Master_Thesis_Kasper_Urbrand_Nielsen_And_Morten_Riise_Jensen_2015

I
An Empirical Analysis of Mobile Commerce Continuance Intention
- A Moderated Mediation Approach
Master’s thesis
MSc in Economics, Business Administration and Marketing
Business and Social Sciences, Aarhus University
Number of characters (w/o spaces): 182.061
Number of illustrations: 23 (18.400)
Number of pages: 91
Authors:
Kasper Urbrand Nielsen
Study no.: KN92599
Exam no.: 410758
Morten Riise Jensen
Study no.: 20119084
Exam no.: 513707
Supervisor:
Athanasios Krystallis Krontalis, PhD
Department of Business Administration, Aarhus University
Aarhus, 1st of August 2015,
Kasper Urbrand Nielsen Morten Riise Jensen
_____________________________ _____________________________

II
Acknowledgements
Initially, we want to show our great appreciation with our supervisor, Athanasios Krystallis
Krontalis, who in this entire process was of great assistance, and was able to push us beyond
our comfort zone. Also, we want to express our gratitude to our respective families for their
constant support and understandings throughout this entire process. Finally, we’d like to
thank all the participants who willingly helped us collect our data through our
questionnaires. Because of the high response rate, we were able to donate 622 DKK to The
Danish Cancer Society (Kræftens Bekæmpelse)1.
1
Appendix 17

III
Abstract
The features of smartphones increase continuously, and operates, to a higher degree,
as personal assistants that affects nearly every aspect of one’s every-day life. Therefore, in
a world where people feel more stressed, it thus seemed inevitable that smartphones
became a popular option for both business owners and consumers to connect. However,
one third of Danish m-commerce users have cancelled a purchase in an m-commerce
environment due to unsatisfactory experiences, whereas many don’t return. Therefore, the
purpose of this thesis was to determine how m-vendors could improve their strategies in a
way that could increase the likelihood of maintaining existing m-commerce users. Thus, the
study set out to investigate the mechanisms influencing the relationship between prior
experiences and the intention to continue using m-commerce in the future.
ECM-IS posits that prior experiences are indeed affected by post-usage expectations
toward future usage, why this thesis aims to investigate this fact. Therefore, drawing on the
ECM-IS, a conceptual model was established with the extensions of cognitive belief
constructs of the TAM, as well as trust and flow, with the intent to identify the underlying
mechanisms influencing this relation by the use of mediation techniques. Furthermore,
acknowledging the differences in users’ perceptions, the thesis finally analyzed the
conditions of these proposed mediation effects established. Findings were based on valid
responses collected from Danish m-commerce users through quantitative surveys with 187
and 125 responses respectively, using non-probability sampling techniques.
Findings were that prior experiences had a large significant impact on satisfaction,
though this effect was partially mediated by the post-usage expectations of perceived ease
of use, perceived usefulness, trust and flow. Furthermore, these post-usage expectations
were highly influencing users’ continuance intention to use m-commerce. However, the
effects of perceived ease of use and trust were fully mediated by satisfaction, meaning that
m-vendors must be able to fully satisfy their users to yield the effects of these cognitive
beliefs. Perceived usefulness and flow turned out partially mediated by satisfaction,
meaning that the effects would diminish, but not vanish, if users are not satisfied. In
addition, the mediation effect caused by satisfaction between flow and continuance
intention was moderated by users’ tendency of impulsive behaviours. Evidently, highly
impulsive users are driven by sudden urges and current stimulus, whereas less impulsive
users use prior experiences as heuristics for future behaviour, why the necessity of a
satisfactory experience is relatively more important for their intention to reuse the system.
Keywords: M-Commerce, ECT, TAM, Trust, Flow, Impulsiveness, Self-Efficacy, Continuance
Intention, Confirmation, Satisfaction, Moderated Mediation, Mediation.

IV
CONTENTS OF THE THESIS
- CHAPTER I - INTRODUCTION...............................................................................9
1.1. INTRODUCTION.....................................................................................................................10
1.1.2. PROBLEM STATEMENT .................................................................................................................. 13
1.1.3. DELIMITATIONS ........................................................................................................................... 15
1.1.4. STRUCTURE................................................................................................................................. 15
- CHAPTER II - THEORETICAL FRAMEWORK .........................................................17
2.1. MAJOR RESEARCH MODELS.....................................................................................................18
2.1.1. EXPECTANCY CONFIRMATION THEORY (ECT).................................................................................... 19
2.1.2. EXPECTANCY CONFIRMATION MODEL – INFORMATION SYSTEMS (ECM-IS) .......................................... 20
2.2.3. TECHNOLOGY ACCEPTANCE MODEL (TAM) ..................................................................................... 22
2.2. MODEL EXTENSIONS ..............................................................................................................24
2.2.1. TRUST ........................................................................................................................................ 25
2.2.2. FLOW......................................................................................................................................... 28
2.3. THE MODERATING ROLE OF PERSONAL TRAITS .............................................................................32
2.3.1. SELF-EFFICACY............................................................................................................................. 32
2.3.2. IMPULSIVENESS............................................................................................................................ 35
- CHAPTER III - CONCEPTUAL MODEL & HYPOTHESES..........................................38
3.1. HYPOTHESIS DEVELOPMENT.....................................................................................................39
3.1.1. STUDY ONE................................................................................................................................. 39
3.1.1.2. Creating Satisfaction through Parallel Mediation........................................................... 39
3.1.1.2. TAM................................................................................................................................. 40
3.1.1.3. Trust ................................................................................................................................ 42
3.1.1.4. Flow................................................................................................................................. 44
3.1.2. CREATING CONTINUANCE INTENTION THROUGH SINGLE MEDIATION.................................................... 46
3.1.2.1. TAM................................................................................................................................. 47

V
3.1.2.2. Trust ................................................................................................................................ 48
3.1.2.3. Flow................................................................................................................................. 50
3.2.1. STUDY TWO ................................................................................................................................ 52
3.2.1.1. Moderating Role of Self-Efficacy..................................................................................... 52
3.2.1.2. Moderating Role of Impulsiveness.................................................................................. 53
- CHAPTER IV – METHODOLOGY .........................................................................56
4.1. RESEARCH DESIGN.................................................................................................................57
4.2. INSTRUMENT DEVELOPMENT ...................................................................................................57
4.2.1. STUDY ONE................................................................................................................................. 57
4.2.2. STUDY TWO ................................................................................................................................ 59
4.3. DATA COLLECTION PROCEDURE ................................................................................................60
4.3.1. Pilot Test............................................................................................................................. 60
4.3.2. STUDY ONE................................................................................................................................. 61
4.3.3. STUDY TWO ................................................................................................................................ 62
- CHAPTER V - DATA ANALYSIS & RESULTS..........................................................63
5.1. DATA ANALYSIS INSTRUMENTS.................................................................................................64
5.2. SAMPLE CHARACTERISTICS.......................................................................................................68
5.3. REGRESSION ASSUMPTIONS.....................................................................................................71
5.3.1. RELIABILITY ................................................................................................................................. 71
5.3.2. VALIDITY..................................................................................................................................... 73
5.3.3. NORMAL DISTRIBUTIONS OF RESIDUALS........................................................................................... 74
5.3.4. HOMOSCEDASTICITY..................................................................................................................... 75
5.3.5. INDEPENDENCE OF ERRORS............................................................................................................ 76
5.4. MODEL FIT ..........................................................................................................................77
5.4.1. STUDY ONE................................................................................................................................. 77
5.4.2. STUDY TWO ................................................................................................................................ 78
5.5. HYPOTHESIS RESULTS.............................................................................................................79
5.5.1. STUDY ONE................................................................................................................................. 79
5.5.1.1. Creating Satisfaction ....................................................................................................... 80

VI
5.5.1.2. Creating Continuance Intention...................................................................................... 81
5.5.2. STUDY TWO ................................................................................................................................ 84
5.5.2.1. Moderating Role of Personal Traits ................................................................................ 85
5.5.3 SUMMARY OF HYPOTHESES TESTING ................................................................................................ 89
- CHAPTER VI - RECAPITULATION ........................................................................90
6.1. DISCUSSION.........................................................................................................................91
6.1.1. ASSESSING RQ1 .......................................................................................................................... 91
6.1.2. ASSESSING RQ2 .......................................................................................................................... 95
6.1.3. ASSESSING RQ3 .......................................................................................................................... 98
6.1.4. ASSESSING RQ4 .......................................................................................................................... 99
6.2. IMPLICATIONS ....................................................................................................................100
6.2.1. MANAGERIAL IMPLICATIONS........................................................................................................ 100
6.2.2. LITERATURE IMPLICATIONS .......................................................................................................... 103
6.3. CONCLUSIVE REMARKS .........................................................................................................104
6.4. LIMITATIONS......................................................................................................................105
6.5. FURTHER RESEARCH.............................................................................................................106
REFERENCES .............................................................................................................................108

VII
List of Figures
Figure 1.1 Thesis Structure
Figure 2.1 Expectancy Confirmation Theory
Figure 2.2 Expectancy Confirmation Model – Information Systems
Figure 2.3 Technology Acceptance Model
Figure 3.1 The Conceptual Model of Study One
Figure 3.2 The Conceptual Model of Study Two
Figure 5.1 The Conceptual Model – Results of Study One
Figure 5.2 The Conceptual Model – Preliminary Results of Study Two
Figure 5.3 Moderating Effect of Impulsiveness
Figure 5.4 Moderating Effect of Self-Efficacy
List of Tables
Table 2.1 Constructs and their Origin
Table 3.1 Hypotheses
Table 4.1 Origins of Measurement Items
Table 5.1 Demography Analysis
Table 5.2 Share of M-Commerce Activities
Table 5.3 Reliability and Validity Analysis
Table 5.4 Model Fit
Table 5.5 Mediation Analysis of Study One – Part I
Table 5.6 Mediation Analysis of Study One – Part II
Table 5.7 Mediation Effects Assessment – Study One
Table 5.8 Multiple Regression Results – Study Two
Table 5.9 Conditional Indirect Effect at Different Levels of Impulsiveness
Table 5.10 Summary of Hypothesis Testing

VIII
List of Appendices
Appendix 1 Questionnaire Items
Appendix 2 Independent t-test for Online and Offline Respondents
Appendix 3 Independent t-test for Early and Late Respondents
Appendix 4 Demography Analysis – Study One
Appendix 5 Demography Analysis – Study Two
Appendix 6 Scale Reliabilities – Study One
Appendix 7 Scale Reliabilities – Study Two
Appendix 8 Normality of Residuals for Satisfaction – Study One
Appendix 9 Normality of Residuals for Continuance Intention – Study One
Appendix 10 Normality of Residuals for Continuance Intention – Study Two
Appendix 11 Homoscedasticity Analysis
Appendix 12 Homoscedasticity-Consistent Regression Results
Appendix 13 Independence of Errors – Study One
Appendix 14 Independence of Errors – Study Two
Appendix 15 Questionnaire – Study One
Appendix 16 Questionnaire – Study Two
Appendix 17 Donation for The Danish Cancer Society (Kræftens Bekæmpelse)

Chapter I – Introduction
Page 9 of 155
- Chapter I -
INTRODUCTION
”There is nothing more difficult for a truly creative painter than to paint a rose, because
before he can do so he has first to forget all the roses that were ever painted.”
- Henri Matisse

Page 10 of 155
1.1. Introduction
In the recent years there has been an evident aggressive growth in mobile device users
and 3G/4G mobile internet subscriptions sold, which is consequently reflected in an
exponential growing market share of mobile commerce (m-commerce). Indeed, the
concept has increased in popularity to a degree, where experts predict a bright future for
the concept. A report from Digi-capital (2014) estimate a nearly 300% increase to a market
share on 516 billion dollars in 2017. This trend is also apparent in Denmark. From 2012 to
2014 the share of Danish households having a smartphone has risen from 50% to 73% (Dst,
2015), along with an increase in mobile wireless internet subscriptions by 24.5% from 3.2
million subscriptions in 2013 to 4 million in 2014, while also the average amount of internet
data used per subscription has increased from 3.4 GB in 2012 to 10 GB in 2014
(Erhvervsstyrelsen, 2014). Realizing the potential in Danish m-commerce, the Danish
telecommunication industry followed up this diffusion by investing intensively in improving
the telecommunication network with an average investment rate on 19.2% between 2008
and 2012, compared to an overall investment rate on 12.8% for Europe (Erhvervsstyrelsen,
2013). In addition, in the effort of improving the Danish digital infrastructure, the Danish
government has supported the telecommunication industry with several initiatives (Emv,
2015), resulting in superior network coverage and price levels compared to international
standards.
As the share of mobile internet users has increased, so has the share of users who have
conducted a purchase using a mobile internet connection. Indeed, from 2012 to 2014, the
share of Danish consumers who have purchased goods or services through a mobile device
(tablet or smartphone) increased from 19% to 33% (DIBS, 2015). And though the
penetration rate of m-commerce has yet to reach the same level as those of Asia or the
U.S., the business opportunities and values of Danish m-commerce are still projected to
experience a significant increase in the coming years, due to the favorable conditions in the
Danish digital infrastructure, and the increasing saliency in socio-demographic factors
among generation Y (DIBS, 2015). The shopping aspect of m-commerce can be described as
“any monetary transactions related to purchases of goods or services through internet

Page 11 of 155
enabled mobile phones or over the mobile wireless telecommunication network” (Wong et
al., 2012, p. 25), and though it has many similarities to conventional online shopping, it
delivers unique values through measures not possible for other shopping methods.
One of these measures is that m-commerce breaks geographical boundaries, and
empowers users with ubiquity and immediacy, allowing them to search for information and
purchase products or services from anywhere at any time (Tiwari and Buse, 2007). Online
stores therefore collide with offline stores in real time, as consumers are equipped with the
unique opportunity to compare products and prices from multiple sources directly, while
visiting physical stores (Mahatanankoon et al., 2005), which further allows users to make
more informed decisions. Moreover, as opposed to regular computers, mobile devices are
designed to be “always on” and in constant connection with the internet, which allows for
convenient and rapid access to online stores. Additionally, the built-in GPS feature enables
mobile-vendors (m-vendors) to distribute special product offerings based on various
information not accessible by other shopping channels. This be the physical location of the
user (Tiwari and Buse, 2007), which grants an opportunity for m-vendors to customize
marketing and offers based on users’ online check ins on social medias, announced
participation in events etc. (Mahatanankoon et al., 2005). However, though stopping
through mobile devices seems promising, it is not without downsides.
M-vendors who are currently delivering m-commerce services are generally suffering
from low profits, shallow user bases and severe problems with high discontinue rates (Hung
et al., 2012; Lu, 2014), since m-shoppers are volatile and may not return, once they leave
(Chong, 2013). This is problematic, due to the technological disparities between mobile
devices and computers that force businesses to invest significant resources to develop
software to comply with a mobile platform (Chong, 2013). This adjustment, in turn, seems
important to retain users. In fact, every third Danish mobile shopper have cancelled a
purchase process initiated through their mobile device within the past six months, due to
an unsatisfying experience with the shopping channel (DIBS, 2015). These unsatisfactory
experiences are often attributable to the fact that mobile devices have small screens,
inconvenient input, low multimedia processing power and poor connectivity (Lee, 2014). In

Page 12 of 155
addition, m-shoppers operate in an online environment, which prevents them from
assessing reliable indications of actual product quality, while the lack of face-to-face contact
and wireless electronic nature of the operation makes the purchase subjective to concerns
about money, and personal information being distributed to third parties without their
consents. Thus, both greater mistrust and risk is likely to present itself within the context of
m-commerce (San‐Martin and López‐Catalán, 2013). Moreover, previous studies estimate
that the costs of attracting new users are five times the costs of retaining existing ones
(Schefter and Reichheld, 2000). It therefore seems vital that m-vendors manage to reduce
these negative measures, thereby reducing customer churn (Chong, 2013; Luqman et al.,
2014). Also, as system users are independent individuals that are likely to have different
perceptions and orientations, these may induce considerable implications for their
respective behaviours. In fact, previous research have clearly demonstrated that users’
decisions to accept a system is not based on the same set of criteria (Cheng, 2014; Hsu et
al., 2012). For instance, the process of shopping via a mobile device, and maneuvering
mobile applications, often requires a certain level of user skills and technological
comprehension. This may, in the eyes of some users, be a somewhat challenging task, why
these may find themselves constrained by their beliefs in their own abilities. At the same
time, m-vendors have traditionally continuously focused marketing activities, with the
intent to persuade users to conduct unplanned purchases (San‐Martin and López‐Catalán,
2013). However, certain users require a more comprehensive assessment of the market,
why failure to evaluate product alternatives, price differences etc., may easily lead to an
unsatisfactory experience (Rook and Fisher, 1995). Therefore, the intentions to continue
using m-commerce services might likely be dependent upon personal predispositions. This
thus speaks to the fact that m-vendors may have to customize their marketing strategies to
better fulfill individual users’ needs.
Past research have mainly been focusing on investigating salient factors that facilitate
users’ initial intention to adopt m-commerce, leaving research on continuance intention
much more limited (Luqman et al., 2014). This is evident, despite researchers for long have
called for attention to this matter (Choi et al., 2008). The primary issue in the exiting
research has been to assess, whether the determinants recognized in adoption studies

Page 13 of 155
retain their significance in also explaining post-adoption behaviours. Researchers have, in
this relation, turned for assistance in the Expectancy Confirmation Theory (ECT), and found
evidence indicating that by providing an experience that lives up to users’ expectations will
increase their satisfaction, which in turn increases their intention to return (Chong, 2013;
Hung et al., 2012; Lee, 2014). However, as research on post-adoption behaviours are still in
the introduction phase (Groß, 2015), insufficient information is available to make accurate
inferences about the specific nature of expectations that m-vendors need to accommodate
in order to satisfy their users. Moreover, since current research have been focusing on
identifying antecedents of continuance intention, little effort has been put into investigating
how the impact of antecedents is determined by providing an overall satisfactory
experience. Furthermore, in a shopping context, some researchers have investigated the
moderating role of culture (Zhang et al., 2012), innovativeness (Yang, 2012), and
psychographics (Molina-Castillo et al., 2008) in consumers intention to adopt m-commerce.
However, no studies have yet examined the intervening effect of consumers’ adherence to
buy impulsively, as well as the level of mobile self-efficacy among determinants driving
users’ continuance intention.
A severe limitation also worth mentioning in current literature is that research amongst
European consumers are very limited. The majority of studies published within the area is
conducted in either East Asia or the U.S, why generalization of results is often constrained
by cultural barriers (Groß, 2015; Luqman et al., 2014). In addition, m-commerce is an area
in constant motion. Mobile technologies, as well as telecommunication networks are
evolving rapidly, and as consumers gain more and more experience, new perceptions and
needs may quickly emerge (Pappas et al., 2014). Further research in this field is therefore
needed.
1.1.2. Problem Statement
As evident from the preceding introduction, the improvement in telecommunications
networks, as well as in mobile technologies, have ramped up the sales of 3G/4G enabled
mobile devices and users, are becoming more willing to accept these devices as commercial
tools. Now, m-vendors need to ensure market growth by understanding how they can retain

Page 14 of 155
their users. However, current research on m-commerce post-adoption behaviours is still in
its infancy (Groß, 2015; Luqman et al., 2014) and are lacking a better understanding of the
mechanisms that cause users to continue using m-commerce systems. Gaining insight into
processes that stimulate the individual user’s intention to return will inevitable empower
m-vendors to recognize and deliver the needed downstream activities, and increase the
possibility of receiving a satisfying return on investment. The main purpose of this study will
therefore be to understand:
What are the underlying mechanisms causing users to continue using m-commerce,
and are these dependent upon users’ personal traits?
In order to investigate causalities that influence the individual users’ post-adoption
behaviour, this study draws on ECT (Oliver, 1980) and empirically tests three research
models that focus on post-adoption beliefs (Bhattacherjee, 2001). The purpose is to
understand how a fulfillment of users’ expectations can increase their satisfaction with m-
commerce, and to understand the mediating role of satisfaction in driving users’
continuance intention. Furthermore, the study seeks to reveal if the mediating role of
satisfaction is consistent for users with different personal traits. As such, the research
models are developed to test the following research questions:
RQ1: By what extent do perceived ease of use, perceived usefulness, trust and flow explain
the relationship between user expectancy confirmation and user satisfaction?
RQ2: By what extent does user satisfaction explain the effects of perceived ease of use,
perceived usefulness, trust and flow on continuance intention?
RQ3: Assuming a mediational effect between flow and continuance intention through user
satisfaction, by what extent is this effect dependent on users’ tendency to buy impulsively?
RQ4: Assuming a mediational effect between flow and continuance intention through user
satisfaction, by what extent is this effect dependent on users’ degree of self-efficacy?

Page 15 of 155
1.1.3. Delimitations
In the area of m-commerce there are various definitions, including different devices, as
well as several unrelated activities that both include monetary and non-monetary activities.
Most users can, however, be divided into two categories: mobile users, with the focus on
communication purposes (i.e. using the device for gaming, texting, calling etc.) and mobile
shoppers, with a distinct focus on purchasing (buying products or services) (Hung et al.,
2012). This study focuses on the latter, and investigates the aspects of m-commerce that
includes the use of mobile phone applications or mobile phone browsers (e.g. Safari,
Firefox, Chrome) to purchase products or services (tickets, bets, travels, physical products,
software, subscriptions) from electronic retail stores. This therefore excludes activities, such
as browsing for information or scanning QR codes. By the same token, smartphone users
are through proximity payment technology (RFID, NFC) essentially offered the opportunity
to use their smartphones as mobile wallets to conduct on the spot payments, by swiping
their smartphones over a terminal (Zhou, 2013a). The same option that is available via
applications such as “Mobilepay” or “Swipe”. This naturally spawns a different usage
context, hence the aspect of using the smartphone as a payment method will not be
addressed. Also, it does not deal with the costs of having access to online services, i.e.
internet subscription fees. Furthermore, m-commerce literature have failed to have an
explicit focus on smartphones and tablets, despite the fact that statistics show that the
majority of m-commerce activities are conducted through these devices (DIBS, 2015;
eMarketer, 2013). Thus, little is essentially known about the influence of these devices on
m-commerce behaviours (Groß, 2015). In an effort to shed some light on the area, this study
will only focus on smartphones. Tablets are, due to their sizes, not considered truly mobile,
and are therefore excluded.
1.1.4. Structure
This study is split into six chapters (figure 1.1). The first chapter provides an introduction
along with the problem statement. Chapter two discusses the theoretical foundation as well
as prior research findings. Chapter three covers hypothesis development, and presents the
conceptual models. Chapter four discusses instrument development and data collection

Page 16 of 155
procedures. Chapter five describes the statistical techniques and methods used to test the
conceptual models, followed by analyses of the data and presentation of results. Chapter
six covers a discussion of results, implications, limitations of the study and suggestions for
future research.
FIGURE 1.1. THESIS STRUCTURE
SOURCE: OWN MAKING

Chapter II – Theoretical Framework
Page 17 of 155
- Chapter II -
THEORETICAL FRAMEWORK
“You can’t sell anything, if you can’t tell anything.” - Beth Comstock

Page 18 of 155
This chapter comprises three parts. First, an examination of relevant theory and
frameworks within m-commerce, followed by model extensions that could possibly
enhance the explanatory value for these models and finally, the last part of the chapter
reviews findings within personal traits in the context of moderation.
2.1. Major Research Models
Several different research models have been used when studying consumer behaviour.
And although there are a differences as to which products consumers’ purchase and from
the channels, from where they purchase, there is generally a uniform approach. For
instance Fishbein and Ajzen's (1975) Theory of Reasoned Action (TRA) and Ajzen's (1985)
further evolved Theory of Planned Behaviour (TPB) that has been applied to investigate
consumer behaviour among consumers shopping in both stationary shops, (Irianto, 2015),
consumers shopping through electronic commerce (e-commerce) (George, 2004; Lim and
Dubinsky, 2005) and m-commerce (Kim, 2010; Lin and Wang, 2006). Still, researchers have
persistently tried to develop extensions and entire models trying to accommodate the
aspects of more specific consumer situations; as for instance the Unified Theory of
Acceptance and Use of Technology (UTAUT), Decomposed Theory of Planned Behaviour
(DTPB) or the Technology Acceptance Model (TAM) that, based on the characteristics of
TRA and TPB, has been customized to accommodate the alleged differences between
consumers considering buying products in general, and consumers considering adopting a
certain technology (Chuttur, 2009; Davis, 1989). Given its results in many studies, TAM has
thus heavily been applied to investigate consumers’ intention to adopt m-commerce (Lee,
2014; Thong et al., 2006; Zhou, 2014a). In the context of maintaining customers, the range
of theories is more limited. However a commonly applied theory is the ECT (E.g. Chong,
2013; Kim, 2010; Lin and Wang, 2006).

Page 19 of 155
2.1.1. Expectancy Confirmation Theory (ECT)
Measuring the intention to purchase a product, or to use a system, is a very important
aspect in conquering a market. However, an equally important aspect, if not more so, is the
aspect of measuring a consumer’s willingness to repeat that same behaviour (Thong et al.,
2006). Note that consumers are, everything else being equal, more valuable for companies
if they re-purchase rather than abandoning the company after first buy (Anderson and
Sullivan, 1993). Hence, the more frequently consumers buy, the more profit suppliers will
yield from these. Given the importance, many researchers have investigated the subject in
depth. Oliver (1980), however, was the first to pioneer with an esteemed framework that is
still highly applied in contemporary research (Bhattacherjee, 2001; Thong et al., 2006). A
prerequisite for this model is that the consumer has already purchased the good or service
before (Oliver, 1980). The framework then proposes that the consumer’s intention to
repurchase is a comprehensive composition of (1) the perceived performance of the
product or service, (2) the (/dis)confirmation of the initial expectations prior to the purchase
and (3) the level of satisfaction (Bhattacherjee, 2001; Oliver, 1980).
Initially, a consumer develops expectations toward a certain good. If these expectations
are sufficiently high, they will eventually likely lead to a purchase (figure 2.1). Subsequently,
consumers will then evaluate their perception of the good’s performance, vis-à-vis their
initial expectations, after which this difference will lead to a (/dis)confirmation of their initial
expectations. If the initial expectations are confirmed, the probability of repurchasing
intention are said to increase significantly, due to a higher degree of satisfaction. (Oliver,
1980).

Page 20 of 155
FIGURE 2.1. EXPECTANCY CONFIRMATION THEORY
SOURCE: OWN MAKING, BASED ON OLIVER (1980)
However, a heavily debated paradox in the ECT, is that the model is not constructed as
perpetual as illustrated. The model thus ignores the antecedents of enhanced experience
with the product and other cognitive processes, which might lead to possible changes in
future expectations (Bhattacherjee, 2001). Consequently, the initial expectations toward a
certain product will, according to the ECT, remain the same regardless of how many times
the consumer will repurchase. This might induce a problem, as the model might generate
deceptive results if not adjusted for different changes in the market, new innovations,
consumer perceptions etc. (Bhattacherjee, 2001; Lee, 2014).
2.1.2. Expectancy Confirmation Model – Information Systems (ECM-IS)
To offset some of the aforementioned shortcomings in the ECT, Bhattacherjee (2001)
realized the need to expand the model with a link of cognition to better explain the model
within an IS context. In his modified framework (figure 2.2), he added the cognition of post-
consumption expectation, represented by perceived usefulness, trying to accommodate for
users’ needs to reevaluate expectations toward the IS, since he claims that post
expectations are essential in the case of system usage, where expectations can change over
time. This setting further complies with the definition of expectation in the ECT, holding that
expectations equal the sum of a user’s beliefs, since perceived usefulness is a cognitive
belief salient to IS usage (Bhattacherjee, 2001). However, an important change to notice is
that while the ECT investigates both pre- and post-consumptions variables, the ECM-IS only
focuses on post-consumption variables, since the effects of the relative match between pre-
consumption expectations and perceived performance is already accounted for in the

Page 21 of 155
confirmation and satisfaction construct. This thus alters the ECT paradigm, since ECM-IS
posits that users using IS will keep updating their expectations as they get more familiar
with the system (Hong et al., 2006). In essence, ECM-IS drives on the assumption that
perceived performance is fully mediated through confirmation, and that post-consumption
expectations are modified through direct experience with the IS, which in turn functions as
vital predictors of users’ satisfaction formation and continuance usage intention
(Bhattacherjee, 2001).
This rationale has predominantly been accepted in contemporary research, and has thus
been adopted in later studies and further expanded with other cognitive beliefs. So, though
the findings within m-commerce are rather variegated, due to the various definitions of the
term, it seems there are congruent findings from the model to explain continuance
intention. E.g. Kim (2010), who proposed a variation of the model to explain users’
continuance intention to use mobile data services, Akter et al. (2013) finding the relation
within mobile applications. Or Hong et al. (2006), whose study found that the very same
model significantly explained the users’ continuance intention within mobile internet in
general. These studies represent different aspects of m-commerce, and thereby indicate
congruity. Thus, it seems that the model is applicable for studies aiming to understand the
continuance intention within m-commerce in general regardless of its nature, since the
model holds that positive confirmation, corresponding cognitions and hence also
satisfaction, will impact the continuance intention (Bhattacherjee, 2001).
FIGURE 2.2. EXPECTANCY CONFIRMATION MODEL - INFORMATION SYSTEMS
SOURCE: OWN MAKING, BASED ON BHATTACHERJEE (2001)

Page 22 of 155
2.2.3. Technology Acceptance Model (TAM)
In the TRA framework from 1975, Fishbein and Ajzen proposed that the actual behaviour
of a given person could be explained by examining the person’s intention and the reasons
for it. Thus, the stronger the intention to a certain behaviour, the more likely the person
will actually behave as intended (Chuttur, 2009). As an extension to the TRA, Davis (1989)
developed the TAM (figure 2.3) to better understand users’ behavioural intention to adopt
an IS (Davis et al., 1989). The model is thus, as opposed to the TRA and TPB, a much more
customized framework (Bhattacherjee, 2001; Thong et al., 2006), why the model has shown
highly significant results throughout the years, and therefore is widely credited as well as
highly applied in contemporary studies (Chuttur, 2009).
In the model, Davis (1989) suggested that there are two variables that best enhance the
understandings of a user’s attitudes towards the intention to adopt an IT. (1) Perceived
usefulness and (2) perceived ease of use. Perceived usefulness has been widely debated,
and was first introduced in a factor analysis by Schultz and Slevin (1975), stating that a
system that does not enhance a user’s performance in his/her job delivery, is not likely to
be well received by the user (Schultz and Slevin, 1975, cited in Longe et al., 2010). As a result
of this analysis, Davis defined perceived usefulness as “the degree to which a person believes
that using a particular system would enhance his or her job performance” (Davis, 1989, p.
320), meaning that the system must deliver enhanced usability. Therefore, perceived
usefulness has heavily been associated with the positive influence on users’ overall
satisfaction with a system, given that perceived usefulness represents a behavioural reward
(Davis, 1989), and thereby an extrinsic motivation related to the confirmation of initial
expectations (Hung et al., 2012; Kim, 2010).
Davis refers to the definition of “ease”, as the freedom from difficulty or great effort and
thus perceived ease of use is defined as “the degree to which a person believes that using a
particular system would be free of effort” (Davis, 1989, p. 320). Thus, the theory builds on
the premise that achieving a desired result easier or faster, possibly enhances the

Page 23 of 155
satisfaction with a system, and in turn, also the likelihood of the customer switching from
one system to another or to adopt a system in general (Bhattacherjee, 2001; Hong et al.,
2006; Thong et al., 2006).
FIGURE 2.3. TECHNOLOGY ACCEPTANCE MODEL
SOURCE: OWN MAKING, BASED ON DAVIS (1989)
However, though the model is considered valid, and has been applied extensively in
contemporary research, researchers have equally criticized the model for its shortcomings.
E.g., Legris et al. (2003) noted that a severe problem with the TAM is that analyses are
conducted with self-reported use data, meaning that the model is subject to bias as
opposed to objective actual-use data. This argument came in the aftermath of a study in La
Presse Montréal in 2000, where researchers had observed that only 67% of users of a public
restroom in New Orleans actually washed their hands after using the toilet. However, in
comparison, the same researchers conducted a survey among 1201 Americans, where 95%
answered that they always wash their hands after using the toilet (La Presse Montréal,
2000; cited in Legris et al., 2003). Also, researchers have pointed out the relatively low
explained variance of the model in a general context, indicating that there are other
variables of significant influence, why the model should be extended with other variables
(Legris et al., 2003; Yang and Yoo, 2004).

Page 24 of 155
2.2. Model Extensions
Hong et al. (2006) extended the applicability of ECM-IS, when studying consumers
continuance intention to use mobile internet. Their initial purpose was, however, to
investigate the explanatory power of ECM-IS by running three separate models; TAM, ECM-
IS and an extended version of ECM-IS that also includes the construct of perceived ease of
use. What they found were that the extended version of ECM-IS were almost as robust as
TAM, while having a higher overall explanatory power compared to the original versions of
ECM-IS and TAM. This study therefore choose to adopt the ECM-IS with both construct from
TAM integrated in the framework. Moreover, it is likely that the mechanisms causing users
to continue shopping via their smartphones, are not solely driven by their perceptions of
the interaction being free of effort and useful. Thus, in order to gain a more comprehensive
understanding of the process that stimulate repetitive behaviours, some additional
extensions are needed.
Previous studies have identified several determinants of users continuance intention to
use m-commerce services: Mobile affinity, mobile device experience, demographics,
frequency of mobile use (Bigné et al., 2007), trust, habit (Lin and Wang, 2006), subjective
norms (Kim, 2010), perceived cost, perceived enjoyment (Chong, 2013) and flow (Zhou,
2013a). However, including all the constructs identified by existing literature in one model
would be illogical because of the risk of ‘overfitting’ the model, thus possibly causing
difficulties of isolating the variables’ individual effects. Thus, based on the literature review,
two additional intrinsic motivators are derived. This being trust and flow, as they are
believed to contain a high likelihood of increasing the explanatory power of the research
models. Also, literature have called for further research to acknowledge the importance of
intrinsic motivators in an m-commerce within the area of shopping (Hung et al., 2012). In
addition, the latest study of Zhou (2014a) further encourages the inclusion of flow, as he
finds that flow followed by satisfaction, was the main factor driving users’ continuance
intention to use mobile internet sites. The same result appeared in his earlier work in
relation to the use of mobile payment services (Zhou, 2013a). Furthermore, the consistent

Page 25 of 155
significant influence of trust on m-commerce behaviours has also made it an object of
interest. Numerous studies have regarded trust a critical antecedents of intention to adopt
m-commerce (Gitau and Nzuki, 2014; Tsu Wei et al., 2009; Vasileiadis, 2014). Other research
have further confirmed that despite a repeating interaction with m-commerce, trust would
stay relevant and remain among determining factors (Chong, 2013; Hung et al., 2012; Zhou,
2013a).
2.2.1. Trust
In literature, trust is regarded as a broad concept since it have been accommodated by
a range of different definitions depending on perspective and research context (Gefen et
al., 2003a; McKnight et al., 2002). This study will, however, treat trust as a set of trusting
beliefs hold by the user, and therefore describe trust as the willingness of users to leave
themselves vulnerable to the actions of others, which is based on the expectations toward
the other party’s future behaviour (Mayer et al., 1995). Thus it is the belief that the trusted
party will not take advantage of the situation and will behave in a dependable, ethically and
socially appropriate manner (Gefen et al., 2003a). In this perspective, users’ level of trust
within m-vendors is believed to emerge from their perceptions of specific attributes offered
by the m-vendor, which subsequently influence their attitudes and behavioural intentions.
This approach does also allow for a more rigorous integration with different behavioural
theories such as TAM and ECM, since it essentially follows the same logic as TRA (Fishbein
and Ajzen, 1975), stating that individuals’ beliefs indirectly influence behavioural intentions
through attitude, and are therefore more aligned with the theoretical foundation of these
models.
Mayer et al. (1995) argue that individuals’ level of trust is reflected by their beliefs
concerning the other parties’ benevolence, ability and integrity. Although Mayer et al.
(1995) discuss trust within organizations, their operationalization of trust is also heavily
applied within IS research (McKnight et al., 2002). However, some researchers have also
included the dimension of predictability as a part of trust, while arguing that individuals who

Page 26 of 155
perceive the other person to be predictable in his/her behaviours, may also be more willing
to depend on that person (Gefen and Straub, 2004; McKnight et al., 1998). At the same
time, it seems reasonable to believe that m-vendors who demonstrate predictable and
consistent behaviours, e.g. who always deliver goods or services on time, would be
considered more trustworthy.
According to the definition of Mayer et al. (1995), ability can be defined as the extent to
which the user presumes m-vendors to possess the sufficient knowledge and competencies
in order to fulfil their task. Benevolence is referred to as user caring, motivation to act in
their users’ best interest and their willingness to put their users’ interests above their own.
Integrity is defined as the m-vendors’ distance to any deceptive behaviour and their ability
to keep promises. Finally, predictability is, according to Gefen et al. (2003a), related to the
users’ perception about the m-vendors’ behavioural consistency.
When investigating the saliency of trust within IS research, it becomes evident that
several researchers regard trust as a particularly critical element within online exchanges
(Gefen et al., 2003a; McKnight et al., 2002), as it is suggested that online users generally
stay away from e-vendors, they do not trust (Jarvenpaa et al., 2000; Liu et al., 2005). Some
researchers even argue that facilitating trust is essential for e-vendors to succeed within e-
commerce (Gefen, 2002; Kim et al., 2008). Several studies have noted that the saliency of
trust generally increases in situations where consumers are facing uncertain situations (Lin
et al., 2014; Siegrist et al., 2005). M-commerce is arguably an area that is attached with
many uncertainties. In fact, Vasileiadis (2014) suggests that the inherent nature of m-
commerce with its constant and rapidly evolving state, is associated with a relatively higher
degree of risks perceived by users compared to e-commerce and traditional offline
channels. Specifically, the possibility of tracking users’ location and users’ preferences raises
a major privacy concern that questions the benevolence of m-vendors (Joubert and Van
Belle, 2013). Users may also suffer from a lack of trust in the technology they use. Users’
access m-commerce services via smartphones from wireless connections in different places,
which evokes not only transactions concerns, but also privacy concerns, since users may

Page 27 of 155
feel vulnerable to hackers and malicious software (i.e. viruses, malware etc.) (Ghosh and
Swaminatha, 2001). A commonly known phenomenon is blue-snarfing, where intruders
hack the Bluetooth system in the smartphone, and thereby gain access to personal
information. Also, the wireless 3G/4G connection is often unstable, and users may therefore
be worried about the consequences of a lost connection during a transaction with the m-
vendor (Lim, 2003). Additionally, as similar to e-commerce, m-commerce includes the
process of purchasing products/services in a virtual environment that consequently inhibits
the users from accessing reliable indication on product/service quality. Research have
demonstrated that intangibility is closely correlated with perceived risk (De Ruyter et al.,
2001). By the same token, users may also be concerned about expenses they may endure if
they cancel, or need to return a product (Vasileiadis, 2014). Thus, having favourable
perceptions concerning the m-vendors’ trustworthiness, diminish the importance of
perceived risks (Lin et al., 2014) and increase the willingness to enter a vulnerable position,
despite risks of receiving a negative outcome (Mayer et al., 1995).
Several studies conducted within the m-commerce domain have empirically confirmed
the saliency of trust in influencing users’ satisfaction (Lin and Wang, 2006; San‐Martin and
López‐Catalán, 2013). For example, Lee and Chung (2009) incorporated trust within the
DeLone and McLean’s IS success model, and demonstrated that users’ degree of trust was
a strong predictor of users’ degree of satisfaction with mobile banking services. Chong
(2013) extended the ECM, revealing that trust was a key construct explaining users’
satisfaction with Chinese m-commerce services. Researchers have also found a direct
connection between trust and users’ behavioural intentions. For instance, the exploratory
analysis conducted by Sadi and Noordin (2011) found trust to be an important construct
driving users’ intention to adopt m-commerce in Malaysia. The longitudinal study of Lin et
al. (2014) integrated trust in the ECM-IS and Valence Theory, uncovering that pre- and post-
trust were critical factors affecting intention to use mobile banking, since pre-trust
diminished the saliency of perceived risk and enhanced the degree of perceived benefit,
while post-trust, through a confirmation of trusting beliefs, would further influence future
behaviours. Additional research have found a significant relationship between trust and

Page 28 of 155
continuance intention to use mobile payment services (Zhou, 2013a) and m-health (Akter
et al., 2013). Factors such as system quality, information quality, service quality, perceived
risks and confirmation of expectations have been identified to influence users’ trust in m-
commerce (Akter et al., 2013; Alsajjan, 2014; Lee and Chung, 2009; Vasileiadis, 2014; Zhou,
2013a).
2.2.2. Flow
The origin of flow theory is found in the research papers of Csikszentmihalyi within the
human psychology domain, where he developed this theory by studying and interviewing
individuals that exhibited a high commitment and devotion toward an activity. E.g.
professional chess players playing chess or rock climbers climbing a mountain
(Csikszentmihalyi, 1975). From his results, he conceptualized a particularly and extremely
gratifying state of mind that occurred when an individual participated in an activity with
total immersion, while experiencing a range of different positive characteristics, such as:
loss of self-consciousness, loss of time sense, sense of effortless control of the situation,
total centring of attention or an embracement of the autotelic nature of the activity
(Csikszentmihalyi, 1997, 1975). Achieving such state of mind is what Csikszentmihalyi
described as gaining “flow experience”, or as his subjects verbalized as “being in the flow”
(Csikszentmihalyi, 1975). He defined this phenomena as the “holistic sensation present
when we act with total involvement” (Csikszentmihalyi, 1975 p. 43). Hereby,
Csikszentmihalyi – simply put - names the feeling that occur when we are fully immersed
and engulfed in an activity, which in turn fills us with enjoyment and fulfilment.
The relevancy of flow theory in an m-commerce context emerges from the work of
Hoffman and Novak (1996), as they extended the applicability of the flow construct in order
to study and explain online experiences in a computer-mediated environment. Though
fitted to an online context, the definition proposed by Hoffman and Novak, (1996) still
draws on the fundamentals of the flow construct, as they define online flow experience as:
“The state occurring during network navigation, which is: (1) characterized by a seamless
sequence of responses facilitated by machine interactivity, (2) intrinsically enjoyable, (3)

Page 29 of 155
accompanied by a loss of self-consciousness, and (4) self-reinforcing” (Hoffman and Novak,
1996 p. 57). This means that users could gain a state of mind similar to what
Csikszentmihalyi (1975) identifies by navigating in a network, as for instance web surfing or
browsing a webpage (Hoffman and Novak, 1996). However, in order for users to experience
online flow, Hoffman and Novak (1996), similar to Csikszentmihalyi (1975), propose that
two primary conditions need to be present. First, the users must fully focus their attention
on the interaction to a degree where they filter out any background noise and irrelevant
thoughts and secondly, strike a balance between skills and challenges. Within the
conceptual framework of Hoffman and Novak (1996), the degree of users attention is
viewed as a consequence of content characteristics (interactivity, vividness) and
involvement, whilst the degree of involvement is determined by the navigation process
characteristics (goal-driven, experiential-driven). Additionally, website performance and
prior website experiences have also been suggested to be factors contributing to online
flow experience (Skadberg and Kimmel, 2004). Similar to Csikszentmihalyi (1975), Hoffman
and Novak (1996) emphasize the importance of the ratio between users’ skill levels and the
challenges faced by users when navigating a network. If the users’ skill level surpasses the
challenges they face, they will experience boredom. If the challenges faced by users surpass
their skill level, they will experience anxiety. Only when users possess a high perceived level
of skills and sense of control congruent with an equally high level of perceived severity of
the task at hand that evokes arousal, they will experience online flow (Novak et al., 2000).
The interest in the flow construct is a consequence of its affect. The possibility of
harvesting intrinsic rewards, such as enjoyment and fulfilment when users experience flow
is suggestively correlated with a range of influential factors affecting online behaviours
(Hoffman and Novak, 2009, 1996; Novak et al., 2000; Siekpe, 2005; Skadberg and Kimmel,
2004). For example, Skadberg and Kimmel (2004) demonstrated that flow had a significant
positive influence on users learning abilities when browsing websites, meaning that the
presence of flow would increase the information processed by the user. In the meantime,
Korzaan (2003) revealed that flow was leading to a more exploratory behaviour when
shopping online and therefore also increased time spend on the website. Both studies
conclude that the outcome of flow was significantly related to users’ attitude toward the

Page 30 of 155
activity. Novak et al. (2000) also further validated and empirically tested their 1996
framework by finding flow to be stimulating positive affects (e.g. satisfaction), and argued
that flow experience can mitigate users’ price sensitivity. In essence, the influential role of
flow in an online context builds on the augmentation that elements of the flow construct
are vital precursor for a pleasant and enjoyable online experience (Hoffman and Novak,
2009; Koufaris, 2002; Siekpe, 2005). Researchers propose that the ratio between failure and
success of online marketers are mediated by their ability to facilitate and create exciting
online experiences that promote online flow (Bilgihan et al., 2014; Hoffman and Novak,
1996). Bilgihan et al. (2014) opined that unsatisfactory online experiences are globally
accountable for a substantial loss in revenue. This proposition is well grounded since several
empirical studies have provided evidence that highlight the magnitude of online flow, and
its ability to influence users’ attitudes and behavioural intentions. For instance, some
studies concluded that flow significantly influenced attitude towards online shopping
(Korzaan, 2003), satisfaction with online shopping (Hsu et al., 2012; Rose et al., 2012;
Sharkey et al., 2012), attitude with instant messaging (Lu et al., 2009), attitude towards
website search engines (Chung and Tan, 2004) satisfaction with online financial services
(Lee et al., 2007a; Xin Ding et al., 2010), satisfaction with e-learning systems (Cheng, 2014)
while other research found flow to also stimulate intention to purchase online (Sharkey et
al., 2012; Siekpe, 2005) and intention to re-visit webpage (Koufaris, 2002; Nel et al., 1999;
Siekpe, 2005).
Although research of flow in an m-commerce context is limited, the research that do
exist provide similar results. The studies mainly published by Tao Zhou clarified the
importance of flow in this setting by finding flow to be significantly related to both users’
satisfaction with and continuance intention to use mobile payment system (Zhou, 2013a),
mobile internet sites (Zhou, 2014a, 2013b, 2011), mobile social network services (Gao and
Bai, 2014; Zhou et al., 2010), intention to use mobile TV (Zhou, 2013c) and continuance
intention to use mobile internet sites (Zhou, 2014b). Thus, studies seem to agree that the
intrinsic rewarding state created by experiencing flow can play a significant role in attitude
formation and satisfaction evaluation. Also, as found in the very basics of Csikszentmihalyi
(1975), users experiencing flow will be likely to engage in repetitive behaviours, as they will

Page 31 of 155
be drawn to re-experience such gratifying state produced by flow.
In literature, flow is illustrated as an elusive concept. Researchers seem to agree on the
conceptual definition of flow provided by Csikszentmihalyi (1975). However, the intuitive
translation into a more universal operational definition seems much more complex, why
antecedents, as well as consequences of flow, differ depending on research context and
researcher (Obadă, 2013). Consequently, this have also created diversity in relation to
measurement approaches. For example, Korzaan (2003) measures flow as a direct
unidimensional construct by providing subjects with a narrative description, followed by
three questions. Other research employ a derived unidimensional measurement that
aggregates antecedents of flow into an overall measurement (Skadberg and Kimmel, 2004;
Zhou, 2014b), while Koufaris (2002) approaches flow as a multidimensional construct,
consisting of three separate dimensions. For simplicity reasons, and based on the
recommendation of Hoffman and Novak (2009), approaching flow as a derived
unidimensional construct seems more fitted with the purpose of this study. Hoffman and
Novak (2009) note that a serious disadvantage of this approach is the fact that it smears the
distinctions between consequences and antecedents, meaning that justifying which items
to reflect flow become somewhat more challenging. However, in order to align this study
with the definition provided by Hoffman and Novak (1996), the antecedents chosen to
represent flow are based on the primary antecedents suggested in their framework. This
being perceived control, enjoyment and focused attention. Within an IS context, perceived
control captures the extent of users’ perceived level of control over the their actions and
over the environment, in which they interact (Koufaris, 2002). Perceived enjoyment reflects
users’ level of intrinsic enjoyment or pleasure associated with the interaction, whereas
focused attention reflects users’ immersion and measures their ability to focus their
attention on the interaction at hand (Koufaris, 2002; Zhou, 2014b). These factors are also
congruent with frequently used measures of flow in m-commerce research (Zhou, 2014a,
2013a)

Page 32 of 155
2.3. The Moderating Role of Personal Traits
Research have repeatedly shown that individual characteristics should not be ignored
when trying to gain a deeper understanding of underlying factors that drives users’
intention to use technologies (Khedhaouria et al., 2014; Venkatesh et al., 2003). Whether it
be a product, service or shopping method, individuals tend to differ in the amount of value
they place on the attributes related to the giving object or activity, which ultimately affects
their behavioural intentions. Thus, what is an important attribute for one individual is not
necessarily important for another. Research within IS have often accused factors such as
age, gender and experience for playing a noticeable part in users’ distribution of value
between attributes related to e-commerce (Pappas et al., 2014; Venkatesh et al., 2003).
However, recent research in m-commerce suggest that also personal traits, such as users
adherence to buy impulsively (San‐Martin and López‐Catalán, 2013) and users’ level of self-
efficacy (Yang, 2012) play an intervening role in users’ perception of m-commerce services.
Users’ degree of impulsiveness and self-efficacy will therefore be assessed.
2.3.1. Self-Efficacy
In a social cognitive learning perspective, human functioning is viewed as product of a
dynamic interrelationship among behaviours, environment and personal factors (cognitive,
affective and biological events) (Bandura, 1986). A concept that is coined reciprocal
determinism, meaning that humans are able to interpret the result of their own behaviours
that may influence their surrounding environment and cognition that subsequently
facilitate future behaviours (Wood and Bandura, 1989). It takes an inside out and outside in
approach to learning and behavioural changes, since it runs on the idea that learning and
change in behaviour can be extracted purely from expectancies (e.g. beliefs) and through
vicarious learning (Bandura, 2001; Bandura et al., 1961), as opposed to traditional learning
theories (e.g. classical conditioning theory). Social cognitive theory is founded on a human
agency perspective (Bandura, 1989). In this sense, in order for individuals to function
successfully within the reciprocal framework, they exercise certain capabilities, which allow

Page 33 of 155
them to contribute to their own motivation, behaviour and development course (Bandura,
1986). Such capabilities are referred to as: symbolize, forethought, vicarious learning, self-
regulation and self-reflection, whereas the self-reflection capability is regarded as a critical
factor in the social cognitive theory, since individuals, through self-reflection, analyse their
own cognition and self-beliefs (Bandura, 1999). Within the self-reflection construct lays the
concept of self-efficacy, which is also a precursor for self-regulation that operates in the
very foundation of the human agency perspective (Bandura, 1999, 1993).
Self-efficacy can, within a computer usage context, be defined as users’ judgment about
his or her capability to undertake a behaviour with confidence in successfully achieving a
desired outcome (Compeau and Higgins, 1995, p. 191; based on Bandura, 1986). It is,
according to Bandura (1997), an important personal factor that functions as an intrinsic
motivator, and is hypothesized to influence, and to be influenced by, the individual’s
behaviour and environment. The self-efficacy construct therefore offers a connection
between self-perception and individual behaviour (Chii and Braun, 1995). Digging deeper
into the mechanism of self-efficacy, it reveals the potential saliency in the current research
domain. According to self-efficacy theory, beliefs about one’s self-efficacy influence human
functioning by affecting how individuals feel, think, motivate themselves and behave
(Bandura, 1997). Thus, also assisting individuals in deciding which activities to pursuit, how
much effort to allocate and their degree of persistency (Bandura, 1991). It is suggested that
individuals with high self-efficacy will tend to view difficult tasks as challenges that should
be mastered rather than threats to be avoided (Bandura, 1994). According to Bandura
(1977), the most influential source, from which individuals judge their level of self-efficacy,
is derived from direct authentic experiences. A sequence of successful experiences raises
self-efficacy appraisals, whereas failures lower them (Bandura, 1977). Thus, in the centre of
the self-efficacy theory lies the belief that individuals’ behaviours are often better explained
by their expectancies and beliefs about their own capabilities, more than what they are
actually capable of doing (Bandura, 1997, 1994, 1986).

Page 34 of 155
According to Compeau and Higgins (1995) the construct of self-efficacy is, within a
computer usage context, measured as a derived unidimensional construct with three
distinct but interrelated dimensions: strength, generalizability and magnitude. Self-efficacy
strength reflects the level of confidence the user has in accomplishing difficult computing
tasks. Generalizability reflect the degree to which the expectation is generalizable to a
specific domain, while users with a high self-efficacy magnitude imagine themselves to be
capable of accomplishing difficult computing tasks with little or no support from others. This
operational definition is directly in line with the most-often applied definition within m-
commerce research (Trivedi and Kumar, 2014; Wang et al., 2006; Yang, 2012, 2010).
Individuals may arguably perceive technologies as daunting challenges, since a
successful use may often be believed to require a certain degree of skill level and mental
effort. It is therefore expected that individuals’ decisions to use technologies may be guided
by their level of self-efficacy, which consequently forms different perceptions and
behaviours. In fact, users’ level of self-efficacy in technology use is apparently of great
significance in nurturing and promoting the use of technologies. For example, users’
perceived level of self-efficacy have been found to be a significant predictor of users’
decision to use online shopping services (Vijayasarathy, 2004), online music services
(Bounagui and Nel, 2009), continue using websites (Wangpipatwong et al., 2008) and more
importantly, to influence users’ decision to use m-commerce services (Trivedi and Kumar,
2014; Wang et al., 2006). The study of Hernández et al. (2010) reported that online shopping
frequency affected users level of self-efficacy, while Compeau and Higgins (1995) found that
users with a high level of self-efficacy used computers more frequently and experienced
less computer anxiety. More relevant to this study is the diversities in users’ perceptions of
determinant factors driving users’ behavioural intentions. Current m-commerce literature
seem to have devoted less attention to examine the moderating effect of self-efficacy. The
study of Yang (2012) however, shed some light on the area, finding that users’ level of self-
efficacy positively moderated the relationship between enjoyment and attitude towards
mobile shopping. Yang’s study also revealed that increased self-efficacy led to increased
control and consequently a higher intention to adopt mobile shopping. In the meantime,

Page 35 of 155
Jaradat and Faqih (2014) demonstrated that users with a high level of self-efficacy were
more likely to perceive mobile payment as useful, and thus more likely to use it when
compared to users with a low level of self-efficacy. In summary, the evidence presented
above indicate that mechanisms of self-efficacy play an important role in motivating users
to engage in m-commerce activities, and high sense of self-efficacy strengthen positive
perception and orientation of m-commerce.
2.3.2. Impulsiveness
The concept of impulsiveness has been heavily debated throughout the years, and has
thus been subject for changes in definitions along the way. In the early fifties, impulsiveness
was primarily regarded as signs of immaturity, primitivism, foolishness and other similar
social deviations (Park and Choi, 2013), whereas the concept has evolved into a more
complex construct in contemporary literature. However, the consensus of general
characteristics of impulsiveness remain the same; that impulsiveness covers purchases with
low or non-existing prior planning, which in turn is concluded to be irrational buying
behaviour (Etzioni, 1986; Park and Choi, 2013). The interesting aspects of impulsiveness,
and the aspects that cause divergence between different studies are, however, what
activates this behaviour and which consequences, it has. Cobb and Hoyer (1986) regard
impulsive behaviour as the decision of buying a product made inside the store. Thus, the
consumer is assumed to have no intentions or plans for buying the product in question
before entering the store, but simply consciously experienced a latent need being brought
to life when being presented with the product. That is, a need that the consumer had not
previously recognized.
This study, however, adopts the definitions of impulsiveness from Rook and Fisher (1995
p. 306), defining that buying impulsively is “a consumer’s tendency to buy spontaneously,
non-reflectively, immediately, and kinetically”, which is acknowledged by several studies
within e-commerce (e.g. Chih et al., 2012; Parboteeah et al., 2009) as well as m-commerce
(e.g. San‐Martin and López‐Catalán, 2013), though findings within the context of m-

Page 36 of 155
commerce are rather limited. Rook (1987) and Rook and Fisher (1995) recognize that
impulsive purchasing behaviour refers to a more specific range of phenomena rather than
just unplanned purchase, thereby challenging the theories of Cobb and Hoyer (1986) by
distinguishing between the two. The difference, according to Rook (1987), is that impulsive
purchasing behaviour occurs when consumers experience sudden powerful and persistent
urge to buy something immediately, as opposed to unplanned purchasing that is
characterized as more ordinary and tranquil. Impulsive behaviour is thus also, to a greater
extent, caused by emotional feelings for the product in the current cognitive state. An
apparent weighting factor of enhancing the chances of impulsive behaviour among
consumers, is to increase the intensity of advertising of the product (Arens and Rust, 2012),
provided that these create positive feelings to the brand in question, and that these
associative feelings act as cues for rewards if purchasing the products. According to
Parboteeah et al. (2009), these feelings might likely be induced by the consumer’s
enjoyment in a given situation when exposed to a product within e-commerce. Their
findings were that perceived enjoyment was in fact the primary explanatory variable of the
urge to buy impulsively. These findings are furthermore supported by Chih et al. (2012),
finding that exposing consumers to hedonic consumption needs in e-commerce will
enhance the positive affects for the customers. That is, hedonic consumptions needs being
exposure to product characteristics that enhance the positive affect (Chih et al., 2012) vis-
à-vis the consumers’ mood (Parboteeah et al., 2009) or associative feelings (Arens and Rust,
2012). In fact, several studies (e.g. Chih et al., 2012; Flight et al., 2012; Rook and Fisher,
1995) suggest that failure to induce positive affects, will severely impair the chances of
consumers repurchasing. Rook and Fisher (1995) found that, in the case of consumers
experiencing negative affects, the general buying behaviour was significantly impaired, and
some impulsive consumers even managed to reject the need for impulsive shopping when
their normative evaluations were sufficiently negative.
In a virtual store context, such as m-commerce, there’s generally tradition for
persuading users to engage impulsively (San‐Martin and López‐Catalán, 2013). However,
acting impulsively severely increases the chances of overall dissatisfaction with the process,

Page 37 of 155
as there are none or low planning prior to the purchase, and a low degree of other
considerations (Rook and Fisher, 1995). However, though the outcome of a negative affect
has proven to be imperative for most peoples’ intention to engage in a similar behaviour
prospectively, George and Yaoyuneyong (2010) stress that highly impulsive people are more
risk tolerant, and are therefore less likely to let prior bad experiences impact future
behaviour.
Construct Conceptualization Object Study
Flow
A holistic sensation present when we
act with total involvement
Human psychology
Csikszentmihalyi
(1997)
Perceived Ease
of Use
The degree to which a person believes
that using a particular system would be
free of effort
User acceptance of
IT
Davis (1989)
Perceived
Usefulness
The degree to which a person believes
that using a particular system would
enhance his or her job performance’
User acceptance of
IT
Davis (1989)
Trust
The willingness of a party to be
vulnerable to the actions of another
party based on the expectation that the
other will perform a particular action
important to the trustor, irrespective of
the ability to monitor or control that
other party
Organizational Trust Mayer et al. (1995)
Confirmation
Users’ perception of the congruence
between expectation of a system and
its actual performance
Continuance
Intention of IT
Bhattacherjee (2001)
Satisfaction
Users’ affect with (feeling about) prior
use
Continuance
Intention of IT
Continuance
Intention
Users’ intention to continue using a
system
Continuance
Intention of IT
Self-Efficacy
Belief about one’s ability to perform a
specific behaviour with confidence in
achieving positive task outcomes
Information systems
Yang (2010),
adopted from
Compeau and
Higgins (1995)
Impulsiveness
A consumer’s tendency to buy
spontaneously, non-reflectively,
immediately, and kinetically
Buying behaviour
Rook and Fisher
(1995)
TABLE 2.1. CONSTRUCTS AND THEIR ORIGINS

Chapter III – Conceptual Model & Hypotheses
Page 38 of 155
- Chapter III -
CONCEPTUAL MODEL & HYPOTHESES
“Yes, I sell people things they don't need. I can't, however, sell them something they don't
want. Even with advertising. Even if I were of a mind to.“ - John O'Toole

Page 39 of 155
3.1. Hypothesis Development
In this chapter, a hybrid conceptual model is developed based on the ECM-IS and TAM,
given that literature have found significant convergence between the explanatory powers
of some variables in both contexts. Furthermore, the model is extended with trust and flow,
as both variables have frequently been positively associated with important user beliefs
regarding continuance usage intention of IS.
3.1.1. Study One
The hypotheses development of study one is two-fold. The first set of hypotheses
concerns the mechanisms of inflicting satisfaction. The second set regards the measures to
enhance the likelihood of users’ continuance intention.
3.1.1.2. Creating Satisfaction through Parallel Mediation
The relationship between confirmation and satisfaction
According to the ECM-IS, users’ degree of satisfaction is derived from a function of
expectations and expectancy confirmation, meaning that users’ evaluation of m-commerce
performance weighted against their initial pre-adoption expectations, will determine the
users’ degree of satisfaction with m-commerce usage (Bhattacherjee, 2001; Oliver, 1980).
Confirmation is therefore positively connected with satisfaction because a confirmation of
expectations of the m-commerce interaction would mean a fulfilment of expected benefits
(Bhattacherjee, 2001). For instance, mobile banking users attributed their satisfaction with
the banks’ ability to provide accurate, complete and relevant information (Lee and Chung,
2009). Similarly, mobile users are more likely to be satisfied with mobile carriers, if they are
able to provide a high network quality and competent customer service (Lim et al., 2006).
Additionally, in consumer research, Oliver and DeSarbo (1988) found that consumers with
high initial expectations would be more likely to form a high degree of satisfaction, if their
expectations were confirmed and subsequently, Brown et al. (2012) found, by drawing on

Page 40 of 155
prospect theory, that negative disconfirmation consequences outweigh the consequences
of positive confirmation in relation to users’ attitude formation towards an intranet-system.
Thus, it seems reasonable to say that that users’ degree of satisfaction depends on the
degree of (/dis)confirmation of expectations. Several m-commerce studies have supported
this link between confirmation and satisfaction (Kim, 2010; Lee, 2014; Lin et al., 2014; Thong
et al., 2006)
3.1.1.2. TAM
The relationship between confirmation and TAM
The variable of perceived usefulness covers the extent to which degree a system
enhances the task performance for a user (Davis, 1989). In an m-commerce context it thus
refers to the users’ enhanced performance by using m-commerce as opposed to doing the
same task without it (Chong, 2013; Lee, 2014). However, the definition is slightly different
when combined with the ECT, since the user, as opposed to the situation in the original
TAM, has already used the system before. Lin et al. (2014) and Bhattacherjee (2001) refer
to the variable as a post-adoption expectation, meaning that the already experienced
usefulness will affect the perceived usefulness for future use. Researchers (e.g. Chong,
2013; Hong et al., 2006; Thong et al., 2006) have heavily hypothesized that confirmation of
pre-consumption expectations will impact the cognitive attitudes regarding future use. Such
cognitive attitudes have, in a technological context, often been hypothesized to be aspects
of the TAM. E.g. Thong et al. (2006), who found significant correlations between
confirmation and perceived usefulness as well as perceived ease of use. In other contexts,
there have been somewhat more divergent results as to what these correlation represent.
For instance in the context of mobile internet, Hong et al. (2006) found relatively little
correlation between confirmation and perceived ease of use as well as an almost non-
existing correlation between confirmation and perceived usefulness. Chong (2013) on the
other hand, found highly significant correlations between confirmation and the two
discussed variables in his study regarding m-commerce. The reason for these differences

Page 41 of 155
could possibly be explained by the difference in context, which in turn impedes a congruent
continuity of results.
The relationship between TAM and satisfaction
Davis et al., (1989) found that perceived usefulness and perceived ease of use had a
major significant impact on users’ affect/attitude toward the adoption of technology in the
TAM (Davis et al., 1989). However, though the concept was developed for adoption of a
technology, Thong et al. (2006) argue that since satisfaction is a type of affect, perceived
usefulness and perceived ease of use can equally be applied as indicators for the satisfaction
of an already tested technology. Thus, the more useful (Bhattacherjee, 2001; Lin et al.,
2014) and free of effort (Chong, 2013; Hong et al., 2006) m-commerce has been for the
respective user so far, the higher the degree of post-adoption expectation will be, and thus
also the satisfaction with m-commerce in respect to the ECT.
The mediating role of TAM
In general, it seems there are substantial evidence in literature suggesting the effects
between confirmation and satisfaction. There have, however, been divergent views as to
what nature the correlation has. Several studies have thus hypothesized different cognitive
components to represent the extent of correlations between the two variables. One of the
more frequent assumptions is that elements from the TAM represent comprehensive
explanatory value in the correlations, as for instance Thong et al. (2006), who acknowledge
the importance of including elements from the TAM to better explain the elements that
influence the variance of satisfaction, though no actual mediation effects were examined.
That particular study was, however, investigated in depth, as Al-Jabri (2015) suggested the
necessity to clarify the nature of correlations between external variables (i.e. training and
communications) and user satisfaction. By implementing perceived ease of use and
perceived usefulness as mediators, which in his study are called ease of use and benefits,
he identified partial mediation, suggesting that perceived ease of use and perceived

Page 42 of 155
usefulness both play a significant role when explaining the significant relation between
training and satisfaction (Jabri, 2015). Burton-Jones and Hubona (2006) proposed a similar
layout, using the same mediators between system experience and usage volume within IS
acceptance, though he failed to identify the effects (Burton-Jones and Hubona, 2006).
As the ECT builds upon the premise that the users’ have already used the system at least
once before, theory has it that they have a certain degree of experience, and that this
experience will evolve in the event of multiple usage. Moreover, as the experience changes
so do the relating variables (Bhattacherjee, 2001), why confirmation in the ECT, to some
degree, can be compared to both system experience and system training, why we propose:
H1a: Users’ extent of Perceived Ease of Use will mediate the relationship between
Confirmation and Satisfaction (CON  PEOU  SAT)
H1b: Users’ extent of Perceived Usefulness will mediate the relationship between
Confirmation and Satisfaction (CON  PU  SAT)
3.1.1.3. Trust
The relationship between confirmation and trust
The definition of trust operationalized in this study reflects the users’ confidence in m-
vendors’ trustworthiness, which allows them to enter a vulnerable position in which they
are dependent on the actions of the m-vendors (Mayer et al., 1995). It is the users’
expectations that m-vendors can be relied upon to fulfil their promises and will not engage
in opportunistic behaviours (Gefen et al., 2003a). Gefen and Straub (2004) suggest that e-
service providers can improve customers’ trusting beliefs by providing trustworthy
attributes (integrity, predictability, ability and benevolence) that are consistent with
customers’ expectations. Individuals’ level of trust increases when the other party displays
behaviours or other indicators that matches their expectations (Johnson and Grayson, 2005;

Page 43 of 155
Zhou, 2013a). It is therefore reasonable to expect that regardless of prior experiences, a
confirmation of m-vendors’ trustworthiness will lead to favourable trusting beliefs.
Empirical evidence supporting this relation are found in m-commerce (Chong, 2013; Lin and
Wang, 2006) and in mobile apps (Akter et al., 2013).
The relationship between trust and satisfaction
The underlying mechanism in which trust suggestively influence satisfaction is based
directly on the fundamental logic of Ajzen (1985) and Oliver (1980), expecting that positive
trusting beliefs will lead to attitude/satisfaction. Pavlou and Fygenson (2006) opined that
favourable trusting beliefs create positive perceptions about the outcome of e-vendors’
actions that consequently evoke positive attitudes. Specifically, trust mitigate risk
perceptions, e.g. beliefs about being exploited or mistreated by the e-vendor, which in turn
positively affect users’ attitude (Jarvenpaa et al., 2000). Ziaullah et al. (2014) even advocate
that trust is fundamental for creating satisfied and committed customers. Meanwhile, in
the longitudinal study of Kim et al. (2009), it was concluded that users’ trust within e-
vendors was a critical factor that not only guided their initial purchase decisions, but also
impacted their evaluations of the experiences and further shaped their long-term decisions.
Additional m-commerce research have provided support for the correlation between trust
and satisfaction (Chong, 2013; Lin and Wang, 2006; San‐Martin and López‐Catalán, 2013).
It is therefore reasonably to believe that a fulfilment of users’ trusting beliefs will positively
influence their satisfaction with m-vendors.
The mediating role of trust
The studies of Hung et al. (2012) and Thong et al. (2006) encouraged researchers to look
beyond the extrinsic constructs of TAM and thereby also investigate intrinsic variables,
through which the effect of confirmation on satisfaction could be mediated. Building on
these studies, and as indicated by the finding presented so far, this study infers that the
effects of a positive confirmation of one’s expectations, e.g. m-vendors behave as expected

Page 44 of 155
of them, will translate into a higher degree of trust that subsequently positively influence
users’ degree of satisfaction. In an offline banking context, the study of Carlander et al.
(2011) found that the positive effect derived from delivering an acceptable degree of service
quality positively influenced customers’ degree of satisfaction. This effect was, however,
fully explained by customers’ degree of trust in the bank. Therefore, based on the presented
evidence, it is expected that:
H1c. Users’ extent of Trust will mediate the relationship between Confirmation and
Satisfaction (CON  TRU  SAT)
3.1.1.4. Flow
The relationship between confirmation and flow
Recall that flow is referred to as the pleasant feeling and enjoyment derived from a
match between skill level and challenge that allowed for immersion and acting with a sense
of total control (Hoffman and Novak, 1996). Thus, by incorporating the flow construct within
the ECM-IS framework, it implies that a confirmation of users’ expectations towards m-
commerce will have a direct impact on their degree of flow elicited from the interaction.
Support for this proposition is found within an e-learning context. Specifically, Cheng (2014)
connected the flow construct with ECM theory, and demonstrated that a confirmation of
nurses’ expectations towards e-learning systems were in fact directly related to the amount
of flow experienced. Similar, Sørebø et al. (2009) found a significant positive relationship
between expectancy confirmation and teachers’ level of intrinsic motivation towards e-
learning.
The relationship between flow and satisfaction
According to Csikszentmihalyi and LeFevre (1989, p. 816), individuals engaging in
activities in which they experienced flow reported to “feel more active, alert, concentrated,

Page 45 of 155
happy, satisfied, and creative”, and according to Hoffman and Novak (1996 p. 58), in a
computer-mediated environment, “resulting in a state of mind that is extremely gratifying”.
This makes it reasonable to believe that flow experience may in fact influence users’ degree
of satisfaction with m-commerce. By the same token, the m-commerce activities
investigated in this study bears both hedonic and utilitarian motives, e.g. shopping, ticketing
etc., why users may expect to obtain a pleasant and enjoyable experience in order to be
fully satisfied. Relating flow theory to an m-commerce context, it implies that some
predefined conditions might be necessary in order for users to gain an ‘optimal experience’
(Hoffman and Novak, 1996). For example, users are required to possess a certain degree of
knowledge and skills prior to the m-commerce interaction before flow is to be experienced,
meaning that the user’s knowledge about m-commerce and restrictions of the smartphone
will not impede the m-commerce experience. E.g. small screens, small bottoms, insufficient
understanding of mobile security, complex interfaces or general unfamiliarity with mobile
use might lead an unskilled user to feel a lack of control. Research have showed that
experienced customers tend to feel more in control when shopping online and are therefore
more satisfied (Pappas et al., 2014). Furthermore, if the users are vulnerable to distractions
while using m-commerce, it may prevent them from fully focusing on the interaction that
may lead to dissatisfactory. In online banking, Lee et al. (2007) argued that lacking face-to-
face contact and the general distracting environment associated with using a computer, e.g.
pop-ups, e-mails, instant messages etc., diminished the customers’ ability to focus on the
interaction and therefore caused the customer to be dissatisfied. This study therefore
expects to find that users experiencing flow through m-commerce interactions will
consequently show positive perceptions towards the m-commerce experience that may
account for a significant proportion of users’ overall evaluation of the m-commerce
experience, i.e. influence their degree of satisfaction. Previous studies within m-commerce
were also found to deliver evidence for this relationship, in mobile internet sites (Zhou,
2014a, 2013b, 2011) and in mobile payment services (Zhou, 2013a)

Page 46 of 155
The mediating role of flow
In essence, a realization of expected benefits of m-commerce usages, i.e. a confirmation
of users expectations would spawn a positive psychological state (Bhattacherjee, 2001) and
increase the level of flow perceived by users (Cheng, 2014), which would further translate
into an increased level of satisfaction (Hsu et al., 2012; Novak et al., 2000; Zhou, 2014a).
Hence, it is suspected that flow might facilitate some of the effect between confirmation
and satisfaction, meaning that users’ flow experience will only impact their degree of
satisfaction, if their initial expectations regarding flow towards m-commerce are confirmed,
while users whose expectations are not confirmed, will be unlikely to form a high degree of
satisfaction. This leads to the following hypothesis:
H1d: Users’ extent of Flow will mediate the relationship between Confirmation and
Satisfaction (CON  FLO  SAT)
3.1.2. Creating Continuance Intention through Single Mediation
The relationship between satisfaction and continuance intention
In accordance with ECM-IS users decision to re-use m-commerce should be guided by
their initial degree of satisfaction with the system (Bhattacherjee, 2001). A more intuitive
explanation for this relationship is found within the construct of satisfaction. According to
Choi et al. (2008), satisfaction within m-commerce is represented by an aggregation of
positive, negative, or indifferent feelings accumulated through multiple interactions with
m-commerce. This is, however, similar to traditional offline satisfaction, meaning that
satisfaction is expressed through an affective state, which is suggestively influencing
behavioural intentions directly and indirectly through attitude (Oliver, 1980). Marketing
literature generally agree that consumers’ degree of satisfaction holds a high explanatory
power in relation to one’s decision to patronize goods or services (Oliver and Bearden, 1985;
Swan and Trawick, 1981). Numerous studies within IS also support this relationship (Chen

Page 47 of 155
and Chou, 2012; Halilovic and Cicic, 2013; Mohamed et al., 2014). Consequently, this study
infers that an equivalent relationship is present within an m-commerce context. Previous
m-commerce research do also support this relation (e.g. Hong et al., 2006; Hung et al., 2012;
Kim, 2010).
3.1.2.1. TAM
The relationship between perceived ease of use and continuance intention
In the context of continuance usage intention, research stress that a technology that is
perceived easier to use will lead to a higher probability of continuing using a technology, as
opposed to a technology that is perceived more advanced (Davis, 1989; Thong et al., 2006).
This claim is furthermore supported by Hong et al., (2006), claiming that users’ perception
of easiness will continuously be enhanced, as the users will gain more experience from the
usage of the system, hence creating more familiarity with the system (Hong et al., 2006;
Thong et al., 2006). Therefore, given fact that users will gain more experience by using the
system more frequently, Bhattacherjee (2001) advocated for the necessity to include
additional attitudes in the context of continuance intention, given that this situation will
indeed expand the paradigms of usage (Bhattacherjee, 2001). Perceived ease of use has
thus frequently been subject for postulations regarding its proclaimed direct impact on
continuance intention. However, though several studies have found a significant direct
impact in various extents (e.g. Hong et al., 2006; Thong et al., 2006), other studies have
failed to identify this correlation (e.g. Chong, 2013; Zhou, 2011). Thus, it seems there are
inconsistent evidence whether or not this link exists. However, realizing the absence of
impact, Zhou (2011) proved that his definition of perceived ease of use had a significant
impact on continuance intention if mediated by satisfaction. Though there are
inconsistencies, it’s our assumptions that:
H2a. Users’ extent of Satisfaction will mediate the relationship between Perceived Ease
of Use and Continuance Intention (PEOU  SAT  CI)

Page 48 of 155
The relationship between perceived usefulness and continuance intention
Though perceived usefulness was originally intended to prove the impact on the
intention to adopt a technology, comprehensive research indicate that the variable is also
viable to impact the intention of continuance usage, as people generally strive for rewards
and to utilize a situation as much as possible no matter the timing (Bhattacherjee, 2001).
Consequently, the linkage between perceived usefulness and continuance intention has
been proposed in various studies. For instance in the context of IS (Bhattacherjee, 2001),
data services (Kim, 2010) and within m-commerce (Chong, 2013; Lin et al., 2014). However,
the study of this impact has shown divergent results, why this proclaimed linkage is rather
questionable.
When demonstrating the direct correlations between perceived usefulness and
continuance intention, Bhattacherjee (2001) further found that satisfaction with an IS
mediated an indirect influence between the two variables. Lin and Wang (2006) support
this thesis, claiming that satisfaction with an IS is determined by two aspects: confirmation
and perceived value, thus leading us to hypothesize that:
H2b. Users’ extent of Satisfaction will mediate the relationship between Perceived
Usefulness and Continuance Intention (PU  SAT  CI)
3.1.2.2. Trust
The relationship between trust and continuance intention
According to Mayer et al. (1995) when the degree of trust exceeds a threshold value of
perceived risks, the person will be motivated to enter a vulnerable position, even though
risks are present. Researchers within IS therefore generally agree that trust can directly
influence behavioural intentions since it, as mentioned earlier, diminishes risk perceptions
(Kim et al., 2008). Liu et al. (2005) proposed an online “privacy-trust-behaviour model” in

Master_Thesis_Kasper_Urbrand_Nielsen_And_Morten_Riise_Jensen_2015

Master_Thesis_Kasper_Urbrand_Nielsen_And_Morten_Riise_Jensen_2015

Recommended

Recommended

More Related Content

Similar to Master_Thesis_Kasper_Urbrand_Nielsen_And_Morten_Riise_Jensen_2015

Similar to Master_Thesis_Kasper_Urbrand_Nielsen_And_Morten_Riise_Jensen_2015 (20)

Master_Thesis_Kasper_Urbrand_Nielsen_And_Morten_Riise_Jensen_2015