Professional Documents
Culture Documents
Acknowledgement
First I would like express my gratitude to Athanasios Krystallis, my supervisor, for his
guidance and help during the entire working process. I also want to thank my Danish
family for their constant support and for their help with the sample collection. Last but
not least, I would like to express my appreciation to my parents, my sister and my
husband who are always there to give me the inspiration to achieve my goals.
Daniela Vera
CONTENTS
1.
2.
Introduction ........................................................................................................... 1
1.1
1.2
1.3
Contribution.................................................................................................... 5
1.4
1.5
Structure ......................................................................................................... 6
2.1.1
2.1.2
2.1.3
2.1.4
2.1.5
2.1.6
2.2
3.
2.2.1
Trust ...................................................................................................... 15
2.2.2
2.2.3
2.2.3.1
2.2.3.2
3.1.1
3.1.2
3.1.3
3.1.4
3.1.5
3.1.6
3.1.7
3.1.8
3.2
4.
5.
Methodology........................................................................................................ 33
4.1
4.2
4.3
Data Analysis................................................................................................ 37
5.2
5.3
5.3.1
5.3.1.1
Reliability ....................................................................................... 41
5.3.1.2
Validity .......................................................................................... 42
5.3.2
5.4
5.3.2.1
5.3.2.2
5.3.2.3
5.3.2.4
Predictive Relevance....................................................................... 45
5.4.1
Bootstrapping ........................................................................................ 46
5.4.2
6.
5.4.2.1
5.4.2.2
5.4.3
5.4.4
Discussion ........................................................................................................... 54
6.1
7.2
Managerial Implications................................................................................ 61
7.3
Limitations ................................................................................................... 62
7.4
8.
References ........................................................................................................... 64
9.
Appendices .......................................................................................................... 70
List of Figures
Figure 1: Thesis Structure ............................................................................................. 6
Figure 2: The Technology Acceptance Model ............................................................... 9
Figure 3: The Theory of Planned Behavior .................................................................. 10
Figure 4: The Decomposed Theory of Planned Behavior ............................................. 13
Figure 5: The Conceptual Model ................................................................................. 31
Figure 6: The Conceptual Model Results of Path Significances ................................ 53
List of Tables
Table 1: Peterson et al. (1997) Product and Service Classification Grid ....................... 20
Table 2: Construct Definitions .................................................................................... 32
Table 3: Sources of Questionnaire Items ..................................................................... 34
Table 4: Sample Characteristics .................................................................................. 38
Table 5: Products and Services Purchased Online ....................................................... 39
Table 6: Products and Services Selected to Answer the Survey ................................... 40
Table 7: Reliability Results Composite Reliability and Cronbachs Alpha ................ 41
Table 8: Validity Results AVE ................................................................................. 42
Table 9: Variance Explanation Results ........................................................................ 44
Table 10: Effect Size Results ...................................................................................... 45
Table 11: Predictive Relevance Results ....................................................................... 46
Table 12: Tests of PLS Paths with Bootstrap ............................................................... 47
Table 13: Products and Services Classification in This Study ...................................... 48
Table 14: Test of Homogeneity of Variances Product Types .................................... 49
Table 15: ANOVA Behavioral Intention and Product Types .................................... 49
Table 16: Behavioral Intention and Product Type Statistics ......................................... 50
Table 17: Low and High Trust Groups ........................................................................ 50
Table 18: Test of Homogeneity of Variances Trust Groups ...................................... 51
Table 19: ANOVA Behavioral Intention and Trust Groups ...................................... 51
Table 20: Robust Tests of Equality of Means Trust Groups ...................................... 51
Table 21: Summary of Results of Hypotheses Testing ................................................. 52
List of Appendices
Appendix 1: Questionnaire Measures and Sources ...................................................... 71
Appendix 2: Web-Based questionnaire ........................................................................ 72
Appendix 3: Purchase Frequencies .............................................................................. 77
Appendix 4: Danish Online Shoppers.......................................................................... 80
Appendix 5: Latent Variables Correlations .................................................................. 82
Appendix 6: Latent Variables Squared Correlations and AVE ..................................... 83
Appendix 7: Cross Loadings ....................................................................................... 84
Appendix 8: SmartPLS Output Path Analysis ........................................................... 88
Appendix 9: SmartPLS Output Bootstrapping .......................................................... 89
Appendix 10: P-P Plots and Q-Q Plots ........................................................................ 90
Appendix 11: Post Hoc Tests ...................................................................................... 91
Introduction
1. INTRODUCTION
Introduction
product or service without incurring in time and transportation costs; moreover, online
stores are available to consumers anytime and anywhere. Another benefit is that
consumers have easy access to product and service information; and many online stores
provide tools for product comparison and help in making purchasing decisions. Through
online stores, consumers have access to products in foreign countries that are not
available in their own countries. Lastly, online stores have reduced operation costs
compared to traditional brick-and-mortar stores, cutting on labor and store rental costs
which allows for lower prices offered to consumers (Euromonitor International, 2012b).
Further evidence is found in a Danish survey which revealed that the three main reasons
for shopping online were lower prices in online stores compared to physical stores,
convenience and quick price comparisons (Euromonitor International, 2012c).
Although online stores have many advantages, they also have disadvantages over brickand-mortar stores. One of the most important disadvantages is the fact that consumers
cannot touch, feel, taste or smell the products; this prevents consumers from assessing
product quality and increases risk perceptions. Other disadvantages are related to
delivery delays, security and privacy concerns which can affect consumers trust on
online stores.
A web retailers storefront is its website, meaning that when interacting with a web
retailer, consumers become IT users and face new challenges such as navigating the
website, information overload, unfriendly user interfaces, or complex ordering
processes. These challenges reduce consumers perceptions of control and confidence
over online activities.
Online retail sales show that online shopping remains popular for certain types of
products, for instance according to a study sponsored by the European Parliament, EU
consumers mainly buy clothes and travel related products and services online, whereas
computers and electronic products are the least likely to be bought online (Muller et al.,
2011).
Despite the fact that the figures show increasing online sales, many online consumers
use information gathered online to make purchases offline, this is evidenced by the high
abandon rates of shopping carts (Kiang et al., 2011). Consumers use online stores to
gain market knowledge, they learn about price levels and product differences, yet they
Introduction
dont make the final transaction with the online store (Broekhuizen and Huizingh,
2009).
Online Consumer Behavior Research
Given the widespread proliferation of online shopping, online consumer behavior has
become an important topic among researchers, this is illustrated by the great number of
publications on different fields such as information systems, marketing, management
and psychology (Cheung et al., 2005).
Researchers have been exploring online consumer behavior for many years and two
widely accepted views stand out in e-commerce literature: consumer-oriented and
technology-oriented view. The consumer-oriented view places focus on consumers
salient beliefs about online shopping, whereas the technology-oriented view studies the
impact of website design and usability on consumers behavior (Zhou et al., 2007). The
findings support both views and it is apparent in the literature that both views
complement each other.
Prior research shows that there are numerous factors that affect online consumer
behavior, nonetheless there are mixed findings in literature and many factors that
influence online consumer purchasing behavior have yet to be explored, especially
considering the dynamics of technology and consumer needs, which are constantly
evolving, and as a result significant factors five years ago may differ today as
consumers become more experienced internet users.
Furthermore, most of the previous online shopping research is focused on one specific
type of product such as books (Gefen et al., 2003, Lin, 2007), clothing (Ha and Stoel,
2009, Hansen and Mller Jensen, 2009, Kim and Kim, 2004, Tong, 2010, Kim et al.,
2003, Yoh et al., 2003), groceries (Hansen et al., 2004), financial services (McKechnie
et al., 2006, Suh and Han, 2003) and car insurance (Broekhuizen and Huizingh, 2009).
Previous research has also investigated product characteristics and online behavior
using a conventional product classification scheme, exploring how search, experience,
and credence goods vary in their impact on purchase intentions (Brown et al., 2003,
Girard et al., 2003, Korgaonkar et al., 2006, So et al., 2005, Soopramanien et al., 2007).
Few studies explore different product types and online purchasing intentions using a
classification scheme that takes into account the specific features of internet (Ian and
Introduction
Sui Meng, 2000, Vijayasarathy, 2002) and the findings show mixed results which call
for further investigation.
Introduction
RQ4. How do attitude, subjective norm and perceived behavioral control impact online
purchase intentions and consequently, online purchasing?
RQ5. To what extent, if any, is product type related to online purchase intention?
1.3 Contribution
The results from this study will make a positive contribution to the online consumer
behavior literature by providing a deeper understanding of consumer beliefs about
online shopping and how these impact attitude and intention to shop online. The
knowledge from this study will provide valuable information for retailers about the
relevant factors that drive consumers to shop online and the products and services that
are more likely to be purchased online. The information could help retailers adapt their
strategies to fit customer needs and attract and retain customers. From the marketing
point of view, gaining useful insight into online consumer behavior is fundamental and
the knowledge from this study could help create marketing strategies tailored to respond
to online consumer specific requirements and needs.
The figures and statistics show that e-commerce is full of opportunities for businesses of
any size and at the same time low barriers to entry are making the market more
competitive (Datamonitor, 2011), thus understanding online consumer behavior and
what drives them to shop online is crucial for any business that wants to be competitive.
Introduction
1.5 Structure
This paper is divided into seven sections: introduction, literature review, conceptual
model and hypotheses, methodology, data analysis and results, discussion, summary
and conclusions (Figure 1).
After the current introductory part, the literature review is presented in the second
section, where relevant theoretical models applied in the e-commerce literature are
discussed as well as prior research findings. The third section presents the conceptual
model and hypotheses. The fourth section outlines the research methodology, where
instrument development and data collection procedures are discussed. The fifth section
includes data analysis and results. The sixth section provides discussions concerning the
results. Finally, summary of findings, managerial implications, limitations and further
research are presented.
1. Introduction
2. Literature Review
4. Methodology
6. Discussion
Literature Review
2. LITERATURE REVIEW
This section consists of three main parts. The first part reviews major conceptual
research models from both the perspective of traditional consumer behavior and
information systems. The second part reviews research findings and extensions to the
major conceptual models in the literature. The third part reviews literature relating to
product classifications and research findings regarding purchase intention in the context
of different product types.
Literature Review
shopping by studying web site design and content as well as system usability (Zhou et
al., 2007).
Literature Review
influenced by PEOU and not the other way around, the rationale behind it is that easyto-use technology is more useful than hard-to-use technology and useful technology
may not necessarily be easy to use (Figure 2).
TAM has been widely adopted in information systems (IS) research and it has been
successfully applied as a theoretical framework to predict online purchasing behavior
(Chen and Tan, 2004, Hernndez et al., 2010, Pavlou, 2003, Vijayasarathy, 2004),
moreover, researchers have applied TAM to predict online purchasing behavior in the
context of books (Gefen et al., 2003, Lin, 2007), clothing (Ha and Stoel, 2009, Tong,
Literature Review
2010) and financial services (Suh and Han, 2003). Although TAM has proven to have
valid constructs, the explanatory power of TAM has been enhanced by the aggregation
of other constructs into the research model, as the model has been seen as too
parsimonious by researchers (Venkatesh and Davis, 2000). One of the extensions of
TAM was proposed by Venkatesh and Davis (2000), referred as TAM2, the model
includes subjective norm as it was found to have significant influence on PU and
behavioral intention.
Figure 3).
10
Literature Review
11
Literature Review
particular system would enhance his or her job performance (Davis 1989, p. 320),
Taylor and Todd (1995) suggest that PU, as defined in TAM, is equivalent to Rogers
relative advantage, since both constructs refer to a relative improvement in performance
and their measures have been operationalized in terms of their relative impact on
performance. According to Rogers (1995), complexity represents the degree to which an
innovation is perceived to be difficult to understand, learn or operate. Taylor and Todd
(1995) suggest that PEOU (the degree to which a person believes that using a particular
system would be free of effort) is analogous to Rogers complexity construct, although
in an opposite way. Compatibility refers to the degree to which the innovation fits with
the potential adopters existing values, previous experiences and current needs (Rogers,
1995).
Previous studies have suggested the decomposition of subjective norm into two
dimensions: interpersonal influence and external influence (Bhattacherjee, 2000, Hsu
and Chiu, 2004, Lin, 2007). Interpersonal influence refers to word-of-mouth influence
by friends, family, colleagues, while external influence is related to mass media reports,
experts opinions and other non-personal information.
Ajzen (1991) decomposed the PBC component into two dimensions: self-efficacy and
facilitating conditions. The dimension of self-efficacy is defined as an individuals
perception of his or her individual capabilities; in the context of online shopping it
refers to consumers self-assessment of his or her capabilities to shop online. The
second dimension, facilitating conditions, is concerned with external resource constrains
that may influence on engaging a particular behavior, such as time, money and
technology; in the context of online shopping the issue of technology constrains is
related to the availability of supporting internet equipment (Ajzen, 1991, Ajzen, 2002,
Lin, 2007).
DTPB has been successfully applied as research model in online shopping to predict
purchasing behavior, repurchase intention and as a model to understand the relation of
two behaviors such as getting information and actual online purchasing (Chen, 2009,
Hsu and Chiu, 2004, Lin, 2007, Pavlou and Fygenson, 2006).
12
Literature Review
13
Literature Review
It can be said that TAM is more parsimonious than TPB, thus it can be useful in
research focused on achieving an overall understanding of behavior (Lin, 2007). On the
other hand, DTPB sacrifices parsimony but provides better insight into the determinants
behavioral intention and actual behavior (Lin, 2007). Therefore, it is reasonable to
conclude that DTPB is appropriate for the purpose of the present study.
14
Literature Review
studies have explored this factor, the relation between product types and online behavior
is also discussed in detailed.
2.2.1 Trust
Trust can be defined as the expectation that other individuals or companies with whom
one interacts will not take advantage of a dependence upon them. It is the belief that the
trusted party will behave in an ethical, dependable, and socially appropriate manner and
will fulfill their expected commitments. (Gefen et al., 2003, p. 308). According to
McKnight et al. (2002), trust is defined as the belief that allows consumers to willingly
become vulnerable to web retailers after having taken the retailers characteristics into
consideration. These definitions are consistent with the three trusting beliefs that are
used most often in literature (Bhattacherjee, 2000, McKnight et al., 2002, Pavlou, 2003,
Pavlou and Fygenson, 2006): competence, integrity and benevolence. Competence is
the belief in the trustees ability to perform as expected by the trustor. Integrity is the
belief that the trustee will be honest and keep its promises. Benevolence is the belief
that the trustee will not act opportunistically.
Trust is a central element in exchange relationships that are characterized by uncertainty
and vulnerability (McKnight et al., 2002). Prior research confirms that trust plays a
relevant role in consumer behavior, both online and offline. (Chen, 2009). However, the
importance of trust increases in the online context because perceptions of uncertainty
may be especially significant on an e-commerce environment, where certain cues that
evoke trust cannot be fully assessed (e.g. product characteristics, physical store, sales
person) (Jarvenpaa et al., 2000, Pavlou, 2003, Verhagen et al., 2006). Furthermore, lack
of trust has been credited as one of the main reasons preventing consumers from
engaging in e-commerce (Jarvenpaa et al., 2000, Monsuw et al., 2004).
Several researchers have demonstrated that trust positively influences attitudes towards
online purchasing (Chen and Tan, 2004, Delafrooz et al., 2011, Ha and Stoel, 2009,
Kim, 2012, Ling et al., 2010, Suh and Han, 2003, Zarmpou et al., 2012). For instance,
Jarvenpaa et al. (2000) empirically showed that trust has a significant effect on
consumers attitudes toward online purchasing in multiple cultures. Suh and Han (2003)
integrated trust with TAM constructs and found comparable results, namely, trust is a
significant factor influencing attitude toward online shopping Pavlou and Fygenson
15
Literature Review
16
Literature Review
attitude toward purchasing from an online store, while security did not have a
significant effect.
Vijayasarathy (2004) extended TAM with subjective norm, privacy risk and security
risk in order to predict consumers online purchase intentions, the empirical test proved
all constructs significant but privacy concerns.
The concept of risk has been widely studied in literature from different perspectives, but
in general terms it can be said that trust and risk are interwoven (Jarvenpaa et al., 2000),
as trust is needed in uncertain situations and this means assuming risks and becoming
vulnerable to trusted parties. Thus, trust plays a central role in helping consumers
overcome perceptions of risk; this applies to the e-commerce context, when a web
retailer can be trusted to show competence, benevolence, and integrity, there is much
less risk involved in engaging and interacting with the web retailer since trust makes
consumers comfortable sharing personal information and making purchases. According
to Pavlou (2003), trust is one of the most effective tools for reducing uncertainty and
risks.
Several studies incorporated both perceived risk and trust in their research model while
exploring online consumer behavior, their findings support the significance of both
constructs, however, trust shows stronger significance in influencing attitude and
purchase intention (Chen, 2009, Jarvenpaa et al., 2000, Pavlou, 2003, Verhagen et al.,
2006). Furthermore, if a consumer decides to trust on a web retailer, the decision to
transact inseparably entails an interaction with its website interface, hence, it is fair to
say that trust in a web retailer implicitly encompasses trust in the integrity of the
transaction medium (i.e. web retailers infrastructure) (Pavlou, 2003). Web retailers can
affect trust in their infrastructure by facilitating encrypted transactions, installing
firewalls, using authentication mechanisms, and establishing privacy seals and
disclosures. By implementing these privacy and security mechanisms it can be argued
that both, seller performance risks, privacy and security risks can be decreased.
Based on the findings regarding trust and perceived risks and the fact that both
constructs are interwoven, the present study will focus on trust as a major determinant
of online consumer behavior, assuming that seller performance risk and privacy and
security risks are implicitly considered when a consumer decides to trust on a web
retailers competence, benevolence and integrity.
17
Literature Review
18
Literature Review
19
Literature Review
consumers to download music and movie clips as well as demo software and games for
evaluation before purchase (Kiang et al., 2011).
Based on the special characteristics of the internet Peterson et al. (1997) proposed a
specially designed classification system for online products and services. The system
includes a three-dimension classification scheme that distinguishes online and offline
channel impacts: cost and frequency of purchase, value proposition, and degree of
differentiation. The cost and frequency dimension ranges from inexpensive and
frequently purchased to expensive and infrequently purchased goods. According to
Peterson et al. (1997), individuals tend to avoid purchasing inexpensive and frequently
purchased goods online.
between tangible and intangible products. The third dimension, degree of differentiation
is related to the degree of product customization that creates competitive advantage. The
three dimensions are illustrated in Table 1.
Dimension 1
Dimension 2
Dimension 3
There are several different product classifications and contributions in the context of
online shopping, for example De Figueiredo (2000), based on information asymmetry
between sellers and buyers in e-commerce, classified products into a spectrum of four
categories that include: commodity (its quality is easily determined by its description),
quasi-commodity, look and feel, and look and feel with variable quality. De Figueiredo
(2000) found that the biggest increase in e-commerce occurred in product categories
which are near the commodity product side of the spectrum.
Perceived risk is an important element in online shopping, as mentioned previously in
this paper, and the effect of perceived risk may be subject to product characteristics
(Zhou et al., 2007). Bhatnagar et al. (2000) found that product risk is higher for
20
Literature Review
21
In this section, a conceptual model based on TPB is developed. The conceptual model
draws upon the idea credited to Taylor and Todd (1995) that TPB beliefs can be
decomposed into multidimensional constructs, additionally trust beliefs are integrated
into the model based on the literature review and empirical findings that support trust as
a major determinant of consumer online behavior; finally product type and its influence
on behavioral intention is incorporated in the analysis.
22
23
with intention. There is support for the dual role of PBC by Taylor and Todd (1995),
Lin (2007) and Pavlou and Fygenson (2006), who argued that neglecting PBC could
lead to an incomplete study of consumer online behavior. Hence, the following
hypotheses are proposed:
H4: A consumers PBC over online purchasing positively influences his/her
behavioral intention to purchase online (PBC BI).
H5: A consumers PBC over online purchasing positively influences his/her
actual online purchase (PBC B).
24
25
26
27
the internal notion of perceived ability, while the second component is related to
external constrains. Prior research has found significant links between these constructs
and PBC, either using the construct controllability or facilitating conditions as
similar concepts (Chen, 2009, Hsu and Chiu, 2004, Lin, 2007, Pavlou and Fygenson,
2006, Taylor and Todd, 1995).
Self-Efficacy
Self-efficacy is defined as an individuals self-confidence in his/her ability to perform a
behavior. In the context of the present study, self-efficacy is defined as a consumers
self-assessment of his/her capabilities to shop online as proposed by Vijayasarathy
(2004).
It is expected that higher levels of self-efficacy will cause higher levels of PBC, thus the
following hypothesis is presented:
H12: A consumers positive self-efficacy positively influences his/her perceived
behavioral control over online purchasing (SE PBC).
Facilitating Conditions
Facilitating conditions reflect the availability of resources needed to engage in a
behavior. Online purchasing requires resources such as time and money, thus more
resources available to the consumer lead to higher PBC. Hence, the following
hypothesis is proposed:
H13: A consumers positive facilitating conditions positively influence his/her
perceived behavioral control over online purchasing (FC PBC).
28
Papatla, 2010, Girard et al., 2003, Ian and Sui Meng, 2000). Some researchers focusing
on online purchase determinants also explored the influence of product types. Brown et
al. (2003) and So et al. (2005) found significant results between search and experience
products and online purchase intentions; Cha (2011) compared the factors that facilitate
or hinder purchase intention of real and virtual items and found significant results; and
Vijayasarathy (2002) whose study was based on Peterson et al. (1997) product
classification, suggests that tangibility of the product has a significant effect on
intention to shop online.
This study employs two of the dimensions of the product classification proposed by
Peterson et al. (1997): cost and value proposition. The third dimension, namely degree
of differentiation is not included due to practical reasons. The first dimension can be
distinguished between low cost products (e.g. CDs/DVDs) and high cost products (e.g.
refrigerator). Value proposition is an indication of the tangibility of the product and can
also be differentiated at two levels, tangible (e.g. clothing) and intangible (e.g.
software). The combination of the two dimensions yields the following four product
types:
(1) Low cost, tangible
(2) Low cost, intangible
(3) High cost, tangible
(4) High cost, intangible
Based on the aforementioned literature review, the following hypothesis is presented:
H14: A consumers behavioral intention to purchase online differs by product
type (Product type BI).
29
30
31
Construct
Definition
Reference
Perceived
Usefulness
Davis, 1989
Perceived
Ease of Use
Davis, 1989
Compatibility
Rogers, 1995
Trust
Interpersonal
Influence
External
Influence
Self-efficacy
Facilitating
Conditions
Behavioral
Intention
Attitude
Subjective
Norm
Perceived
Behavioral
Control
The belief that the trustee will act cooperatively to fulfill the
trustors expectations without exploiting its vulnerabilities
Influence by friends, family members, colleagues, superiors,
and experienced individuals known to the potential adopter
Mass media reports, expert opinions, and other non-personal
information considered by individuals in performing a behavior
An individual's self-confidence in his/her ability to perform a
behavior
Reflects the availability of resources needed to engage in a
behavior, such as time, money or other specialized resources
Motivational factors that capture how much effort a person is
willing to make in order to perform a behavior
An individuals evaluation of the outcome resulting from
performing a behavior
An individuals perception of normative social pressure to perform
a behavior
An individuals self-assessment of his or her capabilities to
perform a behavior
Pavlou and
Fygenson, 2006
Bhattacherjee,
2000
Bhattacherjee,
2000
Taylor and
Todd, 1995
Taylor and
Todd, 1995
Ajzen, 1991
Ajzen, 1991
Ajzen, 1991
Ajzen, 1991
32
Methodology
4. METHODOLOGY
This section presents the research methodology of this study. First, the research
approach is outlined; this is followed by instrument development, sampling and data
collection method.
33
Methodology
Table 3 summarizes the constructs and their respective sources and Appendix 1 presents
the detailed item information.
Constructs
Items
Sources
Compatibility (COM)
Trust (TR)
Bhattacherjee (2000)
Bhattacherjee (2000)
Self-Efficacy (SE)
Perceived Behavioral
Control (PBC)
Attitude (A)
The questionnaire was developed in English. The approach to testing the conceptual
model was based on the one used by Taylor and Todd (1995) to test a TPB model with
decomposed belief structures.
In order to gather the right group of respondents for the purpose of the study, the
questionnaire was established to start with a screening question: During the last 12
months, have you purchased any product/service using the internet?, those individuals
who answered no were excluded because they did not fit the target population.
Self-report method is used in this study where respondents are asked to report their
previous purchase experiences and behaviors. Respondents are asked to indicate which
products/services they have purchased online; they had the option to mark up to three
products/services. Actual purchasing behavior was measured with a single item, Please
indicate how many times you have purchased each product/service in the last 12
months, where a 7-point Likert scale included: Once, 2-3 times a year, 4-5 times a
year, once per 1 or 2 months, 2 times a month, 3 times a month, more than 3 times a
month. Prior research measured actual behavior in a similar way by asking purchase
frequency during a determined period of time with a Likert scale either indicating
34
Methodology
number of times purchased or frequency ranging from seldom to often (Barkhi et
al., 2008, Chen and Tan, 2004, Pavlou, 2003, Suh and Han, 2003).
At this point, respondents are requested to choose one product/service (from those
selected in the first part of the survey) in order to answer the rest of the questionnaire
with that product/service in mind. A program replaced the name of the product/service
selected by the respondent as this approach ensured that the respondent was consistent
in his or her answers to the questions.
All measurements used 7-point Likert scales ranging from strongly disagree to
strongly agree, except for item INT1 regarding internet experience with a scale
ranging from limited to significant, item INT2 where respondents indicated how
many hours per week they use the internet 2, the scale for item SE2 ranged from
difficult to easy and attitude 7-point Likert scales ranging from bad idea to good
idea, foolish idea to wise idea, dislike to like and unpleasant to pleasant.
The questionnaires last section included demographic questions such as gender, age,
country, education and marital status.
Based on the data set, INT2 was later transformed into groups, using SPSS -Recode variable functionforming a 7-point Likert scale in order to execute the analysis.
35
Methodology
36
This section presents data analysis and discusses the results. Sample characteristics are
presented first, followed by measurement validation. Then model assessment, path analysis
and hypotheses testing are provided. Finally, the results are discussed.
37
independent latent variables impacting it. The conceptual model proposed in this study
has no formative indicators and the largest number of independent latent variables that
impact the same dependent variable is four. Thus, the minimum required sample size for
this study is 40. The number of usable responses collected in the present study is 138 ,
therefore satisfying the minimum sample requirement in PLS analysis. For the control
variables, the approach by Liang et al. (2007) was used, where the control variables are
connected with the independent variable on SEM PLS analysis.
Gender
Age
Education
Marital status
Continent
Male
Female
<21
21-29
30-39
40-49
50-59
60 and over
No formal schooling
Primary school
High school
College/University
Post-graduate degree
Single
Married
Separated
Divorced
In a relationship
<10
11-20
21-30
31-40
41-50
51-60
>61
Europe
North America
South America
Other
Frequency
Percent
75
63
2
30
49
34
14
9
1
3
13
77
44
23
71
5
3
36
35
41
26
16
7
7
6
98
13
13
14
54
46
1
22
36
25
10
7
1
2
9
56
32
17
51
4
2
26
25
30
19
12
5
5
4
71
9
9
10
38
mainly Denmark, with the rest of the sample including North America, South America
and other countries from Asia. Table 4 summarizes some of the demographic attributes
of the respondents.
Since a screening question was used, all respondents have at least purchased online once
(during the last twelve months at the time of the survey), however it appears as most of
them have experience shopping online as the majority indicated three products and/or
services. Among the most frequently purchased items are travel services, clothing and
books. Table 5 presents the products and services purchased online by the respondents.
Products/services purchased online Frequency Percent
Travel
Clothing
Books
Tickets
Software
Electronics
CDs/DVDs
Personal care products
Sports equipment
Subscriptions
Appliances
Furniture
TOTAL
66
52
51
47
34
34
24
21
13
11
6
4
363
18
14
14
13
9
9
7
6
4
3
2
1
100
39
Frequency
Percent
Travel
40
29
Clothing
22
16
Books
20
14
Tickets
20
14
Electronics
13
Software
CDs/DVDs
Sports equipment
Subscriptions
Appliances
By using the snowball sampling technique there is a risk that the sample is not
representative of the population of online shoppers. Therefore non-response bias was
assessed by comparing the studys respondents with online shoppers in previous studies
that did not use student populations as their only target in the following dimensions:
gender, age and education (Aljukhadar and Senecal, 2011, Broekhuizen and Huizingh,
2009, Burkolter and Kluge, 2011, Verhagen et al., 2006, Vijayasarathy, 2002). In prior
studies gender distribution tends to be balanced, additionally, online shoppers tend to be
well educated and with a mean age of 33 years. Gender distribution and education in the
sample appear to be similar to prior studies samples; however respondents are slightly
older in the present study, t-tests showed no significant differences (p > 0.05). Since
half the sample comes from Denmark; current Danish online consumer demographics
were considered. According to a publication by Posten Norden Distanshandeln i Norden
2010 (Distance Trade in the Nordic Countries in 2010) a typical Danish online
consumer is a man between 30 and 49 years old (Euromonitor International, 2012c). In
addition, according to Danmarks Statistik, 41 percent of online shoppers are highly
educated and 51 percent of shoppers are men; the only parameter that differs slightly
with this studys sample is age, namely the younger 16 to 19 group, which represents 18
percent of online shoppers according to Danmarks Statistik (Appendix 4). Based on
these results, the sample appears to be representative of the population, although the
effect of non-response bias cannot be discounted entirely.
40
5.3.1.1 Reliability
All constructs consist of more than one item, except actual behavior (B), which was
measured by purchase frequency of the respondents selected product/service.
Cronbachs alpha was used to assess internal consistency, since it provides an estimate
for the reliability based on the indicator intercorrelations (Henseler et al., 2009). Alpha
coefficients range from 0 to 1 where higher coefficients indicate higher reliability. The
accepted value of Cronbachs alpha is 0.70, whereas a value below 0.6 indicates a lack
of reliability (Nunnally et al., 1967). Table 7 shows that all constructs present alpha
coefficients higher than 0.70, except for trust (TR) with 0.68 and internet experience
(INT) with 0.57.
# of
Indicators
Composite
Reliability
Cronbachs
Alpha
Attitude (A)
Behavior (B)
4
1
0.91
1.00
0.87
1.00
0.97
0.93
Compatibility (COM)
External Influence (EI)
3
3
0.92
0.88
0.87
0.79
Enjoyment (ENJ)
0.91
0.86
0.92
0.82
0.92
0.83
3
2
3
0.75
0.91
0.93
0.57
0.81
0.89
0.86
0.74
Self-Efficacy (SE)
0.91
0.81
0.93
0.86
Trust (TR)
0.82
0.68
Construct
41
measure. The composite reliability takes into account that indicators have different
loadings, and can be interpreted in the same way as Cronbachs alpha. The accepted
value for composite reliability is 0.70 or higher (Henseler et al., 2009). The composite
reliability values are shown on Table 7, the values for all constructs are above the cutoff
level. The averaged composite reliability for all constructs is 0.90 showing high
reliability. Therefore, it can be said that the measurement instrument of this study is
reliable.
5.3.1.2 Validity
For the assessment of validity, convergent and discriminant validities are used.
Convergent validity means that a set of indicators represents one and the same
underlying construct, which can be analyzed through their unidimensionality.
Discriminant validity is a complementary concept, meaning that each indicator should
not have a stronger connection with constructs other than the one it attempts to reflect.
Construct
AVE
Attitude (A)
0.72
Behavior (B)
1.00
0.93
Compatibility (COM)
0.80
0.71
Enjoyment (ENJ)
0.78
0.85
0.85
0.51
0.84
0.82
0.67
Self-Efficacy (SE)
0.84
0.88
Trust (TR)
0.61
Fornell and Larcker (1981) suggest using the average variance extracted (AVE) as a
criterion of convergent validity. AVE measures the amount of variance that a latent
variable captures from its indicators relative to the amount due to measurement error.
(Chin, 2010). An AVE value of at least 0.5 indicates sufficient convergent validity,
meaning that a latent variable is able to explain more than half of the variance of its
42
indicators on average (Henseler et al., 2009). AVE is only applicable for mode A
(outward-directed) reflective constructs or latent variables. The entire measurement
instrument in the present study is reflective, thus AVE is applicable to all constructs.
AVE values are shown in Table 8, all values are greater than 0.50, achieving convergent
validity.
There are two measures of discriminant validity: The Fornell-Larcker criterion and the
cross loadings (Henseler et al., 2009). The Fornell-Larcker criterion indicates that a
latent variable shares more variance with its assigned indicators than with any other
latent variable, in other words, the AVE of each latent variable should be greater than
the latent variables highest squared correlation with any other latent variable. The
second measure of discriminant validity takes into account the loading of each indicator,
where it is expected to be greater than all of its cross-loadings (Henseler et al., 2009).
Although the Fornell-Larcker criterion assesses discriminant validity on the construct
level, the cross loadings allow this evaluation on the indicator level (Chin, 2010). The
results of both Fornell-Larcker criterion (Appendix 6: Latent Variables Squared
Correlations and AVE) and cross loadings (Appendix 7: Cross Loadings) suggest that
all construct measurements have adequate discriminant validities.
43
R2
Attitude (A)
0.52
Behavior (B)
0.10
0.38
0.66
0.47
44
According to Cohen (1988) values of 0.02, 0.15, and 0.35 can be interpreted as small,
medium, and large effects at the structural level, respectively.
The
values were calculated manually for each latent variable. Table 10 presents the
results of effect size; all effects on behavior are small, attitude has a medium effect on
behavioral intention, PEOU appears to have a medium effect on attitude while the rest
of variables have a small effect on attitude. All effects on subjective norm and PBC are
above 0.15, representing medium effects.
Effects
Effects on Behavior:
Behavioral Intention
0.03
PBC
0.01
Enjoyment
0.03
Internet experience
0.01
0.23
Subjective Norm
0.02
PBC
0.04
Effects on Attitude:
Trust
0.01
Perceived Usefulness
0.04
0.20
Compatibility
0.04
0.22
External influence
0.16
Effects on PBC:
Self-efficacy
0.24
Facilitating Conditions
0.39
45
Q2
Attitude (A)
0.35
0.09
0.36
0.55
0.41
Behavior (B)
Behavioral Intention (BI)
Perceived Behavioral Control (PBC)
Subjective Norm (SN)
46
0.05 requires t-value > 1.645. Appendix 9 shows the graphical bootstrap output with tvalue for each path and Table 12 summarizes the results.
Hypotheses
Path
Coeff.
T Statistics
H1:
0.193
2.553**
H2:
0.440
4.963**
H3:
0.126
1.321
H4:
0.194
2.260*
H5:
PBC Behavior
-0.078
1.000
H6:
0.180
2.051*
H7:
0.404
4.608**
H8:
Compatibility Attitude
0.193
1.986*
H9:
Trust Attitude
0.111
1.266
H10:
0.421
4.238**
H11:
0.353
4.210**
H12:
Self-efficacy PBC
0.394
3.685**
H13:
0.491
5.279**
Note: one-tail
The results support the proposed relationships between behavioral intention and
behavior (H1) (t = 2.553, p < 0.01); attitude and behavioral intention (H2) (t = 4.963, p
< 0.01); perceived behavioral control and behavioral intention (H4) (t = 2.260, p <
0.05); perceived usefulness and attitude (H6) (t = 2.051, p < 0.05); perceived ease of use
and attitude (H7) (t = 4.608, p < 0.01); compatibility and attitude (H8) (t = 1.986, p <
0.05); interpersonal influence and subjective norm (H10) (t = 4.238, p < 0.01); external
influence and subjective norm (H11) (t = 4.210, p < 0.01); self-efficacy and perceived
behavioral control (H12) (t = 3.685, p < 0.01); facilitating conditions and perceived
behavioral control (H13) (t = 5.279, p < 0.01). The results do not support the proposed
relationships between subjective norm and behavioral intention (H3) (t = 1.321, p >
0.05); perceived behavioral control and behavior (H5) (t = 1.000, p > 0.05); trust and
attitude (H9) (t = 1.266, p > 0.05).
With regard to the control variables included in the model, shopping enjoyment (ENJ) is
significantly related with the behavior, namely online purchasing (t = 1.948, p < 0.05)
while internet experience (INT) is not significantly related. The total effect size of both
control variables is small, only 0.05; this is also the case if behavioral intention is used
as the focal variable for the control variables.
47
ANOVA was conducted to test if behavioral intention differs by product type (H14).
In order to obtain the dependent variable behavioral intention, measures of behavioral
intention BI1 and BI2 were transformed into one variable using SPSS Transform
Compute variable Mean function.
The products selected by the respondents were grouped with SPSS Variable
transformation function -recode into different variables. Products were grouped according
to the classification proposed by Peterson et al. (1997), Table 13 illustrates the results.
Product Classification
Product/Service
Frequency
Entertainment tickets,
subscriptions
22
Software, travel
48
54
Electronics, appliances
14
48
transformation (reflected). After the transformation, the value of skewness was -0.895
and for kurtosis -0.540. The values are in the interval (-1+1), thus it can be stated that
the variable does not deviate significantly from normality. P-P plots and Q-Q plots also
confirm that the distribution is closer to normal as the observed standardized residuals
are closely located around the 45 degree line (Appendix 10).
Homogeneity of variance
The second assumption regarding equal variance between the groups on the dependent
variable was assessed using Levenes Test of Homogeneity of Variance. The result is
shown in Table 14. The p-value is 0.092, meaning that the hypothesis of equal variances
cannot be rejected, thus, the homogeneity of variance assumption is satisfied.
Levene Statistic
df1
df2
Sig.
2.189
134
.092
df
Mean Square
Sig.
Between Groups
1.453
.484
1.716
.167
Within Groups
37.835
134
.282
Total
39.288
137
49
significant at the 0.026 level. It can be concluded that online purchase intention with
regards to tangible products differs when it comes to low vs. high cost. Therefore, H14
which postulated differences in intentions by product type is partially supported. In
addition, it can be inferred from Table 16 that among tangible products, purchase
intention tends to be higher for low cost products than high cost products.
N
Mean
Std. Deviation
22
6.182
1.24924
48
6.240
1.07678
54
6.398
1.03877
14
5.571
1.55486
Total
138
6.225
1.15745
Mean
Std. Deviation
Low trust
74
5.912
1.25610
High trust
64
6.586
0.91528
Total
138
6.225
1.15745
50
closer to normality. Equal variance between the groups on the dependent variable was
assessed using Levenes Test of Homogeneity of Variance; Table 18 illustrates that the
p-value is 0.000, meaning that the assumption of equal variances is not satisfied. As a
consequence it will be referred to the p-value of the Welch-test which can be conducted
if two groups have unequal variances.
Levene Statistic
df1
df2
Sig.
14.459
136
.000
The ANOVA result in Table 19 reveals that differences between low and high trust
groups in regard to behavioral intention differ significantly at the p-value of 0.000 (F =
16.341, p = 0.000). As the assumption of homogeneity was not fulfilled it is referred to
the Welch-test.
Sum of Squares
df
Mean Square
Sig.
Between Groups
4.069
4.069
15.714
.000
Within Groups
35.218
136
.259
Total
39.288
137
The Welch-test in Table 20 has a p-value of 0.000 and is significant. As 0.000 < 0.05,
H0 is rejected and H1 is accepted. There is evidence to support that consumers with
high levels of trust show higher intentions to purchase online, than consumers with
lower levels of trust on web retailers.
Welch
Statistica
df1
df2
Sig.
16.341
133.878
.000
51
Hypotheses
H1:
H2:
H3:
H4:
H5:
H6:
H7:
H8:
H9:
H10:
H11:
H12:
H13:
H14:
Relationship
Result
BI B
Supported
A BI
Supported
SN BI
Rejected
PBC BI
Supported
PBC B
Rejected
PU A
Supported
PEOU A
Supported
COM A
Supported
TR A
Rejected
II SN
Supported
EI SN
Supported
SE PBC
Supported
FC PBC
Supported
Product type
BI
Partially
Supported
52
53
Discussion
6. DISCUSSION
54
Discussion
convenience and time saving offered by web retailers are more willing to purchase
online. In addition to time saving, consumers who like to shop from home, avoiding big
crowds or maintain busy schedules may show a more positive attitude towards online
shopping.
Trust: The impact of trust on attitude toward online purchasing has been empirically
supported in the e-commerce literature (Chen and Tan, 2004, George, 2004, Ha and
Stoel, 2009, Jarvenpaa et al., 2000, Pavlou and Fygenson, 2006, Suh and Han, 2003).
However, unlike most studies, this study did not provide empirical evidence to support
that trust positively impacts attitude. A plausible explanation for this unexpected finding
is related to the participants prior experience purchasing online. In a study comparing
potential customers versus repeat customers in the online context, Gefen et al. (2003)
found that there are two separate populations regarding trust beliefs, since for repeat
customers trust is influenced by prior experience with web retailers and for potential
customers, trust is mainly based on their disposition to trust. The participants in this
study have experience shopping online and may not had have bad experiences from
their past online purchasing activities, therefore, trust does not play a significant role in
future interactions with web retailers as many may believe that payment by credit card
online might be as safe as payment in a physical store. Another possible reason for the
finding in this study could be due the fact that half of the sample is formed by Danish
respondents, who are probably familiar with Trust Pilot, which is an open communitybased platform for sharing real reviews of shopping experiences, Trust Pilot is ranked
among the top visited Danish websites (Alexa, 2012). Danish consumers may be also
familiar with the quality stamp E-mark that was developed with the support of the
Danish Ministry of Science with purpose of ensuring legal and ethical operations
(Euromonitor International, 2012c). Further, it is fair to say that many online stores
have established good reputations and in turn, many online consumers may choose to
transact with those online stores. To sum up, positive reviews from websites like Trust
Pilot and the E-mark quality stamp together with positive prior experiences shopping
online may have an effect on trusting beliefs which can render them not critical in future
interactions.
Low vs. High Trust groups and Behavioral Intention: Given the unexpected result
regarding the significance of trust in this study, additional analysis were performed in
order to get a better understanding of the role of trust on online behavioral intention.
55
Discussion
Many researchers have proposed the relationship between trust and attitude, as in this
study; nevertheless e-commerce literature has also found significant results when
analyzing the direct impact of trust on purchase intention (Gefen et al., 2003, Pavlou,
2003), thus additional analysis were performed in this study to be able to understand the
role of trust on purchase intentions. The results show that consumers who have high
levels of trust on web retailers express significantly more willingness to purchase online
than those consumers who have lower levels of trust. The results are somewhat aligned
with the extant literature that supports the relevant role of trust in consumer online
behavior. However, the dual role of trust as an attitude predictor and as an important
factor impacting purchase intention needs to be examined.
56
Discussion
behavior; however the presence of facilitating conditions may not, per se, encourage the
behavior.
57
Discussion
58
Discussion
tangibility. The results show that cost plays a significant role for tangible products, as
consumers indicate higher purchase intentions for low cost, tangible products when
compared to high cost, tangible products. No significant differences were found in
relation to intangible products, regardless of cost differences. When compared to
previous research, the results are mixed. Ian and Sui Meng (2000) found that low cost
and frequently purchased goods are more likely to be purchased online than high cost
and infrequently purchased goods; the study also found that intangible products are
related to higher willingness to buy than tangible products. Vijayasarathy (2002) found
that, cost does not have an impact on purchasing intentions while product tangibility
does, his study found that intentions to shop online where higher for intangible products
than tangible products. One possible explanation for the results in this study is that the
internet penetration and spread of web retailers has increased considerably in the last
decade, thus web retailers are dealing with a competitive environment where free
delivery and easy returns are now common, and as a result, low cost tangible products
show similar high levels of purchasing intention as intangible goods. Moreover, internet
shoppers are able to try out the demonstration version of computer software, or be given
trial periods of online newspapers, video and music samples, before making a purchase
decision. This reduces uncertainty in purchase decision and stimulates purchases.
However it appears that perceptions of risk related to high cost, tangible products (i.e.
financial and performance risks), are still making an impact on consumers, in other
words, web retailers of high cost tangible products still face challenges of narrowing the
sensory gap that exists between their products and online consumers.
59
This section presents the summary and conclusion of the study. First a summary of
findings is presented, then managerial implications, limitations and finally
recommendations for future research are presented.
60
this, as control beliefs become particularly relevant considering the virtual and
impersonal nature of online shopping, furthermore, engaging in online purchasing
implies the use of technology and consumers who are confident about their capabilities
are more willing to make purchases online.
It was hypothesized that online purchase intention would differ by product type. In this
study products and services were categorized based on Peterson et al. (1997)
classification scheme, using cost and tangibility as two major characteristics. The results
showed that consumers are more willing to purchase low cost, tangible products and are
less willing to purchase high cost, tangible products over the internet, while no
significant differences were found in relation to intangible products, regardless of their
cost.
This study draws from previous research and presents a relatively comprehensive, yet
parsimonious model to describe and predict online consumer behavior. The empirical
findings not only offer more insight into the factors that impact online purchasing
behavior but also further empirically support the theory of planned behavior in the ecommerce context.
61
advantages such as useful product information and ease to compare products and prices
online.
Higher levels of trust are associated with higher willingness to shop online, maintaining
clear shipping and return policies, as well as a secure check out process is fundamental.
Marketers must communicate consumers that online shopping can be convenient, safe
and simple to use.
With regard to product types, the findings suggest that high cost, tangible products are
associated with higher levels of risk, thus consumers are less willing to purchase this
type of products online. High costs imply a potential financial risk and consumers are
often limited in terms of touch, feel and smell when purchasing tangible products over
the internet, as a result performance risks are enhanced. The more marketers know
about the path to purchase for a particular product type, the more relevant they can
make their messages, for example providing consumers with broad and useful
information for research-heavy, high cost, tangible products such as electronics.
7.3 Limitations
The main limitation of the present study is related to the sampling method, as a result of
using the snowball technique, the chances of having participants with similar traits are
higher (people tend to associate with those similar to them, thereby sharing the survey
with those contacts in their network and so forth). Although there were no significant
differences regarding gender distribution, education and age of the participants
compared to similar prior studies samples, participants in this study were slightly older
in average, hence non-response bias cannot be entirely discounted.
This study explored some of the key factors affecting online purchasing, where
measurement of purchasing behavior was based on purchasing frequencies, which can
be considered a limitation. First, self-report methods have disadvantages related to the
accuracy of the information provided by the participants, since recalling information
from the last twelve months can be challenging. Second, in order to get information of
the actual behavior, longitudinal studies or lab experiments provide more accurate and
real information.
62
63
References
8. REFERENCES
ADAMS, J., KHAN, H. T. A., RAESIDE, R. & WHITE, D. 2007. Research methods
for graduate business and social science students, Sage.
AJZEN, I. 1991. The theory of planned behavior. Organizational behavior and human
decision processes, 50, 179-211.
AJZEN, I. 2002. Perceived Behavioral Control, SelfEfficacy, Locus of Control, and the
Theory of Planned Behavior. Journal of applied social psychology, 32, 665-683.
ALEXA. 2012. Trust Pilot, Statistics Summary [Online]. Available:
http://www.alexa.com/siteinfo/trustpilot.com [Accessed November 2012].
ALJUKHADAR, M. & SENECAL, S. 2011. Segmenting the online consumer market.
Marketing Intelligence & Planning, 29, 421-435.
BARKHI, R., BELANGER, F. & HICKS, J. 2008. A model of the determinants of
purchasing from virtual stores. Journal of Organizational Computing and
Electronic Commerce, 18, 177-196.
BENEDICKTUS, R. L., BRADY, M. K., DARKE, P. R. & VOORHEES, C. M. 2010.
Conveying Trustworthiness to Online Consumers: Reactions to Consensus,
Physical Store Presence, Brand Familiarity, and Generalized Suspicion. Journal
of Retailing, 86, 322-335.
BHATNAGAR, A., MISRA, S. & RAO, H. R. 2000. On risk, convenience and internet
shopping behavior. Communications of the ACM, 43, 98-105.
BHATTACHERJEE, A. 2000. Acceptance of e-commerce services: the case of
electronic brokerages. Systems, Man and Cybernetics, Part A: Systems and
Humans, IEEE Transactions on, 30, 411-420.
BROEKHUIZEN, T. & HUIZINGH, E. K. R. E. 2009. Online purchase determinants:
Is their effect moderated by direct experience? Management Research News, 32,
440-457.
BROWN, M., POPE, N. & VOGES, K. 2003. Buying or browsing?: An exploration of
shopping orientations and online purchase intention. European Journal of
Marketing, 37, 1666-1684.
BURKOLTER, D. & KLUGE, A. 2011. Online consumer behavior and its relationship
with socio-demographics, shopping orientations, need for emotion, and fashion
leadership. Journal of Business, 2.
CHA, J. 2011. Exploring the Internet as a Unique Shopping Channel to Sell Both Real
and Virtual Items: a Comparison of Factors Affecting Purchase Intention and
Consumer Characteristics. Journal of Electronic Commerce Research, 12, 115124.
CHEEMA, A. & PAPATLA, P. 2010. Relative importance of online versus offline
information for Internet purchases: Product category and Internet experience
effects. Journal of Business Research, 63, 979-985.
64
References
65
References
GEORGE, J. F. 2004. The theory of planned behavior and Internet purchasing. Internet
research, 14, 198-212.
GIRARD, T., KORGAONKAR, P. & SILVERBLATT, R. 2003. Relationship of Type
of Product, Shopping Orientations, and Demographics with Preference for
Shopping on the Internet. Journal of Business & Psychology, 18, 101-120.
HA, S. & STOEL, L. 2009. Consumer e-shopping acceptance: Antecedents in a
technology acceptance model. Journal of Business Research, 62, 565-571.
HAENLEIN, M. & KAPLAN, A. M. 2004. A beginner's guide to partial least squares
analysis. Understanding statistics, 3, 283-297.
HAIR, J., SARSTEDT, M., RINGLE, C. & MENA, J. 2012. An assessment of the use
of partial least squares structural equation modeling in marketing research.
Journal of the Academy of Marketing Science, 40, 414-433.
HANSEN, T. & MLLER JENSEN, J. 2009. Shopping orientation and online clothing
purchases: the role of gender and purchase situation. European Journal of
Marketing, 43, 1154-1170.
HANSEN, T., MLLER JENSEN, J. & STUBBE SOLGAARD, H. 2004. Predicting
online grocery buying intention: a comparison of the theory of reasoned action
and the theory of planned behavior. International Journal of Information
Management, 24, 539-550.
HENSELER, J., RINGLE, C. M. & SINKOVICS, R. R. 2009. The use of partial least
squares path modeling in international marketing. Advances in international
marketing, 20, 277-319.
HERNNDEZ, B., JIMNEZ, J. & MARTN, M. J. 2010. Customer behavior in
electronic commerce: The moderating effect of e-purchasing experience. Journal
of Business Research, 63, 964-971.
HSU, M.-H. & CHIU, C.-M. 2004. Predicting electronic service continuance with a
decomposed theory of planned behaviour. Behaviour & Information Technology,
23, 359-373.
IAN, P. & SUI MENG, P. 2000. Factors influencing the types of products and services
purchased over the Internet. Internet research, 10, 102-113.
IBM CORP. 2010. IBM SPSS Statistics for Windows. 19.0 ed. Armonk, NY: IBM
Corp.
INTERNET RETAILER. 2012. Forrester Research: US Smartphone Commerce
Forecast [Online]. Available: http://www.internetretailer.com/trends/webtechnology/ [Accessed November 2012].
JARVENPAA, S. L., TRACTINSKY, N. & VITALE, M. 2000. Consumer trust in an
Internet store. Information Technology and Management, 1, 45-45.
JAYAWARDHENA, C., WRIGHT, L. T. & DENNIS, C. 2007. Consumers online:
intentions, orientations and segmentation. International Journal of Retail &
Distribution Management, 35, 515-526.
KEISIDOU, E., SARIGIANNIDIS, L. & MADITINOS, D. 2011. Consumer
characteristics and their effect on accepting online shopping, in the context of
different product types. Int. Journal of Business Science and Applied
Management, 6.
66
References
KIANG, M. Y., YE, Q., HAO, Y., CHEN, M. & LI, Y. 2011. A service-oriented
analysis of online product classification methods. Decision Support Systems, 52,
28-39.
KIM, E. Y. & KIM, Y. K. 2004. Predicting online purchase intentions for clothing
products. European Journal of Marketing, 38, 883-897.
KIM, J. 2012. An empirical study on consumer first purchase intention in online
shopping: integrating initial trust and TAM. Electronic Commerce Research, 12,
125-150.
KIM, Y. K., KIM, E. Y. & KUMAR, S. 2003. Testing the behavioral intentions model
of online shopping for clothing. Clothing and Textiles Research Journal, 21, 3240.
KORGAONKAR, P., SILVERBLATT, R. & GIRARD, T. 2006. Online retailing,
product classifications, and consumer preferences. Internet research, 16, 267288.
LEE, K. S. & TAN, S. J. 2003. E-retailing versus physical retailing: A theoretical model
and empirical test of consumer choice. Journal of Business Research, 56, 877885.
LIAN, J.-W. & LIN, T.-M. 2008. Effects of consumer characteristics on their
acceptance of online shopping: Comparisons among different product types.
Computers in Human Behavior, 24, 48-65.
LIANG, H., SARAF, N., HU, Q. & XUE, Y. 2007. Assimilation of enterprise systems:
The effect of institutional pressures and the mediating role of top management.
MIS Quarterly, 31, 59-87.
LIN, H.-F. 2007. Predicting consumer intentions to shop online: An empirical test of
competing theories. Electronic Commerce Research and Applications, 6, 433442.
LING, K. C., CHAI, L. T. & PIEW, T. H. 2010. The Effects of Shopping Orientations,
Online Trust and Prior Online Purchase Experience toward Customers Online
Purchase Intention. International Business Research, 3, P63.
LING, K. C., DAUD, D. B., PIEW, T. H., KEOY, K. H. & HASSAN, P. 2011.
Perceived Risk, Perceived Technology, Online Trust for the Online Purchase
Intention in Malaysia. International Journal of Business and Management, 6,
167-182.
LOWENGART, O. & TRACTINSKY, N. 2001. Differential effects of product category
on shoppers selection of Web-based stores: a probabilistic modeling approach.
Journal of Electronic Commerce Research, 2, 12-26.
MCKECHNIE, S., WINKLHOFER, H. & ENNEW, C. 2006. Applying the technology
acceptance model to the online retailing of financial services. International
Journal of Retail & Distribution Management, 34, 388-410.
MCKNIGHT, D. H., CHOUDHURY, V. & KACMAR, C. 2002. Developing and
Validating Trust Measures for e-Commerce: An Integrative Typology.
Information Systems Research, 13, 334-359.
67
References
68
References
VERHAGEN, T., SELMAR, M. & YAO-HUA, T. 2006. Perceived risk and trust
associated with purchasing at electronic marketplaces. European Journal of
Information Systems, 15, 542-556.
VIJAYASARATHY, L. R. 2002. Product characteristics and Internet shopping
intentions. Internet research, 12, 411-426.
VIJAYASARATHY, L. R. 2004. Predicting consumer intentions to use on-line
shopping: the case for an augmented technology acceptance model. Information
& Management, 41, 747-762.
WEN, C., PRYBUTOK, V. R. & XU, C. 2011. An Integrated Model for Customer
Online Repurchase Intention. The Journal of Computer Information Systems, 52,
14-23.
YOH, E., DAMHORST, M. L., SAPP, S. & LACZNIAK, R. 2003. Consumer Adoption
of the Internet: The Case of Apparel Shopping. Psychology & Marketing, 20,
1095-1118.
ZARMPOU, T., SAPRIKIS, V., MARKOS, A. & VLACHOPOULOU, M. 2012.
Modeling users acceptance of mobile services. Electronic Commerce Research,
1-24.
ZHOU, L., DAI, L. & ZHANG, D. 2007. Online shopping acceptance modela critical
survey of consumer factors in online shopping. Journal of Electronic Commerce
Research, 8, 41-62.
69
Appendices
9. APPENDICES
APPENDICES
70
Appendices
71
INT1: What is your level of experience with the use of Internet? (limited/significant)
Shopping
Enjoyment
(ENJ)
Perceived
Usefulness
(PU)
Perceived
Ease of
Use (PEOU)
INT2: On average how many hours per week do you spend on the Internet?
INT3: Using the Internet is very important to me.
ENJ2: Under the right circumstances, shopping is fun.
ENJ3: For me, shopping is a pleasurable activity.
PU2: Shopping online for product x provides access to useful shopping information.
PU3: Shopping online saves me time when purchasing product x.
PEOU1: Shopping online for product x to me is clear and easy to understand.
PEOU2: I find shopping online for product x easy to do.
PEOU3: It would be easy for me to become skilled at shopping online for product x.
COM1: Shopping online for product x fits well with my lifestyle.
Compatibility
(COM)
COM2: Shopping online for product x fits well with my shopping needs.
COM3: Shopping online for product x is compatible with the way I like to shop.
Sources
Barkhi et al.
(2008)
Jarvenpaa et al.
(2000)
Lin (2007)
adapted from
Davis (1989)
Lin (2007)
adapted from
Davis (1989)
Vijayasarathy
(2004) adapted
from Taylor and
Todd (1995)
Interpersonal II1: My friends or family think that shopping online for product x is a good idea.
Influence (II) II2: My friends or family encourage me to shop online for product x.
EI1: I have read/seen news reports which say that online shopping provides a good way of
purchasing product x.
External
Influence (EI) EI2: The popular press adopts a positive view towards online shopping for product x.
EI3: Mass media reports have influenced me to try online shopping to purchase product x.
SN1: People who are important to me would recommend that I purchase product x
Subjective online.
Norm (SN)
SN2: Most of the people who I value would purchase product x online.
SE1:If I wanted to, for me to purchase product x online would be ...(difficult/easy)
Self-Efficacy
(SE)
Facilitating
Conditions
(FC)
Gefen et al.
(2003)
Bhattacherjee
(2000)
Bhattacherjee
(2000)
Pavlou and
Fygenson (2006)
Lin (2007)
adapted from
Taylor and Todd
(1995)
Lin (2007)
adapted from
Taylor and Todd
(1995)
Pavlou and
Fygenson (2006)
Pavlou and
Fygenson (2006)
Pavlou and
Fygenson (2006)
Appendices
72
Appendices
Appendix 2: Web-Based questionnaire (cont.)
73
Appendices
Appendix 2: Web-Based questionnaire (cont.)
74
Appendices
75
Appendices
Appendix 2: Web-Based questionnaire (cont.)
76
Appendices
Once
Appliances
Books
CDs/DVDs
Clothing
Electronics
Furniture
Personal care
Software
Sports equipment
Subscriptions
Tickets
Travel
0
10
12
14
25
30
35
10
15
20
Clothing
Electronics
Personal care
Software
Sports equipment
Tickets
Travel
0
10
15
20
77
Appendices
Appendix 3: Purchase Frequencies (cont.)
10
12
2 times a month
Books
Clothing
Software
Tickets
0
3 times a month
Books
Clothing
Electronics
Software
Tickets
0
78
Appendices
Appendix 3: Purchase Frequencies (cont.)
Books
Software
79
Appendices
80
Appendices
81
Appendices
82
Appendices
AVE (highlighted)
83
Appendices
A1
A2
A3
A4
BI1
BI2
COM1
COM2
COM3
EI1
EI2
EI3
ENJ1
ENJ2
ENJ3
FC1
FC2
II1
II2
INT1
INT2G
INT3
PB4
PBC1
PBC2
PEOU1
PEOU2
PEOU3
PU1
PU2
PU3
SE1
SE2
SN1
SN2
TR1
TR2
TR3
BI
COM
0.911
0.847
0.876
0.751
0.547
0.549
0.477
0.459
0.528
0.294
0.363
0.015
0.013
0.155
0.204
0.559
0.424
0.397
0.235
0.343
-0.045
0.408
0.101
0.424
0.473
0.613
0.615
0.528
0.492
0.411
0.339
0.670
0.512
0.233
0.269
0.415
0.292
0.486
0.056
0.047
0.075
0.173
0.200
0.227
0.210
0.192
0.208
-0.097
-0.021
-0.075
0.174
0.197
0.168
0.104
0.123
-0.083
-0.053
0.050
0.093
0.203
1.000
0.120
-0.081
0.150
0.089
0.085
0.103
0.112
0.200
0.016
0.118
0.023
0.056
0.071
0.031
0.066
0.556
0.416
0.474
0.467
0.968
0.964
0.427
0.351
0.465
0.170
0.255
0.067
0.038
0.091
0.094
0.464
0.372
0.329
0.201
0.364
0.036
0.402
0.221
0.443
0.378
0.414
0.348
0.329
0.296
0.270
0.311
0.533
0.623
0.260
0.327
0.277
0.220
0.234
0.482
0.398
0.537
0.432
0.443
0.459
0.912
0.883
0.886
0.242
0.358
0.110
-0.079
0.065
0.045
0.548
0.473
0.285
0.200
0.273
0.143
0.443
0.228
0.483
0.399
0.461
0.502
0.410
0.342
0.406
0.417
0.400
0.328
0.335
0.362
0.490
0.475
0.454
84
Appendices
Appendix 7: Cross Loadings (cont.)
A1
A2
A3
A4
BI1
BI2
COM1
COM2
COM3
EI1
EI2
EI3
ENJ1
ENJ2
ENJ3
FC1
FC2
II1
II2
INT1
INT2G
INT3
PB4
PBC1
PBC2
PEOU1
PEOU2
PEOU3
PU1
PU2
PU3
SE1
SE2
SN1
SN2
TR1
TR2
TR3
EI
ENJ
FC
II
0.303
0.222
0.197
0.251
0.215
0.185
0.271
0.247
0.279
0.871
0.902
0.744
0.200
0.160
0.165
0.169
0.147
0.419
0.593
0.058
0.005
0.161
-0.070
0.078
0.150
0.243
0.197
0.081
0.296
0.279
0.171
0.155
0.190
0.618
0.474
0.440
0.160
0.264
0.065
0.017
0.060
0.353
0.101
0.063
0.045
-0.051
0.038
0.172
0.137
0.207
0.818
0.898
0.927
0.014
-0.008
0.017
0.058
0.005
-0.068
0.184
0.205
-0.074
-0.074
0.179
0.122
0.127
0.073
0.186
0.124
-0.002
0.022
0.017
0.078
0.296
0.075
0.205
0.486
0.437
0.549
0.305
0.450
0.420
0.516
0.553
0.420
0.153
0.247
-0.014
-0.061
0.001
0.069
0.910
0.930
0.306
0.224
0.373
0.049
0.316
0.124
0.786
0.576
0.463
0.468
0.490
0.368
0.358
0.267
0.608
0.624
0.283
0.295
0.323
0.299
0.496
0.342
0.350
0.264
0.178
0.333
0.209
0.212
0.191
0.285
0.567
0.555
0.237
0.132
-0.048
0.039
0.244
0.275
0.913
0.935
0.125
-0.033
0.179
-0.073
0.215
0.265
0.213
0.138
0.123
0.377
0.321
0.229
0.175
0.294
0.586
0.568
0.232
0.084
0.194
85
Appendices
A1
A2
A3
A4
BI1
BI2
COM1
COM2
COM3
EI1
EI2
EI3
ENJ1
ENJ2
ENJ3
FC1
FC2
II1
II2
INT1
INT2G
INT3
PB4
PBC1
PBC2
PEOU1
PEOU2
PEOU3
PU1
PU2
PU3
SE1
SE2
SN1
SN2
TR1
TR2
TR3
INT
PBC
PEOU
PU
0.406
0.227
0.333
0.223
0.371
0.373
0.461
0.364
0.351
0.124
0.228
-0.064
0.021
0.134
0.152
0.331
0.273
0.194
0.087
0.626
0.545
0.914
0.196
0.342
0.221
0.315
0.297
0.301
0.346
0.266
0.209
0.418
0.325
0.210
0.197
0.149
0.259
0.243
0.453
0.400
0.535
0.236
0.473
0.398
0.427
0.494
0.390
0.099
0.194
-0.033
-0.135
-0.047
-0.033
0.652
0.736
0.333
0.156
0.405
0.033
0.300
0.035
0.937
0.893
0.457
0.420
0.488
0.344
0.291
0.253
0.662
0.666
0.345
0.313
0.228
0.262
0.433
0.604
0.489
0.603
0.499
0.363
0.420
0.454
0.477
0.436
0.248
0.238
-0.027
0.073
0.128
0.222
0.500
0.466
0.298
0.046
0.305
0.027
0.356
0.121
0.474
0.443
0.905
0.916
0.888
0.393
0.368
0.461
0.589
0.433
0.108
0.139
0.359
0.304
0.576
0.480
0.482
0.420
0.364
0.370
0.313
0.407
0.420
0.422
0.312
0.357
0.076
0.105
0.118
0.180
0.401
0.356
0.402
0.319
0.205
0.005
0.399
0.161
0.359
0.308
0.511
0.405
0.409
0.887
0.878
0.662
0.436
0.341
0.291
0.288
0.314
0.391
0.319
86
Appendices
A1
A2
A3
A4
BI1
BI2
COM1
COM2
COM3
EI1
EI2
EI3
ENJ1
ENJ2
ENJ3
FC1
FC2
II1
II2
INT1
INT2G
INT3
PB4
PBC1
PBC2
PEOU1
PEOU2
PEOU3
PU1
PU2
PU3
SE1
SE2
SN1
SN2
TR1
TR2
TR3
SE
SN
TR
0.622
0.562
0.592
0.398
0.629
0.590
0.331
0.386
0.350
0.212
0.259
-0.043
-0.040
0.010
0.062
0.672
0.573
0.335
0.152
0.533
0.049
0.381
0.074
0.732
0.579
0.519
0.458
0.541
0.408
0.353
0.258
0.915
0.916
0.258
0.218
0.234
0.262
0.365
0.241
0.206
0.250
0.207
0.350
0.250
0.345
0.267
0.376
0.450
0.596
0.405
0.092
0.031
0.008
0.238
0.324
0.529
0.606
0.090
0.087
0.224
0.042
0.326
0.319
0.167
0.130
0.050
0.281
0.304
0.157
0.153
0.313
0.941
0.931
0.303
0.090
0.201
0.496
0.353
0.494
0.417
0.282
0.318
0.516
0.487
0.588
0.364
0.368
0.206
0.112
0.268
0.301
0.458
0.446
0.234
0.191
0.199
0.102
0.261
0.075
0.365
0.379
0.532
0.511
0.437
0.291
0.371
0.405
0.371
0.311
0.209
0.289
0.829
0.652
0.842
87
Appendices
88
Appendices
89
Appendices
90
Appendices
91
Mean
Difference
(I-J)
Std. Error
High cost,
-0.00785
intangible
Low cost,
-0.0985
tangible
High cost,
0.26138
tangible
High cost, intangible
Low cost,
0.00785
intangible
Low cost,
-0.09065
tangible
High cost,
0.26923
tangible
Low cost, tangible
Low cost,
0.0985
intangible
High cost,
0.09065
intangible
High cost,
.35988*
tangible
High cost, tangible
Low cost,
-0.26138
intangible
High cost,
-0.26923
intangible
Low cost,
-.35988*
tangible
*The mean difference is significant at the 0.05 level.
0.13681
0.954
-0.2784
0.2627
0.1344
0.465
-0.3643
0.1673
0.18166
0.153
-0.0979
0.6207
0.13681
0.954
-0.2627
0.2784
0.10541
0.391
-0.2991
0.1178
0.1614
0.098
-0.05
0.5885
0.1344
0.465
-0.1673
0.3643
0.10541
0.391
-0.1178
0.2991
0.15936
0.026
0.0447
0.6751
0.18166
0.153
-0.6207
0.0979
0.1614
0.098
-0.5885
0.05
0.15936
0.026
-0.6751
-0.0447
(J) Product
types
Sig.
95% Confidence
Interval
Lower
Upper
Bound
Bound