Behavioral intention to use mobile stock trading , evidence from vietnams securities investors

This implies that to facilitate the behavioral intention to use mobile stock trading, securities firms need to consider securities investors’ technological perceptions, risk perceptionso

International School of Business -

NGUYEN PHUC BINH

BEHAVIORAL INTENTION TO USE

MOBILE STOCK TRADING:

EVIDENCE FROM VIETNAM’S SECURITIES INVESTORS

MASTER OF BUSINESS (Honours)

Ho Chi Minh City – 2015

International School of Business -

Nguyen Phuc Binh

BEHAVIORAL INTENTION TO USE

MOBILE STOCK TRADING:

EVIDENCE FROM VIETNAM’S SECURITIES INVESTORS

ID: 22130010

MASTER OF BUSINESS (Honours)

SUPERVISOR: Dr TRAN PHUONG THAO

Ho Chi Minh City – 2015

ACKNOWLEDGMENTS Firstly, I would like to express my deepest appreciation to my supervisor Dr Tran Phuong Thao for her professional guidance, valuable advices, continuous encouragement, and support that made this thesis possible.

I would like to extend deep senses of gratitude to Prof Nguyen Dinh Tho, Dr Tran

Ha Minh Quan, and lecturers who have taught and transferred me valuable knowledge and experiences during my time at the International School of Business, special thanks to all of my dear friends in Mbus4 class who gave me useful materials, responses and experiences to conduct this study.

I would also like to express my grateful thanks to my managers, my friends, and

my colleagues who participated in filling the questionnaires and/or helped send the questionnaires to their peers; to securities investors, and provided valuable information and comments for this study.

Personally, I wish to express my deep gratitude to my colleagues and friends working at Vietcombank Fund Management (VCBF); Saigon Securities Inc (SSI); VNDirect Securities Corporation (VND); Hochiminh City Securities Corporation (HSC); Maybank Kim Eng Securities Limited; Mirae Asset Wealth Management Securities (VN) JSC, and Vietcapital Securities Corporation (VCSC).

Ho Chi Minh City, Vietnam,

January 27, 2016

Nguyen Phuc Binh

ABSTRACT The purpose of this study is to investigate the determinants of securities investors’ behavioral intention towards using mobile stock trading Based on a modified UTAUT (Unified Theory of Acceptance and Use of Technology) with multi-facet perceived risks and privacy concerns, a comprehensive research model was proposed An empirical survey with a valid sample of 244 securities investors was conducted in Vietnam to test the research model The analysis results of SEM indicated that three enablers of adopting mobile stock trading in Vietnam are UTAUT constructs (i.e., performance expectancy, effort expectancy and social influence), and the inhibitors are the perceptions of risk (i.e security risk, economic risk) and privacy concerns This implies that to facilitate the behavioral intention to use mobile stock trading, securities firms need to consider securities investors’ technological perceptions, risk perceptions

of this type of trading, and concerns on the disclosure of personal information upon logging in the application The findings of this study not only have important implications for mobile commerce research, but also provide insights for securities firms and developers of mobile stock trading systems.

Keywords: mobile stock trading (m-trading), UTAUT, security risk, economic risk, privacy concerns, behavioral intention to use.

TABLE OF CONTENTS

ACKNOWLEDGMETS i

ABSTRACT ii

TABLE OF CONTENT iii

LIST OF FIGURES v

LIST OF TABLES vi

LIST OF ABBREVIATIONS vii

CHAPTER 1: INTRODUCTION 1

1.1 Background of the study 1

1.2 Research gap 2

1.3 Research objectives and research questions 4

1.4 Research methodology and research scope 4

1.5 Research structure 5

CHAPTER 2: LITERATURE REVIEW& HYPOTHESES DEVELOPMENT 6

2.1 Theoretical background 6

2.1.1 Unified Theory of Acceptance and Use of Technology (UTAUT) 6

2.1.2 The extended UTAUT. 7

2.2 Behavioral intention, risk perceptions and privacy concern 8

2.2.1 Behavioral Intention. 8

2.2.2 Risk Perception. 9

2.2.3 Privacy Concern. 9

2.3 Hypotheses Development 10

2.3.1 Hypotheses Derived From UTAUT 11

2.3.2 Hypotheses Derived From Risk Perceptions 12

2.3.3 Hypotheses Derived From Privacy Concern 13

2.4 Conceptual model 15

2.5 Chapter summary 15

CHAPTER 3: METHODOLOGY 16

3.1 Research design 16

3.1.1 Research process 16

3.1.2 Measurement scales. 17

3.2 Measurement refinement.… 20

3.3 Sample 21

3.4 Data analysis and interpretation 22

3.4.1 Reliability measure 22

3.4.2 Validity measure by EFA (Exploratory Factor Analysis). 22

3.4.3 CFA & SEM 23

3.5 Pilot test 23

3.5.1 Cronbach’s Alpha. 23

3.5.2 Exploratory factor analysis. 24

3.6 Chapter summary 26

CHAPTER 4: RESEARCH FINDINGS 27

4.1 Data description 27

4.2 Confirmatory factor analysis (CFA) 28

4.2.1 Saturated model. 28

4.2.2 Composite reliability and variance extracted. 29

4.3 The structural equation model analysis (SEM) 29

4.4 Discussion 32

4.4.1 UTAUT constructs. 32

4.4.2 Perceived risks. 33

4.4.3 Privacy concerns. 34

4.5 Chapter summary 35

CHAPTER 5: CONCLUSION, IMPLICATION& LIMITATIONS 36

5.1 Key findings 36

5.2 Managerial Implications 38

5.2.1 UTAUT constructs. 38

5.2.2 Perceived risks. 40

5.2.3 Privacy concerns. 40

5.3 Research contribution 41

5.4 Limitation and further study 42

REFERENCE… 44

Appendix A: List of in-depth interview participants 52

Appendix B: In-depth interview’s refinement measurement scale 53

Appendix C: Questionnaire (English Version) 58

Appendix D: Questionnaire (Vietnamese Version) 61

Appendix E: Discriptive Statistics 64

Appendix F: CFA results 65

Appendix G: In-depth interview for “security risk” and “privacy concerns” 66

LIST OF FIGURES

2.2 Basically generalized UTAUT model of extant researches 8 2.3 Basically generalized extended UTAUT model of extant researches 8

LIST OF TABLES

4.2 The results for reliability and variance extracted test 29 4.2 Summary of results of testing hypotheses by SEM 32

11 UTAUT Unified theory of acceptance and use of technology

CHAPTER 1: INTRODUCTION 1.1 Research Background

“Information technology is not the cause of the changes we are living through But without new information and communication technologies none of what is changing our lives would be possible” (Castells, 1999, p.2) In the most recent decades, we have witnessed mobile technology, an important and most innovated component information communication technology (ICT), has achieved a virtual revolution It has transformed a pure means of communication through a full set of mobile services in finance, shopping, payment, social networking, remote diagnostics, tourism, and etc., which added billions of dollars to the world economy Therefore, bringing new technologies to improve life and business performance seems not new at all (Chong et al., 2010) Actually, mobile technology has helped many financial institutions in various economic sectors with many opportunities to reach new and potential markets, and customers The development of mobile applications has provided financial service providers with new channels of reaching their customers and in turn its customers have more chance to access their financial information than ever before (TFSA, 2013) One of the outstanding advantages of mobile technology is its ability to give users instant, new, and useful information at appropriate time regardless of locations (Liang

& Yeh, 2009) Without being so, mobile commerce will lose its significant economic value As

mobile-commerce uninterruptedly grows, brick-and-mortar or online financial services will be demanded to be available in mobile devices by the consumers (TFSA, 2013).

Typically, securities trading, a component of financial service has also been sharply changed since first taking shape in the world due to the widespread adoption of ICT Since introduced in 1995, online trading which is a newly innovated mode of securities trading has increased dramatically (Li, Lee & Cude, 2002) Upon implementation of online trading, no longer do securities investors need to get out of their houses to the trading floors located at busy streets in big cities thanks to the Internet-based trading systems that covered every corner of the world Online securities trading opened a new chapter in the establishment and development of global securities market, and it is expected to continue being a valued choice for investors (American Banker, as cited in Li et al., 2002) Online trading (i.e website-based trading and mobile-based trading) as a substitute to the traditional trading (i.e direct trading

at floors, phone-based trading) has unique characteristics which help the investors handle their portfolio without or less being contingent on brokers as well as preserve the personal privacy Online trading helps brokerage firms reduce costs by eliminating human interaction

(Bakos et al., 2000) With online trading, buying or selling securities is only a single click on a computer mouse (Li et al., 2002).

Although the Vietnamese financial market in general; and specifically the securities market have not been significantly developed, it was categorized as an attractive emerging market Obviously, more and more foreign prominent financial institutions such as Citibank, HSBC, ANZ bank, Commonwealth bank, Bank of Tokyo, Franklin Templeton, Blackhorse, AIG, Prudential, Manulife, … have made entry in Vietnam market and have found success This proves the true potential of the Vietnam’s financial market In the integration with global playground, the Vietnamese financial institutions have been also adopting and developing online services for years Currently there is a very limited number of financial institutions in Vietnam (only banks develop mobile banking and securities brokerage firms develop mobile securities trading) conducting their services on the basis of mobile technology There are some reasons that people believe in the wide and rapid adoption of mobile commerce in Vietnam According to MOIT (2014), Vietnam is in the group of countries having high rate of mobile purchase, and the potential for mobile shopping has been realized and has great opportunities for growth Ngo et al (2015) indicated that researches into mobile shopping in Vietnam too few to profoundly understand the determinants of usage adoption.

Despite the benefits of mobile technology in contemporary finance, there are potential risks in association with financial transactions via mobile devices, such as hacking, misusage, and privacy concerns Those disadvantages of mobile devices that are vulnerable to lose consumers’ money and prestige are why the usage of e-finance (i.e e-banking and electronic securities trading) has not widely adopted yet That is why behavioral intention to use mobile technology, the antecedent of actual usage, in financial service is widely studied over the world recently in tandem with the striking development of smart-phones and its applications.

Corp., FPT Securities Corp., MB Securities Corp., VPbank Securities Corp., BaoViet Securities Corp., etc), even developed smart-phone-based application for its clients to conduct securities trading everywhere However, there are some limitations while carrying out mobile-based transactions due to its nature (Tai & Ku, 2013) as well as careful consideration of customers with regard to the risk assessment (Zhou, 2012), thus only few Vietnamese securities investors actually use M-Trading In other words, the questions on rationales of securities investors willing or reluctant to use M- Trading and factors influencing their behavioral intention to use M-Trading in Vietnam have received increasing concerns of researchers.

In the literature, many prior studies on ICT’s adoption in financial industry has been conducted up to now such as internet banking (Wang & Shan, 2012; Chong et al., 2010; Kim et al., 2007; Laforet & Li, 2005), mobile banking (Yu, 2012; Aboelmaged & Gebba, 2013), internet securities trading (Teo et al 2004, Ramayah et al 2009, Singh et al 2010), online securities trading (Abroud, Choong & Muthaiya, 2010); M-Trading (Tai &

Ku, 2013) Remarkably, in Vietnam’s context, there have been also many researches online banking (Chong et al., 2010); e-banking (Nguyen Duy Thanh & Cao Hao Thi, 2011; Nguyen et al., 2014); e-payment (Nguyen & Lin, 2011), mobile transfer of money (Le Van Huy & Tran Nguyen Phuong Minh, 2011); mobile-learning (Ngo & Gwangyong, 2014), personal internet banking (Hoang, 2015), mobile payment (Pham & Liu, 2015), mobile shopping (Ngo et al., 2015), but only few on e-trading of securities were conducted.

Because Vietnamese investors' behavioral intention to use M-Trading have not been well indicated, which factors deter or encourage their adoption remain unknown, a better understanding of their behavioral intention would have great practical implications, not only for brokerage firms seeking to manage more effectively the implementation of M-Trading and improve their services, but also for the authorities of the Vietnam’s State Securities Commission that would have proper policies on administering securities market Obviously, understanding why securities investors are willing or reluctant to use M-Trading by developing and empirically examining a comprehensive model of securities investors’ behavioral intention to use M-Trading is strongly needed.

In Vietnam, M-Trading as other types of mobile commerce is defined as conducting transactions using mobile devices such as smart phones, tablets, PDAs (Personal Digital Assistants), and other mobile devices (except for laptops) As same as Internet shopping, it requires Internet access In this study, the theoretical background of the research is developed with the concept of literature review of the unified theory of acceptance and use of technology (UTAUT) that combines key constructs from Technology Acceptance Model

(TAM) Recently, the UTAUT model or extended UTAUT model have successfully employed to explain behavioral intention to usemobile securities trading (Tai & Ku, 2013), users’ intention and behavior of mobile banking (Yu, 2012), and intention to use internet banking (Yee et al., 2015) Therefore, to fulfill the gap in the context of Vietnam, this study is about to employ extended UTAUT model (Venkatesh et al., 2003) liaised with multi-facet perceived risks (Tai &

Ku, 2013) and privacy concerns (Zhou, 2012) to investigate the influential level of these factors

on Vietnamese’s behavioral intention to use M-Trading.

1.3 Research objectives and research questions

The objective of this study is to investigate the factors influencing securities investors’ behavioral intention Particularly, the study aims at answering the following questions:

Question 1: which factors based on the modified UTAUT model influence securities investors’ behavioral intention to use M-Trading in Vietnam?

Question 2: Which factors of perceived risks influence securities investors’ behavioral intention to use M-Trading in Vietnam?

Question 3: Whether do the privacy concerns affect securities investors’ behavioral intention to use M-Trading in Vietnam or not?

1.4 Research methodology and research scope

This study uses questionnaires to collect data The survey is originally developed in English and then translated into Vietnamese In-depth interviews are then conducted with eight people in order to modify the Vietnamese version of the questionnaire before the survey

is implemented in mass The next step is analyzing the collecting data The data of this research is processed using SPSS software with three main stages First, Cronbach’s Alpha is used to test the reliability of the measurement scale Then, the validityof the measurement scale will be checked by Exploratory Factor Analysis (EFA) Finally, structural equation model (SEM) and path analysis are employed as the main method for investigating the relationships among factors in the research model Ho Chi Minh City is the largest city and the metropolitan area in Vietnam It is also the economic center of Vietnam and accounts for a large proportion

of the economy of Vietnam Moreover, Ho Chi Minh City has been chosen to conduct the survey for this study since it is one of the biggest cities in Vietnam in which the largest securities exchange of Vietnam is located (Ho Chi Minh City Securities Exchange

– HOSE) The research’s subjects are Vietnamese individual securities investors with the age range above 18 years old They might either be used to use internet securities trading, and or internet/mobile banking service, or never use any e-financial services because the study’s goal

is to find out the behavioral intention to use M-Trading instead of actual use Moreover,

since investigating the behavioral intention, potential investors at the same age level will be also invited to participate.

1.5 Research Structure

The research is divided into five chapters.

The first chapter introduces about background, research problems, research questions, research purpose, scope of research and research structures.

The second chapter covers literature review of the previous research and shows hypotheses, as well as the conceptual model of the research.

The third chapter presents the research process, sampling size, measurement scale, main survey, and data analysis method.

The fourth chapter concentrates on preparation data, descriptive data,

assessment measurement scale and hypotheses testing.

The fifth chapter points out research overview, research findings, managerial implications, research limitations and directions for future research.

CHAPTER 2: LITERATURE REVIEW & HYPOTHESES DEVELOPMENT

The chapter 2 is to present the theories associated with behavioral intention to use mobile technology, the acceptance of mobile-based financial services and the models testing the adoption of mobile technology over the world Moreover, a conceptual model is built resulting from the hypotheses generating from extant literature, simultaneously, its constructs and relationship hypothesized among these constructs are also discussed.

2.1 Theoretical background

Studies on ICT users’ acceptance and use have been conducted extensively because ICT have been in wide usage, and several models originated from different theoretical disciplines (i.e psychology, sociology and information systems) have been developed to explain the acceptance and usage One stream developed on motivational model has focused on how extrinsic and intrinsic motivations influence the acceptance (Davis et al., 1992) Another stream based on TAM model to explore the role of perceived usefulness and perceived ease of use on usage intentions and actual usage (Davis, 1989) Venkatesh et al (2003) indicated that the UTAUT model is able to explain sixty nine percent of intention to use ICT (technology acceptance) while other previous model explained approximately forty percent of technology acceptance Since UTAUT is a comprehensive model which has been adopted by several previous studies

to successfully predict users’ usage intention toward e-commerce, e-financial services,

an extended UTAUT (incorporating financial risk, economic risk, functional risk, and privacy concerns) serves as a robust basis for doing the same in Vietnam context.

2.1.1 Unified Theory of Acceptance and Use of Technology (UTAUT)

UTAUT is developed by Venkatesh et al (2003) to integrate eight theories, which include the technology acceptance model (TAM), innovation diffusion theory (IDT), the motivational model, the theory of reasoned action (TRA), the theory of planned behavior (TPB), a model combining the TAM and TPB, the model of PC utilization and social cognitive theory UTAUT proposes that four constructs including performance expectancy, effort expectancy, social influence, and facilitating conditions affect user adoption of an ICT The theory postulates that four core constructs – performance expectancy, effort expectancy, social influence, and facilitating conditions – are direct determinants of ICT behavioral intention and ultimately behavior (Venkatesh et al., 2003).

Performance expectancy is similar to perceived usefulness and relative advantage The construct of performance expectancy is aggregated from five performance-related constructs: perceived usefulness, extrinsic motivation, job-fit, relative advantage and outcome

expectations Effort expectancy is similar to perceived ease of use and complexity It is similar to constructs included in previous models or theories, namely, perceived ease of use, complexity, and ease of use Social influence is similar to subjective norm This construct proposes that people’s ICT acceptance behavior is affected by if they believe others expect them to be willing or reluctant to a certain behavior Facilitating conditions are similar to perceived behavioral control In UTAUT, facilitating conditions are integrated thirty two factors used in eight competing models into five constructs and empirically identified that behavioral intention and facilitating conditions were two direct determinants of adoption behavior This construct reflects that users need to be equipped with mobile internet knowledge in order to use ICT system Without owning these knowledge and resources, they cannot adopt ICT system.

2.1.2 The extended UTAUT

The UTAUT model without modification cannot be applied to the research on user’s acceptance of mobile commerce since all ICT adoption theories or models, including UTAUT, were developed for PC and/or fixed line Internet systems/applications Since 2003,

among many studies citing UTAUT, very few employ all if its constructs (Williams et al., 2011) Extant researches have used extended UTAUT to explain user adoption in online securitiesing in the financial market (Wang & Yang, 2005), internet banking (Yee et al., 2015; El-Qirem, 2013; Yu, 2012), in health information technology (Kijsanayotin et al., 2009), in digital library (Nov & Ye, 2009), and in e-government services (Suha & Anne, 2008) Further, the UTAUT has also been employed to examine user adoption of mobile

services, such as mobile banking (Yu, 2012), mobile wallet (Shin, 2009), mobile payment (Kim et al., 2009), and mobile technologies (Park et al., 2007) These studies mainly focused on employing the UTAUT or revising UTAUT by combining TAM’s construct, or

by adding up a few independent variables as the role of enabler of intention adoption.

Reactions to use ICT Intention to use ICT Actual use of ICT

Figure 2.2: Basically generalized model of extant researches

However, other than most previous studies employing UTAUT only adopted a single construct to evaluate users’ risk perceptions, Zhou (2012) examined location-based services usage from the perspectives of UTAUT and perceived risk, privacy concerns and trust Tai &

Ku (2013) investigated the determinants of securities investors’ intention towards using mobile securities trading by developing an extended UTAUT model incorporating symmetry axis of which usage intention influenced by UTAUT’s constructs and perceived risks.

Positive Effects to use Usage Intention of Negative effects to use mobile-based services mobile-based services mobile-based services

Demographics

Figure 2.3: Basically generalized extended UTAUT model of extant researches

Since (extended) UTAUT is a comprehensive model which has been adopted by several previous studies to successfully predict users’ usage intention toward mobile- based service, but very few studies were implemented as the same manner in Vietnam.

An extended UTAUT (incorporating enablers as performance expectancy, effort expectancy, and social influence; and inhibitors such as financial risk, economic risk, functional risk, and privacy concerns) will serve as a robust basis for an empirically testing the behavioral intention to use M-Trading in Vietnam.

2.2 Behavioral intention, risk perceptions and privacy concerns

2.2.1 Behavioral Intention

Ajzen (1991) argues that ‘‘Intentions are assumed to capture the motivational factors that influence a behavior They are indications of how hard people are willing to try, of how much of an effort they are planning to exert in order to perform the behavior’’ (p 181) It is similar to attitude towards behavior (TRA, TPB, DTPB) and extrinsic and intrinsic motivation (MM) derived from previous models or theories Usage intention to adopt/accept ICT system measures an individual’s relative strength of intention to perform a behavior (Fishbein & Ajzen, 1975) It indicates a person’s motivation to perform a specific behavior,

and is viewed as the antecedent of actual behavior Consistent to all models portraying from psychological theories, which argue that individual behavior is predictable and influenced by

individual intention, UTAUT argued and demonstrated usage intention to have significant

influence on ICT usage (Venkatesh et al., 2003; Venkatesh & Zhang, 2010).

2.2.2 Risk Perception

Concerning the acceptance of mobile-based mode in financial services, Tai & Ku (2013) indicated that risk perceptionsare important determinant of behavioral intention Perceived risks are usually considered as one of the potential barriers (Chen, 2008; Luo et al., 2010; Hsu et al., 2011) Noticeably, mobile users see risk in prospect uncertainty arising from data input errors, software failures, connection loss, and privacy loss (Mallat

et al., 2008; Cruz et al., 2010; Koenig-Lewis et al., 2010) M-Trading is a variant of based financial service, of which while using, the users are required to sign up with certain information Hence, there is an anticipated risk of exposure to opportunistic hackers who can access their trading accounts, delete data or make unauthorized trades As a result, investors may elect to forgo the potential benefits of using M-Trading.

mobile-Many previous researches have found that the intention to use mobile-based financial services is influenced by users’ perception of risk For instance, Mallat (2007) indicated that perceived risk is the main obstacle of the adoption mobile payment sytems Futher, Mallat et

al (2008) found that perceived risk is a key determinant of using mobile ticketing service Consistent to Mallat (2007) and Mallat et al (2008), Cruz et al (2010) and Koenig-Lewis et al (2010) indentified high perceived risk as a key inhibitor of mobile banking.

In the perspective of mobile-based financial services, users perceive risk from several facets, the most common of which are security risk, economic risk, and functional risk (Cruz et al., 2010; Koenig-Lewis et al., 2010; Wessels & Drennan, 2010) Perceived security risk

of mobile financial services lies in the perception of potential harm due to electronic fraud or hacker attacks Perceived economic risk arises from the perception of possible economic loss due to transaction error or faulty operation Perceived functional risk lies

in the perception of possible lack of service reliability or accessibility.

2.2.3 Privacy concerns

Information privacy refers to the claim of ICT’s users to determine for themselves when, how, and to what extent their information is communicated to others (Malhotra et al., 2004), and privacy information concernsexhibit ICT users’ concern on what extent their personal information to be disclosed (Li, 2011) Due to the differences in culture, regulatory laws, past experiences, and personal characteristics, ICT’susers exhibit dissimilar degrees of concerns on information privacy (Malhotra et al., 2004) The users with high levels of

privacy concerns believe that service providers generally tend to behave opportunistically with their personal information Therefore, in response to a request from securities firms for personal information, the securities traders will likely to refuse to provide personal information (Dinev & Hart, 2006) and/or to provide incorrect personal information (Teo

et al., 2004) From the UTAUT perspective, privacy concerns are viewed as usage inhibitors (Bansal et al., 2010) Besides, Malhotra et al (2004) indicated that privacy concerns of Internet users including collection, control, and awareness in addition to the general concern and specific concern proposed by Li (2011) Previous researches evidenced that privacy concerns significantly influences perceived risk (Zhou, 2012, Junglas et al., 2008, Bansal et al., 2010) Moreover, privacy concerns has significant effects

on user adoption of instant messaging (Lowry et al., 2011); web-based healthcare services (Bansal et al., 2010); electronic health records (Angst & Agarwal, 2009); software firewalls (Kumar et al., 2008); and ubiquitous commerce (Sheng et al., 2008).

2.3 Hypothesis development

Venkatesh et al.’s (2003) UTAUT is adopted as a primary theoretical framework to examine securities investors’ acceptance of M-Trading However, since the M-Trading context differs in some ways from the traditional ICT context, not any single constructs of UTAUT may fit the specific M-Trading’s context Hence, it is necessary to integrate the risk perceptions into the extended UTAUT model to propose our research model Because M-Trading in Vietnam is still in its infancy, as the matter of fact, there is very limited number of securities investors having actually used this mobile application Therefore, this study considers behavioral intention to use M-Trading as a dependent variable, and excludes from the proposed model two constructs pertaining to UTAUT, i.e use behavior and experience.

The developer of UTAUT model also stated that “facilitating conditions” construct becomes non-significant in predicting intention when both “performance expectancy” construct and “effort expectancy” construct exist in the research model Furthermore, facilitating conditions reflect that users have ability and resources necessary to use M-Trading (Venkatesh et al 2003) This means securities traders need to be equipped with mobile internet knowledge in order to use M-Trading, and need to pay communication fees and service fees associated with that usage (Zhou, 2012) However, in the real settings of Vietnam, only few securities investors have used M-Trading and brokerage firms are providing its clients with this application on the fee-free basis, so facilitating conditions, which is the antecedent of use behavior and has no significant association with behavioral intention, is excluded from the research model Besides, since this study is to investigate M-

Trading’s adoption in a voluntary usage context, the UTAUT’s moderating variable

“voluntariness” is also not included.

Extant researches on mobile-based financial services usage behavior have had findings which showed that user’s concerns about risk issues are key determinants for the adoption (Tai & Ku, 2013; Laukkanen & Kiviniemi, 2010; Luo et al., 2010) Perceived risk

is regarded as a person’s awareness of prospective uncertainty and adverse consequences

of engaging in a given activity (Forsythe et al., 2006; Littler & Melanthiou, 2006; Bland et al., 2007 ; Im et al., 2008) In spite of perceiving the benefit of a given service, people’s intention to adopt the service may be hesitant due to their perceived risks with respect to service’s usage M-Trading’s platform is an ICT artifact composed of mobile Internet, mobile devices and mobile systems, so upon implementing a securities transactions through mobile devices, some negative results including increased data entry errors; electronic data interception and unstable wireless connections may be occurred while not being found in other traditional formats Securities investors’ intention to use M-Trading may be impeded resulted from their perceptions of the risks.

Many previous studies employing UTAUT model found that privacy concerns strongly impacts on perceived risk in general While finding M-Trading’s adoption in Taiwan, Tai & Ku (2013) specified perceived risks as security risk; economic risk; and functional risk, and included in UTAUT model but failed to explore the relationship amongst privacy concerns and these three types of risks This research tries to fill the gap Taking the context in which M-Trading occurs into account, this study incorporates perceived risks and privacy concerns of which influences the perceptions

of risk, into UTAUT model to produce a more precise explanation of the antecedent of securities investors’ adoption or resistance of M-Trading.

2.3.1 Hypotheses Derived From UTAUT

In this study, behavioral intention is an endogenous variable In M-Trading’s context, this construct is conceptualized as the extents to which securities investors believe that M-Trading will improve their transaction performance Performance expectancy is defined as the extent to which an individual believes that usage of certain ICT system will help improve their performance (Venkatesh et al., 2003) Performance expectancy is the instrumental value of using M-Trading such as the improvement of trading efficiency, the increment of convenience in trading Such benefits will influence the behavioral intention

to use M-Trading Thus, the following hypothesis is proposed:

Hypothesis 1: Securities investors with high performance expectancy for Trading will have greater behavioral intention to use it.

M-Effort expectancy is defined as the extent to which individuals believe that learning to use a certain ICT system will not require significant effort (Venkatesh et al., 2003) Effort expectancy of using M-Trading is the users’ evaluation of how much effort is required to learn how to use and engage with the system Therefore, behavioral intention to use M-Trading is anticipated to increase if the investors believe that M-Trading is easy to handle Extant studies have broadly found that effort expectancy for using an ICT system is a significant antecedent

of behavior intention to use the ICT system (Venkatesh & Morris, 2000; Wang et al., 2009; Deng et al., 2011) Thus, the following hypothesis is proposed:

Hypothesis 2: Securities investors with high effort expectancy for M-Trading will have greater behavioral intention to use it.

Social influence is defined as the degree to which an individual perceives that its important peers expect his/her to employ a certain ICT system (Venkatesh et al., 2003) Social influence is conceptualized as the extent in that securities investors are encouraged to use M-Trading by their peers Since M-Trading is still too new to be popularly used in Vietnam, the users are only expected to be influenced by their peers’ perceptions of the quality and capabilities of M-Trading Moreover, usage intention indicates that users will follow their experience, preference and external environment

to collect information, evaluate alternatives, and make usage decision (Zeithaml, 1988; Dodds et al., 1991) Prior studies have found that social influence is an important predictor of usage intention to use a certain information system (Baron et al 2006; Wang et al 2009) It is expected that people’s behavioral intention to use a given ICT- based service is influenced by their peers’ opinion of that service (Karahanna et al 1999; Venkatesh & Davis, 2000) Thus, the following hypothesis is proposed:

Hypothesis 3: Securities investors who perceive a high degree of positive social influence (i.e., supportive of M-Trading) from their peers will have a greater behavioral intention to use M-Trading.

2.3.2 Hypotheses Derived From Risk Perceptions

Tai & Ku (2013) proved three-facet perceived risks (i.e security risk, economic risk, and functional risk) positively influence behavioral intention to use M-Trading, and Dai et al (2014) also indicated multi-dimensional perceptions of risk are among the most critical variables in the study of online shopping In this study, security risk, economic risk, and functional risk are investigated its effects on behavioral intention

to use M-Trading in Vietnam.

Security risk is securities traders’ perception of prospective harm caused by electronic fraud or hacker attacks while using M-Trading Security risk is found the main obstacle to the

adoption of mobile financial services and has been suggested being the greatest challenge to

the mobile financial service provider (Luarn & Lin, 2005; Misra & Wickamasinghe, 2004; Mallat et al., 2008) Miyazaki & Fernandez (2001) identified security risk (i.e potential fraud, misrepresentation) as a key concern for Internet users Previous researches have indicated that many people believe that they are susceptible to identity theft while using

mobile financial services (Mallat, 2007; Wessels & Drennan, 2010). Thus, the following hypothesis is proposed:

Hypothesis 4: Securities traders perceiving high security risk in M-Trading will have less behavioral intention to use it.

Economic risk is investors’ perception of the possibility of economic loss due to transaction error or incorrect operation when using M-Trading Extant researchers found that people resists using financially mobile-based services (Koenig-Lewis et al., 2010; Wessels & Drennan, 2010; Hsu et al., 2011) Other than the traditional trading channels such as website-based and telephone-based securities trading, while using M-Trading, data entry via small touch screen with limited display resolution may bring about input errors and typos easily made and difficult to detect In addition, if using other formats, the investors or their securities-brokers can manually verify the accuracy of the transaction information, in contrast, rarely is such safeguards available in M-Trading system, leading

to feelings of uncertainty and fear Thus, the following hypothesis is proposed:

Hypothesis 5: Securities investors who perceive a high economic risk for Trading will have less behavioral intention to use it.

M-Functional risk is securities investors’ perceived possibility of service unavailability or malfunction Many researchers found that many people forgo using mobile financial services due to concerns for such failures Shen et al (2010) and Wessels & Drennan (2010) found that many people are frightened that there would have occurrence of a failure of service systems or disconnection from the mobile Internet while conducting financial transactions via mobile devices Moreover, many resisters of system usage tend to suppose that mobile devices, mobile operating systems and networks are intrinsically unstable and worry that transactions may be interrupted, ceased, or delayed (Mallat et al., 2008; Cruz et al., 2010; Koenig-Lewis et al., 2010) Thus, the following hypothesis is proposed:

Hypothesis 6: Securities investors who perceive a high functional risk for Trading will have less behavioral intention to use it.

M-2.3.3 Hypotheses Derived From Privacy Concerns

Upon using M-Trading, users’ personal information (i.e., username, pass-code, location, verified code, account number) needs signing up to log in the system This

disclosure may arouse investors’ concern about their privacy They may worry about their

mobile-based application developers’ practice on information collection, storage and usage For instance, the securities traders are doubtful about their personal information being shared with other third parties without their prior approval or knowledge by the service providers, and may be anxious about the potential losses associated with information disclosure, such as information leakage and sales Due to this concern, securities traders’ behavioral intention to

use M-Trading may be impeded Extant researches (Bansal et al., 2010, Sheng et al., 2008,

Miyazaki & Fernandez, 2001) indicated this negative relationship between privacy concerns and ICT’s behavioral intention Hence, the following hypothesis is proposed:

Hypothesis 7: Securities investors with high privacy concerns in M-Trading will have less behavioral intention to use M-Trading.

Noticeably, privacy concerns and security risk are indicated as two clearly distinct constructs (Miyazaki &Fernandez, 2001; Román 2007; Román &Cuestas

2008, Riquelmi & Román, 2014) However, there exists interactive influence on each other (Belanger et al 2002, as cited in Riquelmi & Román, 2014; Schlosser et al 2006;

Hu et al 2010) For example, a high concern for personal information privacy would directly produce negative attitudes toward the security of smart-phone application, and the securities investors who lack knowledge about online security and the third party security identification would worry about disclosing personal information during the process of mobile-based trading Miyazaki & Fernandez (2001) concluded that both privacy concerns and security risk are the major obstacles in the development of online shopping Accordingly, the following hypothesis is proposed:

Hypothesis 8: Privacy concernsare positively correlated to security risk.

In addition, Malhotra et al (2004) indicated that Internet users with a high degree of information privacy concerns are likely to be high perceptions of risk Nepomuceno et al (2012) indicated that the perceptions of risk are increased by privacy concerns when North American households conduct purchases in an online environment Other previous researches have also indicated the effect of privacy concerns on perceived risks (Zhou, 2012; Eastlick et al., 2006; Bansal et al., 2010) Thus, the following two hypotheses about multi-facet perceived risks are proposed:

Hypothesis 9: Privacy concerns are positively correlated to economic risk.

Hypothesis 10: Privacy concerns are positively correlated to functional

risk 2.4 Conceptual model

Based on the hypotheses above, the below research model (Figure 2.4) is

proposed and evaluated empirically in M-Trading’s settings.

Security Risk

Economic Risk

Functional Risk

CHAPTER 3: METHODOLOGY This chapter presents a detailed account of a research methodology of this study First, it starts with sample description, followed by data collection method and research process Then, measurement scales are presented to develop questionnaire After that, in-depth interview is conducted to help measurement scales clearer and understandable Beside, this chapter also aims at explicating the research approach choice and presenting the purposes of using that method.

3.1 Research design

3.1.1 Research process

The study includes two main steps, (1) pilot survey and (2) main survey Both qualitative and quantitative surveys were used in the pilot test, and the quantitative survey was used in the main survey The main respondents comprise individual securities investors in Ho Chi Minh City Based on the previous research and the Vietnamese context, the draft questionnaire consisted of eight measurement scales, which were performance expectancy, effort expectancy, social influence, security risk, economic risk, functional risk, privacy concerns and behavioral intention Then, the draft questionnaire was translated from English into Vietnamese.

The pilot test used qualitative and quantitative method was conducted from May to July of 2015 In order to test logic of the questionnaires prior collection data on large cover, a pilot test will be carried out with a small group consisting of two securities broking professionals, and four senior securities investors who are familiar with website-based trading and used to use mobile financial applications via smart-phone (i.e VCB/ACB mobile banking, Paypal) All of them have much knowledge and many experience years in the securities (equity) trading field in Vietnam Firstly, the aim of the pilot test is to explain to all of them; moreover, the questionnaires and relative documents will be also sent to them After that, a discussion with them will be conducted to define which items would be eliminated or which items would be added up or be revised to be suitable to Vietnam context After adjusting the first questionnaire table, the questionnaire will be delivered to a small sample size of fifteen convenient colleagues and clients to recognize whether any item is still unclear to understand,

or is susceptible to misunderstand After receiving all feedbacks, the final version of questionnaires is available for the main survey The quantitative pilot test was used to evaluate the items resulted from in-depth discussions prior to conduct main survey The items were evaluated through assessing reliability composite (Cronbach Alpha) and exploring factor analysis (EFA) The main survey was conducted in Ho Chi Minh City from June to July of

2015 With the prior notice, the questionnaire was emailed to all staffs of VNDirect

Securities Corporation (VNDS), dozens of brokers of Saigon Securities Inc (SSI), dozens

of brokers of Ho Chi Minh City Securities Corporation (HSC) and some securities market professionals in Mirae Asset Corporation; in VPBank Securities Corp.; Vietcapital Securities Corp Especially, thanks to the securities brokers in VNDS, SSI and HSC in association with this survey, the questionnaire were delivered to the securities investors via email Any and all results were automatically returned to the author Besides, dozens of in- depth interviews at trading floors were also conducted The items then were analyzed by CFA and the hypotheses were tested by SEM (see Figure 3).

Literature

Review

Cronbach Alpha

- Eliminate corrected item - total correlation

- Evaluate Cronbach’s Alpha

Exploratory Factor Analysis (EFA)

- Eliminate the variables with low EFA

- Evaluate the validity and the correlation

among variables to identify underlying factors

or define number of extracted factors

Second Pilot Scales

Pilot Quantitative Study (n = 120)

Confirmatory Factors Analysis (CFA)

- Model fit, item loadings ³0.05

- Composite reliability, extracted variances, uni-dimensionality

test, convergent validity and discriminant validity

Structural Equation Model (SEM)

- Theoretical model test

- Model fit, Component fit

Figure 3: Research process (adopted from Nguyen Dinh Tho & Nguyen Thi Mai Trang, 2007)

3.1.2 Measurement scales

As mentioned above, the draft questionnaire includes eight measurement scales, which were (1) performance expectancy, (2) effort expectancy, (3) social influence, (4) security risk, (5) economic risk, (6) functional risk, (7) privacy concerns, and (8) behavioral intention All of the variables related to the factors in research model are measured using five-point Likert scale, ranging: 1 (strongly disagree), 2 (disagree), 3 (neutral), 4 (agree), and 5 (strongly agree).

Performance expectancy

Performance expectancy mentions the degree that an individual believes that usage

of M-Trading will help improve their securities-trades’ performance Basing on the research model developed by Tai & Ku (2013), this construct comprises four items:

1 Using M-Trading would enhance my securities trading efficiency PE1

3 Using M-Trading would increase the convenience of securities trading PE3

4 Using M-Trading would enable me to accomplish securities trading more quickly PE4

Effort expectancy

Effort expectancy mentions the degree that an individual believes that learning

to use M-Trading will not require significant effort Basing on the research model developed by Tai & Ku (2013), this construct comprises four items:

3 It would be easy for me to become skillful at using M-Trading EE3

Social influence

Social influence is defined as the degree to which an individual perceives that its important peers expect his/her to use M-Trading Basing on the research model developed by Tai & Ku (2013), this construct comprises four items:

1 I feel people around me would encourage me to use M-Trading SI1

2 People who are important to me would think that i should use M-Trading SI2

4 In my environment, people encouraged me to use M-Trading SI4

Perceived risks

According to Tai & Ku (2013), perceived risks including security risk, economic risk, and functional risk reflect the users’ worries and concern on the usage of M-Trading Risks

associated with usage of M-Trading comprise: disclosure of personal information, loss

of account, hacker attacks, malfunction, and transaction error There are twelve items measuring these three constructs, four items for each construct.

Security risk (adopted from Tai & Ku, 2013 )

1 I would not feel secure conducting securities trades via M-Trading systems

2 I am worried that others might be able to access my M-Trading account.

3 I would not feel secure sending sensitive information across M-Trading systems.

4 I would not feel totally safe providing personal information over M-Trading systems.

Coding SR1 SR2 SR3 SR4

Economic risk (adopted from Tai & Ku, 2013 )

1 I am uneasy about using M-Trading because I may lose money due to incorrect

1 M-Trading systems may not perform well because of the limited processing power of

mobile devices.

2 M-Trading systems may not perform well because of system failure.

I am uneasy about using M-Trading because securities transactions may fail due to

3 the unstable nature of mobile devices, mobile operating systems or mobile

networks.

4 I am concerned that M-Trading services cannot meet my needs due to poor

functionality or system malfunctions.

Coding FR1 FR2 FR3 FR4

Privacy concerns

Privacy concerns reflect M-Trading users’ concern on personal information disclosure to securities firms Basing on the research model developed by Zhou (2012), this construct comprises four items:

Privacy concerns (adopted from Zhou, 2012 )

1 I am concerned that the information I disclosed to the service provider could be

misused

2 I am concerned that a person can find private information about me on Internet.

3 I am concerned about providing personal information to the service provider, because

of what others might do with it.

4 I am concerned about providing personal information to the service provider, because

it could be used in a way I did not foresee.

Coding PC1 PC2 PC3 PC4

Behavioral intention

Behavioral intention mentions the users’ intention to use M-Trading Basing on the research model developed by Tai & Ku (2013), this construct comprises four items:

Behavioral intention(adopted from Venkatesh et al., 2003 ) Coding

3.2 Measurement refinement

In qualitative study, all observed items of draft questionnaire were translated into Vietnamese Accordingly, the in-depth interviews with six participants were undertaken with draft questionnaire List of participants is depicted in Appendix A All the comments are taken note in the Appendix B Changes for the Vietnamese version was made for the purpose of accuracy and clarity Although most of the scales were used widely in the previous research, this study was important before launching the quantitative survey due to the differences in the research setting: applying in the Vietnamese context In the end of this study, the modification and revision of questionnaire survey are noticed in Appendix B The official questionnaire surveys are shown in Appendix C for English version and Appendix D for Vietnamese version.

Measurement scales after being modified through in-depth interviews includes thirty four items as depicted as Table 3.1

Table 3.1: Final measurement scales

Performance Expectancy

1 I think that using M-Trading would enhance my securities trading efficiency

2 I feel M-Trading is useful

3 I think that using M-Trading would increase the convenience of securities trading

4 I think that using M-Trading would enable me to accomplish securities trading

more quickly

Effort Expectancy

5 With my ability, learning how to use M-Trading would be easy for me.

6 I expect that M-Trading would be displayed understandably and easy to

utilize as same as website-based trading

7 I would attempt to use M-Trading skillfully.

I would find M-Trading easy to use as same as website-based trading or other

8 mobile-based

applications Social Influence

9 I feel people around me would encourage me to use mobile-based financial

applications (banking, securities trading, electronic payment)

10 People who are important to me would think that I should use mobile-based

financial applications (banking, securities trading, electronic payment)

I will use mobile-based financial applications (banking, securities trading,

11 electronic payment) to be correspondent to my peers since they used/are

about to use

PC1 PC2 PC3 PC4 EE1 EE2 EE3 EE4

SI1 SI2 SI3

The mass media often mobile-based financial applications (banking, securities

12 trading, electronic payment) are often covered by the mass media, I use it

on trial basis

My school, my company and community encourage me to use mobile-based

13 financial applications (banking, securities trading, electronic payment)

Security Risk

14 I would not feel secure conducting securities trades via mobile securities trading

systems.

15 I am worried that others might be able to access my M-Trading account.

16 I would not feel secure sending sensitive information across mobile securities trading systems.

17 I would not feel totally safe providing personal information over M-Trading

20 I (would) feel uneasy about using M-Trading because I may lose money due to

system processing errors.

21 When transaction errors occur, I (would) be concerned that the securities broker

may not compensate my loss.

Functional Risk

22 M-Trading systems may not perform well because of the limited processing power

of mobile devices.

23 M-Trading systems may not perform well because of system failure I

(would) feel uneasy about using M-Trading because securities transactions

24 may fail due to the unstable nature of mobile devices, mobile operating

systems or mobile networks.

25 I (would) be concerned that M-Trading services cannot meet my needs due to

poor functionality or system malfunctions.

Privacy concerns

26 I am concerned that the information I disclosed to the service provider could be

misused.

27 I am concerned that a person can find private information about me on Internet.

28 I am concerned about providing personal information to the service provider,

because of what others might do with it.

29 I am concerned about providing personal information to the service provider, because

it could be used in a way I did not foresee.

Behavioral Intention

30 I intend to use M-Trading in the future.

31 I predict I would use M-Trading in the future

32 I plan to learn skillfully the usage of M-Trading in the future.

33 I will refer M-Trading to other people.

SI4 SI5

SR1 SR2 SR3 SR4 ER1 ER2 ER3 ER4

FR1 FR2 FR3 FR4

PC1 PC2 PC3 PC4

UI1 UI2 UI3 UI4 3.3 Sample

Due to limited time, the convenience sampling approach was conducted in Ho Chi Minh City The method of self-administered survey was employed for this study, which consisted of seven factors and thirty three variables The survey was conducted in Ho Chi Minh City The sample was selected using a non-probability sampling technique – convenience sample The main respondents comprises individual securities investors in Ho

Chi Minh City, the biggest securities exchange in Vietnam, and especially brokers are also taken into consideration since the individual investors seek for guidance, advice from broker/investment adviser to meet their financial goals.

securities-The reliable and validity of variables will be tested by using Cronbach’s Alpha and EFA, CFA, after that the SEM was applied to test model and hypotheses First of all, the sample size was required to have enough quantity for the analysis The minimum sample size was 100 and not less than five times of items (Hair et al 2010), thus: n > 100 and n = 5k (where k is the number of items).

Thus, the minimum sample size was 5x33 = 165 samples.

With expectation to obtain a sample size of about 300, about 400 questionnaires were delivered to participants After data collection, total 321 responses were collected; the response rate was approximately 80.5 percent Then, total no questionnaire was eliminated because they were invalid (respondents just chose one option for all questions or their answers were implausible) Finally, 244 questionnaires were used as valid data for this research In comparison with minimum sample size, this number of data was satisfactory.

3.4 Data analysis and interpretation

Total 244 responses were used for data analysis Both SPSS 22 and Amos 22are used to test the model In the first part, while Cronbach’s Alpha tested the reliability for each measurement component separately, EFA tested the validity for all items scale Considering the convergent and discriminated validity, the inappropriate items would be eliminated if necessity In second part Amos 22 for CFA and SEM with purpose of enhancing the value of the model was employed.

3.4.1 Reliability measure

In order to assess reliability of each of scales with particular sample, as well as consider the internal consistency of the scales, it is necessary to use Cronbach’s Alpha coefficient which should be above 0.6 (Devellis, 2003) Also, the corrected item - total correlation values should be at least 0.3 to ensure each of items was measuring the same from the scale as a whole (Pallant, 2011).

3.4.2 Validity measure by EFA (Exploratory Factor Analysis)

In order to evaluate the validity and the correlation among variables to identify underlying factors or define number of extracted factors, EFA was applied with the oblique approach using the Promax method However, some requirements of EFA should be satisfied (Pallant, 2011):

- The minimum of sample size should be at least 100 and rate of observations per items of models should be five cases for each of the items, so that meant the minimum required

sample size should be at least 5m = 5x33 = 165 cases (where m: quantity of items from the conceptual model).

- The correlations of r of the correlation matrix should show at least 0.3.

- Kaiser-Meyor-Olkin (KMO) test must be equal or above 0.6 (Tabachnick & Fidell, 2007).

- Barllett’s test of sphericity should have significant less than 5%.

- To extract factors, the eigenvalue of factors must be greater than 1 (Kaiser,

1956) 3.4.3 CFA & SEM

The CFA results would indicate the model fit if CMIN/DF was less than 3 with value larger than 5%, GFI, RFI, and CFI were larger than 0.9, and RMSEA was smaller than 10% Based on composite reliability (CR), the author evaluated the measurement scale’s reliability and used average variance extracted (AVE) to conclude the convergent validity and the correlation between items (r) to identify the discriminated validity Then structural equation modeling (SEM) was used to test thehypothesized model and was applied to estimate path coefficients for each proposedrelationship in the structural model.

p-3.5 Pilot test

Prior to conduct the official survey, the items of conceptual model’s constructs were evaluated by pilot quantitative study This pilot evaluation was carried out by quantitative survey with a convenient sampling size at 120 (n = 120) Two tools employed to do pilot quantitative study were reliability evaluation by Cronbach’s Alpha and Explotary factors analysis (EFA).

3.5.1 Cronbach’s Alpha

The Cronbach’s Alpha coefficient for internal consistency reliability test was used for each scale in this research model Cronbach’s Alpha reliability coefficient normally ranges between 0 and 1 George & Mallery (2003) provide the following rules of thumb: _ >0.9 – Excellent, _ >0.8 – Good, _ >0.7 – Acceptable, _ >0.6 – Questionable, _ >0.5 – Poor, and _ <0.5 – Unacceptable The results of Cronbach’s Alpha coefficients for each scale are presented as following table 3.2.

Table 3.2: Cronbach’s Alpha test

Items Scale Mean if Scale Variance if Corrected Item- Cronbach's

Item Deleted Item Deleted Total Correlation Alpha if Item

Performance Expe ctancy (PE1 – PE4) Cronbach’s Alpha = 0.848

3.5.2 Exploratory Factor Analysis

After testing Cronbach’s Alpha coefficient, the measures were continued to be analyzed by EFA analysis method (extraction method: Principal Axis Factoring and rotation method: Promax).

The Kaiser-Meyor-Olkin (KMO) is 0.8706, greater than the requirement at 0.6

by Tabachnick & Fidell (2007) In addition, Barllett’s test of sphericity’s significance was less than 5%.

Table 3.3: EFA – KMO and Bartlett's Test KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling Adequacy 0.8706 Bartlett's Test of Sphericity Approx Chi-Square 2062.052

All factors have eigenvalues greater than 1, and that there are seven factors extracted and total variance extracted was 64.367% (greater than requirement of 50%) Thus, variance extracted matched requirement.

Table 3.4: EFA - Total Variance Explained

Initial Eigenvalues Extraction Sums of Squared Rotation Sums of

3.6 Chapter summary

In summary the survey was designed based on previous measurement scales Specifically, scales of performance expectancy and effort expectancy consist four items for each, and social influence consist five items adopted from Venkatesh et al (2003) Scale of security risk, economic risk, and functional risk include for items for each adopted from Tai

& Ku (2013) Modifications for the Vietnamese version and English version of

questionnaire were conducted due to the necessary of accuracy and clarity In total, measurement scales included thirty four items had been used to formulate the questionnaire survey, which had been delivered to four hundred respondents Total

321 responses were collected; the response rate was approximately 80.5 percent The final 244 questionnaires were used as valid data for this research Both SPSS 22 and Amos 22 are used to test the measurement and theoretical model As the result, all items satisfied the requirement for reliability (Cronbach’s Alpha), and the validity and the correlation among variables also met the requirement through EFA analysis.

CHAPTER 4: RESEARCH FINDINGS Chapter 4 presents the analysis results, which included respondents’ demographics, descriptive statistic, confirmatory factor analysis, structural equation modeling, and the explanation for the finding results While respondents’ demographic made the review of sample’s characters, the descriptive statistic tested the normal distribution of variables Then, CFA examined the reliability and validity of the first order constructs, the second order construct and the final measurement model SEM was used to test the conceptual model Based on the analysis’s results, the explanation for finding research was discussed.

4.1 Data description

The collected data was analyzed by using the SPSS – Statistical software package.

This part aims to provide the general information of respondents.

Table 4.1: Descriptive Statistics

18 to 35 years old, were representative for 8-x and 9-x generation These generations were considered the innovation group, who very fast adoptednew technology from all over the world The differentiation among age groups might affect the result of data analysis.

Remarkably, all of respondents have educational level at graduates and post-graduates (100%) This means that the higher educational level is, the easier new technology is accepted Moreover, 98.7% of respondents are using smart-phone and 65.6% out of them are

using mobile-based financial applications (mobile payment, mobile banking) Besides, 75% of respondents conducted online financial trades (securities, gold, forex).

Descriptive statistic was run to test whether variables distribute in normal distribution or not The variables had Skewness value within (-0.062 to 0.132), and the value of Kurtosis was within (-1.111 to 1.517) Examining the individual factor as shown in Appendix E, all items had values of Kurtosis and Skewness smaller than |2|,

so 33 observed items were moved to the next step of analysis.

4.2 Confirmatory Factor Analysis (CFA)

4.2.1 Saturated model

Figure 4.1: The saturated model (standardized)

In order to test the discriminant validity of all constructs of research model, a

saturated model needed generating Anderson & Gerbing (1988) denoted that saturated model

can be defined as one in which all parameters related to the constructs to one another are estimated.

The Chi-square value of the model was 554.036, chi-square normalized by degree of freedom (χ 2 /df) was 1.189 less than 3, RMSEA value was 0.028, TLI, CFI, GFI and AGFI value were respectively 0.98, 0.978, 0.884 and 0.860 All of the factor loadings for the items were greater than 0.5 As the result, these scales had convergent validity and the model achieved a good fit to the data (Figure 4.1) Besides, the measurement scales were also satisfactory with uni-dimensionality requirement due to

no correlation among their items error.

For the discriminant validity, the correlations among constructs were lower than 1 and significant (p < 05), so they achieved the discriminant validity.

4.2.2 Composite reliability and variance extracted

The reliability of measurement scales were evaluated by Cronbach’s Alpha of Reliability Coefficient (α), Composite Reliability (P CR ) and Average Variance Extracted (P AVE ), where in, (Ρ CR ) and (Ρ AVE ) are calculated by the below formulas:

(developed by Joreskog (1971) and Fornell & Larcker (1981)) Table 4.1 showed that almost indices satisfied requirement for Cronbach’s Alpha (α ≥ 0.5), Composite Reliability (P CR ≥ 0.5), Average Variance Extracted(P AVE ≥ 0.5) except for

EE construct had slightly lower Average Variance Extracted(P VC = 0.46) but acceptable Accordingly, the measurement scales of these constructs matched reliability requirement.

Table 4.2: The results for reliability and variance extracted test

Figure 4.2: structural equation model& path analysis

In this study, there are ten hypotheses needed testing The theoretical model was tested using SEM approach.

Model estimation results in agood fit between the model and data: Chisquare/df = 1.3, p = 0.000, GFI= 0.869, TLI= 0.964, CFI= 0.968, RMSEA= 0.035 (Figure 4.2) The estimation results of the relations among constructs in model were statistical significant with 95% confidence interval (p <0.05) except for FR construct with an extremely high p- value (p = 0.312) It demonstrated that theoretical model fit to data (Figure 4.3).

The results showed that at the significant level of 1%, EE has a substantial contribution to BI (β = 0.29) and SI had positively significant direct effect on BI (β = 0.172) In addition, at the confidence interval 95%, PE (β = 0.16) had positively significant direct effect on BI Hence, all UTAUT’s constructs adopted in the conceptual model in this study were fit Once again, we can confirm that those three constructs explained well the ICT’s behavioral intention in the context of Vietnam Thus, hypotheses 1 through 3 were all supported.

Concerning the inhibitors of ICT’s behavioral intention set in the conceptual model comprising of PC, SR, ER, and FR, these four constructs had negatively effect on BI