Page 1
CONSUMER SATISFACTION AND REPURCHASE INTENTION FROM
CROSS-BORDER E-COMMERCE: A TRUST-RISK-BASED STUDY
BY
MR. NATTHAKORN KHAYAIYAM
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF MASTER OF SCIENCE PROGRAM
(MANAGEMENT INFORMATION SYSTEMS)
MANAGEMENT INFORMATION SYSTEMS
FACULTY OF COMMERCE AND ACCOUNTANCY
THAMMASAT UNIVERSITY
ACADEMIC YEAR 2018
COPYRIGHT OF THAMMASAT UNIVERSITY
Ref. code: 25615802037480DCY
Page 2
CONSUMER SATISFACTION AND REPURCHASE INTENTION FROM
CROSS-BORDER E-COMMERCE: A TRUST-RISK-BASED STUDY
BY
MR. NATTHAKORN KHAYAIYAM
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF MASTER OF SCIENCE PROGRAM
(MANAGEMENT INFORMATION SYSTEMS)
MANAGEMENT INFORMATION SYSTEMS
FACULTY OF COMMERCE AND ACCOUNTANCY
THAMMASAT UNIVERSITY
ACADEMIC YEAR 2018
COPYRIGHT OF THAMMASAT UNIVERSITY
Ref. code: 25615802037480DCY
Page 4
(1)
Thesis Title CONSUMER SATISFACTION AND REPURCHASE
INTENTION FROM CROSS-BORDER
E-COMMERCE: A TRUST-RISK-BASED STUDY
Author Mr. Natthakorn Khayaiyam
Degree Master of Science Program
(Management Information Systems)
Major Field/Faculty/University Management Information Systems
Commerce and Accountancy
Thammasat University
Thesis Advisor Professor Siriluck Rotchanakitumnuai, Ph.D.
Academic Years 2018
ABSTRACT
The waves of cross-border e-commerce (CBEC) is growing very fast resulting
big opportunities as well as threats to firms. Understanding consumer perception
towards cross-border e-commerce is crucial as it would result in consumer loyalty and
sustainable profit. This study is formulated based on a question that how trust and
perceived risk of the consumer influence repurchase from CBEC in both pre and post
purchase amongst Thai e-shoppers. The data was collected via a designated online
questionnaire distributed via social media and sharable link. Structural Equation Model
(SEM) and Confirmatory Factor Analysis was conducted to analyse the data together
with Paired T-test for analysing differences of trust and perceived risk in pre-to-post
purchase and domestic against cross-border e-commerce. The SEM result suggests that
post-purchase trust and perceived risk are highly influence to consumer satisfaction
and repurchase intention. Surprisingly, consumer tend to expose themselves to risk
before shopping; however, higher trust and lower risk are shaped firmly after they
experienced shopping from CBEC. Moreover, consumer do not perceived e-seller trust
and risk much difference between CBEC and domestic e-commerce. This study extends
the current Expectation-disconfirmation Theory and knowledge of trust and perceived
risk towards cross-border e-commerce. Furthermore, the implication of this study helps
Ref. code: 25615802037480DCY
Page 5
(2)
firms to emphasize on the trust and perceived risk in CBEC challenge within Thailand
and for firm who would like to expand to cross-border business.
Keywords: e-commerce, cross-border e-commerce, trust, perceived risk, repurchase
intention, Expectation-disconfirmation theory.
Ref. code: 25615802037480DCY
Page 6
(3)
ACKNOWLEDGEMENTS
I would like to express my deep gratitude to Professor Siriluck
Rotchanakitumnuai my research supervisors, for her patient guidance, enthusiastic
encouragement, and useful critiques of this work. I would like to thank Assistant
Professor Mathuspayas Thongmak, the chairman of my research, for her advice and
assistance in reviewing my research. Moreover, I would like to thank you Assistant
Professor Chatpong Tangmanee from Department of Statistics, Chulalongkorn
University, for his valuable guidance on statistical analysis and Structural Equation
Modelling used in this research. My grateful thanks are also extended to officers in
Master of Science Program in Management Information Systems, Thammasat Business
School, who always help and offer me the advices needed for research. I would like
to thank to Library of Thammasat University who provides necessary tools for thesis
development and
Finally, I wish to thank my parents and colleagues for their support and
encouragement throughout my study.
Mr. Natthakorn Khayaiyam
Ref. code: 25615802037480DCY
Page 7
(4)
TABLE OF CONTENTS
Page
ABSTRACT (1)
ACKNOWLEDGEMENTS (3)
LIST OF TABLES (7)
LIST OF FIGURES (8)
LIST OF ABBREVIATIONS (9)
CHAPTER 1 INTRODUCTION 1
1.1 Background of e-commerce 1
1.1.1 Electronics Commerce and Cross-border Electronic 1
Commerce
1.1.2 Global E-commerce Economy 2
1.1.3 Global Cross-border E-commerce Economy 3
1.1.4 Southeast Asia and Thailand Cross-border E-commerce 3
Economy
1.2 Rational of the study 5
1.3 Research questions 6
1.3 Objectives 6
CHAPTER 2 REVIEW OF LITERATURE 8
2.1 Theory 8
2.1.1 Trust-based Consumer Decision Making Model 8
2.1.2 Expectation-Confirmation Theory 9
Ref. code: 25615802037480DCY
Page 8
(5)
2.2 Related research 10
2.2.1 Trust 10
2.2.2 Perceived Risk 13
2.2.3 Interaction between Trust and Perceived Risk 16
2.2.4 Incorporating Trust in Expectation-disconfirmation theory 16
CHAPTER 3 RESEARCH METHODOLOGY 19
3.1 Conceptual model 19
3.2 Variable definition 20
3.3 Research hypotheses 21
3.4 Population and samples 25
3.5 Research Instruments and Data Collection 26
3.6 Data analysis Process 34
CHAPTER 4 RESULTS AND DISCUSSION 38
4.1 Descriptive statistics 38
4.1.1 Variables Descriptive Statistics and Normality of Residuals 38
4.1.2 Measurement Model Validity and Reliability 42
4.1.3 Samples demographics 45
4.2 Paired t-test for Mean Difference 49
4.3 Confirmatory Factor Analysis 52
4.4 Structural Equation Model 54
4.5 Hypothesis testing summary 57
CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS 61
5.1 Conclusion 61
5.2 Benefit 62
5.2.1 Theoretical benefits 63
Ref. code: 25615802037480DCY
Page 9
(6)
5.2.2 Practical benefits 63
5.3 Research limitation 64
5.4 Recommendation for future research 64
REFERENCES 65
APPENDICES
APPENDIX A QUESTIONNAIRE 72
APPENDIX B CONFIRMATORY FACTOR ANALYSIS DETAIL 79
APPENDIX C MEASUREMENT MODEL RESIDUAL VARIANCE 82
APPENDIX D MEASUREMENT MODEL CORRELATION 84
BIOGRAPHY 86
Ref. code: 25615802037480DCY
Page 10
(7)
LIST OF TABLES
Tables Page
3.1 Measurement items in pre-purchase and post-purchase stage 28
3.2 Model Validity and Reliability Analysis Criteria 35
3.3 SEM Recommended Fit Index and Criteria 36
3.4 SEM Fit Index Combination Criteria. 37
4.1 Factor Descriptive Statistics 38
4.2 Assessment of Multivariate Normality 41
4.3 Construct Summary 42
4.4 Cross-border E-commerce Variable Correlations for Discriminant
Validity
44
4.5 Measurement Model Validity and Reliability Analysis Result 45
4.6 Gender 46
4.7 Monthly Income 47
4.8 Highest Education Obtained 47
4.9 Occupation 48
4.10 Pair Summary for Paired T-test of Mean Difference 49
4.11 Paired Samples Test of Mean between Pre-purchase and
Post-purchase from Cross-border E-commerce and Domestic
E-commerce.
51
4.12 Confirmatory Factor Analysis Coefficients and Squared Multiple
Coefficients.
52
4.13 Structural Model Goodness of Fit Evaluation Summary. 54
4.14 Standardized Coefficients Effect in Structural Model. 56
4.15 Paired T-test Hypothesis Testing Result. 57
4.16 SEM Hypothesis Testing Result. 58
B.1 Confirmatory Factor Analysis Result. 79
C.1 Residuals Variance Summary. 82
D.1 Measurement Model Correlation Matrix. 84
Ref. code: 25615802037480DCY
Page 11
(8)
LIST OF FIGURES
Figures Page
1.1 Online Retail Market Size Year-over-year. 2
1.2 Global E-Commerce Market Share. 3
1.3 Top 5 Shopping Application in Southeast Asia in 2017. 4
2.1 Trust-based Consumer Decision Making Model. 9
2.2 Expectation-Confirmation Decision Theory. 10
2.3 McKnight and Chervany ‘s Trust dimensions. 11
2.4 Featherman and Pavlou’s Risk Facets. 14
2.5 Lanktan et al.’s integrating technology Trust in EDT. 17
3.1 Proposed research model. 19
4.1 Histogram of sample age distribution. 46
4.2 Frequency of participants’ purchase via cross-border e-commerce. 48
4.3 Structural Model Standardised Coefficients. 55
B.1 Structural Model and CFA 79
Ref. code: 25615802037480DCY
Page 12
(9)
LIST OF ABBREVIATIONS
Symbols/Abbreviations Terms
a
Cronbach’s alpha (alpha)
b Standardised regression coefficient
B Unstandardised regression coefficient df Degree of freedom c2 Chi’s Squared
R2 R-Sqaured SMC Squared Multiple Correlation (Multi-
variate R-sqaured)
Ref. code: 25615802037480DCY
Page 13
1
CHAPTER 1
INTRODUCTION
1.1 Background of E-Commerce
1.1.1 Electronics Commerce and Cross-border Electronic Commerce
Electronic commerce (EC) is a form of exchange communication
through network technology. It includes data management, system and data security,
and system communication. This exchange is a form of paperless transaction for data,
product, or service exchange between parties. The term electronic commerce was
referring to electronic communication sending purchase order or business document
back in 1970s; however, the term was then developed into a form of commercial
exchange through websites since the development of world wide web as a new
business model.
Based on the relationship of the business to the parties, there are
four types of e-commerce principally – B2B, B2C, C2B, and C2C. Firstly, B2B or business-
to-business e-commerce is an e-commerce that both parties are businesses that
exchange data, services, or product between each other. It can be considered as a
form of a firm and its suppliers. Secondly, B2C or business-to-consumers e-commerce
refers to a form of electronic commerce that a business provides services or products
to end-user consumers per order. It is widely adopted and can been seen everywhere
nowadays. Thirdly, C2B consumer-to-business e-commerce is an exchange of data,
product, or services that created by the end-users offer to a firm. Lastly, C2C consumer-
to-consumer is an electronic business that the consumers exchange data, products, or
services which they produce them by themselves. The platform from maybe
constructed by a vendor to facilitate the transaction and it may be called a market
place.
The cross-border e-commerce (CBEC) is not a big different to the
aforementioned term of e-commerce. It expands the term to cover an aspect of
borderless commerce, thanks to the growing and expanding of internet access and
communication technologies in the last decades. The advanced technology in
Ref. code: 25615802037480DCY
Page 14
2
communication and information exchange provide a solid foundation to the growth of
e-commerce firms such as Amazon, Alibaba, etc. to gain more investment power that
they can penetrate other market outside their home countries alone.
1.1.2 Global E-commerce Economy
Electronic commerce, or E-Commerce, is inarguably growing and
expanding throughout the world. Market value of E-Commerce triplicates from about
one million US dollars in 2012 to 3 million USD in 2017 globally. Electronic commerce
offers customer various benefit on shopping such as variety of product, fast searching,
decreasing asymmetric of product information, etc. (Amaral, 2017).
Figure 1.1 Online Retail Market Size Year-over-year (Euromonitor International, 2018)
There are various of players in global E-Commerce but major firms
which are Amazon.com, Alibaba Group, JD.com, and eBay take about a half of the
market share. (Euromonitor International, 2018) These companies have been
penetrated their domestic market and cross-border market.
Ref. code: 25615802037480DCY
Page 15
3
Figure 1.2 Global E-Commerce Market Share (DHL Express, 2016)
1.1.3 Global Cross-border E-commerce Economy
International electronic trade or cross-border E-Commerce has been
enriched thank to technology advancement and connected world. It is today’s growth
rocker in E-Commerce with USD 300 million market size in 2017 and average growth
of 20 percent per year (DHL Express, 2016). Approximately 56 percent of cross- border
E-Commerce is sold via Alibaba, Amazon, and eBay whereas 34 percent of order made
in China (International Post Corporation, 2018). China apparently becomes a big market
for E-Commerce and Chinese firms are expanding their market to nearby region
especially South East Asia where they consider it as a potential market for E-
Commerce. The expansion can be notice obviously in recent years that major Chinese
firms bought quite a number of shares from current and new e-commerce platform
providers such as Lazada, Shopee, or even expanding their services to the region such
as Alibaba that starts a unified service for shopping experience from the purchasing to
payment gateway service.
1.1.4 Southeast Asia and Thailand Cross-border E-commerce
ASEAN, Association of South East Asian Nations, has announced that
they will boost digital economy throughout the region which will benefit cross-border
E-Commerce. The ministry of trade and industry of Singapore said that they would
Ref. code: 25615802037480DCY
Page 16
4
support advanced payment and transaction to help small and medium firms to expand
their E-Commerce; however, there are challenges for less developed counties such as
Myanmar, Cambodia, and the Philippines on logistic and internet penetration
(Iwamoto, 2018). However, between the support of local government, E-Commerce
giants from around the world are already on the market such as e-Bay, Amazon, and
Alibaba Group; especially Chinese E-Commerce firms that are penetrating in South East
Asian such as Lazada, and Shopee (SEA Limited, 2017).
Figure 1.3 Top 5 Shopping Application in Southeast Asia in 2017 (SEA Limited, 2017)
Lazada is apparently the biggest online shopping platform that
operates in South East Asia providing over 2,500 brands and 100,000 sellers to serve
560 million customers in the Thailand, Vietnam, Indonesia, Malaysia, the Philippines,
and Singapore. In 2017, Alibaba acquired approximately 81 percent of its stake making
Alibaba to be the major stakeholder and key player in this region (SEA Limited, 2017).
ASEAN is seen as a potential market for many investors due to it low market saturation,
low competitors, and high adoption of information and communication technology.
Cross-border E-Commerce in Thailand has been growing in
accordance to the region trend. In Thailand, E-Commerce has been adopted for many
years with the linear increasing sales from 2,865 million USD in 2015 to 4,239 million
Ref. code: 25615802037480DCY
Page 17
5
USD in 2017 (Priest, 2017). There are major players in Thailand market which are
Lazada, weloveshopping, Tarad, Zalora. Acquired by Alibaba, Lazada is closely
supported by its major stakeholder to advance its strategy towards “Thailand 4.0” and
remain its position as market leader in Thailand (SEA Limited, 2017). Although E-
Commerce competition in Thailand is not aggressive but there are opportunities and
high potential to grow. In 2017, Shopee, a part of SEA Group (former Garena), has
started their E-Commerce business in the region including Thailand which warming
competition in Thailand market (SEA Limited, 2017) while JD.com, Alibaba’s fierce
competitor, will join the market very soon after they have prepared logistics and
branding in the region (Priest, 2017).
In the customer perspective, the E-Commerce benefits their
shopping. Consumers across markets are motivated to shop cross-border for
fundamental reasons - product availability, a more attractive offering (including price),
and trust. Fashion and electronics are long-known cross-border top sellers, but
consumers now crave more on other product categories such as beauty and cosmetics,
pet care, food and beverage, and sporting goods (DHL Express, 2016).
1.2 Rational of the study
Cross-border e-commerce becomes a new economy of global trade due
to supporting factors on information technology and communication technology. The
advancement of the technologies provides a huge opportunity for large capital firms
to expand their business to go cross border to make more competitiveness in the
trading business; however, cross-border e-commerce has received relatively less
research attention than in domestic e-commerce context (Xiao, Wang, & Liu, 2018)
while the business has been changing very fast. Exploring the theory and factors that
are applicable to cross-border e-commerce context is necessary. Nevertheless, the
research on the e-commerce settings has been widely study and well settled where
trust and risk play an important role in the adoption of e-service. There are many
theories and factors influencing adoption of e-commerce on both pre-purchase and
post purchase stage in which this study aims to explore them under cross-border
Ref. code: 25615802037480DCY
Page 18
6
e-commerce context to expand current knowledge on e-commerce exposing adoption
factor to better business decision on their investment. Studying on Thailand’s
population is also an interesting aspect due to a high growth in both digital commerce
adopters and Thailand is part of South East Asia region which is a fast-growing region
in e-commerce market (Priest, 2017). Thais are also a well digital adopters with
supportive infrastructure in term of internet speed, devices, and purchasing power.
1.3 Research questions
1.3.1 Do consumer trust expectation and risk expectation influence
consumer disconfirmation and later trust and risk perception towards cross-border e-
commerce or not?
1.3.2 Do disconfirmation of consumer trust and risk expectation, consumer
satisfaction, post-purchase trust and risk perception influence repurchase intention
from cross-border e-commerce or not?
1.3.3 Does consumer trust affects perceived risk at the pre-purchase stage
as well as post-purchase stage or not?
1.3.4 Do consumers expect trust and risk from cross-border e-commerce
sellers and domestic e-commerce sellers differently across the pre-purchase and post-
purchase stage or not?
1.4 Objective
1.4.1 To study consumer trust expectation and risk expectation that
influence consumer disconfirmation and later trust and risk perception towards cross-
border e-commerce.
1.4.2 To study disconfirmation of consumer trust and risk expectation,
consumer satisfaction, post-purchase trust and risk perception that influence
repurchase intention from cross-border e-commerce.
1.4.3 To study consumer trust affecting perceived risk at the pre-purchase
stage as well as post-purchase stage.
Ref. code: 25615802037480DCY
Page 19
7
1.4.4 To study consumers difference perception of trust and risk from
cross-border e-commerce sellers and domestic e-commerce sellers across the pre-
purchase and post-purchase stage.
Ref. code: 25615802037480DCY
Page 20
8
CHAPTER 2
REVIEW OF LITERATURE
2.1 Theory
In this study, we explore previous literatures regarding the consumer
decision making and intention to purchase product. There are theory and model that
are interesting in the context of cross-border e-commerce which are Trust-based
Consumer Decision Making Model (TCDM) and the Expectation-Confirmation Theory
(ECT). The TCDM model features trust and risk as the main construct that influence
intention to buy or adopt services while the ECT evaluates repurchase intention of the
consumers in collaboration between confirmation of pre-purchase expectation and
satisfaction after purchase.
2.1.1 Trust-based Consumer Decision Making Model
The Trust-based consumer decision making theory propose three
main factors that influence purchase intention which are Trust, Perceived Risk, and
Perceived Benefits. The model was developed from the previous proposed Valence
Framework resulting in the three aforementioned independent variables (Kim, Ferrin,
& Rao, 2008; Naovarat, 2015). The model diagram showing independent variables and
dependent variables is shown in the Figure 2.1.
Trust in the model refers to the consumer trust on the E-Commerce
platform they are using with the risk and uncertainty for making purchase. If the
consumer has more trust in the platform, they will feel more confidence. Risk refers
to the customer perception that online purchasing has some risks or uncertainties that
they need to accept when making a purchase. There are risks of monetary loss,
shipment loss, or unable to deliver in the field of e-commerce that are significant
factors for customer’s purchase decision making. Lastly, Benefit or perceived benefit
refers to consumers perception towards E-Commerce benefiting them in form of time
efficacy e.g. fast navigation comparing to physical store shopping and the convenience
in the shopping experience. The benefits may be preceded by a cognitive stage leading
Ref. code: 25615802037480DCY
Page 21
9
an overall perception of benefit in form of monetary savvy, time savvy, and effort
savvy. These three constructs have effect over customer intention to making a
purchase on online scenario.
Figure 2.1 Trust-based Consumer Decision Making Model (Kim et al., 2008)
2.1.2 Expectation-Confirmation Theory
The expectation-confirmation theory has been widely adopted and
many fields of study, especially marketing field. The theory extensively studies the
post-purchase experience involving customer satisfaction and continuance of using
services or repurchase based on their satisfaction of prior purchase, usage of services,
or adoption. The theory proposes that the expectation which is a pre-purchase stage
affects the confirmation or disconfirmation of the belief which can be resulted in actual
purchase or adoption (Oliver, 1980). Satisfaction in the theory refers to the customer
overall satisfaction after adoption, purchase, or usage of product or services.
Although ECT has been widely adopted in marketing field; however,
many researchers discuss that the theory ignores viability of the post-experience
expectation i.e. the customer changes their perception after period of time after
making purchase from negative to positive or vice versa due to intermediate effect of
environment possibly social influence and marketing campaign (Bhattacherjee, 2001).
Ref. code: 25615802037480DCY
Page 22
10
Also, the pre- purchase stage influenced by the external factors and the post-purchase
influenced by the consumers perception themselves are also discussed.
Figure 2.2 Expectation-Confirmation Decision Theory (Oliver, 1980).
2.2 Related research
We have reviewed the previous literatures in the context of e-commerce
and cross-border e-commerce adoption. There are constructs that are related to the
ECT and TCDM model; however, the literatures expand more knowledge in each
variable in both ECT and TCDM model such as the multi-dimension of trust and risk
which is intensively studied.
2.2.1 Trust
The definition of trust is complicate due to its abstract and complex
factor nature. Trust has been addressed in many previous researches with vast
definition and stages (Bonsón Ponte, Carvajal-Trujillo, & Escobar-Rodríguez, 2015).
Several theories regarding trust are diverse into many stages of the interaction between
trustee and trustor (Stouthuysen, Teunis, Reusen, & Slabbinck, 2018). Trust has been
studied in many fields, yet the diverse conceptualisation of trust is ununified. Fisher
and Zoe refers to McKnight et al. previous research that an important consideration in
exchange relationships is determining the parties with whom an individual is willing to
interact. Trust is of central importance in this decision (Fisher & Zoe Chu, 2009). Trust
reflects the willingness of a party to be vulnerable to the actions of another party
Ref. code: 25615802037480DCY
Page 23
11
based on positive expectations regarding the other party’s motivation and/or
behaviour so that in the field of electronic commerce, trust formulate a belief of the
consumer on the seller to the extent of positive belief. Due to the nature of the
electronic commerce, consumers experience uncertainty of the online purchase;
therefore, trust plays an important role as a solution to uncertainty/risk (Kim et al.,
2008). Similarly, the model of dyadic trust in organizational relationships proposed by
Mayer et al. states that if the trustor perceives a trustee’s ability, benevolence, and
integrity to be sufficient, the trustor will develop trust (an intention to accept
vulnerability) toward the trustee; besides, trust can alleviate the effect of risk creating
a favourable condition for the trustee to accept the vulnerabilities.
Figure 2.3 McKnight and Chervany’s Trust dimensions. (McKnight, Choudhury, & Kacmar,
2002)
Nevertheless, in the context of B2C e-commerce, a consumer and
an online vendor would only be fully familiar after the consumer transacts with the
vendor and analyses the results. On McKnight and Chervany’s previous research in the
Ref. code: 25615802037480DCY
Page 24
12
area of electronic commerce proposes a decomposition and conceptualisation of trust
and the antecedents of trust that trust constructors can be differentiated into two
dimensions which are Institutional trust and dispositional trust (Harrison McKnight &
Chervany, 2001).
2.2.1.1 Dispositional Trust
Disposition to trust means the extent to which one displays
a consistent tendency to be willing to depend on others in general across a broad
spectrum of situations and persons (McKnight et al., 2002). There are two sub-
constructs in the dispositional trust which are Faith in humanity and Trusting Stance.
Faith in humanity means that one assumes others are usually competent, benevolent,
honest/ethical, and predictable. Trusting stance means that, regardless of what one
assumes about other people generally, one assumes that one will achieve better
outcomes by dealing with people as though they were well-meaning and reliable.
Therefore, trusting stance is like a personal choice or strategy to trust others. Because
it involves a choice that is presumably based on a subjective calculation of the odds
of success in a venture, trusting stance derives from the calculative, economics-based
trust research stream.
2.2.1.2 Institutional-based Trust
Institutional trust refers to an individual’s beliefs about the
structural safety or favourability of the conditions. Institutional trust ‘‘generalizes
beyond a given transaction and beyond specific sets of exchange partners”.
“Favourable conditions” refers to the legal, regulatory, business, and technical
environment perceived to support success. Structural assurance means that one
believes that protective structures - guarantees, contracts, regulations, promises, legal
recourse, processes, or procedures are in place that are conducive to situational
success. These two characteristic groups of trust formulate the trusting belief and
trusting intention which is the belief act on their benefits. The trusting belief according
to (Harrison McKnight & Chervany, 2001), can be classified into four categories -
Competence, Benevolence, Integrity, and Predictability belief.
Ref. code: 25615802037480DCY
Page 25
13
(1) Competence
Competence is connected to objective perception (e.g.
reliability, technical capabilities, know-how, and skills) of the seller/vender in the
context of e-commerce. Thus, customer review or feedback from the past experience
informs first-time online shopper the reliability and ability of the online vendor.
(2) Benevolence
Benevolence refers to the belief that the opposite side of the
communication cares about and act in one’s interest. Benevolence reflects the specific
relationship between trustor and trustee. It is not the overall trustee kindness in
general, but it is a subjective belief towards individual relationship between trustor
and trustee.
(3) Integrity
Integrity means that one believes that the other party makes
good-faith agreements, tells the truth, acts ethically, and fulfils promises. This would
reflect the belief that the Internet vendor will come through on its promises and
ethical obligations, such as to deliver goods or services or to keep private information
secure. Thus, integrity is more about the character of the trustee in the same direction
towards the trustor than about the trustor-trustee relationship.
(4) Predictability
Predictability means that one believes the other party’s
actions (good or bad) are consistent enough that one can forecast them in a given
situation. Those with high trusting belief-predictability would believe that they can
predict the Internet vendor’s future behaviour in a given situation.
2.2.2 Perceived Risk
Risk is one of the most important factors to study in cross-border e-
commerce due to its uncertainty in transaction and experience. Cross-border online
shopping is an unfamiliar and uncertain activity for consumers, more so than domestic
electronic commerce (Lesma & Okada). Risk is a situation that that may be resulted in
uncertainty or negative consequence (Naovarat, 2015). Many researches in the past
have studied the relationship between Trust and Perceived Risk. Perceived Risk has a
negative impact on customer trust in online transaction. It is an important factor in
Ref. code: 25615802037480DCY
Page 26
14
online purchasing scenario. The consumer’s will reluctantly be making decision to
purchase comparing to traditional channel of purchase due to lack of touch and feel
or product trial. Risk may be classified into various type such as Jacoby and Kaplan’s
seven type of risk which are financial, performance, physical, psychological, time,
social, and opportunity cost risk (Kim et al., 2008). Likewise, risk can be grouped into
System-dependent uncertainty and Transaction-specific uncertainty according to study
from (Rouibah, Lowry, & Hwang, 2016).
Figure 2.4 Featherman and Pavlou’s Risk Facets (Featherman & Pavlou, 2003)
There are many previous research areas that intensively study on the
perceived risk as well as consumer trust in different context e.g. information system,
marketing, and psychology. In the previous research from Featherman and Pavlou on
the consumer risk on an aspect of consumer behaviour, they intensively study the
various aspect of risk and classify them into 7 facets of risk which are performance risk,
financial risk, time risk, psychological risk, social risk, privacy risk, and overall risk
(Featherman & Pavlou, 2003).
Ref. code: 25615802037480DCY
Page 27
15
2.2.2.1 Performance Risk
Performance Risk is the possibility of the product
malfunctioning and not performing as it was designed and advertised and therefore
failing to deliver the desired benefits. In the area of e-commerce, this performance risk
is quite important for the representation of the product is on the website only leading
lack of touch and feel of the product compare to traditional channel of purchase.
2.2.2.2 Financial Risk
Financial risk is the potential monetary outlay associated with
the initial purchase price as well as the subsequent maintenance cost of the product.
In the context of e-commerce, it is the risk that the transaction payment may fail
leading to losses of money. This scenario is different to traditional channel transaction
that the consumer pays directly to the vendor and receive the product without delay.
2.2.2.3 Time Risk
Time risk refers to the consumers perception that they are
wasting time for doing the shopping e.g. searching for website information or products
information and product search before making a purchase. It may occur at the pre-
purchase navigation and at the purchase stage when making a wrong decision or
transaction on the website. The consumers may feel that it is wasting of their time
refunding, returning, or petitioning to the vender to change or claim if the product or
service will not fit to their need. This is the projection of uncertainties on the after-
sales services.
2.2.2.4 Psychological Risk
Psychological risk refers to the self-reflection uncertainties
which introduce projected loss of self-esteem. This facet of risk cases the consumers
to potentially perceive that the e-service or product does not reflect their self-images
or identity and/or not achieving their purchase goal from e-service is frustrating and
deteriorate their self-respect or appreciation towards themselves.
2.2.2.5 Social Risk
Social risk is an uncertainty that the consumer perceived from
external agents that using e-service will lead to social perception changes towards
them such as negative feeling from the friends or family, getting lost of trend, lower
Ref. code: 25615802037480DCY
Page 28
16
respect from the others, or potential perception as a foolish from others adopting or
use the e-service.
2.2.2.6 Privacy Risk
Privacy risk is a concern of potential losses of personal
information and transaction information without notice and permission granted
occurring during the usage of e-service. The riskiest potential loss from the privacy risk
is the fraudulent transaction by digital thief inside the same e-service platform and
outside the platform.
2.2.2.7 Overall Risk
Overall risk is a perception of overall aspect of risk in general
that the customer perceived from the e-service usage. The risk assesses as the overall
feeling of the consumers regardless of specific aspect of risk.
2.2.3 Trust and Perceived Risk Interaction
Incorporating trust and risk in e-commerce context has been
observed in many previous papers; however, the direction of effect is quite
controversial whether trust affects perceived risk, perceived risk affects trust, or they
both affect each other at the same time.
Pavlou’s study on Trust and Risk interaction towards Intention
suggest that Trust affects Risk (P. A. Pavlou, 2003) based on TAM (Technology
Acceptance Model) in accordance to the study by Kim et. al. which conduct
longitudinal study based on EDT suggests the same effect direction (Kim, Ferrin, & Rao,
2009) as well as the work done by Zhu et. al. in later year (Zhu, Neal, Lee, & Chen,
2009). These studies are significant statistically affirming the researcher’s belief that
Trust affects Risk. However, there are recent studies that suggest that Perceived Risk
affect Consumer Trust. These studies are conducted in marketing perspective towards
online purchase which the researchers consider Risk as an independent variable that
influencing consumer trust (Pappas, 2016; Rouibah et al., 2016).
2.2.4 Incorporating Trust and Perceived Risk in Expectation-
Disconfirmation Theory
In previous study on the usage of EDT with technology adoption,
there are several models proposed by the researchers to incorporate the trust into
Ref. code: 25615802037480DCY
Page 29
17
EDT model. Trust theory in technology and e-commerce setting has been intensive
studied by McKnight. In their paper regard the trust and technology adoption and
continuance intention, they proposed the model in Figure 2.5 showing relationships of
trust expectation, trust disconfirmation, satisfaction, perceived performance, and
trusting intention in the aftersales stage to the technology continuance as the outcome
(N. Lankton, McKnight, & Thatcher, 2014).
Figure 2.5 Lanktan et al.’s integrating technology Trust in EDT (N. Lankton et al., 2014)
Based on Lanktan’s study on trusting belief as an expectation in EDT
model, it is interesting that the proposed Technology Trusting Expectation does not
affect the Technology Trusting Disconfirmation, but the construct positively affects the
Technology Trusting Performance which is a post-evaluation belief after they
experience the technology.
Another study was conducted to observe the influence on Trust and
Risk in post-purchase experience by Mou et. al. incorporating the aforementioned
constructs in Expectation Disconfirmation Theory. Although Trust and Risk has various
antecedent to examine, post-adoption/behavioural action’s perception of Trust and
Risk are not very influential to customer intention to repurchase or continuance
intention even though the previous study shows strong relationship of Trust to
intention to use in TAM model (Gefen, Karahanna, & Straub, 2003). This conflict of
relationship is due to strong and dominant effect of customer satisfaction (Mou, Shin,
Ref. code: 25615802037480DCY
Page 30
18
& Cohen, 2015) in the Expectation Disconfirmation Theory while TAM does not
incorporating Satisfaction into the Theory.
Ref. code: 25615802037480DCY
Page 31
19
CHAPTER 3
RESEARCH METHODOLOGY
3.1 Conceptual Model
Based on Trust-based Consumer’s Decision-making Model (TCDM) and the
Expectation-confirmation Theory (ECT), we adapt the ECT and Trust-based Consumer’s
Decision-making Model and propose the following conceptual model as shown in the
Figure 3.1 to study on the cross-border e-commerce pre and post purchase stage
leading to consumer repurchase intention.
Figure 3.1 Proposed research model
The model incorporates Trust and Risk as consumer expectation (Trusting
Expectation and Risk Expectation). The trust and risk are evaluated again at the post-
purchase stage as a combination with satisfaction and disconfirmation to have direct
effect on consumer repurchase intention.
Ref. code: 25615802037480DCY
Page 32
20
3.2 Variable Definition
3.2.1 Trusting Expectation and Trusting Performance
Trust in this study refers to the customer belief towards cross-border
e-commerce and domestic e-commerce. The measurements aim to evaluate
trustworthiness of e-sellers based on consumer’s trusting belief in previous studies on
trust towards information systems (Gefen et al., 2003; McKnight, Carter, Thatcher, &
Clay, 2011; McKnight & Chervany, 2014) and trust in electronic commerce (Bonsón
Ponte et al., 2015; Kim, 2014; Kim et al., 2008; McKnight et al., 2002; Mukherjee, Arnott,
& Nath, 2007; Sfenrianto, Wijaya, & Wang, 2018; Stouthuysen et al., 2018) where
measurements are evaluating interpersonal trusting belief on overall trustworthiness
of the e-sellers, act on consumer benefit (Integrity), willingness to support
(Benevolence), and information quality (Competence) provided by e-sellers.
3.2.2 Perceived Risk Expectation and Perceived Risk Performance
In this study, Perceived Risk in 4 different facets which are financial
risk, delivery time risk, privacy risk, and overall risk are measured in accordance to
previous studies on risk dimensions towards information system and e-commerce area
(Featherman & Pavlou, 2003; Mou, Cohen, Dou, & Zhang, 2017; Mou et al., 2015;
Pappas, 2016; Rouibah et al., 2016). The financial risk is the risk of possibility to lose
money during transaction while the delivery risk concerns timeliness of product deliver,
and the privacy risk is the risk that their personal information will be exposed due to
using the website/application. These risks show high importance in previous studies,
but they were measured in e-commerce context and cross-border e-commerce in e-
sellers’ perspective (Guo, Bao, Stuart, & Le-Nguyen, 2018). This study will comparatively
measure the four risks in cross-border and domestic context as well as pre-purchase
as an expectation (Perceived Risk Expectation) and post-purchase setting (Perceived
Risk Performance) as a performance perception according to Expectation-
disconfirmation Theory.
3.2.3 Expectation Disconfirmation
Expectation Disconfirmation in this study refers to the confirmation
of overall expectation of past experience (Oliver, 1980), confirmation of trusting belief
Ref. code: 25615802037480DCY
Page 33
21
(N. Lankton et al., 2014), and this study propose of disconfirmation of perceived risk
expectation from purchasing via cross-border e-commerce. The measurements are
based on Expectation Disconfirmation Theory within the context of e-commerce (Kim
et al., 2008, 2009), but they are adapted to cross-border e-commerce in this study.
3.2.4 Satisfaction
Satisfaction has been studied for many years. In the previous
literatures in marketing and psychology researchers defines the satisfaction differently.
In this study, we define consumer’s satisfaction based on the Expectation-Confirmation
Theory (Bhattacherjee, 2001; Fang, George, Shao, & Wen, 2016; Kim et al., 2008; Mou
et al., 2017) that the term refers to the customer feeling of contented or pleased after
experienced product and services from cross-border e-commerce.
3.2.5 Repurchase Intention
The Repurchase Intention in this study is a tendency that the
consumer is likely to repurchase from cross-border e-commerce again. This construct
is found in various studies on e-service or information system continuance and loyalty
(Chong, 2015; Mou et al., 2017) as well as repurchase intention in accordance to the
Expectation-Confirmation Theory (Ambalov, 2018; Bhattacherjee, 2001; Kim et al., 2008;
Shang & Wu, 2017).
3.3 Research hypothesis
3.3.1 Relationship of Trusting Expectation and Perceived Risk
Expectation
Based on previous study on trust and risk interaction, it is very
controversial on examination of their relationship; however, the meta-analysis on this
relationship suggest that Trust negatively affecting Risk has a stronger effect than the
other paths (Mou et al., 2015). We derived this ground analysis to propose our
hypothesis as the following.
H1 Trusting Expectation from cross-border e-commerce has negative
effect on Perceived Risk Expectation from cross-border e-commerce.
Ref. code: 25615802037480DCY
Page 34
22
3.3.2 Relationship of Trusting Expectation, Perceived Risk Expectation,
Expectation Disconfirmation.
Trust Expectation and Perceived Risk Expectation are considered
different perspective of consumer expectation that influenced by disconfirmation of
the consumer belief based on EDT model and the previous study on mobile
commerce (Chong, 2015). They both affect the consumer intention to purchase in the
context of Theory of Planned Behaviour as well (Park, Lee, & Ahn, 2014) Based on
previous study of Trust and Risk in the EDT model (Chong, 2015; Mou et al., 2015), the
path of the performance stage is not significantly influencing the consumer intention
due to stronger influence of customer satisfaction, but the study on Trusting Belief
towards Expectation-Disconfirmation in EDT shows a significant relationship. Therefore,
we propose the hypothesis of their relationship in accordance to Expectation-
disconfirmation theory as the following.
H2a Trusting Expectation from cross-border e-commerce has positive
effect on Expectation Disconfirmation from cross-border e-commerce.
H2b Perceived Risk Expectation from cross-border e-commerce has
negative effect on Expectation Confirmation from cross-border e-commerce.
H2c Trusting Expectation from cross-border e-commerce is higher
than domestic e-commerce on average.
H2d Perceived Risk Expectation from cross-border e-commerce is
less than domestic e-commerce on average.
3.3.3 Relationship of Trusting Performance, Perceived Risk
Performance, and Expectation Disconfirmation.
The performance stage of the consumers’ perception of Risk and
Trust are formed at the post-purchase stage. Based on the study on EDT and Trust,
the path of Trusting Performance influences the Disconfirmation of Expectation in the
post-purchase stage (Ambalov, 2018; Kim et al., 2009; N. Lankton et al., 2014). This
study includes the Perceived Risk Performance into the EDT model as well in order to
observe both Trust and Risk in the post-purchase stage. Therefore, the hypothesis
regarding this cross-border e-commerce constructs’ interaction are proposed as the
following.
Ref. code: 25615802037480DCY
Page 35
23
H3a Trusting Performance from cross-border e-commerce has
negative effect on Expectation Confirmation from cross-border e-commerce.
H3b Perceived Risk Performance from cross-border e-commerce has
negative effect on Expectation Confirmation from cross-border e-commerce.
H3c Trusting Performance from cross-border e-commerce is higher
than domestic e-commerce on average.
H3d Perceived Risk Performance from cross-border e-commerce is
less than domestic e-commerce on average.
3.3.4 Relationship of Trusting Performance and Perceived Risk
Performance from cross-border e-commerce.
Based on the meta-analysis on this relationship, this study follow the
concept that Trust negatively affecting Risk which has a stronger effect (Mou et al.,
2015). The Trusting Performance and Perceived Risk Performance are the post-
purchase perception which share the same properties to the pre-purchase perception
except that this is an updated belief after the consumer experienced e-service in
accordance to Expectation Disconfirmation context (N. Lankton et al., 2014; P. A.
Pavlou, 2003). Therefore, this study proposes the hypothesis over their relationship as
the following.
H6 Trusting Performance has negative effect on Perceived Risk
Performance
3.3.5 Relationship of Expectation Disconfirmation and Satisfaction
from cross-border e-commerce.
The Expectation-disconfirmation Theory suggest that the Expectation
Disconfirmation/Confirmation has positive effect on consumer satisfaction (Oliver,
1980) after they have made a purchase. Previous study on Expectation-disconfirmation
Theory and consumer satisfaction also confirm that the relationship between these
latent variables is affirmative (Ambalov, 2018; Bhattacherjee, 2001; N. Lankton et al.,
2014; N. K. Lankton, McKnight, Wright, & Thatcher, 2016).
H7 Expectation Disconfirmation has positive effect on Satisfaction.
Ref. code: 25615802037480DCY
Page 36
24
3.3.6 Relationship between Pre-purchase and Post-Purchase of
Consumer Trust and Perceived Risk from cross-border e-commerce.
In this study, we believe that the perception of pre-purchase risk and
pot-purchase trust are different. The customers have an updated belief of trust
towards the seller after they experienced the purchase via cross-border e-commerce
in accordance to previous study of Trust in EDT context (Ambalov, 2018; Bhattacherjee,
2001; N. Lankton et al., 2014; N. K. Lankton et al., 2016; Shang & Wu, 2017). The
proposed Perceived Risk Expectation and Perceived Risk Performance share the same
concept of Trust that the study can observe consumer perception of risk on pre-to-
post purchase as well. Therefore, this study proposes the following hypothesis.
H4a Trusting Expectation (pre-purchase) has positive influence
towards Trusting Performance (post-purchase).
H4b Trusting Performance (post-purchase) is increased from Trusting
Expectation (pre-purchase).
H5a Perceived Risk Expectation (pre-purchase) has positive influence
towards Perceived Risk Performance (post-purchase).
H5b Perceived Risk Performance (post-purchase) is decreased from
Perceived Risk Expectation (pre-purchase).
3.3.7 Satisfaction and Repurchase Intention from cross-border e-
commerce.
Customer satisfaction influence Intention to repurchase or
continuance usage of e-services in various study on their relationships (Lee, Park, &
Han, 2011; Marinkovic & Kalinic, 2017; Pham & Ahammad, 2017; Sfenrianto et al., 2018;
Shang & Wu, 2017). In this study, the observation of their relationships is based
principally on the Expectation-Disconfirmation Theory under e-service context (Bonsón
Ponte et al., 2015; Kim et al., 2008, 2009; Zhu et al., 2009) which define satisfaction as
a consumer overall satisfactorily experience after usage of e-service and propose the
following hypothesis.
H8 Satisfaction has positive effect on Repurchase Intention.
Ref. code: 25615802037480DCY
Page 37
25
3.4 Population and samples
3.4.1 Population
Based on Statista, 12.1 million Thailand’s online shopper is estimated
in 2017 and the number of digital shoppers is estimated to 13.9 million in 2021. These
12.1 million e-commerce shoppers can be estimated as a population of total online
shoppers including both online shopper who experience and not experience cross-
border e-commerce; however, no evidence from any sources that classify number
cross-border shoppers is addressed; therefore, the population for our study is unknown
within the 12.1 million users.
3.4.2 Samples
Sample size for analysis is calculated based on Cohen’s guideline for
power analysis (Cohen, 1988). This study is a new context of e-commerce therefore
the effect size is assumed to be trivial (f2 = 0.05 i.e. at least 5% variance explained in
the multiple linear regression model). Moreover, this study sets up 0.95 confidence
interval with 95 power of test (1 – Type II error probability = 0.95) based on Cohen’s
formula and guideline.
! =$(1 − ())
()
Using R’s pwr package, the package calculates the result with
aforementioned parameters to retrieve 416 minimum sample size for the regression
analysis (u = 6, f 2 = 0.05, Power Test (b) = 0.95, Significant Level (a) =0.05).
In term of Student’s T-test, Cohen’s guideline for small effect size is
used for two-tailed and two-sided hypothesis testing (Effect Size(d) = 0.2, Power Test
(b) = 0.95, Significant Level (a) = 0.05, Type = Two-tailed) in which the result in 326
minimum sample size which is the same to power table provided from Cohen’s
guideline.
Due to the two-methods analysis used in this study – regression and
t-test analysis, sample size to be used for the full model will be based on minimum
sample size for regression analysis which is 416 samples.
Ref. code: 25615802037480DCY
Page 38
26
3.5 Research Instruments and Data Collection
This study is a quantitative exploratory research aiming to quantify factors
influencing repurchase intention and difference perception of Trust and Risk between
pre-purchase and post-purchase across cross-border and domestic purchase;
therefore, the questionnaire is used with Likert-scale questions to ask each respondent
on the main continuance model and the attitude change model on Trust and Risk.
The pre-purchase constructs consist of Perceived Risk Expectation and
Trusting Expectation. These two variables are measured two times, domestic context
and cross-border context, on the same measurement item. The post-purchase Trusting
Performance and Risk Performance are measured on the same setting, then, they are
compared on consumer Trust and Risk perception between domestic e-commerce and
cross-border e-commerce. Cross-border pre-purchase and post-purchase perception of
Trust and Risk are evaluated as well in order to observed how the consumer
perception changes; however, the study of Expectation on pre-purchase stage is based
on the recall of previous experience.
3.5.1 Research Instrument Development
The questionnaire is developed to collect the sample’s response
regarding their perception on cross-border e-commerce purchase in electronic form
via SurveyMonkey. The questionnaire consists of 3 parts as described below.
Part 1 provides introductory information survey purpose, e-
commerce and cross-border e-commerce purchase, and direction for the respondents
to understand the subject and scope of the survey they are answering. This part also
asks the respondent whether they have experienced cross-border e-commerce
purchase or not in order to screening out ineligible sample from the analysis.
Part 2 consists of interval measurement items in Likert-scale to
observe sample attitudes towards pre-purchase and post-purchase experience based
on Trust, Risk, and Expectation-disconfirmation Theory. Each item requires the
respondent to provide opinion in scale of 1 to 5 which 1 is Strongly Disagree, 2 is Agree,
Ref. code: 25615802037480DCY
Page 39
27
3 is Neutral, 4 is Agree, and 5 is Strongly Agree. The measurement items can be referred
on Table 3.1.
Part 3 contains items to collect respondent demographic information
which are gender, highest education level, monthly income, and frequency of online
purchase.
This study conducts a pre-test with 30 sample surveys once before
the actual data collection. The pre-test samples are analysed to ensure reliability of
the research tools using Cronbach’s Alpha as measurement (a > 0.8); factor analysis
is conducted (factor loading > 0.7) as well to ensure factor correlation and grouping
within each construct are acceptable statistically.
3.5.2 Data Collection
The questionnaire was transformed into online form on
SurveyMonkey.com, and the questionnaire was distributed online to participants. The
participant is voluntarily to answer the questions without any interference from the
researcher.
Ref. code: 25615802037480DCY
Page 40
Table 3.1
Measurement items in pre-purchase and post-purchase stage. Construct Factor Abbr. Original Measurement Item Adapted Measurement Item Adapted from
Trusting Expectation/
Trusting Performance
(Cross-border and
Domestic e-commerce)
CBTRE1, CBTRP1
DMTRE1, DMTRP1
This site is trustworthy. This seller is trust worthy. (Kim et al., 2008)
(Gefen et al., 2003)
(P. A. Pavlou, 2003)
This Web retailer is trustworthy.
CBTRE2, CBTRP2
DMTRE2, DMTRP2
If I required help, LegalAdvice.com
would do its best to help me.
The seller would be willing to
help when there is question or
problem.
(McKnight et al., 2002)
CBTRE3, CBTRP3
DMTRE3, DMTRP3
This Web retailer is known as one
that keeps promises and
commitments.
The seller would keep promises
and commitments for their
services.
(P. A. Pavlou, 2003)
(Pavlou & Fygenson, 2006)
CBTRE4, CBTRP4
DMTRE4, DMTRP4
This Web vendor would be honest
in providing accurate information
about this product.
would provide product and
services information correctly.
(P. A. Pavlou, 2003)
(Pavlou & Fygenson, 2006)
28
Ref. code: 25615802037480DCY
Page 41
Table 3.1
Measurement items in pre-purchase and post-purchase stage. (Cont’d) Construct Factor Abbr. Original Measurement Item Adapted Measurement Item Adapted from
Perceived Risk Expectation/
Perceived Risk Performance
(Cross-border and
Domestic e-commerce)
CBPRE1, CBPRP1
DMPRE1, DMPRP1
Purchasing from this Website
would involve more product risk
(i.e. not working, defective
product) when compared with
more traditional ways of shopping.
Purchasing from this Website
would involve more product risk
(i.e. not working, defective
product).
(Kim et al., 2008)
CBPRE2, CBPRP2
DMPRE2, DMPRP2
I would be concerned that I may
suffer from monetary loss due to
the seller’s fraudulent acts.
Purchasing from this seller
would involve more financial risk
(i.e. fraud, hard to return).
(Hong, 2015)
(Kim et al., 2008)
(Featherman & Pavlou,
2003) Purchasing from this Website
would involve more financial risk
(i.e. fraud, hard to return) when
compared with more traditional
ways of shopping.
29
Ref. code: 25615802037480DCY
Page 42
Table 3.1
Measurement items in pre-purchase and post-purchase stage. (Cont’d) Construct Factor Abbr. Original Measurement Item Adapted Measurement Item Adapted from
Perceived Risk Expectation/
Perceived Risk Performance
(Cross-border and
Domestic e-commerce)
CBPRE3, CBPRP3
DMPRE3, DMPRP3
Purchasing from this seller
would involve uncertainty on
product delivery.
New item
CBPRE4, CBPRP4
DMPRE4, DMPRP4
How would you rate your overall
perception of risk from the seller?
Overall, purchasing from this
seller would be risky.
(Kim et al., 2008)
(Featherman & Pavlou,
2003) On the whole, considering all sorts
of factors combined, about how
risky would you say it would be to
sign up for and use XXXX?
30
Ref. code: 25615802037480DCY
Page 43
Table 3.1
Measurement items in pre-purchase and post-purchase stage. (Cont’d) Construct Factor Abbr. Original Measurement Item Adapted Measurement Item Adapted from
Expectation
Disconfirmation
(Cross-border e-commerce)
DCE1 Overall, most of my expectations
from using this website were
confirmed.
Overall, most of my
expectations were confirmed.
(Mou et al., 2017)
DCE2 My experience with using this
website was better than what I had
expected.
My experience with purchasing
from this seller was better than
what I had expected.
(Mou et al., 2017)
DCE3 Overall, purchasing from this
seller was not risky as expected.
New item
DCE4 My experience with using this
website was better than what I had
expected.
Overall, purchasing from this
seller is trust worthy as
expected.
(Kim et al., 2008)
(Pham & Ahammad, 2017)
31
Ref. code: 25615802037480DCY
Page 44
Table 3.1
Measurement items in pre-purchase and post-purchase stage. (Cont’d) Construct Factor Abbr. Original Measurement Item Adapted Measurement Item Adapted from
Satisfaction
(Cross-border
e-commerce)
SAT1 I am satisfied with the purchase experience from
this website (e.g., ordering, payment procedure).
I am satisfied from purchasing
from this seller.
(Pham & Ahammad, 2017)
(Sfenrianto et al., 2018)
Overall, I am quite satisfied with my experience
dealing with e-sellers.
SAT2 My experience with using this website was better
than what I had expected.
Overall, I am quite satisfied
with my experience dealing
with e-sellers.
(Kim et al., 2008)
(Featherman & Pavlou, 2003)
SAT3 I have good impression with the service provided by
e- sellers.
I have good impression with
the service provided by e-
sellers.
(Pham & Ahammad, 2017)
(Kim et al., 2009)
(Featherman & Pavlou, 2003) How do you feel about your overall experience of
the purchase through this Website?
I am satisfied with my overall experiences of online
shopping at this website.
32
Ref. code: 25615802037480DCY
Page 45
Table 3.1
Measurement items in pre-purchase and post-purchase stage. (Cont’d) Construct Factor Abbr. Original Measurement Item Adapted Measurement Item Adapted from
Satisfaction
(Cross-border e-commerce)
SAT4 My experience with using this website
was better than what I had expected.
Overall, how would you rate
your experience purchasing from
this seller?
(Kim et al., 2008)
(Featherman & Pavlou,
2003)
Repurchase Intention
(Cross-border e-commerce)
CON1 I intend to continue using Microsoft
Access to create databases.
I intended to continue
purchasing from cross-border
seller again in the future.
(Featherman & Pavlou,
2003)
CON2 I plan to continue using Microsoft
Access after this class.
I plan to purchase from cross-
border seller again in the future.
(N. Lankton et al., 2014)
CON3 I intend to continue using [Microsoft
Access/MySNW.com].
I intend to repurchase from
cross-border seller in near
future.
(N. Lankton et al., 2014)
CON4 In the near future, I intend to continue
using Microsoft Access.
I intend to continue purchasing
from cross-border seller.
(N. Lankton et al., 2014)
33
Ref. code: 25615802037480DCY
Page 46
34
3.6 Data Analysis Process
IBM SPSS, a software package, is used to conduct descriptive statistics,
reliability test, normality test, correlation analysis, student’s t-test, and IBM AMOS is
used to conduct CFA and SEM on the data collected. The process of data analysis is
based on two aspects which are descriptive statistics, and inferential statistics.
3.6.1 Descriptive Statistics
The questions regarding consumer’s profile (e.g. age, gender, income,
purchase frequency) are analysed using descriptive statistics consisting of valid
percentage, frequency, mean, and standard deviation.
3.6.1.1 Variable Distribution and Normality Test
The analysis purpose is to observe the data distribution to
ensure that the data is normally distributed aligned to regression assumption. Due to
large sample size (between 30 – 1,000 samples in the test) (Hair, Black, Babin, &
Anderson, 2014), the Kolmogorov-Smirnov nonparametric test of normality is
incorporated in to the analysis with acceptable Lilliefors Significance Correction (Sig. P
< 0.05) (Hair et al., 2014; Nau, 2018).
3.6.1.2 Instrument Validity and Reliability Analysis
The research instrument reliability is analysed using
descriptive statistic tool, Cronbach’s alpha. The pilot group responses are analysed to
ensured that the factors will have internal integrity where Cronbach’s alpha for
extraversion subscale should be higher than 0.8 and each items alpha should be
sufficiently high which alpha is or higher than 0.7 (α >= 0.7) is considered acceptable
while alpha is or higher than 0.8 (α >= .8) is considered excellent (Cerny & Kaiser, 1977;
Kaiser, 1974).
3.6.1.3 Discriminant Validity Analysis
Measurement Model in SEM required good covariation between
factors in each construct. The validity and reliability analysis of measurement model
is conducted. The validity of measurement model consists of convergence validity
which can be evaluated using Average Variance Extracted, Construct Validity which can
Ref. code: 25615802037480DCY
Page 47
35
be evaluated by achieving fitness indexes (refer to Fit Index summary in Table 3.3 and
Table 3.4), and discriminant validity which can be observed through correlation matrix.
The reliability of the instrument can be evaluated via Internal Reliability, Construct
Reliability, and Average Variance Extracted (Ahmad, Zulkurnain, & Khairushalimi, 2016).
Table 3.2
Model Validity and Reliability Analysis Result.
Test Result
Validity
Convergence Validity The value of AVE should be greater or equal
to 0.5 in order to achieve this validity.
Construct Validity The construct validity is achieved when the
Fit Indexes achieve the level of acceptance.
Discriminant Validity The correlation between each pair of latent
exogenous construct should be less than
0.85.
Reliability
Internal Reliability The Cronbach’s Alpha value is 0.6 (α >= 0.6)
or higher. Alpha of 0.8 or higher (α >= .8) is
considered excellent (Cerny & Kaiser, 1977;
Kaiser, 1974).
Construct Reliability CR ≥ 0.6 is required.
Average Variance Extracted AVE ≥ 0.5 for each construct.
Note. From (Ahmad et al., 2016; Bagozzi & Yi, 1988; Cerny & Kaiser, 1977; Hooper, Coughlan, & R. Mullen, 2007;
Kenny, 2015)
3.6.2 Inferential Statistics
In term of inferring the population properties, inferential statistics is
used to analyse the attitude measurement data in Likert-scale with defined 0.05
significant level for hypothesis test within 95% confidence interval.
3.6.2.1 Student t-test
The test is used to quantify difference in mean between pre
and post-purchase perception on Trust and Risk in which the function Paired Student’s
Ref. code: 25615802037480DCY
Page 48
36
T-test is selected. The paired sample t-test with lower p-value (p < 0.05) within 95%
confident interval indicates significant difference in mean (two-tailed)(Student, 1908).
The expectation from student t-test is that there should be no difference between
pre-purchase and post-purchase construct as well as the domestic against cross-border
perception of trust and risk.
3.6.2.2 Structural Equation Model (SEM) and Confirmatory Factor
Analysis (CFA)
Confirmatory Factor Analysis (CFA) and Structural Equation
Model (SEM) is used to determine model structure based and factor covariation to
make a casual conclusion from the model. The goodness of fit of the model can be
determined by various fit index which can be classified in to 3 major groups which are
Absolute Fit Index, Comparative Fit Index, and Other Indices; however, it is
recommended to use Comparative Fit Index or combination of indices to evaluate the
goodness of fit rather than using Absolute Fit Index due to the AFI is highly sensitive
to sample size meaning that the index will performs better if the sample size increases
(Hooper et al., 2007; Hu & Bentler, 1999; Kenny, 2015; Schreiber, Nora, Stage, Barlow,
& King, 2006). The recommended indices to be used in this study are Standardized
Root Mean Square Residuals (SRMR), Root Mean Square Error of Approximation
(RMSEA), Bentler-Bonett Index or Normed-fit Index (NFI), Tucker-Lewis’s Non-normed
Fit Index (TLI or NNFI), and Comparative Fit Index (CFI). The criteria for each fit index
are shown on Table 3.3 below.
Ref. code: 25615802037480DCY
Page 49
37
Table 3.3
SEM Recommended Fit Index and Criteria.
Fit Index Acceptable Threshold Level
Comparative Fit Index
NFI NFI > 0.95
Table 3.3
SEM Recommended Fit Index and Criteria. (Cont’d)
TLI TLI > 0.95
CFI CFI > 0.95
Other Fit Index
SRMR SRMR < 0.08
RMSEA RMSEA < 0.07 which RMSEA < 0.03 indicates excellent fit. Note. From (Schreiber et al., 2006)
Apart from using individual fit index to determine goodness
of fit of Structural Equation model, a combination of fit index is also suggested in
determining the goodness of fit (Hu & Bentler, 1999) which are combinations of NNFI
(TLI) with SRMR, RMSEA with SRMR, and CFI with SRMR together. The criteria used in
combination of fit index is provided on Table 3.4 below.
Table 3.4
SEM Fit Index Combination Criteria.
Combination Criteria Acceptable Threshold Level
NNFI (TLI) and SRMR NNFI of 0.96 or higher and an SRMR of .09 or lower
RMSEA and SRMR RMSEA of 0.06 or lower and a SRMR of 0.09 or lower
CFI and SRMR CFI of .96 or higher and a SRMR of 0.09 or lower Note. From (Hu & Bentler, 1999; Schreiber et al., 2006)
Ref. code: 25615802037480DCY
Page 50
38
CHAPTER 4
RESULTS AND DISCUSSION
This study is an exploratory research on cross-border repurchase intention
based on Trust and Perceived Risk as main factors in both pre-purchase and pose-
purchase experience. We develop a questionnaire and conduct a survey to collect 462
samples. There are 440 valid respondents for they had experience on purchasing
tangible products from cross-border e-sellers which is considered as a good
representation of cross-border e-commerce population in Thailand.
4.1 Descriptive Statistics
Regression analysis involves critical assumptions that analyst has to ensure.
In this section, the observation and test result of normality of residuals,
multicollinearity of latent variables, reliability of measurement items, and factor
analysis are explained.
4.1.1 Variables Descriptive Statistics and Normality of Residuals
The descriptive statistics of measurement items for both SEM
analysis and Paired T-test Analysis is shown on Table 4.1.
Table 4.1
Factor Descriptive Statistics.
Latent Variable Factor N Minimum Maximum Mean Std. Deviation
Cross-border
Perceived Risk
Expectation
(CBPRE)
CBPRE1 440 1 5 3.41 0.921
CBPRE2 440 1 5 3.36 0.976
CBPRE3 440 1 5 3.6 0.799
CBPRE4 440 1 5 3.73 0.763
Ref. code: 25615802037480DCY
Page 51
39
Table 4.1
Factor Descriptive Statistics. (Cont’d)
Latent Variable Factor N Minimum Maximum Mean Std. Deviation
Cross-border
Perceived Risk
Performance
(CBPRP)
CBPRP1 440 1 5 3.25 0.76
CBPRP2 440 1 5 3.13 0.81
CBPRP3 440 1 5 3.5 0.8
CBPRP4 440 1 5 3.5 0.68
Cross-border
Trusting
Expectation
(CBTRE)
CBTRE1 440 1 5 3.93 0.48
CBTRE2 440 1 5 3.88 0.62
CBTRE3 440 1 5 3.78 0.68
CBTRE4 440 1 5 3.76 0.66
Cross-border
Trusting
Performance
(CBTRP)
CBTRP1 440 2 5 3.9 0.49
CBTRP2 440 2 5 3.95 0.59
CBTRP3 440 2 5 3.74 0.69
CBTRP4 440 2 5 3.83 0.6
Domestic
Perceived Risk
Expectation
(DMPRE)
DMPRE1 440 1 5 3.37 0.9
DMPRE2 440 1 5 3.4 0.93
DMPRE3 440 1 5 3.88 0.82
DMPRE4 440 1 5 4 0.71
Domestic
Perceived Risk
Performance
(DMPRP)
DMPRP1 440 1 5 3.18 1.11
DMPRP2 440 1 5 3.1 1.24
DMPRP3 440 1 5 3.28 1.14
DMPRP4 440 1 5 3.78 0.85
Domestic
Trusting
Expectation
(DMTRE)
DMTRE1 440 2 5 4.28 0.7
DMTRE2 440 1 5 4.04 0.66
DMTRE3 440 1 5 4.06 0.71
DMTRE4 440 1 5 4 0.73
Ref. code: 25615802037480DCY
Page 52
40
Table 4.1
Factor Descriptive Statistics. (Cont’d)
Latent Variable Factor N Minimum Maximum Mean Std. Deviation
Domestic
Trusting
Performance
(DMTRP)
DMTRP1 440 1 5 4.21 0.8
DMTRP2 440 1 5 4.08 0.69
DMTRP3 440 1 5 3.84 0.88
DMTRP4 440 1 5 4.06 0.82
Expectation
Disconfirmation
(DCE)
DCE1 440 1 5 3.77 0.58
DCE2 440 1 5 3.81 0.64
DCE3 440 1 5 3.75 0.69
DCE4 440 1 5 3.71 0.67
Satisfaction
(SAT)
SAT1 440 1 5 3.92 0.74
SAT2 440 1 5 3.92 0.74
SAT3 440 1 5 3.9 0.72
SAT4 440 1 5 4.06 0.67
Repurchase
Intention
(CON)
CON1 440 1 5 4.01 0.69
CON2 440 1 5 4.02 0.71
CON3 440 1 5 4.03 0.68
CON4 440 1 5 4.01 0.68
In the samples, perception of cross-border trusting performance
(CBTRP 1-4) has minimum Likert-scale value of 2 as well as domestic trusting
expectation item 1 (DMTRE1). The maximum Likert-scale value of all factors from the
samples is 5. The average of the scales in all factor are between 3 – 4. Standard
Deviation of the factor DMPRP 1 – 3 are quite stand out from the other factors (Std.
Deviation > 1) which reflect a wide spread of the distribution.
Apart from descriptive statistic for these factors, the cross-border
factors are observed their normality to comply with multivariate regression assumption
Ref. code: 25615802037480DCY
Page 53
41
that the data should be normally distributed. The assessment detail produced from
IBM AMOS for SEM analysis is provided on Table 4.2.
Table 4.2
Assessment of Multivariate Normality.
Factor Min Max Skewness Skewness
Critical Ratio Kurtosis
Kurtosis
Critical Ratio
CBTRE1 1 5 -1.711 -14.649 8.623 36.920
CBTRE2 1 5 -0.926 -7.932 3.453 14.786
CBTRE3 1 5 -0.623 -5.337 1.727 7.395
CBTRE4 1 5 -0.672 -5.755 1.601 6.856
SAT1 1 5 -1.168 -10.006 3.094 13.249
SAT2 1 5 -0.965 -8.26 2.273 9.730
SAT3 1 5 -0.756 -6.477 1.552 6.646
SAT4 1 5 -0.973 -8.328 3.046 13.043
DCE1 1 5 -1.562 -13.372 3.569 15.28
DCE2 1 5 -1.411 -12.079 3.718 15.919
DCE3 1 5 -1.223 -10.475 2.151 9.211
DCE4 1 5 -1.227 -10.507 2.354 10.078
CON4 1 5 -0.872 -7.469 2.724 11.663
CON3 1 5 -0.647 -5.541 1.592 6.817
CON2 1 5 -0.928 -7.949 2.429 10.402
CON1 1 5 -0.983 -8.419 2.979 12.755
CBPRP4 1 5 -0.98 -8.393 0.296 1.267
CBPRP3 1 5 -0.644 -5.517 -0.05 -0.213
CBPRP2 1 5 -0.208 -1.781 -0.981 -4.198
CBPRP1 1 5 -0.48 -4.112 -0.496 -2.124
CBTRP4 2 5 -0.246 -2.11 0.349 1.495
CBTRP3 2 5 -0.077 -0.657 -0.218 -0.934
CBTRP2 2 5 -0.122 -1.046 0.265 1.137
CBTRP1 2 5 -0.678 -5.807 2.281 9.768
CBPRE1 1 5 0.109 0.929 -0.284 -1.217
Ref. code: 25615802037480DCY
Page 54
42
Table 4.2
Assessment of Multivariate Normality. (Cont’d)
Factor Min Max Skewness Skewness
Critical Ratio Kurtosis
Kurtosis
Critical Ratio
CBPRE2 1 5 0.062 0.527 -0.418 -1.790
CBPRE3 1 5 -0.22 -1.881 0.019 0.081
CBPRE4 1 5 -0.292 -2.498 0.31 1.325
Multivariate Normality 43.122 11.034
It is very important to point that this study samples are not normally
distributed as the normality can be observed via the skewness statistics and skewness
critical ratio which higher than 1.96. This study does not remove any outlier from the
valid samples because this study believes that the responses from the samples are
skewed and not normal by nature. This lead the selection of the algorithm to use in
Structural Equation Model to handle non-normal data which maximum likelihood
estimator should be selected (Gao, Mokhtarian, & Johnston, 2008).
4.1.2 Measurement Model Validity and Reliability
The validity and reliability of the measurement model is analysed
based on the criteria on Chapter 3. Table 4.3 provides a summary of construct including
factor loading, Cronbach’s alpha (a), Average Variance Extracted (AVE), and Construct
Reliability (CR).
Table 4.3
Construct Summary
Construct Measurement Item Factor Loading a AVE CR
Cross-border Perceived Risk
Expectation
(CBPRE)
CBPRE4 0.674 0.839 0.730 0.862
CBPRE3 0.701
CBPRE2 0.781
CBPRE1 0.765
Ref. code: 25615802037480DCY
Page 55
43
Table 4.3
Construct Summary. (Cont’d)
Construct Measurement Item Factor Loading a AVE CR
Cross-border Trusting
Performance
(CBTRP)
CBTRP1 0.747 0.778 0.691 0.910
CBTRP2 0.678
CBTRP3 0.635
CBTRP4 0.702
Cross-border Perceived Risk
Performance
(CBPRP)
CBPRP1 0.654 0.807 0.714 0.881
CBPRP2 0.593
CBPRP3 0.919
CBPRP4 0.688
Repurchase Intention
(CON)
CON1 0.715 0.804 0.712 0.896
CON2 0.712
CON3 0.674
CON4 0.748
Expectation Disconfirmation
(DCE)
DCE4 0.773 0.868 0.785 0.938
DCE3 0.768
DCE2 0.780
DCE1 0.820
Satisfaction
(SAT)
SAT4 0.751 0.813 0.699 0.881
SAT3 0.698
SAT2 0.618
SAT1 0.730
Cross-border Trusting
Expectation
(CBTRE)
CBTRE4 0.669 0.721 0.630 0.876
CBTRE3 0.628
CBTRE2 0.652
CBTRE1 0.571
Note. a = Cronbach’s Alpha; AVE = Average Variance Extracted; CR = Construct Validity
Referring to discriminant validity analysis, the correlation matrix is
observed showing that there is no too highly correlated construct (correlation
coefficients > 0.85) in the model as shown on Table 4.4. Therefore, there is no
multicollinearity evident on the model.
Ref. code: 25615802037480DCY
Page 56
44
Table 4.4
Cross-border E-commerce Variable Correlations for Discriminant Validity
CBTRE CBPRE CBTRP CBPRP DIS SAT CON
CBTRE 1
CBPRE .512** 1
CBTRP 0.083 -.191** 1
CBPRP .189** -0.090 .431** 1
DIS -.115* -.389** .282** .475** 1
SAT -0.006 -.256** .162** .439** .636** 1
CON -.120* -.168** .123** .192** .335** .222** 1
Note. CBTRE = Trusting Expectation; CBPRE = Perceived Risk Performance; CBTRP = Trusting Performance; CBPRP
= Perceived Risk Performance; DIS = Disconfirmation; SAT = Satisfaction; CON = Repurchase Intention
In term of CFA, discriminant validity in each factor is observed to
ensure non-multicollinearity issue on the model. A correlation of 0.85 and higher (or -
0.85 and lower) indicates poor discriminant validity which leads to multicollinearity
issue. The factor correlation on Appendix B’s Table D.1 indicates that all factors in the
model have lower correlation and they are acceptable for CFA and SEM.
Summary of Validity and Reliability of Measurement Model is
provided on the Table 4.5 passing all criteria required which indicates that the model
is valid and reliable to make prediction of the population.
Ref. code: 25615802037480DCY
Page 57
45
Table 4.5
Measurement Model Validity and Reliability Analysis Result
Test Result
Validity
Convergence Validity All items in a measurement model are statistically
significant. (Refer to Table B.1) Other than that, the
value of AVE for all construct is greater than 0.50.
The Convergent Validity was achieved the
required level. (Refer to Table 4.3)
Construct Validity The construct validity was achieved the required
level. (Refer to Table 4.3)
Discriminant Validity The correlation between all constructs are not
lower than -0.85 or higher than 0.85. (Refer to
Table 4.4)
Reliability
Internal Reliability The value of Cronbach Alpha is greater than 0.60.
The internal reliability was achieved the required
level. (Refer Table 4.4)
Construct Reliability The value of CR for all constructs are greater than
0.60. The composite reliability was achieved the
required level.
Average Variance Extracted The value of AVE for all constructs are greater than
0.50. The required level was achieved. (Refer
Table 8)
4.1.2 Samples demographic
Based on 440 samples, descriptive statistics is used to explain
properties of the samples collected. There are six different demographics observed
which are age, gender, occupation, highest education, monthly income, and frequency
of purchase.
Ref. code: 25615802037480DCY
Page 58
46
The histogram on Figure 4.1 below shows the respondent age
distribution mostly fall between late 20s to mid 30s as detail provided which are a
major consumer of e-commerce in Thailand. This statistic is aligned to penetration of
digital adopters in Thailand based on National Statistical Office of Thailand (National
Statistical Office of Thailand, 2019).
Figure 4.1 Histogram of sample gender.
53 percent of total respondents is female, and another 47 percent is male.
This proportion of gender is slightly different to Thailand population which is
approximately symmetric in gender (50 to 50 percent of male and female).
Table 4.6
Gender
Frequency Percent
Female 233 53.0
Male 207 47.0
61 percent of total respondents, a major group, earns less than 35
thousand Thai Baht a month. The second largest group earns around 25 to 50 thousand
Ref. code: 25615802037480DCY
Page 59
47
a month contributed to 33 percent of total respondents. The smallest groups of are
50 to 75 thousand, 75 to 100 thousand, and over 100 thousand monthly income which
contribute 3 percent, 1 percent, and 0.7 percent respectively.
Table 4.7
Monthly Income
Frequency Percent
25K - 50K 149 33.9
50K - 75K 14 3.2
75K - 100K 5 1.1
Less than 25K 269 61.1
Over 100K 3 0.7
74 percent of the respondents obtained bachelor’s degree which is the
major respondents of this study. 10 and 13 percent of the respondents obtain master’s
degree and diploma/certificate respectively. Only 2 percent of the respondents obtain
only high school level. These statistics also represent the distribution of Thai
population regarding the National Statistical Office of Thailand(National Statistical
Office of Thailand, 2019).
Table 4.8
Highest Education Obtained
Frequency Percent
Bachelor’s degree 326 74.1
Diploma/Certificates 58 13.2
High School 9 2.0
Master’s degree 47 10.7
69 percent of the respondents works in private company while a 10 and
15 percent are students and government officers respectively. The smallest group of
Ref. code: 25615802037480DCY
Page 60
48
respondents reported at 5.2 percent is a business owner or a freelancer group as
showed on Table 4.9 below.
Table 4.9
Occupation
Frequency Percent
Business owner/ Freelancer 23 5.2
Employee 307 69.8
Government Officers 66 15.0
Student 44 10.0
In term of frequency of purchase via cross-border e-commerce, the
participants purchase 3 times per month on average; however, there are some
respondents make purchase more than 10 times a month. Nearly 50 respondents make
purchase via cross-border e-commerce only once a month. The distribution of
purchase frequency is showed on Figure 4.2 below.
Figure 4.2 Frequency participants’ purchase via cross-border e-commerce (monthly)
Ref. code: 25615802037480DCY
Page 61
49
4.2 Paired T-test for Mean Difference
In this study, paired t-test are conducted to measure statistical difference
of pre-purchase trusting expectation, trusting performance, and perceived risk
expectation and perceived risk performance between cross-border e-commerce and
domestic e-commerce. Additionally, the paired t-test is used to compare mean
difference between pre-purchase trusting belief and perceived risk (Trusting
Expectation and Perceived Risk Performance) and post-purchase trusting belief and
perceived risk (Trusting Performance and Perceived Risk Performance). The pair
summary of each paired t-test conducted in this study is provided on Table 4.10.
Table 4.10
Pair Summary for Paired T-test of Mean Difference.
Pair Variables
Pair 1 Cross-border Trusting Expectation (CBTRE)
Domestic Trusting Expectation (DMTRE)
Pair 2 Cross-border Perceived Risk Expectation (CBPRE)
Domestic Perceived Risk Expectation (DMPRE)
Pair 3 Cross-border Trusting Performance (CBTRP)
Domestic Trusting Performance (DMTRP)
Pair 4 Cross-border Perceived Risk Performance (CBPRP)
Domestic Perceived Risk Performance (DMPRP)
Pair 5 Cross-border Trusting Expectation (CBTRE)
Cross-border Trusting Performance (CBTRP)
Pair 6 Cross-border Perceived Risk Expectation (CBPRE)
Cross-border Perceived Risk Performance (CBPRP)
A paired t-test on Pair 1 showed statistically insignificant difference
between pre-purchase trusting belief (Trusting Expectation) and post-purchase trusting
belief (Trusting Performance) from cross-border e-commerce (Mean difference = -.018,
SD = .488, p > 0.05) with a 95% confidence interval ranging from -0.63 to .027.
Ref. code: 25615802037480DCY
Page 62
50
A paired t-test on Pair 2 showed statistically significant increase of post-
purchase perceived risk (Perceived Risk Performance) over pre-purchase perceived risk
(Perceived Risk Expectation) from cross-border e-commerce (Mean difference = -.018,
SD = .658, p < 0.05) with a 95% confidence interval ranging from .116 to .240.
A paired t-test on Pair 3 showed statistically significant lower mean of
cross-border trusting belief (Trusting Expectation) over domestic e-commerce trusting
belief (Trusting Expectation) commerce on pre-purchase stage (Mean difference = -
.260, SD = .613, p < 0.05) with a 95% confidence interval ranging from -.317 to -.202.
A paired t-test on Pair 4 showed statistically significant lower mean of
cross-border perceived risk (Perceived Risk Expectation) over domestic perceived risk
(Perceived Risk Expectation) from pre-purchase stage (Mean difference = -.138, SD =
.765, p < 0.05) with a 95% confidence interval ranging from -.209 to -.066.
A paired t-test on Pair 5 showed statistically significant lower mean of
cross-border trusting belief (Trusting Performance) over domestic trusting belief
(Trusting Performance) from post-purchase stage (Mean difference = -.193, SD = .649,
p < 0.05) with a 95% confidence interval ranging from -.254 to -.132.
A paired t-test on Pair 6 showed statistically insignificant difference
between cross-border perceived risk (Perceived Risk Performance) and domestic
perceived risk (Perceived Risk Performance) from post-purchase stage (Mean difference
= .007, SD = .970, p > 0.05) with a 95% confidence interval ranging from -0.82 to .098.
The summary detail is provided on Table 4.11
Ref. code: 25615802037480DCY
Page 63
Table 4.11 Paired Samples Test of Mean between Pre-purchase and Post-purchase from Cross-border E-commerce and Domestic E-commerce.
Paired Differences
t df Sig. (2-tailed) Mean Std. Deviation Std. Error Mean
95% Confidence Interval of the Difference
Lower Upper Pair 1 CBTRE - DMTRE -.260 .613 .029 -.317 -.202 -8.898 439 .000** Pair 2 CBPRE - DMPRE -.138 .765 .036 -.209 -.066 -3.784 439 .000** Pair 3 CBTRP - DMTRP -.193 .649 .030 -.254 -.132 -6.262 439 .000** Pair 4 CBPRP - DMPRP .007 .970 .046 -.082 .098 .172 439 .864 Pair 5 CBTRE - CBTRP -.018 .488 .023 -.063 .027 -.780 439 .436 Pair 6 CBPRE - CBPRP .178 .658 .031 .116 .240 5.681 439 .000** Note. CBTRE = Cross-border Trusting Expectation, CBPRE = Cross-border Perceived Risk Expectation, CBTRP = Cross-border Trusting Performance, CBPRP = Cross-border Perceived Risk Performance, DMTRE = Domestic Trusting Expectation, DMPRE = Domestic Perceived Risk Expectation, DMTRP = Domestic Trusting Performance, DMPRP = Domestic Perceived Risk Performance * p < .01. **p < .001.
51
Ref. code: 25615802037480DCY
Page 64
52
4.3 Confirmatory Factor Analysis
In Confirmatory Factor Analysis, each factor in the model is observed their
critical ratio (t-value) in which C.R. higher than 2.56 indicate significant error variance at
0.01 level. In this study, all factors in the model have very high C.R. indicating that they
are significant in term of CFA. Square Multiple Correlation (SMC) on each latent variable
are positive indicating significant relationship between each measurement items and
its latent variable.
Table 4.12
Confirmatory Factor Analysis Coefficients and Squared Multiple Coefficients.
Observed
Variable
Latent
Construct b B S.E. C.R. SMC
P-
value
CBPRE4
CBPRE
0.674 1.000 0.447
CBPRE3 0.701 1.088 0.082 13.282 0.394 ***
CBPRE2 0.781 1.481 0.139 10.656 0.425 ***
CBPRE1 0.765 1.369 0.130 10.516 0.326 ***
CBTRP1
CBTRP
0.747 1.000 0.558
CBTRP2 0.678 1.084 0.085 12.780 0.459 ***
CBTRP3 0.635 1.182 0.098 12.030 0.404 ***
CBTRP4 0.702 1.135 0.086 13.198 0.493 ***
CBPRP1
CBPRP
0.654 1.000 0.414
CBPRP2 0.593 0.965 0.067 14.417 0.356 ***
CBPRP3 0.919 1.489 0.120 12.450 0.847 ***
CBPRP4 0.688 0.950 0.094 10.152 0.445 ***
CON1
CON
0.715 1.000 0.474
CON2 0.712 1.033 0.082 12.582 0.505 ***
CON3 0.674 0.930 0.077 12.049 0.439 ***
CON4 0.748 1.038 0.080 13.001 0.524 ***
Ref. code: 25615802037480DCY
Page 65
53
Table 4.12
Confirmatory Factor Analysis Coefficients and Squared Multiple Coefficients.
(Cont’d)
Observed
Variable
Latent
Construct b B S.E. C.R. SMC
P-
value
DCE4 DCE 0.773 1.000 0.598
DCE3 0.768 1.020 0.063 16.218 0.589 ***
DCE2 0.780 0.968 0.059 16.537 0.609 ***
DCE1 0.820 0.927 0.052 17.659 0.672 ***
SAT4 SAT 0.751 1.000 0.564
SAT3 0.698 0.998 0.076 13.090 0.487 ***
SAT2 0.618 0.901 0.081 11.182 0.382 ***
SAT1 0.730 1.070 0.079 13.596 0.533 ***
CBTRE4 CBTRE 0.669 1.000 0.447
CBTRE3 0.628 0.957 0.097 9.894 0.488 ***
CBTRE2 0.652 0.918 0.091 10.125 0.617 ***
CBTRE1 0.571 0.616 0.067 9.264 0.586 ***
Note. b = Standardized Coefficient Estimate; B = Unstandardized Coefficients; S.E. = Standard Error; P = p-value;
C.R. = Critical Ratio (t-value); SMS = Squared Multiple Correlation
The Goodness of Fit (GOF) of the model for baseline model (c2 =
518.700, DF = 328) is provided from Confirmatory Factor Analysis. The GOF is evaluated
in accordance to recommendation criteria on Chapter 3 which are now summaries in
Table 4.13 below. The model passes TLI, CFI, and SRMR criteria but does not achieve
the NFI criteria. In term of combination criteria, RMSEA & SRMR and CFI & SRMR are
achieved; however, only the combination of NNFI (TLI) & SRMR is not achieved. In
conclusion, most of the criteria for goodness of fit is achieved to justify that this model
is valid.
Ref. code: 25615802037480DCY
Page 66
54
Table 4.13 Structural Model Goodness of Fit Evaluation Summary.
Fit Index Acceptable Threshold Level Model Result
Model Fit Acceptance
Comparative Fit Index NFI NFI > 0.95 0.901 Not
Accepted TLI TLI > 0.95 0.955 Accepted CFI CFI > 0.95 0.960 Accepted
Other Fit Index SRMR SRMR < 0.08 0.055 Accepted RMSEA RMSEA < 0.07
RMSEA < 0.03 indicates excellent fit.
0.037 Accepted
Combination Criteria NNFI (TLI) and SRMR NNFI (TLI) ³ 0.96 and SRMR £
0.09.
Not
Accepted RMSEA and SRMR RMSEA £ and SRMR £ 0.09.
Accepted
CFI and SRMR CFI ³ 0.96 and SRMR £ 0.09 Accepted
4.4 Structural Equation Model
Result from SEM estimation, this model converges properly with Maximum
Likelihood Estimator providing overall model summary and CFA result as discussed on
the previous section. The model also identifies direct effect, indirect effect, and total
effect as shown on Table 4.14 below. The discussion of direct and indirect effect
within the model are discussed on 4.4.1 and 4.1.2.
Ref. code: 25615802037480DCY
Page 67
55
4.4.1 Direct Effect
This study hypothesises that each construct formulates a direct
relationship to the dependent variable; therefore, the direct effect is observed. Trusting
Expectation (CBTRE) (Beta = 0.11, p > 0.05) and Perceived Risk Expectation (CBPRE)
(Beta = 0.04, p > 0.05) are not significantly related to Expectation Disconfirmation (DCE)
and they have very little effect size towards the Expectation Disconfirmation (DCE).
Trusting Expectation (CBTRE) has positive effect on Trusting Performance (CBTRP) on
the post-purchase stage (Beta = 0.58, p < 0.05). Perceived Risk Expectation (CBPRE) has
positive effect on Perceived Risk Performance (CBPRP) on the post-purchase stage (Beta
= 0.62, p < 0.05). Trusting Performance is predictive of Expectation Disconfirmation
(DCE) (Beta = 0.53, p <0.05) as well as the Perceived Risk Performance (CBPRP) (Beta =
-0.43, p < 0.05). The relationship of Expectation Disconfirmation (DCE), Satisfaction
(SAT), and Repurchase Intention (CON) is significant as well where Expectation
Disconfirmation (DCE) is predictive of Satisfaction (SAT) (Beta = 0.80, p < 0.05) and
Satisfaction (SAT) is predictive of Repurchase Intention (CON) (Beta = 0.51, p < 0.05).
Figure 4.3 Structural Model Standardised Coefficients Note. c2 = 518.700; DF = 328; Standardised Root Mean Square Residuals (SRMR) = 0.030; Root Mean Square Error
of Approximation (RMSEA) = 0.037; Normed Fit Index (NFI) = 0.901; Non-normed Fit Index (TLI) = 0.955;
Comparative Fit Index (CFI) = 0.960.
Ref. code: 25615802037480DCY
Page 68
56
4.4.2 Indirect Effect
We hypothesized that the relationship between the belief that
knowledge as isolated facts and course achievement was mediated has an indirect
effect on course achievement, by deep processing. The result (standardized indirect
coefficient = –.06, p > .05) was not statistically significant.
Table 4.14
Standardized Coefficients Effect in Structural Model.
Model Effect Standardized Coefficients (b)
R2 CBTRE CBTRP CBPRE CBPRP DCE SAT CON
CBTRP 0.589 0.000
CBPRE 0.117 0.347
CBPRP -0.300 0.625 0.014
DCE -0.850 0.585 0.490 -0.436 0.459
SAT 0.820 0.589
CON 0.510 0.642
CBTRP
CBPRE
CBPRP -0.140
DCE 0.395 0.131 -0.272
SAT 0.249 0.574 -0.179 -0.349
CON 0.127 0.293 -0.910 -0.178 0.490
CBTRP 0.589
CBPRE 0.117
CBPRP -0.140 -0.300 0.625
DCE 0.310 0.715 -0.224 -0.436
SAT 0.249 0.574 -0.179 -0.349 0.820
CON 0.127 0.293 -0.910 -0.178 0.490 0.510
Ref. code: 25615802037480DCY
Page 69
57
4.5 Hypothesis Testing Summary
Based on SEM, each hypothesis proposed in this study has been evaluated
and summarized as shown on the Table 4.15 for paired T-test and 4.16 for SEM model
hypothesis.
4.5.1 Hypothesis Testing Result
In term of Paired t-test Hypothesis testing, the result shows that the
H2c and H3c are supported as referred to Table 4.18 (p < .05) within 95% confidence
interval and the mean difference is negative in accordance to the research hypothesis.
Table 4.15
Paired T-test Hypothesis Testing Result
Hypothesis Result
H2c Trusting Expectation from cross-border e-commerce is less than
domestic e-commerce on average. Supported
H2d Perceived Risk Expectation from cross-border e-commerce is higher
than domestic e-commerce on average. Not Supported
H3c Trusting Performance from cross-border e-commerce is less than
domestic e-commerce on average. Supported
H3d Perceived Risk Performance from cross-border e-commerce is higher
than domestic e-commerce on average. Not Supported
H4b Trusting Performance (post-purchase) is increased from Trusting
Expectation (pre-purchase). Not Supported
H5b Perceived Risk Performance (post-purchase) is decreased from
Perceived Risk Expectation (pre-purchase). Not Supported
The Structural Equation Model conducted on the cross-border
variables provides a hypothesis testing result as provided on Table 4.16 below. The
hypotheses H1, H2a and H2b are not supported in this study. The discussion on the
hypothesis result is discussed on 4.5.1 to 4.5.5.
Ref. code: 25615802037480DCY
Page 70
58
Table 4.16
SEM Hypothesis Testing Result
Hypothesis Result
H1 Trusting Expectation from cross-border e-commerce has negative
effect on Perceived Risk Expectation from cross-border e-commerce.
Not Supported
H2a Trusting Expectation from cross-border e-commerce has positive effect
on Expectation Disconfirmation from cross-border e-commerce.
Not Supported
H2b Perceived Risk Expectation from cross-border e-commerce has
negative effect on Expectation Confirmation from cross-border e-
commerce.
Not Supported
H3a Trusting Performance from cross-border e-commerce has positive
effect on Expectation Confirmation from cross-border e-commerce.
Supported
H3b Perceived Risk Performance from cross-border e-commerce has
negative effect on Expectation Confirmation from cross-border e-
commerce.
Supported
H4a Trusting Expectation has positive effect on Trusting Performance. Supported
H5a Perceived Risk Expectation has negative effect on Perceived Risk
Performance.
Supported
H6 Trusting Performance has negative effect on Perceived Risk
Performance
Supported
H7 Expectation Disconfirmation has positive effect on Satisfaction. Supported
H8 Satisfaction has positive effect on Repurchase Intention. Supported
4.5.2 Relationship of Trusting Expectation and Perceived Risk
Expectation
The result from hypothesis test shows that consumer’s expectation
of Trust is not influential to Perceived Risk Expectation (H1) which is against the
previous study on trust and risk interaction (Kim et al., 2008, 2009; Mou et al., 2015;
Zhu et al., 2009); however, this relationship is may not relevant in the context of cross-
border e-commerce amongst Thai consumers.
Ref. code: 25615802037480DCY
Page 71
59
4.5.3 Relationship Trusting Expectation, Perceived Risk Expectation,
and Expectation Disconfirmation.
The regression analysis result suggests that the Trusting Expectation
and Perceived Risk Expectation are significantly affect consumer’s Expectation-
disconfirmation of purchasing via cross-border e-commerce. The relationship of
expectation variables in this study reject the result from previous study on the
existence of the expectation construct in the EDT (Ambalov, 2018; Bhattacherjee, 2001;
N. Lankton et al., 2014; N. K. Lankton et al., 2016; Zhang, Lu, Gupta, & Gao, 2015), and
the extension of expectation to Trusting Expectation and Perceived Risk Expectation.
The interesting finding from this expectation variables is that they do not expect the
e-sellers much in term of trust and risk or they have quite vary perception in the
shopper’s mind.
Trusting Performance (H2a) and Perceived Risk Performance (H2b)
which extend the original EDT model (Ambalov, 2018; Bhattacherjee, 2001; Kim et al.,
2009; N. Lankton et al., 2014; N. K. Lankton et al., 2016) replacing the post-purchase
stage’s Performance in this study are significantly affect Expectation-disconfirmation
in the EDT model as well.
4.5.4 Relationship of Trusting Performance, Perceived Risk
Performance, and Expectation Disconfirmation.
Trusting Performance and Perceived Risk Performance is a post-
purchase evaluation of Trust and Risk at the pre-purchase stage. The hypothesis test
result suggests that the Trusting Performance has positive effect over customer
disconfirmation (H3a) while the Perceived Risk Performance negatively influences the
disconfirmation (H5). This result is aligned with the EDT in which the Performance
variables have influence Expectation Disconfirmation (Ambalov, 2018; Bhattacherjee,
2001; N. K. Lankton et al., 2016; Shang & Wu, 2017). Trusting Performance (H3a) and
Perceived Risk Performance (H3b) which extend the original EDT model replacing the
post-purchase stage’s Perceived Performance in this study (N. Lankton et al., 2014) are
significantly affect Expectation-disconfirmation in the EDT model as well.
Ref. code: 25615802037480DCY
Page 72
60
4.5.5 Relationship of Trusting Performance and Perceived Risk
Performance
The result from hypothesis test shows that consumer Trust does not
significant affect Perceived Risk Expectation (H6) in the accordance to the Trusting
Expectation and Perceived Risk Expectation (H1). This is an alignment that may suggest
that trusting belief and perceived risk interaction may not exist in cross-border e-
commerce context; however, in the other social context, these variables may be
related.
4.5.6 Relationship of Disconfirmation, Satisfaction and Repurchase
Intention
Based on the regression analysis and hypothesis testing, the result
supports the hypothesis that the disconfirmation has positive influence on consumer
satisfaction (H7), and the consumer satisfaction has positive effect on continuance
intention (H8). These relationships affirm the EDT variables relationships in previous
study (Ambalov, 2018; Bhattacherjee, 2001; N. Lankton et al., 2014; N. K. Lankton et
al., 2016; Zhang et al., 2015), but the result in this study is in a context of cross-border
e-commerce.
Ref. code: 25615802037480DCY
Page 73
61
CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS
In this chapter, conclusion of the study has provided as well as benefit of
the study, limitation, and recommendation for future study.
5.1 Conclusion
E-commerce has become a new world economy due to advanced
technology nowadays. The connecting world has provided a vast opportunity in
commerce for everybody and the e-commerce has been expanding from domestic
shopping into cross-border shopping; however, there are many barriers that can slow
down the cross-border e-commerce especially risks and trust that the consumer needs
to be evaluate. The domestic sellers have to adapt themselves to the new wave of
commerce transformation and challenges. Understanding the key factor that the
consumer adopt cross-border e-commerce is a key to their competitiveness in both
retaining domestic customer and expanding current business to become cross-border
e-commerce.
This study has conducted a measurement of consumer trusting belief and
perceived risk in pre and post purchase within an Expectation-disconfirmation Theory
in context of cross-border e-commerce. The result from this study shows that trust
and risk in cross-border e-commerce context are significantly important in post-
purchase stage which is the same result in the previous study. This study also confirm
that the expectation of these constructs not is settled well within Expectation-
disconfirmation Theory, and relationships between trust and risk in this study shows
insignificance.
In term of difference perception of pre-to-post purchase trust, the paired
t-test shows that the consumer expects that domestic e-sellers would be more
Ref. code: 25615802037480DCY
Page 74
62
trustworthy than cross-border e-sellers, and their belief is confirmed as showed in the
post-purchase trusting belief which is still lower than domestic e-commerce.
Regarding the difference perception of pre-to-post purchase perceived risk,
the paired t-test shows that the consumer expects that cross-border e-commerce
would be less risky than domestic e-commerce. Interestingly, the consumer perceived
that purchasing from cross-border e-commerce may be risker than expected but the
risk in not different to purchasing via domestic e-commerce. This evaluation of risks is
different to trust for trust is determining on human-like trust while the perceived risks
involves process and systems. This phenomenon imply that cross-border e-commerce
may have more or equally risk over systems and process to domestic e-commerce
while e-seller in cross-border shops are less trustworthy.
The relationship of cross-border variables in the analysis shows that the
effect of trust and risk towards consumer disconfirmation of their initial belief is strong
and stronger in the post-purchase stage. It is very important to firms who would like
to gain higher competitive advantages over the others to pay their attention to building
trust over e-sellers and reducing risks over systems and processes. A proper
management on trust and risk in cross-border e-commerce or domestic e-commerce
will be resulted in higher satisfaction and loyalty of the consumers.
In the context of domestic e-commerce providers, trust and risk are crucial
factors to compete with this new era of colonisation, cross-border e-commerce
expansion, as reflected on t-test that the consumers still believe that domestic e-
sellers are more reliable, trust worthy, and less risky. If the belief is confirmed and
intensified positively, the domestic e-commerce sellers will position themselves
strongly against a high influential cross-border e-commerce firm.
5.2 Benefit
This study provides two aspects of benefits which are theoretical benefit
and the practical benefit for business and practitioners in the area the e-commerce.
Ref. code: 25615802037480DCY
Page 75
63
5.2.1 Theoretical benefit
This study derived the Expectation Disconfirmation Theory to apply
on a context of cross-border e-commerce. The previous studies have been formulating
the EDT model with various constructs; however, in the cross-border e-commerce
setting, the knowledge have not been explored. This study has expanded the
Expectation-disconfirmation Theory to the cross-border e-commerce extent. Moreover,
the previous study on Trust and EDT (Bhattacherjee, 2001; N. Lankton et al., 2014) on
e-service has been reformulated with Perceived Risk in area of information system
(Featherman & Pavlou, 2003; P. A. Pavlou, 2003) into the EDT model providing insight
in how cross-border shoppers perceived towards e-sellers an service providers.
Additionally, the Expectation-disconfirmation Theory with pre-purchase and post-
purchase evaluation driving with Trusting Belief and Perceived Risks in the context
cross-border e-commerce is well settled harmoniously; however, the interaction of
Trust and Risk in previous research (Kim et al., 2008; Mou et al., 2015; P. A. Pavlou,
2003; Visschers & Siegrist, 2008; Zhu et al., 2009) is not confirmed in this study.
5.2.2 Practical benefit
Entrepreneurs and business owner who are running e-commerce and
is suffering from challenges from big e-commerce companies can adopt this study to
improve their business to be able to compete or mitigate the impact of cross-border
expansion. On domestic firms’ perspective, the many big firms running cross-border e-
commerce are challenging domestic e-commerce provider with their vast amount of
investment; however, the result of Trust and Risk in this study suggests the domestic
e-commerce firms to be proactive to raising trust and reducing risks as they are highly
influential to consumer disconfirmation of their prior belief as well as the post-
purchase trust and risk perception. Vice versa, for entrepreneurs who is looking for
expanding their cross-border e-commerce, especially to Thailand, they need to
carefully develop strategies to capture consumer trust and reducing risks especially
risk of deliver and logistic service in Thailand. The product and financial risk are more
concerning factors in post-purchase via cross-border e-commerce i.e. the product
might be defective or bogus, and the process of transaction both payment and refund
Ref. code: 25615802037480DCY
Page 76
64
involve more risk concerns as well which may due to distance effect of sellers and
consumers.
5.3 Limitation
This study was conducted on Thai population only. The result may not be
able to be applied to people in different countries or different social and culture
context. Additionally, the cross-border e-commerce in Thailand is still at a growing
state and not matured as global e-commerce; therefore, the consumer may not
compare them distinctively. Moreover, risks in this study are evaluated as a general
perception in combination of financial risk, delivery risk, and overall risk. This study
measures the pre-purchase expectation as consumer recall of their past-experience
rather than measuring before the actual purchase; therefore, the result of expectation
construct may be distorted by time of answering the questions.
5.4 Recommendation for future study
The result from this study shows that the incorporating trust and risks into
The EDT model is significant; however, the interaction on the trust and risks does not
support the hypothesis. The future research on cross-border e-commerce may observe
their relationships again in a different culture and social setting which may have
different outcome. The perceived risk in this study may be separated and emphasize
independently in the future research to observe them differently such as financial risk,
performance risk (product defection or bogus), and overall risk. There are other factors
that might affect the perception of risk amongst the consumer as well, but they are
not included in this study such as psychological risk and social risk. Moreover, the
research design in the future may conduct twice in a style of longitudinal study
separating the pre-purchase and post-purchase measurement conducting on the same
samples.
Ref. code: 25615802037480DCY
Page 77
65
REFERENCE
Journal Article
Ambalov, I. A. (2018). A meta-analysis of IT continuance: An evaluation of the
expectation-confirmation model. Telematics and Informatics, 35(6), 1561-
1571. doi:10.1016/j.tele.2018.03.016
Bagozzi, R. P., & Yi, Y. J. J. o. t. A. o. M. S. (1988). On the evaluation of structural
equation models. 16(1), 74-94. doi:10.1007/bf02723327
Bhattacherjee, A. (2001). Understanding Information Systems Continuance: An
Expectation-Confirmation Model. MIS Quarterly, 25(3), 351-370.
doi:10.2307/3250921
Bonsón Ponte, E., Carvajal-Trujillo, E., & Escobar-Rodríguez, T. (2015). Influence of
trust and perceived value on the intention to purchase travel online:
Integrating the effects of assurance on trust antecedents. Tourism
Management, 47, 286-302. doi:10.1016/j.tourman.2014.10.009
Cerny, B. A., & Kaiser, H. F. (1977). A Study Of A Measure Of Sampling Adequacy For
Factor-Analytic Correlation Matrices. Multivariate Behavioral Research, 12(1),
43-47. doi:10.1207/s15327906mbr1201_3
Chong, A. Y.-L. (2015). Understanding Mobile Commerce Continuance Intentions: An
Empirical Analysis of Chinese Consumers. Journal of Computer Information
Systems, 53(4), 22-30. doi:10.1080/08874417.2013.11645647
Fang, J., George, B., Shao, Y., & Wen, C. (2016). Affective and cognitive factors
influencing repeat buying in e-commerce. Electronic Commerce Research and
Applications, 19, 44-55. doi:10.1016/j.elerap.2016.08.001
Featherman, M. S., & Pavlou, P. A. (2003). Predicting e-services adoption: a perceived
risk facets perspective. International Journal of Human-Computer Studies,
59(4), 451-474. doi:10.1016/S1071-5819(03)00111-3
Ref. code: 25615802037480DCY
Page 78
66
Fisher, R., & Zoe Chu, S. (2009). Initial online trust formation: the role of company
location and web assurance. Managerial Auditing Journal, 24(6), 542-563.
doi:10.1108/02686900910966521
Gao, S., Mokhtarian, P. L., & Johnston, R. A. (2008). Nonnormality of Data in Structural
Equation Models. 2082(1), 116-124. doi:10.3141/2082-14
Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in Online Shopping:
An Integrated Model. MIS Quarterly, 27(1), 51-90. doi:10.2307/30036519
Guo, Y., Bao, Y., Stuart, B. J., & Le-Nguyen, K. (2018). To sell or not to sell: Exploring
sellers' trust and risk of chargeback fraud in cross-border electronic
commerce. Information Systems Journal, 28(2), 359-383. doi:10.1111/isj.12144
Hong, I. B. (2015). Understanding the consumer's online merchant selection process:
The roles of product involvement, perceived risk, and trust expectation.
International Journal of Information Management, 35(3), 322-336.
doi:10.1016/j.ijinfomgt.2015.01.003
Hu, L. t., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation
Modeling: A Multidisciplinary Journal, 6(1), 1-55.
doi:10.1080/10705519909540118
Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39(1), 31-36.
doi:10.1007/BF02291575
Kim, D. J. (2014). A Study of the Multilevel and Dynamic Nature of Trust in E-
Commerce from a Cross-Stage Perspective. International Journal of Electronic
Commerce, 19(1), 11-64. doi:10.2753/jec1086-4415190101
Kim, D. J., Ferrin, D. L., & Rao, H. R. (2008). A trust-based consumer decision-making
model in electronic commerce: The role of trust, perceived risk, and their
antecedents. Decision Support Systems, 44(2), 544-564.
doi:10.1016/j.dss.2007.07.001
Kim, D. J., Ferrin, D. L., & Rao, H. R. (2009). Trust and Satisfaction, Two Stepping
Stones for Successful E-Commerce Relationships: A Longitudinal Exploration.
Information Systems Research, 20(2), 237-257. doi:10.1287/isre.1080.0188
Ref. code: 25615802037480DCY
Page 79
67
Lankton, N., McKnight, D. H., & Thatcher, J. B. (2014). Incorporating trust-in-technology
into Expectation Disconfirmation Theory. The Journal of Strategic Information
Systems, 23(2), 128-145. doi:10.1016/j.jsis.2013.09.001
Lankton, N. K., McKnight, D. H., Wright, R. T., & Thatcher, J. B. (2016). Research Note—
Using Expectation Disconfirmation Theory and Polynomial Modeling to
Understand Trust in Technology. Information Systems Research, 27(1), 197-
213. doi:10.1287/isre.2015.0611
Lee, J., Park, D. H., & Han, I. (2011). The different effects of online consumer reviews
on consumers' purchase intentions depending on trust in online shopping
malls. Internet Research, 21(2), 187-206. doi:10.1108/10662241111123766
Marinkovic, V., & Kalinic, Z. (2017). Antecedents of customer satisfaction in mobile
commerce. Online Information Review, 41(2), 138-154. doi:10.1108/oir-11-
2015-0364
McKnight, D. H., Carter, M., Thatcher, J. B., & Clay, P. F. (2011). Trust in a specific
technology. ACM Transactions on Management Information Systems, 2(2), 1-
25. doi:10.1145/1985347.1985353
McKnight, D. H., & Chervany, N. L. (2014). What Trust Means in E-Commerce Customer
Relationships: An Interdisciplinary Conceptual Typology. International Journal
of Electronic Commerce, 6(2), 35-59. doi:10.1080/10864415.2001.11044235
McKnight, D. H., Choudhury, V., & Kacmar, C. (2002). Developing and Validating Trust
Measures for e-Commerce: An Integrative Typology. Information Systems
Research, 13(3), 334-359. doi:10.1287/isre.13.3.334.81
Mou, J., Shin, D.-H., & Cohen, J. F. (2015). Trust and risk in consumer acceptance of e-
services. Electronic Commerce Research, 17(2), 255-288. doi:10.1007/s10660-
015-9205-4
Mukherjee, A., Arnott, D. C., & Nath, P. (2007). Role of electronic trust in online
retailing. European Journal of Marketing, 41(9/10), 1173-1202.
doi:10.1108/03090560710773390
Naovarat, S. (2015). Factors that affect intention to purchase of customer and the
role of trust as a moderator in E-Marketplace. Thammasat University,
Ref. code: 25615802037480DCY
Page 80
68
Oliver, R. L. (1980). A Cognitive Model of the Antecedents and Consequences of
Satisfaction Decisions. Journal of Marketing Research, 17(4), 460-469.
doi:10.2307/3150499
Pappas, N. (2016). Marketing strategies, perceived risks, and consumer trust in online
buying behaviour. Journal of Retailing and Consumer Services, 29, 92-103.
doi:10.1016/j.jretconser.2015.11.007
Park, J., Lee, D., & Ahn, J. (2014). Risk-Focused E-Commerce Adoption Model: A Cross-
Country Study. Journal of Global Information Technology Management, 7(2),
6-30. doi:10.1080/1097198x.2004.10856370
Pavlou, & Fygenson. (2006). Understanding and Predicting Electronic Commerce
Adoption: An Extension of the Theory of Planned Behavior. MIS Quarterly,
30(1), 115-115. doi:10.2307/25148720
Pavlou, P. A. (2003). Consumer Acceptance of Electronic Commerce: Integrating Trust
and Risk with the Technology Acceptance Model. International Journal of
Electronic Commerce, 7(3), 101-134.
Pham, T. S. H., & Ahammad, M. F. (2017). Antecedents and consequences of online
customer satisfaction: A holistic process perspective. Technological
Forecasting and Social Change, 124, 332-342.
doi:10.1016/j.techfore.2017.04.003
Rouibah, K., Lowry, P. B., & Hwang, Y. (2016). The effects of perceived enjoyment and
perceived risks on trust formation and intentions to use online payment
systems: New perspectives from an Arab country. Electronic Commerce
Research and Applications, 19, 33-43. doi:10.1016/j.elerap.2016.07.001
Schreiber, J. B., Nora, A., Stage, F. K., Barlow, E. A., & King, J. (2006). Reporting
Structural Equation Modeling and Confirmatory Factor Analysis Results: A
Review. The Journal of Educational Research, 99(6), 323-338.
doi:10.3200/JOER.99.6.323-338
Sfenrianto, S., Wijaya, T., & Wang, G. (2018). Assessing the Buyer Trust and Satisfaction
Factors in the E-Marketplace. Journal of theoretical and applied electronic
commerce research, 13(2), 43-57. doi:10.4067/s0718-18762018000200105
Ref. code: 25615802037480DCY
Page 81
69
Shang, D., & Wu, W. (2017). Understanding mobile shopping consumers’ continuance
intention. Industrial Management & Data Systems, 117(1), 213-227.
doi:10.1108/imds-02-2016-0052
Stouthuysen, K., Teunis, I., Reusen, E., & Slabbinck, H. (2018). Initial trust and
intentions to buy: The effect of vendor-specific guarantees, customer reviews
and the role of online shopping experience ☆. Electronic Commerce
Research and Applications, 27, 23-38. doi:10.1016/j.elerap.2017.11.002
Student. (1908). The Probable Error of Mean. Biometrika, 6(1), 1-25.
doi:10.1093/biomet/6.1.1
Visschers, V. H. M., & Siegrist, M. (2008). Exploring the Triangular Relationship between
Trust, Affect, and Risk Perception: A Review of the Literature. Risk
Management, 10(3), 156-167.
Xiao, Z., Wang, J. J., & Liu, Q. (2018). The impacts of final delivery solutions on e-
shopping usage behaviour. International Journal of Retail & Distribution
Management, 46(1), 2-20. doi:10.1108/ijrdm-03-2016-0036
Zhang, H., Lu, Y., Gupta, S., & Gao, P. (2015). Understanding group-buying websites
continuance. Internet Research, 25(5), 767-793. doi:10.1108/IntR-05-2014-0127
Conference Preceedings
Harrison McKnight, D., & Chervany, N. L. (2001, 2001//). Trust and Distrust Definitions:
One Bite at a Time. Paper presented at the Trust in Cyber-societies, Berlin,
Heidelberg.
Lesma, V. R. B., & Okada, H. (2012/10//). Feedback and trust-related factors of
consumer behavior in cross-border electronic commerce.
Mou, J., Cohen, J., Dou, Y., & Zhang, B. (2017). Predicting Buyers’ Repurchase
Intentions in Cross-Border E-Commerce: a Valence Framework Perspective,
Guimarães, Portugal.
Ref. code: 25615802037480DCY
Page 82
70
Zhu, D., Neal, G. S. O., Lee, Z., & Chen, Y. (2009, 29-31 Aug. 2009). The Effect of Trust
and Perceived Risk on Consumers' Online Purchase Intention. Paper
presented at the 2009 International Conference on Computational Science
and Engineering.
Electronic Media
Amaral, T. P. M., Teresa Garín; Iñigo, García Herguera; Unda, Angel Valarezo. (2017).
Drivers and Barriers to cross-border e-commerce: evidence from Spanish
household.
DHL Express. (2016). THE 21ST CENTURY SPICE TRADEA GUIDE TO THE CROSS-BORDER
E-COMMERCE OPPORTUNITY. Retrieved from Press Release website:
http://www.dhl.com/content/dam/downloads/g0/press/publication/g0_dhl_e
xpress_cross_border_ecommerce_21st_century_spice_trade.pdf
International Post Corporation. (2018). IPC CROSS-BORDER E-COMMERCE SHOPPER
SURVEY. (January, 2018), 24. https://www.ipc.be/services/market-
research/cross-border-shopper-survey
Iwamoto, K. (2018). ASEAN to promote cross-border e-commerce with new
framework. NIKKEI ASIAN REVIEW.
Kenny, D. A. (2015). Measuring Model Fit. Retrieved from
http://www.davidakenny.net/cm/fit.htm
National Statistical Office of Thailand. (2019). Demography Population and Housing
Branch. Retrieved 10/01/2019
http://statbbi.nso.go.th/staticreport/page/sector/en/01.aspx
Nau, R. (2018, 1/06/2018). Regression diagnostics: testing the assumptions of linear
regression. Retrieved from http://people.duke.edu/~rnau/testing.htm
Priest, C. (2017). Commerce 2018 — Internet Trends Report. Retrieved from Medium
website: https://medium.com/@cameronpriest/commerce-2018-internet-
trends-report-cd5ee157bb0d
Ref. code: 25615802037480DCY
Page 83
71
SEA Limited. (2017). FORM F-1 REGISTRATION STATEMENT SEA LIMITED. Washington
D.C., United States: Securities and Exchange Commission Retrieved from
https://www.sec.gov/Archives/edgar/data/1703399/000119312517291352/d36
3501df1.htm
Book
Ahmad, S., Zulkurnain, N., & Khairushalimi, F. (2016). Assessing the Validity and
Reliability of a Measurement Model in Structural Equation Modeling (SEM)
(Vol. 15).
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale, N.J.: L. Erlbaum Associates.
Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2014). Multivariate data
analysis (Seventh edition, Pearson new international edition ed.): Harlow:
Pearson Education Limited.
Hooper, D., Coughlan, J., & R. Mullen, M. (2007). Structural Equation Modeling:
Guidelines for Determining Model Fit (Vol. 6).
Ref. code: 25615802037480DCY
Page 84
APPENDICES
Ref. code: 25615802037480DCY
Page 85
72
APPENDIX A
QUESTIONNAIRE
แบบสอบถามฉบบนเปนสวนหนงของการศกษาหลกสตรปรญญาโท สาขาวชาการบรหาร
ระบบสารสนเทศ (MIS) คณะพาณชยศาสตรและการบญช มหาวทยาลยธรรมศาสตร โดยม
วตถประสงค เพอเกบขอมลผบรโภครวมถงความคดเหนตาง ๆ เพอนำมาเปนขอมลในการศกษา และ
วจยในหวขอ “ความพงพอใจและความตงใจซอซำผานชองทางพาณชยอเลกทรอนกสขามชาตผานปจจยหลกดานความเชอมนและการรบรความเสยง”
ผวจยจงใครขอความอนเคราะหตอบแบบสอบถามนตามความเปนจรงและครบถวน เพอทำ
ใหผลการวจยนสมบรณตามความมงหมายของงานวจย โดยในสวนของขอมลสวนบคคลทไดรบจาก
การทำแบบสอบถาม ผวจยจะเกบรกษาเปนความลบอยางเครงครดและไมเปดเผยตอสาธารณชน ไมวากรณใด ๆ ทงสน และการวเคราะหและนำเสนอขอมลจะทำในภาพรวม ไมระบถงตวตนของ
ผตอบแบบสอบถามเทานน
แบบสอบถาม ประกอบไปดวย 3 สวน ดงน
สวนท 1: นยามและสาระสำคญ
สวนท 2: แบบสอบถามเกยวกบปจจยดานความเชอมน ความเสยง ความพงพอใจและความตงใจซอ
ซำผานชองทางพาณชยอเลกทรอนกสขามชาต
สวนท 3: ขอมลทวไปของผตอบแบบสอบถามและขอมลการซอสนคา
Ref. code: 25615802037480DCY
Page 86
73
สวนท 1 คำนยามและสาระสำคญ
พาณชยอเลกทรอนกสขามชาต (Cross-border E-commerce) ตามแบบสอบถามตอไปน
หมายถงการซอขายสนคาผานชองทางออนไลนทมลกษณะการจดตงเปนรานคา (online store) หรอ
มลกษณะเปนพนทกลางในการใหบรการซอขายสนคาจากผผลตถงผบรโภค (marketplace) โดยการคาผานทางชองทางดงกลาวจะเกดขนผานหนาเวบไซตและ/หรอแอปพลเคชนทใหบรการบน
อปกรณสอสาร การซอสนคาผานชองทางดงกลาวสามารถสงคำสงซอผานทางหนาเวบไซตหรอ
แอปพลเคชนบนอปกรณเคล อนท (โทรศพทมอถอ แทบเลต ฯ) ผานการเช อมตออนเทอรเนตใน
ขณะใชงาน โดยการซอขายดงกลาวผใหบรการมการจดตงบรษทเพอดำเนนกจการดานพาณชยอเลกทรอนกสในประเทศตนเองและ/หรอในตางประเทศ และการซอขายสนคาตองมลกษณะการ
ขนสงสนคาขามพรมแดนระหวางประเทศ ในแบบสอบถามนจะใชคำวา "รานคาออนไลนขาม
พรมแดน" เพอใหงายตอความเขาใจ
การซ อสนคาผานรานคาพาณชยอเลกทรอนกสขามชาต (Cross-border e-commerce
purchase) หมายถงการสงซอสนคาผานรานคาพาณชยอเลกทรอนกสซงมการขนสงสนคาขามเขต
แดนประเทศระหวางตนทางผขายสนคาในประเทศหนงไปสปลายทางผรบในอกประเทศหนง ทงน
ผ ใหบรการรานคาออนไลนอาจจดต งบรษทอย ในประเทศไทยหรอไมอย ในประเทศไทยกได ตวอยางเชน การสงสนคาผานเวบไซต Amazon.com ซงผจำหนายสนคาอยในประเทศสหรฐอเมรกา
และผรบอยในประเทศไทย มการสงคำสงซอจากประเทศไทยผานหนาเวบไซตหรอแอปพลเคชนและ
ทำธรกรรมออนไลน จากนนมการขนสงสนคาจากรานคาใน Amazon ขามพรมแดนสผรบในประเทศไทย
หรอการสงซอสนคาผาน Lazada.com โดยผขายมสนคาเกบไวทประเทศจน มการสงสนคานำเขามาสประเทศไทย เปนตน
Ref. code: 25615802037480DCY
Page 87
74
ทานเคยซอสนคาผานรานคาออนไลนทตองสงสนคาขามพรมแดนหรอไม (มการสงสนคานำเขามาใน
ประเทศไทย เปนสนคาจบตองได)
� เคย � ไมเคย (จบแบบสอบถาม)
สวนท 2 แบบสอบถามเกยวกบปจจยดานความเชอมน ความเสยง ความพงพอใจและความตงใจซอซำ
ผานชองทางพาณชยอเลกทรอนกสขามชาต
Ref. code: 25615802037480DCY
Page 88
75
ปจจยดานความคาดหวงกอนซอสนคา
จากประสบการณการซอสนคาออนไลนขามพรมแดน ใหทานนกถงความรสกกอนการซอสนคาวาทาน
มความคาดหวงตอผขายจากรานคาขามพรมแดนเมอเทยบกบรานคาภายในประเทศในคำถามตอไปน
อยางใดโดยมระดบความเหนดงตอไปน 1 ไมเหนดวยอยางยง
2 ไมเหนดวย
3 เฉย ๆ
4 เหนดวย 5 เหนดวยอยางยง
คำถาม รานคาขาม
พรมแดน รานคาในประเทศ
ระดบความคดเหน 1 2 3 4 5 1 2 3 4 5
มความนาเชอถอ
จะใหความชวยเหลอเตมทเมอมขอสงสยหรอปญหา
จะทำตามขอตกลงในการใหบรการทใหไวตอกน
จะใหขอมลสนคาและบรการครบถวน
จะมปญหาเกยวกบตวสนคา (เชน คณภาพ ตำหนสนคา ใชงานไมได)
จะมความเสยงทางการเงนจากการซอสนคา (เชน คาใชจาย
ในการคนสนคา ถกโกง ภาษ)
จะมปญหาและความไมแนนอนจากการจดสงสนคา
โดยรวมแลวจะมความเสยงจากการซอสนคา
Ref. code: 25615802037480DCY
Page 89
76
ปจจยดานความเชอมนและความเสยงหลงซอสนคา
จากประสบการณการซอสนคาออนไลนขามพรมแดน ใหทานนกถงความรสกหลงการซอสนคาวาทาน
มความคาดหวงตอผขายจากรานคาขามพรมแดนเมอเทยบกบรานคาภายในประเทศในคำถามตอไปน
อยางใดโดยมระดบความเหนดงตอไปน 1 ไมเหนดวยอยางยง
2 ไมเหนดวย
3 เฉย ๆ
4 เหนดวย 5 เหนดวยอยางยง
คำถาม รานคาขาม
พรมแดน รานคาในประเทศ
ระดบความคดเหน 1 2 3 4 5 1 2 3 4 5
มความนาเชอถอ
ใหความชวยเหลอเตมทเมอมขอสงสยหรอปญหา
ทำตามขอตกลงในการใหบรการทใหไวตอกน (ขอตกลงการใชงาน)
ใหขอมลสนคาและบรการครบถวน
รสกมปญหาเกยวกบตวสนคา (เชน คณภาพ ตำหนสนคา ใชงานไมได)
รสกมความเสยงทางการเงนจากการซอสนคา (เชน คาใชจายในการคนสนคา ถกโกง ภาษ)
รสกมปญหาและความไมแนนอนจากการจดสงสนคา (เชน
รอนานเกนไป สงลาชา วนจดสงไมแนนอน)
โดยรวมแลวมความเสยงจากการซอสนคา
Ref. code: 25615802037480DCY
Page 90
77
ปจจยดานความการยนยนความคาดหวง ความพงพอใจ และความตงใจซอซำ
จากประสบการณการซอสนคาออนไลนขามพรมแดน ใหทานนกถงความรสกหลงการซอสนคาวาทาน
มความคาดหวงตอผขายในคำถามตอไปนอยางใดโดยมระดบความเหนดงตอไปน
1 ไมเหนดวยอยางยง 2 ไมเหนดวย
3 เฉย ๆ
4 เหนดวย
5 เหนดวยอยางยง
คำถาม รานคาขาม
พรมแดน รานคาในประเทศ
ระดบความคดเหน 1 2 3 4 5 1 2 3 4 5
โดยรวมแลวเปนไปตามทคาดหวงไว
โดยรวมแลวดกวาทคดไว
โดยรวมแลวไมมความเสยงอยางทคดไว
โดยรวมแลวนาเชอถออยางทคดไว
รสกพงพอใจตอประสบการณการซอสนคา
ไดรบประสบการณทดจากการซอสนคา
โดยรวมรสกประทบใจกบการซอสนคา
โดยรวมทานรสกอยางไรกบประสบการซอสนคา
จะซอสนคาออนไลนขามพรมแดนอกครงในอนาคต
วางแผนวาจะซอสนคาขามพรมแดนอกครงในอนาคต
ตงใจจะซอสนคาออนไลนขามพรมแดนในอนาคต
จะซอสนคาขามพรมแดนตอไป
Ref. code: 25615802037480DCY
Page 91
78
สวนท 3: ขอมลทวไปของผตอบแบบสอบถามและขอมลการซอสนคา เพศ อายปจจบน _________
� ชาย � หญง
ระดบการศกษา
� ปรญญาตร � ปรญญาโท
� ปรญญาเอก � มธยมศกษา / ประกาศนยบตร
� อน ๆ
รายไดตอเดอน
� นอยกวา 25,000 � 25,001 – 50,000
� 50,001 – 75,000 � 75,001 – 100,000
� มากกวา 100,000
อาชพ
� พนกงานประจำบรษทเอกชน � นกเรยน / นกศกษา
� ขาราชการ / พนกงานรฐวสาหกจ � ธรกจสวนตว / อาชพอสระ
� อน ๆ โปรดระบ ____________________________________
ความถในการซอสนคาตอเดอน _________________
ประเภทสนคาทซอ
� เสอผา รองเทา เครองแตงกาย
� เครองสำอาง อปกรณเสรมความงาม อาหารเสรม
� คอมพวเตอร อปกรณอเลกทรอนกส
� เครองใชไฟฟา
� หนงสอ เครองใชสำนกงาน
� อน ๆ โปรดระบ ____________________________________
Ref. code: 25615802037480DCY
Page 92
79
APPENDIX B
CONFIRMATORY FACTOR ANALYSIS DETAIL
B.1 Measurement Model and Structural Model
Figure B.1 Structural Model and CFA
B.2 Confirmatory Factor Analysis
Table B.1
Confirmatory Factor Analysis Result.
Observed
Variable
Latent
Construct
Std.
Estimate Estimate S.E. C.R. P-Value SMC
CBPRE CBTRE 0.117 0.135 0.072 1.880 0.060 CBTRP CBTRE 0.589 0.490 0.060 8.174 *** CBPRP CBTRP -0.300 -0.404 0.071 -5.715 *** CBPRP CBPRE 0.625 0.604 0.073 8.285 *** DCE CBPRE 0.049 0.050 0.068 0.733 0.463
Ref. code: 25615802037480DCY
Page 93
80
Table B.1
Confirmatory Factor Analysis Result. (Cont’d)
Observed
Variable
Latent
Construct
Std.
Estimate Estimate S.E. C.R. P-Value SMC
DCE CBTRE -0.085 -0.100 0.076 -1.319 0.187 DCE CBTRP 0.585 0.827 0.107 7.700 *** DCE CBPRP -0.436 -0.457 0.073 -6.247 *** SAT DCE 0.802 0.778 0.061 12.806 *** CON SAT 0.510 0.496 0.076 6.503 ***
CBPRE4 CBPRE 0.674 1.000 0.447
CBPRE3 CBPRE 0.701 1.088 0.082 13.282 *** 0.394
CBPRE2 CBPRE 0.781 1.481 0.139 10.656 *** 0.425
CBPRE1 CBPRE 0.765 1.369 0.130 10.516 *** 0.326
CBTRP1 CBTRP 0.747 1.000 0.558
CBTRP2 CBTRP 0.678 1.084 0.085 12.780 *** 0.459
CBTRP3 CBTRP 0.635 1.182 0.098 12.030 *** 0.404
CBTRP4 CBTRP 0.702 1.135 0.086 13.198 *** 0.493
CBPRP1 CBPRP 0.654 1.000 0.414
CBPRP2 CBPRP 0.593 0.965 0.067 14.417 *** 0.356
CBPRP3 CBPRP 0.919 1.489 0.120 12.450 *** 0.847
CBPRP4 CBPRP 0.688 0.950 0.094 10.152 *** 0.445
CON1 CON 0.715 1.000 0.474
CON2 CON 0.712 1.033 0.082 12.582 *** 0.505
CON3 CON 0.674 0.930 0.077 12.049 *** 0.439
CON4 CON 0.748 1.038 0.080 13.001 *** 0.524
DCE4 DCE 0.773 1.000 0.598
DCE3 DCE 0.768 1.020 0.063 16.218 *** 0.589
DCE2 DCE 0.780 0.968 0.059 16.537 *** 0.609
DCE1 DCE 0.820 0.927 0.052 17.659 *** 0.672
Ref. code: 25615802037480DCY
Page 94
81
Table B.1
Confirmatory Factor Analysis Result. (Cont’d)
SAT4 SAT 0.751 1.000 0.564
SAT3 SAT 0.698 0.998 0.076 13.090 *** 0.487
SAT2 SAT 0.618 0.901 0.081 11.182 *** 0.382
SAT1 SAT 0.730 1.070 0.079 13.596 *** 0.533
CBTRE4 CBTRE 0.669 1.000 0.447
CBTRE3 CBTRE 0.628 0.957 0.097 9.894 *** 0.488
CBTRE2 CBTRE 0.652 0.918 0.091 10.125 *** 0.617
CBTRE1 CBTRE 0.571 0.616 0.067 9.264 *** 0.586
Ref. code: 25615802037480DCY
Page 95
82
APPENDIX C
Residual Variance
Table C.1
Residuals Variance Summary. Estimate S.E. C.R. P
z1 0.196 0.029 6.74 ***
z2 0.26 0.039 6.75 ***
z4 0.089 0.012 7.273 ***
z5 0.134 0.02 6.879 ***
z3 0.113 0.015 7.447 ***
z6 0.091 0.015 6.04 ***
z7 0.23 0.031 7.34 ***
e8 0.317 0.03 10.563 ***
e7 0.324 0.032 10.006 ***
e6 0.371 0.05 7.478 ***
e5 0.351 0.044 7.918 ***
e13 0.108 0.01 10.816 ***
e14 0.188 0.015 12.139 ***
e15 0.28 0.022 12.693 ***
e16 0.18 0.015 11.739 ***
e17 0.33 0.026 12.569 ***
e18 0.424 0.032 13.384 ***
e19 0.1 0.033 3.024 0.002
e20 0.247 0.026 9.498 ***
e25 0.231 0.021 11.096 ***
e26 0.251 0.023 11.159 ***
e27 0.252 0.021 11.878 ***
e28 0.206 0.02 10.3 ***
Ref. code: 25615802037480DCY
Page 96
83
Table C.1
Residuals Variance Summary. (Cont’d) Estimate S.E. C.R. P
e12 0.183 0.015 12.064 ***
e11 0.197 0.017 11.756 ***
e10 0.163 0.014 11.547 ***
e9 0.114 0.01 10.973 ***
e24 0.198 0.019 10.606 ***
e23 0.269 0.023 11.685 ***
e22 0.337 0.028 12.181 ***
e21 0.257 0.023 11.026 ***
e4 0.242 0.022 10.778 ***
e3 0.276 0.024 11.579 ***
e2 0.223 0.02 11.122 ***
e1 0.154 0.012 12.424 ***
Ref. code: 25615802037480DCY
Page 97
APPENDIX D MEASUREMENT MODEL CORRELATION
Table D.1 Measurement Model Correlation Matrix.
Variable 1 2 3 4 5 6 7 8 9 10 11 12 CBTRE1 1.000 CBTRE2 0.413 1.000 CBTRE3 0.349 0.408 1.000 CBTRE4 0.395 0.411 0.421 1.000 SAT1 0.100 0.153 0.102 0.003 1.000 SAT2 0.079 0.157 0.058 0.014 0.550 1.000 SAT3 0.050 0.115 0.107 0.067 0.498 0.559 1.000 SAT4 0.148 0.163 0.129 0.118 0.538 0.443 0.539 1.000 DCE1 0.223 0.239 0.221 0.179 0.492 0.421 0.431 0.498 1.000 DCE2 0.168 0.200 0.120 0.136 0.406 0.385 0.445 0.444 0.629 1.000 DCE3 0.163 0.207 0.195 0.129 0.477 0.345 0.402 0.461 0.631 0.673 1.000 DCE4 0.202 0.208 0.152 0.138 0.466 0.436 0.382 0.438 0.615 0.617 0.586 1.000 CON4 0.128 0.121 0.075 0.031 0.196 0.133 0.150 0.207 0.235 0.255 0.210 0.211
84
Ref. code: 25615802037480DCY
Page 98
Table D.1 Measurement Model Correlation Matrix (Cont.) Variable 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 CON3 0.505 1.000 CON2 0.494 0.529 1.000 CON1 0.571 0.442 0.495 1.000 CBPRP4 -0.061 -0.076 -0.098 -0.180 1.000 CBPRP3 -0.115 -0.145 -0.172 -0.141 0.458 1.000 CBPRP2 -0.090 -0.090 -0.075 -0.143 0.398 0.537 1.000 CBPRP1 -0.028 -0.117 -0.093 -0.072 0.463 0.588 0.618 1.000 CBTRP4 0.139 0.080 0.168 0.133 -0.086 -0.104 -0.050 0.019 1.000 CBTRP3 0.079 0.012 0.121 0.041 -0.011 -0.003 0.064 0.117 0.481 1.000 CBTRP2 0.182 0.123 0.159 0.164 -0.090 -0.155 -0.149 -0.059 0.480 0.417 1.000 CBTRP1 0.179 0.070 0.128 0.165 -0.080 -0.198 -0.127 -0.082 0.519 0.456 0.521 1.000 CBPRE1 -0.062 -0.070 -0.039 -0.084 0.320 0.395 0.308 0.371 0.201 0.262 0.126 -0.006 1.000 CBPRE2 -0.057 -0.092 -0.059 -0.099 0.262 0.406 0.286 0.337 0.185 0.309 0.123 0.003 0.725 1.000 CBPRE3 -0.046 -0.095 -0.087 -0.131 0.332 0.310 0.279 0.345 0.089 0.168 0.049 0.025 0.531 0.575 1.000 CBPRE4 -0.055 -0.094 -0.066 -0.132 0.376 0.391 0.298 0.380 0.052 0.089 0.060 0.002 0.506 0.502 0.564 1.000
85
Ref. code: 25615802037480DCY
Page 99
86
BIOGRAPHY
Name Mr. Natthakorn Khayaiyam Date of Birth June 28, 1989 Educational Attainment April 2012: Bachelor of Arts in English Work Position April 2017 - Present:
Data Scientist and Engineer Krungthai-AXA Life Insurance
Work Experiences October 2016 - March 2017: Business Support Analyst Deutsche Bank A.G. Bangkok January 2013 - May 2016: Data Specialist and Client Reporting Market Pulse International, Thailand.
Ref. code: 25615802037480DCY