Journal of Internet Banking and Commerce

Transcription

1 Journal of Internet Banking and Commerce An open access Internet journal ( Journal of Internet Banking and Commerce, August 2014, vol. 19, no.2 ( The Adoption of Personal Internet Banking in Vietnam MANH HUNG HOANG, MSc. Faculty of Science and Technology, Assumption University, Thailand Postal Address: VMS Building 6th Floor, Assumption University, Thailand Organizational Website: Mr. Hoang is a PhD Student of Faculty of Science and Technology in Assumption University, Thailand. His areas of interest are Internet and Banking, IT Security, ERP System, Open Source and Database. Abstract Based on previous studies a model of the determinants of an individual s intentions (Behavioral Intention) to use Personal Internet Banking (PIB) services was formulated, tested using data collected by questionnaire from 319 users of PIB in Vietnam, and developed to arrive at a final model. Independent determinants included an individual s personal characteristics (Gender, Age, Level of Education, Internet, Internet Banking, and Position Description), personality traits (Extraversion, Agreeableness, Conscientiousness, Neuroticism, and Openness to ), and level of trust in the bank (Trust). These were proposed to affect Behavioral Intention directly or indirectly by influencing the individual s perceptions of the ease of use and usefulness of PIB ( Ease of Use and ). From a theoretical perspective many of the findings in previous studies were found to be associated with significant correlations rather than significant causal effects which involved only four independent variables (Position Description, Trust, Agreeableness, and Neuroticism). New findings stressed the important mediation roles played by Ease of Use and in the total effect of Trust on Behavioral Intention while all of the personal characteristics, except for Position Description, were found to be less important than in several previous studies. Among the personality traits the direct effect of Agreeableness on Ease of Use and

2 JIBC August 2014, Vol. 19, No Neuroticism on Behavioral Intention had not been reported in previous PIB studies and the same was true for other significant correlations involving Extraversion, Conscientiousness, and Openness to. Based on the final causal model practical implications of the findings are discussed. Keywords: Ease of Use; ; Personal Characteristics, Personality Traits; Behavioral Intention; Trust; Personal Internet Banking; Vietnam; Manh Hung Hoang, 2014 INTRODUCTION Personal Internet Banking (PIB) is one of many services that banks provide as alternative channels for customers to access their bank accounts instead of going to the bank. With PIB, benefits accrue from serving more customers with fewer employees and lessening loads on ATMs (e.g. employees and security staff needed to refill money, maintenance, electricity, and rental of location). PIB frees the customer from the constraints of time and place to conduct their personal banking transactions. Internet banking in Vietnam has appeared later than in other countries in the region. However, the Vietnam government is trying to keep up with the technology. As shown in Table 1 the growth in the number of banks providing PIB services increased from only three in 2004 to 45 out of a total of 50 banks in December 2011 (Ministry of Industry and Trade, 2012). Increasing the use of PIB is one of the efforts to push the development of e-commerce, which is an objective in the mission of the Vietnam government to 2020 (State Bank of Vietnam, 2011). Although the Vietnam government and banks are encouraging business and citizens to use cashless transactions with many promotion programs and projects, the success of PIB is not only dependent on government and bank support but also customer acceptance (Le, 2011). Table 1: Number of banks providing PIB services in Vietnam by year Year Number of Banks Source: Vietnam e-commerce Report 2011 Very few studies of PIB have been conducted in Vietnam and they do not reflect the complete picture of PIB in Vietnam (Wang and Pho, 2009; Chong et al., 2010; Le, 2011; Lin and Nguyen, 2011) because the infrastructure to support PIB in Vietnam has only been implemented quite recently and extra services are still in the experimental stage. This study is motivated by the need to further understand the effects on an individual s attitude to the adoption of PIB in Vietnam and to addressing the lack of current research on this topic in Vietnam.

3 JIBC August 2014, Vol. 19, No Based on the findings from previous studies (Section 3), an extended version of the Technology Acceptance Model (TAM) model (Davis et al., 1989) was developed as the theoretical model (Section 4) for explaining an individual s intention to use PIB in Vietnam. Data collected by questionnaire was prepared and analyzed (Section 5) as a prerequisite to the analysis and development of the theoretical model in order to produce a final model (Section 6). The findings of the study are discussed in Section 7 and conclusions are presented in Section 8. RESEARCH DESIGN AND METHODOLOGY A cross-sectional field study design was used to collect data using a questionnaire designed to measure the variables in the theoretical model. English and Vietnamese language versions of the questionnaire were prepared and reviewed by a focus group of five PIB users representing the target population. Suggested modifications were included in revised versions of the questionnaire that was then administered in a pilot study using a sample of 10 suitable participants. Their responses and comments were noted and any necessary modifications were incorporated into the final versions of the questionnaire and the Vietnamese language version was then used in the full study. A notated English language version of the final questionnaire is included in Appendix A1. The target population for the study was individuals who were at least 18 years of age with at least one month of Internet and PIB experience. A purposive sampling method was decided with a minimum sample size of approximately 400 individuals which represented a 95 percent confidence level and 5 percent precision in what was assumed to be an unknown but very large target population. Also, this sample size satisfied the criteria for the statistical validity of the structural equation modeling (SEM) and other statistical techniques used in the study (Kline, 2005). Sampling was done in stages using educational institutions, organizations, social media, and the researcher s personal contacts. Individuals were invited to complete the questionnaire online at a designated web site where explanations of the purpose of the questionnaire and instructions for its completion were available. RELATED LITERATURE The purpose of the review of literature related to PIB was to identify among recent studies the important variables which have been shown to influence an individual s intentions to use PIB services. In line with the nature of this study the review has focused on previous studies in developing nations which have used survey questionnaires as the means of collecting data and quantitative methods for analysis. An overview of related studies is presented first followed by a discussion of the important variables and their relationships which form the basis for the theoretical model used in this study. An Overview of Previous Studies Table 2 presents an overview of previous studies of Internet banking where the unit of analysis is an individual user and data was collected by questionnaire. The review covered the period from 2008 and variables included in these studies are identified.

4 JIBC August 2014, Vol. 19, No From the table 2 it is seen that over the period there were numerous studies of PIB where the unit of analysis is an individual user. Studies have been conducted of particular user groups, developing countries, and countries with different cultures and religions. Most studies identified factors that affect the adoption of PIB technologies while others focused on resistance factors and barriers may cause a low rate of adoption. Customer satisfaction and intentions to continue to use PIB services were the dependent variables in most of the studies. The majority of the studies tested an explanatory theoretical model while the remainder was exploratory or descriptive in nature. Many researchers have used TAM as a theoretical basis for studying PIB while other theories have received relatively less attention compared to TAM even though they were more recently than TAM (e.g. UTAUT). Model Variables Based on a comprehensive survey of the variables and the theoretical bases used in previous studies (Table 2), 15 important variables were identified and classified into four groups: three variables related to TAM; six variables related to personal characteristics of individuals; trust as an important indicator of an individual s attitude to PIB; and five variables related to an individual s personality traits, which reflect cultural influences measured at the level of an individual. TAM Variables and Ease of Use are in all of the previous studies which adopted TAM. However, in many studies (van Raaij and Schepers, 2008; Aldás- Manzano et al., 2009; Chong et al., 2010; Le, 2011; Kesharwani and Bisht, 2012; Yousafzai and Yani-de-Soriano, 2012) the Attitude construct was removed from TAM while retaining Intention because this two variables are significantly positively correlated when the adoption of a technology is voluntary (Davis et al., 1989; Venkatesh et al., 2003) and PIB is a self-service which operates based on the voluntariness of use by the customer. Also, although Intention and Actual Use in general technology adoption might be different in the context of voluntary PIB usage they have been shown to be significantly correlated ((Prompattanapakdee, 2009). For this reason, many studies removed Actual Use from TAM and instead used Intention as the dependent variable based on the understanding that if usage is voluntary and the individual s intentions to use the system are strong (weak) then actual usage will be high (low). Trust Trust is an important factor in many social interactions involving uncertainty and dependency (Grabner-Kräuter and Faullant, 2008) especially in relation to Internet based applications and PIB in particular. In PIB, Trust is understood as the confidence which the individual has in the bank s ability to provide reliable Internet banking services. Previous studies have identified consistently the significant role of Trust in PIB. Some researchers refer to Credibility instead of Trust but these constructs have a similar meaning. Results from previous studies show that Trust has a direct effect on an individual s perceptions of the usefulness and ease of use as well as their intention to use the system (Aldás-Manzano et al., 2009; Alsajjan and Dennis, 2010; Le, 2011). According to Winley (2011), customers lose their trust for four main reasons: security; the service provider's reputation; privacy; and risks associated with reliability of the services.

5 JIBC August 2014, Vol. 19, No Personal Characteristics Several previous studies of PIB have included variables related to personal characteristics (Gender, Age, Level of Education, Internet, Internet Banking, and Position Description) as having an influence on the individual s intention to use PIB services. Yousafzai et al. (2009) found effects of Age and Gender on and Intention to use PIB. Computer self-efficacy (related to level of education) is also an important factor with effects on and Ease of Use (Le, 2011; Sentosa et al., 2012) and Intention to Use (Chau and Ngai, 2010; Kesharwani and Tripathy, 2012; Sentosa et al., 2012). It seems obvious that because PIB is a self-service technology based on the Internet it will attract more attention from those who often use and work in an Internet environment and this was confirmed by the studies of Chan and Hosein (2010) and Ozdemir and Trott (2009). Ozdemir and Trott (2009) also found that customers with higher incomes in more senior positions with higher levels of education who work longer hours are likely to adopt PIB. However, in Vietnam, because income is often not reported precisely by individuals in questionnaires it was not among the personal characteristics that were included in the theoretical model for the study. Culture Characteristics and the Individual s Personality Hofstede (2006) states that: Culture is the collective programming of the mind that distinguishes the members of one group or category of people from others. Culture characteristics are related to standard operating procedures, unstated assumptions, tools, norms, values, and habits in sampling the environment (Triandis, 2001). Personality is defined as an individual s characteristic pattern of thought, emotion, and behavior, together with the psychological mechanisms hidden or not behind those patterns (Triandis, 2001). As noted by Triandis (2001), when studying the relationship of culture and psychology it is imperative to keep the level of analysis distinct because the results of analyses are often different depending on the level of analysis. Triandis proposed to link the cultural (societal) and individual (psychological) levels of analysis by noting that customs are aspects of culture and habits are aspects of personality and he hypothesize a correspondence between customs, norms, and values on the one hand and habits and patterns of individual behavior on the other hand. In PIB studies several researchers have considered (social) norms as determinants of PIB adoption (Nor et al., 2008; Shi et al., 2008; Lee, 2009; Al-Majali and Mat, 2010; Alsajjan and Dennis, 2010; Zolait, 2010; Abbesi and Haghighy, 2011; Nasri and Charfeddine, 2012). In a study of personality and culture, Hofstede and McCrae (2004) also noted the confusion between levels of analysis. One possibility for relating dimensions of culture and personality factors was measuring the culture dimensions in individuals. This has been tried most often in the case of individualism-collectivism with notably inconsistent

6 JIBC August 2014, Vol. 19, No Table 2: Previous Internet banking studies Country Focus of the Study Variables India Pakistan Malaysia Internet banking adoption under risk and privacy threat Impacts of risk on IB adoption TAM (PU, PEU, Intention), Trust, CC (Social Influence), Risk, Website Design, Behavioral Control TAM (PU, PEU, Intention to Use), PC (Computer selfefficacy), price concern, technology complexity, financial risk, performance risk, physical risk, psychology risk, time-loss risk, CC (social influence) Theoretical Basis TAM TAM References Kesharwani and Bisht, 2012 Kesharwani and Tripathy, 2012 Factor influences IB adoption TAM (PU, PEU), Risk, Consumer Acceptance TAM Safeena et al., 2011 Consumer s perspective on internet banking adoption. TAM (PU, PEU), Risk, Awareness, IB Adoption TAM Safeena et al., 2010 Customer satisfaction on adoption - Akram and Asghar, online banking 2012 Factors obstructed IB adoption Factors manipulate the acceptance of IB Muslim consumer s perception on IB Determinants of internet usage in banks Factors influence IB adoption Factors Affecting the Internet Banking Adoption Factors that influence banking customers intention to use IB Customer perception of e-service quality and adoption of IB Factors influence individual s intention to use IB Technology, Ease to Use, Web Design, Security, Speed, Information Content, Interaction with Internet Banking, Quick response, Cost, Customer, Access, Time Familiarity with technology, Concerns for download speed, Unfriendly website design, fear of government tracking transaction, security, people gaining access/misusing information PU, PUE, Enjoyment, Info on Online Banking, Security and Privacy, Quality of Internet, Customer Acceptance, Characteristics of Innovations (Relative advantage, Compatibility, Complexity, Trialability, and Observability), Time, Communication, Social system PC (Computer self-efficacy), TAM (PU, PEU, Intention), credibility Performance expectancy, Effort expectancy, Social influence, Facilitating condition, Trust importance, complexity, Trialability, Compatibility, perceived risk, utilitarian oriented Internet banking sites, hedonism oriented Internet banking sites DOI Akhlaq and Shah, Qureshi et al., 2008 DOI Anuar et al., 2012 TAM Sentosa et al., 2012 UTAUT Foon and Fah, Suki, 2010 TAM (PU, PEU, Intention to use), Trust TAM Nor et al., 2010 EU, Appearance, Reliability, Customization, - Raman et al., 2008 Communication, Incentive Attitude toward IB, Subjective Norm, Intention to use IB TRA Nor et al., 2008

7 JIBC August 2014, Vol. 19, No Country Focus of the Study Variables Vietnam Thailand Brunei Taiwan China Hong Kong Customers intention to use IB Factors that affect the adoption of online banking Determinants of customer intention to use online banking. Important variables associated with the use of IBS Use of IB Services Factors determine success and affect intention of IB Demographic factors with IB satisfaction Customers intention to use IBS Factors influence IB adoption Reasons for customer continued of use IB Role of trust and risk in IBS usage intention Risk factors with IB adoption Factors impact IB adoption TAM (PU, PEU, Intention of use), PC (Computer selfefficacy),trust TAM (PU, PEU), Government support, Trust, Adoption of IB System quality, Information Quality, Service Quality, Complementary Relationship, Expertise, Trustworthiness, Intention to Use, Satisfaction TAM (PU, PEU, Attitude, Intention, Actual Use), Trust, CC( Personal Relationships, Peer Influence), PC (Age, Level of education, income, position description, Internet experience, IB experience) PC (Age, gender, Level of education, income, Internet experience, IB experience, Position Description), Trust, CC (Personal Relationships, Peer Influence), TAM (PU, PEU, Attitude, Intention, Actual Use) TAM variables, attitude, trust, perceived credibility (security, privacy), antecedent (communication, opportunistic behavior, reputation, reliability, trust in service provider ), factors moderate relationship (user characteristics, customer readiness). PC (Age, Gender, Income, Education Level, Internet experience), Customer Satisfaction TAM (PU, PEU), risk, trust of the Physical bank, Satisfaction with the Physical bank, Specific asset investment in the Physical bank TAM (PU, PEU, Attitude, Intention), benefit, Risk (performance, social, time, financial, time, security), Social Norm, Behavior Control, TAM (PEU, PU, Intention to Continued Use), Cost Saved, Self-Control, Customer Value, Customer Readiness, risk of IBS, Trust in Bank, IBS Competence, Behavioral Intention to use IBS 8 risks factors: Performance, security, financial, privacy, time/convenience, psychological, social, physical Brand Name of IB provider, Credibility, TAM (PU, PEU, Intention to use IB) Theoretical Basis TAM Le, 2011 References TAM Chong et al., 2010 DeLone and Wang and Pho, 2009 McLean Success Model TAM Winley, 2011 TAM TAM and TPB Transaction Cost Analysis and TAM TAM and TPB Prompattanapakdee, 2009 Sattabusaya, 2008 Seyal and Rahim, 2011 Chiou and Shen, 2011 Lee, 2009 TAM Ho and Ko, Zhao et al., Zhao et al., 2008 TAM Cheng and Yeung, 2010

8 JIBC August 2014, Vol. 19, No Country Focus of the Study Variables Australia Turkey Greece Tunisia Measuring IB consumer satisfaction: a core framework Exploring the antecedents to consumers continued and frequent use of IB Determinants of Customers IB acceptance Segmenting IB adoption customers Factors affecting IB adoption IB vs. other banking channels by young consumers Customers resistance behavior toward IB Predict IB adoption by usability, social, psychological aspects Drivers of IB adoption Factors affect the adoption of Internet banking Factors influence IB adoption, Ease of Use, Reliability, Security, Responsiveness, Continuous Improvement Technology factors, Social factors, Value for money factors, Usage factors TAM (PU, PEU, ATT, Behavioral Intentions), Behavioral Control, Playfulness, Risk, Perceptual, related, Socio-economic and situational characteristic Risk, PU, PEU, PC (Mobile Internet, Internet, Income, Working hours) Theoretical Basis References TAM and Ziqi Liao and Tow SERVQUAL Cheung, 2008 Consumer Adapa, 2011 Adoption and DOI TAM Çelik, 2008 TAM Ozdemir et al., 2008 TAM, DOI, and Theory of Risk Ozdemir and Trott, constructs created by 20 subjects - Calisir and Gumussoy, 2008 PC (Gender, Age, Education, Employment, Marital - Patsiotis et al., 2012 status, Income), Human interaction, Enjoyment, economic gains, time and energy savings, usefulness, ease of use, interactivity, timeliness, mutual controllability, responsiveness, awareness, personal capacity, prior experience, perceived risks, privacy concerns, security concern, lack of trial, intentions Relative advantage (usefulness), ease of use (complexity), compatibility, trialability, result demonstrability, visibility and image, voluntariness. Compatibility, Ease of use, Image, Relative advantage, Result demonstrability, Trialability, Voluntariness, Visibility TAM (PU, PEU, Intention to use), Security and Privacy, Attitude, Subjective Norm, Self-efficacy, Government support, Technology support, perceived behavioral control PC (Demographic characteristics, Prior internet Knowledge), Convenience, Security perception, risk, Information on online banking TAM, DOI and Characteristics Innovation TAM and DOI TAM and TPB TAM, TRA, and TPB S. P. Gounaris and Koritos, 2008b S. P. Gounaris and Koritos, 2008a Nasri and Charfeddine, 2012 Nasri, 2011 Factors impact on IB adoptions risks, benefits. - Azouzi, 2009

9 JIBC August 2014, Vol. 19, No Country Focus of the Study Variables Iran Jordan Yemen Mauritius United Kingdom Scotland Spain Portugal Effect of risk on intention of using internet banking Factors affect the IB acceptance Factors influence acceptance of IB: differentiate between private and government IB customers. Factors impact on IB adoptions Success factors of IBS adoption Factors influence IB adoption Factors relating to the adoption and use of IB Factors influence IB adoption Effects of demographics and Technology readiness to IB adoption Perception of young customer on IB Develop and validate a model of trust for IB Determinants of IB use Factors influence brand equity and loyalty in the context of IB Time risk, Financial risk, Performance risk, Social risk, Security risk, Privacy risk TAM (PU, PEU, Attitude, Willingness to use, Actual use), Trust Convenience, Security & Privacy, Cost, Ease of use, Quick transaction, Demographic characteristics PC (Age, income, gender, marriage status, Access to Internet, Internet usage), delivery channel preference, IB usefulness variables, IB characteristics TAM (PU, PEU), Compatibility, Trialability, Trust, Awareness, IB Adoption TAM (Behavior Intention, Attitude, PU, PEU), Compatibility, CC (Subjective Norms, Family Influence, Mass media influence), behavior control, Selfefficacy, Government support, Technology support Intention,Attitude, Subjective Norm, Ease of use, Trialability, Personal Norm, Mass Media Norm, Government Support, Relative Advantage / Compatibility, User s Informational-Based Readiness, Observability, Technology Facilitating Condition, Behavioral Control and Self-efficacy PC (Age, Income), Area, Risk,, Frequency, Type of Person, Internet Location, Constant TAM (PU, PEU, Intention, Actual Use), PC (Age, Gender), TR (Optimism, Innovativeness, Discomfort, Insecurity) Self-efficacy, Risk, Social Influence, PU, PEU, Attitude toward the use of IBS, Behavioral Intention to use of IBS, IB Behavior. Trustworthiness, Security, Privacy, Trust, Risk, Intentions, Ease of Use, Trust, Financial services involvement, Performance, Time loss, Social, Privacy, Security, Risk, IB use Online Risk, Online Benefits, Internet Banking Trust, Brand Awareness/Association, Quality, Internet banking brand equity, Brand loyalty Theoretical Basis References Theory of Hanafizadeh and Risk Khedmatgozar, 2012 TAM Sanayei et al., Mirza et al., 2009b - Mirza et al., 2009a DOI DTPB TPB and Extended TPB AL-Majali and Mat, 2011 Al-Majali and Mat, 2010 Zolait, Khan and Emmambokus, 2011 TAM Yousafzai and Yanide-Soriano, 2012 TRA and TAM Chau and Ngai, Yousafzai et al., 2009 TAM Aldás-Manzano et al., Loureiro and Miranda, 2011

10 JIBC August 2014, Vol. 19, No Country Focus of the Study Variables Poland Austria Estonia Finland USA Malaysia and more than 20 countries Vietnam and Taiwan UK and Saudi Arabia Factor underlying decision to adopt IB Role of trust in IB adoption Innovation adoption of IB Role of familiarity in consumer perception of IB Innovation Resistance of IB adoption Communication strategies with IB resistance customers Factors affect IB adoption Social factor in IB context Role of culture in technology acceptance. Factors influence customers e- payment adoption Cross-market examination about IB acceptance PC (Internet ), Security, Marketing Exposure, Use of IB Products, Type of Internet Connection used, Demographic Characteristics Familiarity Internet, Propensity to trust, Internet trust, attitude, risk internet banking Relative Advantage; Complexity; Risk; and Compatibility Convenience, Security, Status, Auxiliary features, Personal financials, Investment, Exploration, Usage barrier, value barrier, risk barrier, tradition barrier, image barrier Function barriers, Value barriers, Risk barriers, Psychological barriers, Image Barriers. PC (Internet, Internet Usage, Computer Knowledge), Incentive, Security Privacy, TAM (PU, PEU, Attitude, Behavior Intention, Actual Use) Institutional forces (coercive, normative, mimetic), PC (age, gender, income) TAM (PU, PEU, Behavioral Control, Intention to use IB), Subjective Norm, Culture Dimensions TAM (PU, PEU), Risk, Information E- Payment, Personal Innovativeness in Technology Manageability, PU, Trust, Subjective Norms, Attitudinal Intentions Theoretical Basis Regression analysis References Polasik and Wisniewski, Grabner-Kräuter and Faullant, 2008 DOI Eriksson et al., Mäenpää et al., 2008 Theory of Innovation Resistance Theory of Innovation Resistance TAM Social Contagion and Institutional theories Hofstede s dimensions and TAM Laukkanen et al., 2008 Laukkanen et al., 2009 Chan and Hosein, 2010 Shi et al., 2008 Abbesi and Haghighy, 2011 TAM Lin and Nguyen, 2011 TAM and TPB Alsajjan and Dennis, 2010 Note: PC: Personal Characteristics; TAM: Technology Acceptance Model; TPB: Theory of Planned Behavior; TRA: Theory of Reasoned Action; DTPB: Decomposed Theory of Planned Behavior; PU: ; PEU: Ease of Use; DOI: Diffusion of Innovation; UTAUT: Unified Theory of Acceptance and Use of Technology; CC: Culture Characteristics; IBS: Internet Banking Services; ATT: Attitude; TR: Technology Readiness.

11 JIBC August 2014, Vol. 19, No results. In the study by Prompattanapakdee (2009), only Personal Relationships, which was one of the two cultural factors (Personal Relationships and Peer Influence) included in the study, was found to have a significant negative effect on the actual use of PIB and this was contrary to the significant importance of both of these cultural factors in the results of the earlier study of PIB adoption in Thailand by Jaruwachirathanakul and Fink (2005). An alternative approach for examining the effects of culture on the adoption of PIB is to focus attention on cultural influences measured at the level of the individual user rather than the level of the whole society. This is achieved by measuring personality traits of the individual users (Hofstede and McCrae, 2004) and previous studies have used personality traits to predict the use of technology (Bellman, 1998). Jacques et al. (2009) found the effect of personality traits as antecedents of perceived usefulness and perceived ease of use. Personality traits have a significant role in predicting customer behavior in different cultural settings because people from the same culture have similar perceptions and opinions about PIB. One strongly supported theoretical basis for personality trait psychology is the five-factor model (FFM) taxonomy of five traits: Extraversion, Agreeableness, Conscientiousness, Neuroticism, and Openness to. Numerous empirical studies of the FFM show solid evidence for consistency by subjects to describe themselves and others, with stability throughout adult-life spans (Migliore, 2011). THEORETICAL MODEL The theoretical model is based on the findings from previous studies presented in the literature review. The theoretical model includes 15 variables as shown in Figure 1. Personal Characteristics Age Gender Level of Education Internet Internet Banking Position Description Trust Behavioral Intention Personality Traits Extraversion Agreeableness Conscientiousness Neuroticism Openness to Ease of Use Figure 1: Theoretical model.

12 JIBC August 2014, Vol. 19, No The model has 12 exogenous variables which are categorized as: Personal Characteristics (six variables); Personality Traits (five variables); and the single variable (Trust). There are three endogenous TAM variables two of which ( and Ease of Use) are intervening variables and Behavioral Intention which is the dependent variable. Table 3 describes the 34 research hypotheses associated with the direct effects among the variables in Figure 1 with references to previous studies motivated the hypothesis. Table 3: Research hypotheses Research Hypothesis H1: Age has a significant negative direct effect on H2: Age has a significant negative direct effect on Ease of Use H3: Age has a significant negative direct effect on Intention to Use PIB H4: Gender does not have a significant direct effect on H5: Gender does not have a significant direct effect on Ease of Use H6: Gender does not have a significant direct effect on Behavioral Intention H7: Level of Education has a significant positive direct effect on H8: Level of Education has a significant positive direct effect on Ease of Use H9: Level of Education has a significant positive direct effect on Behavioral Intention H10: Internet has a significant positive direct effect on H11: Internet has a significant positive direct effect on Ease of Use H12: Internet has a significant positive direct effect on Behavioral Intention H13: Internet Banking has a significant positive direct effect on H14: Internet Banking has a significant positive direct effect on Ease of Use H15: Internet Banking has a significant positive direct effect on Behavioral Intention H16: Position Description has a significant positive direct effect on H17: Position Description has a significant positive direct effect on Ease of Use H18: Position Description has a significant positive direct effect on Behavioral Intention Reference Prompattanapakdee, 2009; Yousafzai and Yani-de-Soriano, 2012 Prompattanapakdee, 2009; Yousafzai and Yani-de-Soriano, 2012 Prompattanapakdee, 2009; Nasri, 2011; Yousafzai and Yani-de-Soriano, 2012 Prompattanapakdee, 2009 Prompattanapakdee, 2009 Prompattanapakdee, 2009 Prompattanapakdee, 2009 Sattabusaya, 2008; Prompattanapakdee, 2009 Prompattanapakdee, 2009 Ozdemir and Trott, 2009; Prompattanapakdee, 2009; Chan and Hosein, 2010; Le, 2011 Prompattanapakdee, 2009; Chan and Hosein, 2010; Chau and Ngai, 2010; Le, 2011 Ozdemir and Trott, 2009; Prompattanapakdee, 2009; Nasri, 2011; Kesharwani and Tripathy, 2012 Sattabusaya, 2008; Prompattanapakdee, 2009; Chan and Hosein, 2010; Winley, 2011 Sattabusaya, 2008; Prompattanapakdee, 2009; Chan and Hosein, 2010; Winley, 2011 Sattabusaya, 2008; Prompattanapakdee, 2009; Winley, 2011 Sattabusaya, 2008; Prompattanapakdee, 2009 Sattabusaya, 2008; Prompattanapakdee, 2009 Sattabusaya, 2008; Prompattanapakdee, 2009

13 JIBC August 2014, Vol. 19, No Research Hypothesis Reference Sattabusaya, 2008; Prompattanapakdee, H19: Trust has a significant positive direct effect on 2009; Alsajjan and Dennis, 2010; Cheng and Yeung, 2010; Alnsour and Al-Hyari, 2011; Le, 2011; Winley, 2011 H20: Trust has a significant positive direct effect on Prompattanapakdee, 2009; Alnsour and Al- Ease of Use Hyari, 2011; Le, 2011; Winley, 2011 Sattabusaya, 2008; Aldás-Manzano et al., 2009; Prompattanapakdee, 2009; Alsajjan H21: Trust has a significant positive direct effect on and Dennis, 2010; Cheng and Yeung, Behavioral Intention 2010; Chong et al., 2010; Nor et al., 2010; Le, 2011; Sanayei et al., 2011; Winley, 2011 H22: Extraversion has a significant positive direct Devaraj et al., 2008; Lane and Manner, effect on 2011 H23: Extraversion has a significant positive direct Devaraj et al., 2008; Lane and Manner, effect on Behavioral Intention 2011 H24: Agreeableness has a significant positive direct Devaraj et al., 2008 effect on H25: Agreeableness has a significant positive direct Devaraj et al., 2008 effect on Behavioral Intention H26: Conscientiousness has a significant positive Devaraj et al., 2008 direct effect on Behavioral Intention H27: Conscientiousness has a significant positive Huang and Yang, 2010 direct effect on H28: Neuroticism has a significant negative direct Devaraj et al., 2008 effect on H29: Neuroticism has a significant negative direct Devaraj et al., 2008 effect on Behavioral Intention H30: Openness to has a significant Devaraj et al., 2008 positive direct effect on H31: Openness to has a significant Devaraj et al., 2008; Huang and Yang, positive direct effect on Behavioral Intention 2010 Lee, 2009; Alsajjan and Dennis, 2010; Chan and Hosein, 2010; Nor et al., 2010; Safeena et al., 2010, 2011; Chau and Ngai, H32: has significant positive 2010; Cheng and Yeung, 2010; Chong et direct effect on Behavioral Intention al., 2010; Le, 2011; Lin and Nguyen, 2011; Nasri and Charfeddine, 2012; Kesharwani and Bisht, 2012; Kesharwani and Tripathy, 2012 Chan and Hosein, 2010; Chau and Ngai, 2010; Safeena et al., 2011; Cheng and Yeung, 2010; Chong et al., 2010; Safeena H33: Ease of Use has a significant et al., 2010; Le, 2011; Lin and Nguyen, positive direct effect on Behavioral Intention 2011; Kesharwani and Bisht, 2012; Kesharwani and Tripathy, 2012; Nasri and Charfeddine, 2012 Lee, 2009; Cheng and Yeung, 2010; H34: Ease of Use has significant positive Alnsour and Al-Hyari, 2011; Le, 2011; direct effect on Kesharwani and Bisht, 2012; Nasri and Charfeddine, 2012 Note: In each case significant refers to statistical significance at a level of 0.05 or less.

14 JIBC August 2014, Vol. 19, No Table 4 presents the operational definitions and labels used for each of the model variables in Figure 1 with a reference to the source of the definition. Table 4: Definitions and labels for model variables Categories of Variables Personal Characteristics Attitudinal Trust Personality Traits TAM Variables Variable Operational Definition (Label) Age Measured in years. (A) Gender Male or female. (G) Level of Prompattanapakdee, Highest level of formal award. Education (E) Internet Number of months/years using the (IE) Internet. Internet Banking Number of months/years using PIB. (IBE) Position Description (J) Trust (T) Openness to (OE) Conscientiousn ess (CO) Extraversion (EX) Agreeableness (AG) Neuroticism (NE) Ease of Use (PE) The individual s position in their organization in terms of their responsibilities and their span of control. The assured confidence an individual has in the bank s ability to provide reliable internet banking services The degree to which an individual is intellectually curious, open to new ideas, involves imaginative and creative cognition styles. The degree to which individuals control, regulate, and direct their impulses related to decision-making and actionoriented behaviors. The degree to which an individual is assertive, action oriented, and enjoys opportunities for excitement, attention drawing, and talking. The degree to which an individual is concerned for cooperation and social harmony, and behavioral characteristics including being considerate, friendly, generous, helpful, and willing to compromise one s own interests for others. The extent to which an individual displays excessive worry that causes mental distress, emotional suffering, and an inability to cope with day-to-day life activities. The extent to which a person believes that using PIB is free of effort. Reference Prompattanapakdee, 2009 Prompattanapakdee, Prompattanapakdee, 2009 Prompattanapakdee, 2009 Prompattanapakdee, 2009 Le, 2011 Migliore, 2011 Migliore, 2011 Migliore, 2011 Migliore, 2011 Migliore, 2011 Prompattanapakdee, 2009

15 JIBC August 2014, Vol. 19, No Categories of Variables Variable (Label) (PU) Behavioral Intention (BI) Operational Definition The extent to which a person believes that using PIB enhances their banking activities. The strength of an individual s intention to use PIB. Reference Prompattanapakdee, 2009 Prompattanapakdee, 2009 Table 5 describes the type of measure used for each of the model variables, labels used for latent variables, and a reference to an existing measuring instrument that was used. The numerical measurement scales associated with each variable and indicator are shown on the notated questionnaire in Appendix A1. Table 5: Measurement of model variables Variable (Label) Gender (G) Age (A) Level of Education (E) Position Description (J) Internet (IE) Internet Banking (IBE) Type of Measure Measuring Instrument Categorical converted to single interval Prompattanapakdee, scale 2009 Latent Variables (Number of Indicators and Labels) Openness to (OE) 4 indicators: oe1, oe2, oe3, oe4 Conscientiousness (CO) 4 indicators: co1, co2, co3, co4 Donnellan et al., Extraversion (EX) 4 indicators: ex1, ex2, ex3, ex Agreeableness (AG) 4 indicators: ag1, ag2, ag3, ag4 Neuroticism (NE) 4 indicators: ne1, ne2, ne3, ne4 Behavioral Intention (IN) 3 indicators: bi1, bi2, bi3 Ease of Use (PE) 6 indicators: pe1, pe2, pe3, pe4, pe5, pe6 Prompattanapakdee, (PU) 7 indicators: pu1, pu2, pu3, pu4, pu5, 2009 pu6, pu7 Trust (T) 4 indicators: t1, t2, t3, t4 DATA PREPARATION AND PRELIMINARY ANALYSES A sample of 411 completed questionnaires was obtained. There were no missing values for any of the questions and when a randomly selected 10 percent were checked for data entry errors none were detected. Twenty questionnaires were found to include at least one outlier value and these were removed leaving a final sample of 391 questionnaires which satisfied the sample size determined for the study. Principal Component factor analysis was used to test the construct (discriminant and convergent) validity of the measures of the latent model variables. The results of the factor analysis are displayed in Appendix and show that each indicator for a latent variable has a factor loading of magnitude at least 0.4 and an associated eigenvalue of at least 1 (Straub et al., 2004).

16 JIBC August 2014, Vol. 19, No Consequently, all of the indicators demonstrated satisfactory construct validity. The internal consistency reliability of each set of indicators was tested using Cronbach alpha coefficients which are shown in Appendix Table A2. The excellent values of the alpha coefficients could not be improved by the removal of any of the indicators. T-tests were used to determine statistically significant differences (p < 0.05) between males and females among the mean values of the model variables. There were no significant differences due to Gender among the endogenous latent variables, Ease of Use, and Behavioral Intention and the only significant difference between males and females among the endogenous variables was for Age where males on average were 19 months older than females. Consequently, because it was not plausible to propose that Gender was a cause for any of the exogenous variables and it had no influence on the endogenous variables it was removed from the theoretical model. Thus, the research hypotheses H4, H5, H6 in Table 3 associated with direct effects of Gender on, Ease of Use, and Behavioral Intention were not supported. Characteristics of Respondents From Appendix Table A3 it is seen that: 91 percent of respondents is in the age range 18 to 40 years; most respondents (94 percent) have a Bachelor degree or higher; and 29 percent are professionals (e.g. teacher, lawyer, and doctor) followed by those in clerical/officer positions (25 percent). The lowest proportion of respondents is unemployed (4 percent) and the 10 percent of students indicates that young Vietnamese are willing to adopt Internet Banking. The number of male and female respondents represents 57 percent and 43 percent, respectively, and this is close to the proportions of male and female Facebook social network users (56 percent and 44 percent) in Vietnam (We Are Social, 2012). Compared with the proportions of males and females in the population of Vietnam in 2011 (49.5 and 50.5 percent) it is suggested that in Vietnam males are more likely than females to adopt technology and in particular Internet Banking ( From the Appendix Table A4 it is seen that: 92 percent of the respondents have more than two years of experience using the Internet and 69 percent have more than two years of experience using Internet Banking while 30 percent had more than five years of Internet Banking experience. The largest proportion (22 percent) had 2-3 years of Internet Banking which corresponds with the with the fact that in 2011, 45 banks were able to provide Internet Banking services (Table 1) as a result of The Second Payment System and Bank Modernization Project Services project in Vietnam in 2011 (World Bank, 2011). Descriptive Statistics Appendix Table A5 summarizes descriptive statistics for each of the model variables. From Table A5 it is seen that the magnitudes of the values for skewness and kurtosis are within the acceptable limits of 3 and 7, respectively, required for the use of maximum likelihood estimation in subsequent SEM analyses (Kline, 2005).

17 JIBC August 2014, Vol. 19, No For the latent model variables t-tests were used to examine the difference between the mean value of each indicator and the value of 4 which represented a neutral point on their 7-point Likert measurement scales. It was found that all of the model variables and indicators, except for Neuroticism, were significantly greater than the neutral value (p < 0.05). For Neuroticism the means are significantly less than the neutral value (p < 0.05). This indicated that the respondents scored highly on desirable personality traits, lowly on the undesirable trait Neuroticism, and they had very positive responses toward the PIB system and its features. Appendix Table A6 shows the correlations among model variables. Based on the statistically significant correlations (p < 0.05) among the exogenous variables it is seen that: (a) older (younger) respondents had high (low) levels of education and worked in positions with high (low) levels of responsibility; (b) those with high (low) levels of education had high (low) levels of Internet experience; and (c) individuals who worked in positions with high (low) levels of responsibility and those with a high (low) level of Internet experience had high (low) levels of experience with Internet Banking. Older (younger) individuals, those with high (low) levels of education, and those in work positions with high (low) levels of responsibility all displayed low (high) levels of neuroticism although as described above the mean value for this undesirable personality trait among all of the respondents was significantly low. Overall, these significant correlations confirmed normal expectations about the associations among these variables. Considering the significant correlations in Appendix Table A6 in relation to the causal effects proposed in the theoretical model it is observed that: (a) among the causal effects involving the five variables (Age, Level of Education, Internet, Internet Banking, and Position Description) the only significant correlation is that associated with the causal effect of Position Description on Ease of Use; (b) there are significant correlations associated with all of the causal effects involving Trust; (c) there are significant correlations associated with all of the all of the causal effects associated with the five personality traits with the exception of Openness to ; and (d) there are significant correlations associated with all of the causal effects among, Ease of Use, and Behavioral Intention. Significant (non significant) correlations do not represent significant (non significant) causal effects but they do suggest that a causal effect may be significant (non significant). In particular, it is noted that the three personality traits Extraversion, Agreeableness, and Conscientiousness are each significantly correlated with Ease of Use even though the three corresponding direct causal effects are not included in the theoretical model.

18 JIBC August 2014, Vol. 19, No MODEL ANALYSIS AND DEVELOPMENT Internet Banking.001(.001S) Internet.006(.007S).001(.002S).001(.004S.010(.019S).001(.002S).001(.002S) Level of Education.022(.038S).010(.019S) Age.003(.021S).007(.049S).005(.041S) Position Description Trust.052*(.105M).032(.060S).474***(.494M).197***(.219M).021(.042S).170***(.197M).289***(.305M) Ease of Use.491***(.544L) Extraversion.019(.025S) Agreeableness.014(.017S).074(.065S).032(.027S).277***(.266M) Conscientiousness Neuroticism Openness to.023(.028s).003(.003s) -.039(-.041S).032(.044S).038(.050S) -.095*(.154M) Behavioral Intention Figure 2 shows the results of the SEM analysis of the theoretical model using Amos software.

19 JIBC August 2014, Vol. 19, No Internet Banking.001(.001S) Internet.006(.007S).001(.002S).001(.004S.010(.019S).001(.002S).001(.002S) Level of Education.022(.038S).010(.019S) Age.003(.021S).007(.049S).005(.041S) Position Description Trust.052*(.105M).032(.060S).474***(.494M).197***(.219M).021(.042S).170***(.197M).289***(.305M) Ease of Use.491***(.544L) Extraversion.019(.025S) Agreeableness.014(.017S).074(.065S).032(.027S).277***(.266M) Conscientiousness Neuroticism Openness to.023(.028s).003(.003s) -.039(-.041S).032(.044S).038(.050S) -.095*(.154M) Behavioral Intention Figure 2: (a) the unstandardized effects are shown followed by the symbol *, ** or *** if the effect is statistically significant at a level of 0.05, 0.01, or 0.001, respectively. If no symbol is used then this indicates that the effect is not statistically significant; (b) in parentheses the standardized effect is shown with S, M, or L to indicate that the magnitude of the effect is small, medium, or large, respectively, where small corresponds to an effect of magnitude less than 0.1, medium is from magnitude 0.1 to less than 0.5, and large is a magnitude of 0.5 or more. These notations are used throughout subsequent analyses of effects. Table 6 shows the values of the range of fit statistics for the theoretical model in

20 JIBC August 2014, Vol. 19, No Internet Banking.001(.001S) Internet.006(.007S).001(.002S).001(.004S.010(.019S).001(.002S).001(.002S) Level of Education.022(.038S).010(.019S) Age.003(.021S).007(.049S).005(.041S) Position Description Trust.052*(.105M).032(.060S).474***(.494M).197***(.219M).021(.042S).170***(.197M).289***(.305M) Ease of Use.491***(.544L) Extraversion.019(.025S) Agreeableness.014(.017S).074(.065S).032(.027S).277***(.266M) Conscientiousness Neuroticism Openness to.023(.028s).003(.003s) -.039(-.041S).032(.044S).038(.050S) -.095*(.154M) Behavioral Intention Figure 2 as recommended by Kline (2005).

21 JIBC August 2014, Vol. 19, No Internet Banking.001(.001S) Internet.006(.007S).001(.002S).001(.004S.010(.019S).001(.002S).001(.002S) Level of Education.022(.038S).010(.019S) Age.003(.021S).007(.049S).005(.041S) Position Description Trust.052*(.105M).032(.060S).474***(.494M).197***(.219M).021(.042S).170***(.197M).289***(.305M) Ease of Use.491***(.544L) Extraversion.019(.025S) Agreeableness.014(.017S).074(.065S).032(.027S).277***(.266M) Conscientiousness Neuroticism Openness to.023(.028s).003(.003s) -.039(-.041S).032(.044S).038(.050S) -.095*(.154M) Behavioral Intention Figure 2: SEM analysis of the theoretical model. Table 6: Fit statistics for the theoretical model Model N N c NC (χ 2 /df) RMR GFI AGFI NFI IFI CFI RMSEA /779 = Theoretical R 2 : Ease of Use (0.32); (0.49); Behavior Intention Model (0.54) Note: R 2 is the proportion of the variance of each endogenous variable that is explained by the variables affecting it. From Table 6 it is seen that some of the fit statistics are reasonably satisfactory. However, in Figure 3 there are 23 highlighted direct effects that are small in magnitude and not statistically significant. This raises the possibility that their removal from the theoretical model may produce a simpler model with an improved set of fit statistics.

22 JIBC August 2014, Vol. 19, No Internet Banking.001(.001S) Internet.006(.007S).001(.002S).001(.004S.010(.019S).001(.002S).001(.002S) Level of Education.022(.038S).010(.019S) Age.003(.021S).007(.049S).005(.041S) Position Description Trust.052*(.105M).032(.060S).474***(.494M).197***(.219M).021(.042S).170***(.197M).289***(.305M) Ease of Use.491***(.544L) Extraversion.019(.025S) Agreeableness.014(.017S).074(.065S).032(.027S).277***(.266M) Conscientiousness Neuroticism Openness to.023(.028s).003(.003s) -.039(-.041S).032(.044S).038(.050S) -.095*(.154M) Behavioral Intention Figure 2 In section 5 from Appendix Table A6 it was noted that there were three additional plausible direct causal effects that may be included in the theoretical model each with an associated correlation that was statistically significant at a level of 0.05 or less (Extraversion Ease of Use, Agreeableness Ease of Use, and Conscientiousness Ease of Use). Also, as noted above there are 23 direct effects highlighted in Figure 2 that are considered for removal from the theoretical model.