacri-
fices (see Fig. 2); the highest correlation among them is low at 0.48
(see Table 2); VIF values are also low (technicality = 1.63, privacy
risk = 1.23, knowledge of alternative quality = 1.36). Taken together,
this evidence indicates that a more parsimonious second-order repre-
sentation is able to fully capture the predictive power of those first-
order constructs.
To test the validity of the formative construct–app quality, we
followed the established procedures [30,44]. Applying these criteria to
our data, we found that all the measurement items satisfy the validity
requirements (see Table 3). We used them in the structural equation
modeling analysis.
5.2. Model testing
Fig. 2 shows the model testing results. Our model accounts for 44% of
variance in recommendation and 34% of variance in intention to recom-
mend. While intention to recommend has salient effect on recommen-
dation (β = 0.66, p b 0.001), hedonic benefits (β = 0.12, p b 0.05),
satisfaction (β = 0.11, p b 0.05) and app continuance intention (β =
0.40, p b 0.001) were found to be the three antecedents of intention to
recommend. The effects of utilitarian benefits, non-monetary sacrifices
and perceived price on intention to recommend are not significant.
Moreover, the hypotheses regarding the effects of utilitarian benefits
(β = 0.11, p b 0.05), hedonic benefits (β = 0.12, p b 0.05), non-
monetary sacrifices (β = -0.16, p b 0.01) and satisfaction (β = 0.49,

Note:

1. *** p<0.001, ** p<0.01, * p<0.05

2. PRR (Privacy Risk), Knowledge of Alternative Quality (KAQ), App Continuance Intention (ACI), Intention to Recommend (IR)

Fig. 2. PLS Results.

 C. Xu et al. / Decision Support Systems 79 (2015) 171–183
Table 3
Validity of the formative construct–app quality (AQ).
Item Weight p value
AQ1 0.27 b0.001
AQ2 0.17 b0.001
AQ3 0.29 b0.001
AQ4 0.28 b0.001
AQ5 0.30 b0.001
p b 0.001) on app continuance intention are well-supported except the
effect of perceived price on app continuance intention. Last but not least,
although the effect of perceived price on satisfaction were found to be
insignificant, utilitarian benefits (β = 0.32, p b 0.001) and hedonic ben-
efits (β = 0.50, p b 0.001) positively affect satisfaction and non-
monetary sacrifices (β = -0.10, p b 0.05) has negative impact on
satisfaction.
6. Discussion and implications
6.1. Discussion of findings
The present research extends VSL to the mobile app context and
investigates factors that affect users to recommend apps. The results
show that intention to continue to use apps, satisfaction with apps,
and hedonic benefits obtained from using apps directly influence
users' intention to recommend the apps which is a strong predictor
of their recommendation behavior. Utilitarian benefits and hedonic
benefits have positive effects on user satisfaction and app continu-
ance intention: the more utilitarian benefits and hedonic benefits
users obtain, the more satisfied they are with the apps and the stron-
ger their intention to continue to use the apps. As to the specific
value components, app utility and app quality are the main utilitari-
an benefits in the context of mobile apps whereas enjoyment and
app aesthetics are the primary hedonic benefits associated with
app usage. The three sources of non-monetary sacrifices related to
app usage are privacy risk, knowledge of alternative quality and
technicality; they, as a whole, negatively influence user satisfaction
and app continuance intention.
Although not a direct determinant, utilitarian benefits were found to
exert its influence on intention to recommend through satisfaction and
app continuance intention. This result is in line with de Matos and
Rossi's [20] meta-analysis of prior WOM literature which concluded
that few studies found that product quality (a concept similar to utilitar-
ian benefits) is a direct antecedent of WOM. The result here implies that
users do not recommend their apps simply because they are useful and
high in quality; in fact, they recommend apps that deliver such utilitar-
ian benefits when they are satisfied with the apps and are willing to
continue to use them in the future.
Moreover, contrary to our expectation, perceived price does not
have effects on satisfaction, app continuance intention and intention
to recommend. This finding suggests that monetary sacrifices play a
less important role in the post-adoption stage of mobile apps. The
insignificant relationships here could possibly be attributable to the fol-
lowing reasons. First, it may be due to the low app price. Previous liter-
ature [95,96] showed that due to lack of price variation, for low-cost
products, customers regard other factors such as time and effort costs
as more important than monetary sacrifices. Nowadays, a large portion
of mobile apps are free and the price of paid apps is also low. Statistics
showed that 60% of paid apps were priced at $.99; others mostly
range from $.99 to $1.99; few to no apps have prices above $4.99
[26].4 The low price of mobile apps could cause the influences of mone-
tary sacrifices to be minimal when compared to monetary sacrifices
4
It is worth noting that the cost of acquiring apps could be further reduced by using
discounted iTunes and Google Play gift cards to purchase the apps. Such gift cards are very
available on eBay and other gift card selling websites.
179
(e.g., the time and mental efforts spent to deal with the concern of pri-
vacy invasion or to gain familiarity with the app interface), the benefits
VIF associated with using the apps and user satisfaction. Second, it may well
1.56
be that monetary sacrifices is not always perceived as something one
1.16 has to give up in using mobile apps. Previous literature indicated that
1.90 the negative effects of perceived price could be neutralized to be insig-
1.90 nificant if it is positively used for other purposes such as inferring prod-
2.26
uct quality and vendor quality [96]. In this paper, the app users should
know the app quality well, but there are still uncertainties regarding
whether the app vendors are capable of preventing security breach
(e.g., Starbucks app5) and whether they will discontinue the service of
the apps out of sudden (e.g., Evernote Hello app6) and so on. Such un-
certainties, highly relevant to users' interest, are difficult to detect be-
cause users do not have enough information and expertise to discern
any signs of them. When app price is used as a cue to infer those uncer-
tainties (e.g., a paid app could rely on revenues contributed by the users,
thus the vendor is less likely to share user data with third parties for per-
sonal gains), it could attenuate the negative effects of perceived price to
be insignificant. Similarly, we cannot rule out the possibility that individ-
uals who chose a paid app care less about the price partially because they
have a good impression of the product in the first place. As result these
individuals are more likely to be satisfied, to intend to continue to use
the app in the future, and to recommend the app regardless of reasonable
variation in price. This rationale could also be a potential explanation for
perceived price to be insignificant. Third, the price paid by app users is a
source of switching costs [45] which tends to result in customer loyalty
[46]. While loyal customers are more likely to exhibit satisfaction, repur-
chase the products, and recommend them to others, switching costs by
itself can also foster repatronage and encourage recommendations [45].
We speculate that users who have paid for their apps are more satisfied
with the apps and are more willing to continue to use and recommend
them (which are the opposite of our hypotheses and the notion of
VSL). Therefore, when app price is perceived as a form of switching
costs, its negative effects on other VSL constructs are minimized.
Interestingly, we also did not observe non-monetary sacrifices to in-
fluence users' intention to recommend apps. This result and the insignif-
icant relationship between monetary sacrifices and intention to
recommend echoes de Matos and Rossi's [20] meta-analysis of prior
WOM literature which did not identify any types of sacrifices to be the
direct predictors of WOM. A theoretical rationale for this result is pro-
vided by Lam et al.'s [45] work. The classic attitudinal framework
(cognition → affect → behavioral intent or behavior) postulates that
cognition evokes affect which subsequently determines one's behavior-
al intention or behavior. In the context of VSL, customer value as a whole
is regarded as cognition, satisfaction is regarded as affect and customer
loyalty represents behavioral intent or behavior. There are two types of
loyalty behaviors, namely customer retention and customer attraction.
In the consumer context, cognition and affect play different roles in
shaping these two loyalty behaviors. While repatronage pertains to cus-
tomer retention and recommendation to customer attraction, Lam and
his associate found that under the VSL paradigm, recommendation is
solely driven by customers' affective state (satisfaction) and repatronage
is determined by both cognitive evaluation (customer value) and affec-
tive state (satisfaction). Therefore, it is not a surprise that we did not
find the significant relationship between non-monetary sacrifices and
intention to recommend. In fact, our findings with respect to the effects
of all four value components (utilitarian benefits, hedonic benefits,
monetary sacrifices and non-monetary sacrifices) on intention to rec-
ommend are supported by Lam et al.'s [45] work. While other three
value components do not have direct effects on intention to
5
It is reported by USA Today that in May 2015, Starbucks app was hacked; money in
many user accounts were stolen (http://www.usatoday.com/story/tech/2015/05/15/
starbucks-gift-card-hack/27370491/).
6
In February 2015, Evernote announced to discontinue its support for Evernote Hello,
an app that facilitates managing business cards.
180 C. Xu et al. / Decision Support Systems 79 (2015) 171–183
recommend, the only significant relationship between hedonic benefits
and intention to recommend is due to the fact that hedonic benefits
(e.g., having a good time when using an app, enjoying the visually-
appealing interface of an app) reflect users' affective state (which is a
determinant of recommendation according to Lam and his associates).
This research supports the significant effects of satisfaction on app
continuance intention and intention to recommend. App continuance
intention and intention to recommend are two facets of loyalty disposi-
tion. While satisfied customers are not necessarily loyal (which raises
the concern that satisfied users might not necessarily intend to continue
using an app and recommend an app), the Investment Model Theory
and the Commitment-Trust Theory (Li et al., 2006) indicate that this
phenomenon is fundamentally attributable to the intervention of four
factors—alternative quality, opportunistic behavior, communication
quality and investment size. The former two factors were considered
in the research model in the forms of knowledge of alternative quality
and privacy risk. The latter two factors were excluded from the main
analysis because in the a priori focus group interview and pre-test
used to develop and refine the research model, it was found that com-
munication quality and investment size were not appropriate in the
context of this research. Taking into consideration these issues that
might weaken the satisfaction-loyalty relationship, we believe the sig-
nificant associations between satisfaction and app continuance inten-
tion/intention to recommend found in this research are robust.
6.2. Theoretical implications
The theoretical implications of this paper are five-fold. First, this re-
search contributes toward filling a gap in the IS discipline that app rec-
ommendation is a critical success factor of mobile apps, but our
understanding of this phenomenon is limited.
Second, this research contributes to the electronic WOM (eWOM) re-
search. WOM is a type of information exchange in which messages flow
from senders to recipients. This process consists of five elements [69]:
(1) (the sender's) perception as to the value of the source knowledge,
(2) (the sender's) willingness of the source to share that knowledge,
(3) the availability of rich information channels, (4) the willingness of
the recipient to acquire the knowledge, and (5) the absorptive capacity
of the recipient. Cheung and Thadani's [9] meta-analysis of eWOM
found that prior IS research mainly focused on the last three elements
of information flow. This finding is understandable because information
channels are the IT artifacts; IS researchers were interested in how factors
such as argument quality and source credibility influence individuals' in-
tention to acquire eWOM through different information channels. Be-
cause the stimuli associated with the first two elements of information
flow may not be IT artifacts, they were less interesting to IS researchers.
This research shifts the attention from “IT artifacts as the information
channel” to “IT artifacts as the stimuli”. Specifically, it identifies factors
that influence individuals' willingness to initiate eWOM in the first
place. Our results show that senders' perceptions of the value of mobile
apps is formed as a function of app benefits and app sacrifices (corre-
sponding to element 1), and it is the benefits derived from app usage,
user satisfaction and app continuance intention that shape app users' in-
tention to recommend apps (corresponding to element 2). In this sense,
our research enriches IS research in eWOM because it identifies a poten-
tial research area where IT artifacts serve as information exchange stimuli.
Third, given that recommendation resides along the positive end of
the WOM valence continuum, this research represents the efforts from
the IS discipline to fill the gap identified by de Matos and Rossi [20] and
Harrison-Walker [34] that only a few WOM studies examined WOM va-
lence and that this dimension of WOM was only previously measured
indirectly.
Fourth, this research echoes several research calls. It responds to
Chin and Marcolin's [10] request that IS research examine complex
“deep usage” of IT. Moreover, given that mobile apps are a type of
experiential computing technology, this work also responds to Yoo's
[85] call that it is time for IS research to focus on experiential computing.
Fifth, the refined VSL framework is another intellectual contribution
to the IS discipline. To examine app recommendation, this research not
only draws upon VSL but also adapts it to the mobile app context. We
acknowledged the predominant conceptualization that customer
value is customers' overall assessment of what is received (benefit)
and what is given (sacrifice). Thus, we broke down the value construct
that was unanimously treated as a “black box” by past VSL works, ex-
plored specific benefits/sacrifices associated with using mobile apps,
and investigated their individual influences with VSL. To date, little
VSL research has systematically identified specific benefits/sacrifices
that shape customer value. Such oversight is problematic. Because cus-
tomer value is context-based, a new context, such as mobile apps, may
change the compositions of the value construct. Our findings indicate
that the identified four value components behave differently with re-
gard to their linkage with satisfaction, app continuance intention and
recommendation, thus providing support for the necessity to decom-
pose the “black box” of the value construct. Our approach that decom-
poses a construct for deepened insights into its components is in
similar spirit to prior studies [31,83] responding to Benbasat and Barki's
[3] call to pry open the black box of perceived usefulness to refocus on IT
artifact design. Just as prior research identified factors that shape per-
ceived usefulness, our research, by identifying the compositions of per-
ceived value, provides mobile app design guidelines and helps
stakeholders effectively build their value-enhancing programs. Further-
more, the refined VSL framework here offers insights into why satisfied
users are not loyal (e.g., not recommending apps), an increasingly criti-
cal issue in the marketing and IS disciplines that could not be addressed
by the original VSL framework. IS researchers have long questioned IT
post-adoption models such as the Expectation Confirmation Model [4]
in that satisfaction plays a vital role in these theories and in voluntary
IT usage settings, satisfied users may still discontinue to use an IT [46].
This criticism is also applicable to VSL that has satisfaction at the core.
By considering knowledge of alternative quality and privacy risk, the re-
fined VSL framework presented here ensures that the associations be-
tween satisfaction and two loyalty constructs—app continuance
intention and intention to recommend—are robust.
6.3. Practical implications
Given the importance of soliciting recommendation from app users,
the findings of this research could benefit practitioners in several ways.
First, our research identifies two utilitarian benefits and two hedonic
benefits. In cultivating customer value to enhance users' intention to
recommend apps, app developers should stress on the benefits of app
utility, app quality, app aesthetics and enjoyment that the apps can de-
liver in their marketing campaigns.
Second, given the effect of app aesthetics on value formation, app
developers should focus on interface design to ensure that the apps
are visually aesthetic. Aesthetic design can be achieved by a variety of
approaches including the appropriate choice of background color, use
of human images, and so forth [16].
Third, sacrifices associated with app usage influence users' intention
to recommend through satisfaction and app continuance intention.
Therefore, app practitioners should endeavor to minimize three sacrifice
factors—technicality, privacy risk and knowledge of alternative
quality—to maintain user satisfaction and app continuance intention.
Given the distinctive difference between app interfaces and website inter-
faces, creating a user-friendly interface is an important task that merits
app developers' attention. To this end, they need to design an app inter-
face such that it is easy for users to operate. A user guide video can be pro-
vided particularly for users who are new to the apps. To offset users'
privacy concerns, app practitioners could emphasize their privacy policies
in marketing campaigns like TV ads to make the policies publicly known.
 C. Xu et al. / Decision Support Systems 79 (2015) 171–183 181

Practitioners can also compare their apps with competing products in
order to reduce the influence of the alternatives.
6.4. Limitations and future directions
The current study contains several limitations. First, our sample
respondents consist of undergraduate students (with an average
age of 21) and graduate students (with an average age of 31). They
are representative users of mobile apps because Nielson's [59] report
shows that teens and young adults are the major users of
smartphones. Nevertheless, future studies could examine app rec-
ommendation with different segments of users. Second, our respon-
dents represent the U.S. population. Thus, future research could
examine how culture influences app recommendation. Third, cus-
tomer loyalty is a multi-faceted concept where recommendation is
just one of them. Future research could draw upon our refined VSL
model to investigate other facets of mobile app loyalty such as will-
ingness to pay and intention to explore other features. Fourth, our
research model did not consider individual factors such as seeking
reputation, sense of belonging and enjoyment of helping. Individual
factors are significant antecedents of WOM activities or knowledge
sharing (a concept similar to WOM) where the information commu-
nicated could be positive, neutral or negative [e.g., 13,53]. The focus
of this research, however, is mobile app recommendation, a different
dimension of WOM. In order for users to spread words favorable to
apps, there must be salient factors beyond individual factors, and
this is when the effect of product/attributes kicks in. VSL attends to
product factors perceived by app users. This is why we chose VSL to
examine the outcome variable. The refined VSL framework explains
34% of the variance in intention to recommend and 44% of the vari-
ance in recommendation, which are high for social science research
and similar to prior research using alternative models to explain
technology recommendation of [54]. This result indicates that our
model captures the majority of variance in app users' recommenda-
tion intentions and behaviors and individual factors are less impor-
tant in affecting WOM valence. Nevertheless, we suggest future
research examine how VSL works in conjunction with individual fac-
tors in order to facilitate a more comprehensive understanding of the
issue and to enhance the explanatory power of VSL. Fifth, the benefit
and sacrifice components identified in this research are not exhaus-
tive. Although they are key factors in other well-established IT adop-
tion models and the R-square values of the current model are high,
we suggest future research pay attention to any new benefits or sac-
rifices when applying the current model in a new context. Sixth, an
interesting area for future research would be the use of paid apps
so that IS academia could inform practitioners of how to attract indi-
viduals to use paid apps, thus expanding their revenue models. Sev-
enth, this research did not examine negative WOM. The mechanisms
of why and how negative WOM is given are different from positive
WOM; as a result, the determinants of negative WOM are far beyond
satisfaction and customer value [5]. On the other hand, negative
WOM is alien to customer loyalty. Thus, we believe VSL insufficient
to capture the whole notion of this concept. Future research could
examine negative WOM in the mobile app context.

Appendix A. Construct Items

AE1 The interface of my application is aesthetically appealing KAQ1 An alternative application is appealing
AE2 The interface of my application is attractive KAQ2 An alternative application is better than this one
AE3 The interface of my application is aesthetically designed KAQ3 To my knowledge, another application is close to ideal
AE4 The overall look and feel of the application is visually appealing KAQ4 An alternative application is attractive to me
AE5 The interface of my application pleases my senses KAQ5 My needs could easily be fulfilled by an alternative application
PE1 I find using the application to be enjoyable ACI1 I intend to continue using the application
PE2 Using the application is pleasant ACI2 My intentions are to continue using the application rather than
PE3a Using the application is dull using any alternatives
PE4 I have fun using the application ACI3 I plan to keep using this application in the future
PE5 I enjoy using the application ACI4 I expect my use of the application to continue in the future
U1 The application enables me to do it more quickly AQ1 The application is reliable
U2 The application helps me be more effective AQ2 The application can serve a variety of my needs
U3 The application helps me be more productive AQ3 The application is smoothly integrated into my
U4 The application saves me time to use it smartphone (Integration/compatibility)
U5 The application makes it easier to do it AQ4 The application is available whenever I want to use it
U6 The application is useful to me AQ5 The application responds to my request in a timely fashion
S1 I feel very satisfied with the overall experience of T1 I think the application is difficult to use
using the application T2 Learning to operate the application is difficult for me
S2 I am very pleased with the overall experience of T3 My interaction with the application requires a lot of mental effort
using the application T4 It is difficult for me to become skillful at using the application
S3 I am very contended with the overall experience of
using the application IR1 I intend to say positive things about the application
S4 I feel very delighted with the overall experience of IR2 I would say good things about the application
using the application IR3 I would recommend the application to other people
IR4 I intend to encourage other people to use the application
PR1 The application may disclose my personal information to others
PR2 The application may share my personal information to others PP1 I think the fee that I paid for the use of this app is acceptable
PR3 The application may track my habits of mobile phone use PP2 I think the fee that I paid for the use of this app is reasonable
PR4 The application may collect and use my personal information in PP3a I think the fee I paid for the use of this app is high.
an unintended way PP4 I am pleased with the fee that I paid for the use of this app
PR5 The application may cause me to lose control over my privacy
PR6 The application may lead to a loss of my privacy without my knowledge
R1 I say positive things about this application to others frequently
R2 I have spoken favorably to more people this application than
I have spoken most other applications
R3 I seldom miss an opportunity to recommend this application to others
R4 When I recommend this application, I introduce it in detail

Note: U (App Utility), AE (App Aesthetics), PE (Perceived Enjoyment), KAQ (Knowledge of Alternative Quality), PP (Perceived Price), T (Technicality), PR (Privacy Risk), S (Satisfaction), R
(Recommendation), IR (Intention to Recommend), ACI (App Continuance Intention).
a
Denotes items that were dropped after factor analysis.
182 C. Xu et al. / Decision Support Systems 79 (2015) 171–183
References
[1] I. Ajzen, Theory of planned behavior, Organizational Behavior and Human Decision
Processes 50 (2) (1991) 179–211.
[2] E.W. Anderson, M.W. Sullivan, The antecedents and consequences of customer sat-
isfaction for firms, Marketing Science 12 (2) (1993) 125–143.
[3] I. Benbasat, B. Henri, Quo vadis, TAM? Journal of the Association for Information Sys-
tems 8 (4) (2007) 212–218.
[4] A. Bhattacherjee, Understanding information systems continuance: an expectation–
confirmation model, MIS Quarterly 25 (3) (2001) 351–370.
[5] J.G. Blodgett, D.H. Granbois, R.G. Walters, The effects of perceived justice on com-
plainants' negative word-of-mouth behavior and repatronage intentions, Journal
of Retailing 69 (1993) 399–428.
[6] J. Casey, Like VC's, Use PR to Promote Your Mobile Apps, 2012 Available from: http://
mobileorchard.com/like-vcs-use-pr-to-promote-your-mobile-apps/.
[7] L. Chen, T.O. Meservy, M. Gillenson, Understanding information systems continu-
ance for information-oriented mobile applications, Communications of the Associa-
tion for Information Systems 30 (2012) 127–146.
[8] Z. Chen, A.J. Dubinsky, A conceptual model of perceived customer value in e-
commerce: a preliminary investigation, Psychology and Marketing 20 (4) (2003)
323–347.
[9] C.M.K. Cheung, D.R. Thadani, The impact of electronic word-of-mouth communica-
tion: a literature analysis and integrative model, Decision Support Systems 54 (1)
(2012) 461–470.
[10] W.W. Chin, B.L. Marcolin, The future of diffusion research, SIGMIS Database 32 (3)
(2001) 7–12.
[11] W.W. Chin, B.L. Marcolin, P.R. Newsted, A partial least squares latent variable model-
ing approach for measuring interaction effects: results from a Monte Carlo simula-
tion study and an electronic-mail emotion/adoption study, Information Systems
Research 14 (2) (2003) 189–217.
[12] C.-M. Chiu, et al., Re-examining the influence of trust on online repeat purchase in-
tention: the moderating role of habit and its antecedents, Decision Support Systems
53 (4) (2012) 835–845.
[13] C.-M. Chiu, M.-H. Hsu, E.T.G. Wang, Understanding knowledge sharing in virtual
communities: an integration of social capital and social cognitive theories, Decision
Support Systems 42 (3) (2006) 1872–1888.
[14] C.-W. Chu, H.-P. Lu, Factors influencing online music purchase intention in Taiwan:
an empirical study based on the value-intention framework, Internet Research 17
(2) (2007) 139–155.
[15] G.A. Churchill Jr., C. Surprenant, An investigation into the determinants of customer
satisfaction, Journal of Marketing Research 19 (4) (1982) 491–504.
[16] D. Cyr, et al., Exploring human images in website design: a multi-method approach,
MIS Quarterly 33 (3) (2009) A9-539.
[17] D. Cyr, M. Head, A. Ivanov, Design aesthetics leading to m-loyalty in mobile com-
merce, Information & Management 43 (8) (2006) 950–963.
[18] F.D. Davis, Perceived usefulness, perceived ease of use, and user acceptance of infor-
mation technology, MIS Quarterly 13 (3) (1989) 319–340.
[19] F.D. Davis, R.P. Bagozzi, P.R. Warshaw, Extrinsic and intrinsic motivation to use com-
puters in the workplace, Journal of Applied Social Psychology 22 (14) (1992)
1111–1132.
[20] C.A. de Matos, C.A.V. Rossi, Word-of-mouth communications in marketing: a meta-
analytic review of the antecedents and moderators, Journal of the Academy of Mar-
keting Science 36 (4) (2008) 578–596.
[21] T. Dinev, P. Hart, An extended privacy calculus model for E-commerce transactions,
Information Systems Research 17 (1) (2006) 61–80.
[22] Dynmark International, Word of Mouth is One of the ‘Most Effective’ Ways to Pro-
mote Mobile Apps, 2012.
[23] J.R. Edwards, Multidimensional constructs in organizational behavior research: an
integrative analytical framework, Organizational Research Methods 4 (2) (2001)
144–192.
[24] A. Finn, L. Wang, T. Frank, Attribute perceptions, customer satisfaction and intention
to recommend E-services, Journal of Interactive Marketing 23 (3) (2009) 209–220.
[25] M. Fishbein, I. Ajzen, Belief, Attitude, Intention and Behavior: An Introduction to
Theory and Research, Addison Wesley, Reading, MA, 1975.
[26] Flurry, The history of app pricing, and why most apps are freeAvailable from: http://
blog.flurry.com/bid/99013/The-History-of-App-Pricing-And-Why-Most-Apps-Are-
Free2013.
[27] C. Fornell, D.F. Larcker, Evaluating structural equation models with unobservable
variables and measurement error, Journal of Marketing Research 18 (1) (1981)
39–50.
[28] J.E. Gerow, et al., Can we have fun @ work? The role of intrinsic motivation for util-
itarian systems, European Journal of Information Systems 22 (2012) 360–380.
[29] A.K. Ghosh, T.M. Swaminatha, Software security and privacy risks in mobile e-
commerce, Communications of the ACM 44 (2) (2001) 51–57.
[30] O. Götz, K. Liehr-Gobbers, M. Krafft, Evaluation of structural equation models using
the partial least squares (PLS) approach, in: V.E. Vinzi, et al., (Eds.), Handbook of
Partial Least Squares: Concepts, Methods and Applications, Springer, Berlin,
Gernmany, 2010.
[31] Z. Guo, F.B. Tan, K. Cheung, Students' uses and gratifications for using computer-
mediated communication media in learning contexts, Communications of AIS
2010 (27) (2010) 339–378.
[32] I. Ha, Y. Yoon, M. Choi, Determinants of adoption of mobile games under mobile
broadband wireless access environment, Information Management 44 (3) (2007)
276–286.
[33] J.F. Hair, et al., Multivariate Data Analysis with Readings, Prentice-Hall, Englewood
Cliffs, NJ, 1995.
[34] L.J. Harrison-Walker, The measurement of word-of-mouth communication and an
investigation of service quality and customer commitment as potential antecedents,
Journal of Service Research 4 (1) (2001) 60–75.
[35] C.-C. Hsiao, J.-S. Chiou, The impact of online community position on online game
continuance intention: do game knowledge and community size matter? Informa-
tion Management 49 (6) (2012) 292–300.
[36] C.B. Jarvis, et al., A critical review of construct indicators and measurement model
misspecification in marketing and consumer research, Journal of Consumer Re-
search 30 (2) (2003) 199–218.
[37] E. Katz, P. Lazarsfeld, Personal Influence, The Free Press, New York, 2015.
[38] W.J. Kettinger, S.-H.S. Park, J. Smith, Understanding the consequences of information
systems service quality on IS service reuse, Information Management 46 (6) (2009)
335–341.
[39] D.J. Kim, D.L. Ferrin, H.R. Rao, A trust-based consumer decision-making model in
electronic commerce: the role of trust, perceived risk, and their antecedents, Deci-
sion Support Systems 44 (2) (2008) 544–564.
[40] H.-W. Kim, H.C. Chan, S. Gupta, Value-based adoption of mobile internet: an empir-
ical investigation, Decision Support Systems 43 (1) (2007) 111–126.
[41] H.-W. Kim, S. Gupta, J. Koh, Investigating the intention to purchase digital items in
social networking communities: a customer value perspective, Information Man-
agement 48 (6) (2011) 228–234.
[42] A. Kolirin, Amid Too Many Choices, App Consumers Rely on Word-of-Mouth Mar-
keting, 2012 Available from: http://www.emarketingandcommerce.com/article/
amid-too-many-choices-app-consumers-rely-word-of-mouth-marketing/1#.
[43] Y.-F. Kuo, C.-M. Wu, W.-J. Deng, The relationships among service quality, perceived
value, customer satisfaction, and post-purchase intention in mobile value-added
services, Computers in Human Behavior 25 (4) (2009) 887–896.
[44] F. Lai, D. Li, C.-T. Hsieh, Fighting identity theft: the coping perspective, Decision Sup-
port Systems 52 (2) (2012) 353–363.
[45] S.Y. Lam, et al., Customer value, satisfaction, loyalty, and switching costs: an illustra-
tion from a business-to-business service context, Journal of the Academy of Market-
ing Science 32 (3) (2004) 293–311.
[46] D. Li, G. Browne, J. Wetherbe, Why do internet users stick with a specific web site? A
relationship perspective, International Journal of Electronic Commerce 10 (4)
(2006) 105–141.
[47] H. Liang, Y. Xue, Understanding security behaviors in personal computer usage: a
threat avoidance perspective, Journal of the Association for Information Systems
11 (7) (2010) 394–413.
[48] H.-H. Lin, Y.-S. Wang, An examination of the determinants of customer loyalty in
mobile commerce contexts, Information Management 43 (3) (2006) 271–282.
[49] T.-C. Lin, et al., The integration of value-based adoption and expectation–
confirmation models: an example of IPTV continuance intention, Decision Support
Systems 54 (2012) 63–75.
[50] D. Liqiong, M.S. Poole, Affect in web interfaces: a study of the impacts of web page
visual complexity and order, MIS Quarterly 34 (4) (2010) A10-711.
[51] J.-B. Lohmöller, Latent Variable Path Modeling with Partial Least Squares, Germany
Physica-Verlag, Heidelberg, 1989.
[52] H.-P. Lu, K.-L. Hsiao, The influence of extro/introversion on the intention to
pay for social networking sites, Information Management 47 (3) (2010)
150–157.
[53] M. McLure Wasko, S. Faraj, Why should I share? Examining social capital and knowl-
edge contribution in electronic networks of practice, MIS Quarterly 29 (1) (2005)
35–57.
[54] C.L. Miltgen, A. Popovič, T. Oliveira, Determinants of end-user acceptance of biomet-
rics: integrating the “Big 3” of technology acceptance with privacy context, Decision
Support Systems 56 (2013) 103–114.
[55] R.M. Morgan, S.D. Hunt, The commitment-trust theory of relationship marketing,
Journal of Marketing 58 (3) (1994) 20.
[56] MTV Networks, Love 'em or leave 'em: adoption, abandonment and the app-addled
consumer, http://www.prnewswire.com/news-releases/mtv-networks-mobile-
apps-study-reveals-the-life-cycle-of-an-app-from-discovered-to-discarded-
123348433.html2011.
[57] K.B. Murray, A test of services marketing theory: consumer information acquisition
activities, Journal of Marketing 55 (1) (1991) 10–25.
[58] F.F.-H. Nah, K. Siau, H. Sheng, The value of mobile applications: a utility company
study, Communications of the ACM, Association for Computing, Machinery 2005,
pp. 85–90.
[59] Nielsen, Young adults and teens lead growth among smartphone ownersAvailable
from: http://blog.nielsen.com/nielsenwire/online_mobile/young-adults-and-teens-
lead-growth-among-smartphone-owners/2012.
[60] P.A. Pavlou, O.A. El Sawy, From IT leveraging competence to competitive advantage
in turbulent environments: the case of new product development, Information Sys-
tems Research 17 (2006) 198–227.
[61] S. Petter, D. Straub, A. Rai, Specifying formative constructs in information systems
research, MIS Quarterly 31 (4) (2007) 623–656.
[62] E.M. Rogers, Diffusion of Innovations: Simon and Schuster, 2010.
[63] M.J. Schwartz, California targets mobile apps for missing privacy policies[cited 2012
February 14th]; Available from: http://www.informationweek.com/government/
mobile/california-targets-mobile-apps-for-missi/2400126032012.
[64] I. Sinha, W.S. DeSarbo, An integrated approach toward the spatial modeling of per-
ceived customer value, Journal of Marketing Research (JMR) 35 (2) (1998)
236–249.
[65] D. Sirdeshmukh, J. Singh, B. Sabol, Consumer trust, value, and loyalty in relational
exchanges, Journal of Marketing 66 (1) (2002) 15–37.
[66] H.J. Smith, T. Dinev, H. Xu, Information privacy research: an interdisciplinary review,
MIS Quarterly 35 (4) (2011) A27-980.
 C. Xu et al. / Decision Support Systems 79 (2015) 171–183 183

[67] J.-Y. Son, S.S. Kim, Internet users' information privacy-protective responses: a taxon- [89] V.A. Zeithaml, Consumer perceptions of price, quality, and value: a means-end
omy and a nomological model, MIS Quarterly 32 (3) (2008) 503–529. model and synthesis of evidence, Journal of Marketing 52 (3) (1988) 2–22.
[68] Speckyboy Editors, The Aesthetics of a Successful Mobile App: Lessons from Three [90] L. Zhao, et al., Assessing the effects of service quality and justice on customer satis-
Top Grossing Apps, 2012 Available from: http://speckyboy.com/2012/10/25/the- faction and the continuance intention of mobile value-added services: an empirical
aesthetics-of-a-successful-mobile-app-lessons-from-three-top-grossing-apps/. test of a multidimensional model, Decision Support Systems 52 (3) (2012) 645–656.
[69] S.W. Susssman, W.S. Siegal, Informational influence in organizations: an integrated [91] T. Zhou, An empirical examination of continuance intention of mobile payment ser-
approach to knowledge adoption, Information Systems Research 14 (1) (2003) vices, Decision Support Systems 54 (2) (2013) 1085–1091.
47–65. [92] W.B. Dodds, K.B. Monroe, D. Grewal, Effects of price, brand, and store information on
[70] J.C. Sweeney, G.N. Soutar, L.W. Johnson, The role of perceived risk in the quality– buyers' product evaluations, Journal of Marketing Research 28 (3) (1991) 307–319.
value relationship: a study in a retail environment, Journal of Retailing 75 (1) [93] L.A. Catherine, R. Agarwal, Practicing safe computing: a multimedia empirical exam-
(1999) 77–105. ination of home computer user security behavioral intentions, MIS Quarterly 34 (3)
[71] N. Tractinsky, A.S. Katz, D. Ikar, What is beautiful is usable, Interacting with Com- (2010) 613–643.
puters 13 (2) (2000) 127–145. [94] A. Burton-Jones, D.W. Straub, Reconceptualizing system usage: an approach and
[72] C. Tung-Zong, A.R. Wildt, Price, product information, and purchase intention: an em- empirical test, Information Systems Research 17 (3) (2006) 228–246.
pirical study, Journal of the Academy of Marketing Science 22 (1) (1994) 16–27. [95] A. Downs, A theory of consumer efficiency, Journal of Retailing 37 (1961) 6–12.
[73] O. Turel, A. Serenko, N. Bontis, User acceptance of hedonic digital artifacts: a theory [96] H.-W. Kim, Y. Xu, S. Gupta, Which is more important in Internet shopping, perceived
of consumption values perspective, Information Management 47 (1) (2010) 53–59. price or trust? Electronic Commerce Research and Applications 11 (3) (2012)
[74] O. Turel, A. Serenko, N. Bontis, User acceptance of wireless short messaging services: 241–252.
deconstructing perceived value, Information Management 44 (1) (2007) 63–73.
[75] H. van der Heijden, Factors influencing the usage of websites: the case of a generic
Chenyan Xu is an assistant professor of information systems in the Computer Science and
portal in The Netherlands, Information Management 40 (6) (2003) 541–549.
Information Systems Program, School of Business, at the Stockton University. His research
[76] H. van der Heijden, User acceptance of hedonic information systems, MIS Quarterly
interests include social networking sites, human-computer-interaction and business intel-
28 (4) (2004) 695–704.
ligence/analytics. His work has appeared in journals including Communications of the
[77] V. Venkatesh, et al., User acceptance of information technology: toward a unified
ACM, Information and Management, Journal of Computer Information Systems, Informing
view, MIS Quarterly 27 (3) (2003) 425–478.
Science, and Journal of Information Technology Case and Application Research. Chenyan
[78] V. Venkatesh, J.Y.L. Thong, X. Xu, Consumer acceptance and use of information tech-
can be reached by email at chenyan.xu@stockton.edu
nology: extending the unified theory of acceptance and use of technology, MIS
Quarterly 36 (1) (2012) 157–178.
Daniel A. Peak is an associate professor in information technology in the Information
[79] H. Verkasalo, et al., Analysis of users and non-users of smartphone applications,
Technology and Decision Sciences Department, College of Business Administration at the
Telematics and Informatics 27 (3) (2010) 242–255.
University of North Texas. A concert pianist with B.Mus. and M.Mus. degrees in perfor-
[80] Y. Wang, H.-P. Lo, Y. Yang, An integrated framework for service quality, customer
mance, he received his Ph.D. in 1994 from UNT with majors in Information Systems and
value, satisfaction: evidence from China's telecommunication industry, Information
in Finance. He is an editor of the Journal of IT Cases and Applications Research. Dr. Peak
Systems Frontiers 6 (4) (2004) 325–340.
has more than 20 years of IT consulting and planning experience working for executives
[81] J.D. Wells, J.S. Valacich, T.J. Hess, What signal are you sending? How website quality
of Fortune 500 companies, and has won and directed numerous production projects and
influences perceptions of product quality and purchase intentions, MIS Quarterly 35
research grants. He is a member of the Decision Science Institute and Association for Infor-
(2) (2011) A18-373.
mation Systems and has publications in Informing Science, Information and Management,
[82] B.H. Wixom, P.A. Todd, A theoretical integration of user satisfaction and technology
Information Systems Management, Journal of Computer Information Systems, and other
acceptance, Information Systems Research 16 (1) (2005) 85–102.
journals. Dan can be reached by email at daniel.peak@unt.edu
[83] C. Xu, et al., It is not for fun: an examination of social network site usage, Informa-
tion Management 49 (5) (2012) 210–217.
Victor R. Prybutok is a Regents Professor of Decision Sciences in the Information Technol-
[84] Z. Yang, R.T. Peterson, Customer perceived value, satisfaction, and loyalty: the role of
ogy and Decision Sciences Department and Associate Dean for Graduate Programs and Re-
switching costs, Psychology and Marketing 21 (10) (2004) 799–822.
search in the College of Business at the University of North Texas, He received, from Drexel
[85] Y. Yoo, Computing in everyday life: a call for research on experiential computing,
University, his B.S. with High Honors in 1974, a M.S. in Bio-Mathematics in 1976, a M.S. in
MIS Quarterly 34 (2) (2010) 213–231.
Environmental Health in 1980, and a Ph.D. in Environmental Analysis and Applied Statis-
[86] V.Y. Yoon, R.E. Hostler, Z. Guo, T. Guimaraes, Assessing the moderating effect of con-
tics in 1984. Dr. Prybutok is an American Society for Quality certified quality engineer, cer-
sumer product knowledge and online shopping experience on using recommenda-
tified quality auditor, certified manager of quality / organizational excellence, and an
tion agents for customer loyalty, Decision Support Systems 55 (4) (2013) 883–893.
accredited professional statistician (PSTAT®) by the American Statistical Association. Dr.
[87] V. Zeithaml, L. Berry, A. Parasuraman, The nature and determinants of customer ex-
Prybutok has authored over 160 journal articles, sever book chapters, and more than
pectations of service, Journal of the Academy of Marketing Science 21 (1) (1993)
180 conference presentations in information systems measurement, quality control, risk
1–12.
assessment, and applied statistics. Victor can be reached by email at victor.prybutok@
[88] V. Zeithaml, L. Berry, A. Parasuraman, The behavioral consequences of service qual-
unt.edu
ity, Journal of Marketing 60 (1996) 31–46.