ONLINE GAME SERVICE SATISFACTION AND PREFERENCE: AN EMPIRICAL STUDY OF VIETNAMESE ONLINE GAMING INDUSTRY In Partial Fulfillment of the Requirements of the Degree of MASTER OF BUSINESS ADMINISTRATION In International Business By Mr: Minh-Quan Nguyen ID: MBA04031 International University - Vietnam National University HCMC August 2013 ONLINE GAME SERVICE SATISFACTION AND PREFERENCE: AN EMPIRICAL STUDY OF VIETNAMESE ONLINE GAMING INDUSTRY In Partial Fulfillment of the Requirements of the Degree of MASTER OF BUSINESS ADMINISTRATION In International Business by Mr: Minh-Quan Nguyen ID: MBA04031 International University - Vietnam National University HCMC August 2013 Under the guidance and approval of the committee, and approved by all its members, this thesis has been accepted in partial fulfillment of the requirements for the degree. Approved: ---------------------------------------------Chairperson --------------------------------------------Committee member ---------------------------------------------Committee member --------------------------------------------Committee member ---------------------------------------------Committee member --------------------------------------------Committee member Acknowledge I would like to thank my supervisor, Dr. Van-Phuong Nguyen - Deputy Dean, School of Business (IU-VNU), for the patient guidance, encouragement and advice that he has provided throughout my time as his student. I have been extremely lucky to have a supervisor who cared so much about my work, and who responded to my questions and queries so promptly. Thanks to my father for the courage and everything a father could provide as support despite the distance. My heartfelt and never-ending love to Phoenix and friends who never kept away from me when I needed them most. I must appreciate the VNU Central Library for being of gigantic use to me in the process of this work. -i- Plagiarism Statements I would like to declare that, apart from the acknowledged references, this thesis either does not use language, ideas, or other original material from anyone; or has not been previously submitted to any other educational and research programs or institutions. I fully understand that any writings in this thesis contradicted to the above statement will automatically lead to the rejection from the MBA program at the International University – Vietnam National University Ho Chi Minh City. - ii - Copyright Statement This copy of the thesis has been supplied on condition that anyone who consults it is understood to recognize that its copyright rests with its author and that no quotation from the thesis and no information derived from it may be published without the author’s prior consent. © Minh-Quan Nguyen/ MBA04031/ 2011-2013 - iii - Table of Contents Chapter One - Introduction ..................................................................................................................... 1 1.1. A brief description of the research topic ...................................................................................... 1 1.2. Background .................................................................................................................................. 2 1.3. Research purpose ......................................................................................................................... 3 1.4 Research questions ........................................................................................................................ 3 Chapter Two – Literature Review ........................................................................................................... 5 2.1. Online game ................................................................................................................................. 5 2.2. Experiential value......................................................................................................................... 6 2.2.1. Utilitarian Value.................................................................................................................... 6 2.2.2. Hedonic Value ...................................................................................................................... 7 2.3. The Technology Acceptance Model (TAM) .................................................................................. 8 2.4. Transaction Cost Analysis (TCA) ................................................................................................. 10 2.4.1. Bounded rationality ............................................................................................................. 11 2.4.2. Opportunism ....................................................................................................................... 11 2.4.3. Uncertainty .......................................................................................................................... 11 2.4.4. Asset specificity .................................................................................................................. 11 2.4.5. Buying frequency ................................................................................................................ 12 2.4.6. Transaction costs ................................................................................................................. 12 2.5. Service Quality (SERVQUAL)....................................................................................................... 14 2.5.1. Tangibles: Physical facilities, equipment, and appearance of personnel. .......................... 16 2.5.2. Reliability: Ability to perform the promised service dependably and accurately. ............. 16 2.5.3. Responsiveness: Willingness to help customers and provide prompt service. .................. 16 2.5.4. Assurance: Knowledge and courtesy of employees and their ability to inspire trust and confidence..................................................................................................................................... 16 2.5.5. Empathy: Caring, individualized attention the firm provides its customers. ..................... 16 2.6. Satisfaction and Preference: ...................................................................................................... 17 2.6.1 Satisfaction ........................................................................................................................... 17 2.6.2 Preference ............................................................................................................................ 19 Chapter III - Research Methodology .................................................................................................... 20 3.1. Theoretical framework and hypotheses ..................................................................................... 20 3.1.1. Integrated model ................................................................................................................. 20 - iv - 3.1.2. Backgrounds of online game satisfaction ........................................................................... 21 3.1.3. Online game satisfaction and preference ............................................................................ 25 3.1.4. Online game satisfaction as mediator ................................................................................. 26 3.2. Research Approach .................................................................................................................... 27 3.3. Research Design......................................................................................................................... 28 3.4. Research process ........................................................................................................................ 29 3.4.1. Pretesting............................................................................................................................. 29 3.4.2. Questionnaire design ........................................................................................................... 30 3.4.3. Sample Selection and Data Collection Procedure ............................................................... 36 3.4.4. Data Analysis Method ......................................................................................................... 37 Chapter IV – Data Analysis and findings ............................................................................................. 39 4.1. Descriptive Statistics .................................................................................................................. 39 4.2. Reliability Test ........................................................................................................................... 43 4.3. Exploratory factor analysis ........................................................................................................ 49 4.3.1. EFA of Experience value (EXP) with Satisfaction and Channel Preference ...................... 52 4.3.2. EFA of Technology Acceptance Model (TAM) with Satisfaction and Channel Preference. ...................................................................................................................................................... 52 4.3.3. EFA of Transaction Cost Analysis (TCA) with Satisfaction and Channel Preference ....... 53 4.3.4. EFA of Service Quality (SERVQUAL) with Satisfaction and Channel Preference .......... 54 4.4. Confirmatory factor analysis...................................................................................................... 60 4.5. Structural Equation Modeling (SEM) ........................................................................................ 69 4.6. Mediation ................................................................................................................................... 73 Chapter V – Discussion and conclusions .............................................................................................. 75 5.1. Interpretation of Results ............................................................................................................. 75 5.2. Practical implications ................................................................................................................. 80 5.3. Limitation and further research .................................................................................................. 81 References ............................................................................................................................................. 83 Appendix ............................................................................................................................................... 92 Appendix A. Vietnamese questionnaire............................................................................................ 92 Appendix B. Confirmatory Factor Analysis (CFA) ........................................................................ 100 Appendix C. Structural equation modeling (SEM) in AMOS ........................................................ 101 -v- List of tables . 1. TABLE 3.1. CODED ITEMS OF MEASUREMENT SCALE ................................................................................................... 30 . 2. TABLE 4.1.1. GENDER .......................................................................................................................................... 39 . 3. TABLE 4.1.2. AGE ................................................................................................................................................ 40 . 4. TABLE 4.1.3. EDUCATION ...................................................................................................................................... 41 . 5. TABLE 4.1.4. INCOME........................................................................................................................................... 41 . 6. TABLE 4.1.5. ...................................................................................................................................................... 42 . 7. TABLE 4.1.6. EXPERIENCE...................................................................................................................................... 42 . 8. TABLE 4.1.8. CONNECTIVITY .................................................................................................................................. 43 . 9. TABLE 4.2.1. CRONBACH'S ALPHA OF EMP .............................................................................................................. 43 . 10. TABLE 4.2.2. CRONBACH'S ALPHA OF EQU ............................................................................................................ 44 . 11. TABLE 4.2.3. CRONBACH'S ALPHA OF USE ............................................................................................................. 44 . 12. TABLE 4.2.4. CRONBACH'S ALPHA OF UT ............................................................................................................... 45 . 13. TABLE 4.2.5. CRONBACH'S ALPHA OF HD .............................................................................................................. 45 . 14. TABLE 4.2.6. CRONBACH'S ALPHA OF UNC ............................................................................................................ 45 . 15. TABLE 4.2.7. CRONBACH'S ALPHA OF ASSET ......................................................................................................... 46 . 16. TABLE 4.2.7. CRONBACH'S ALPHA OF TIME ........................................................................................................... 46 . 17. TABLE 4.2.8. CRONBACH'S ALPHA OF REL.............................................................................................................. 47 . 18. TABLE 4.2.9. CRONBACH'S ALPHA OF RESP ........................................................................................................... 47 . 19. TABLE 4.2.10. CRONBACH'S ALPHA OF ASS ........................................................................................................... 47 . 20. TABLE 4.2.11. CRONBACH'S ALPHA OF SAT ........................................................................................................... 48 . 21. TABLE 4.2.12. CRONBACH'S ALPHA OF PREF ......................................................................................................... 48 . 22. TABLE 4.3.1. EFA OF EXP................................................................................................................................... 52 . 23. TABLE 4.3.2. EFA OF TAM ................................................................................................................................. 52 . 24. TABLE 4.3.3. EFA OF TCA .................................................................................................................................. 53 . 25. TABLE 4.3.4. EFA OF SERVQUAL ....................................................................................................................... 54 . 26. TABLE 4.3.4. THE RETAINED MEASUREMENT SCALES ................................................................................................. 55 . 27. TABLE 4.4.1: CRITERIA FOR MEASUREMENT MODEL.................................................................................................. 60 . 28. TABLE 4.4.2 GOODNESS-OF-FIT INDICES OF MEASUREMENT MODEL............................................................................. 63 . 29. TABLE 4.4.3 REGRESSION WEIGHTS: (GROUP NUMBER 1 - DEFAULT MODEL) – 1ST ROUND............................................. 64 . 30. TABLE 4.4.4. REGRESSION WEIGHTS: (GROUP NUMBER 1 - DEFAULT MODEL) - 2ND ROUND........................................... 66 . 31. TABLE 4.5.2. LIST OF REJECTED HYPOTHESES ........................................................................................................... 70 . 32. TABLE 4.5.2. REGRESSION WEIGHTS: (GROUP NUMBER 1 - DEFAULT MODEL) – FINAL ROUND ......................................... 72 . 33. TABLE 4.6.1. MEDIATION ANALYSIS....................................................................................................................... 73 . 34. TABLE 5.1.1. RESULT OF HYPOTHESES .................................................................................................................... 75 - vi - List of figures FIGURE 1. INTEGRATED MODEL .................................................................................................................................... 26 FIGURE 2. GENDER .................................................................................................................................................... 40 FIGURE 3. SECOND-ORDER CFA ................................................................................................................................... 62 FIGURE 4. CFA OF MEASUREMENT MODEL STANDARDIZED ................................................................................................ 68 FIGURE 5. STRUCTURAL EQUATION MODELING RESULT AFTER ADJUSTING THEORETICAL MODEL ................................................ 69 FIGURE 6. INTEGRATED MODEL RESULT AFTER ADJUSTING THEORETICAL MODEL ..................................................................... 69 - vii - Abstract Online game’ efficiency, content and service quality has become one of the key aspects among other factors that contribute to online game firms business growth and leading position in the business environment with mass competition. Efficiency, content and service also plays a significant role in service sectors since due to its untouchable nature the features cannot be spelled out for consumers to directly make judgment before decisions are made. In order for businesses to improve and maintain a better positioning in the competitive era, it is necessary to evaluate the performance of the services rendered to their customers. In recent times, online game companies spend a great deal of time and money in configuring high efficiency, content and service to satisfy their customers. Understanding consumer-level interaction with the online game will enhance understanding of consumer behavior, online game design issues, and drivers of consumer satisfaction with and preference for the online game. Customer satisfaction can be evaluated through an assessment of the quality of efficiency, content and service delivered by the online game provider to their customers and the level of efficiency, content and service can also be measured considering customers’ expectations and perceptions. Purpose: This study is aimed to apply the EXP, TAM, TCA, SERVQUAL instrument in assessing Vietnamese gamer perceptions of online game service and the level of satisfaction obtained from the online game services rendered by the Vietnam online game firms. - viii - Method: The convenience sampling technique was used to obtain data from the chosen population to enable an evaluation of perceptions of online game satisfaction and preference in Vietnam. Findings and Conclusion: Firstly, with regards to the relationship between customer satisfaction and customer preference, consistently with prior research, this analysis reconfirms that customer satisfaction is an important component positively affecting customer preference. Secondly, experiential values do not effect to gamer satisfaction and preference, the reason for this finding is that although online games can be entertained by gamers, the cost of playing online game currently seem is the unsolvable problem to the online game firms (Hsu & Lu, 2004). Thirdly, it is very interesting that Time does not effect to game players satisfaction. That is the public concern on video game addiction rising over past decade in Vietnam.. Hence, Vietnamese gamers do not consider Time factor as an effect Satisfaction factor because there is a large number of Vietnamese gamers who have personal issue with game addiction. Fourthly, the service quality (SERVQUAL) factors have significant effects to experiential value factor could be used to approach a new rating system on experiential value by service quality issued by online game firms. Last but not least, assurance (ASS) has a strong positive effect to satisfaction (SAT) (1.249). Hence, this study support strongly the face that the Vietnamese online gamers are very interested in assurance, especially in game items trading that issued by online game producers. Keywords: Online game, Satisfaction (SAT), Preference (PREF), Experiential Value (EXP), Technology Acceptance Model (TAM), Transaction Cost Analysis (TCA), service quality (SERVQUAL) - ix - Chapter One - Introduction Online games or networked games are being rapidly developed with the phenomenal growth of the Internet (Chen, Wang, & Lee, 2009). Due to deep penetration into the consumer market, gaming is considered a prime driver of PC technology (Hsu & Lu, 2004; Von Ahn, 2006) and currently is one of the few profitable e-commerce applications (H.-E. Yang, Wu, & Wang, 2009). With online gaming being a billion dollar industry and game companies making revenue from subscription charges (Adams, 2010), the presence of gamer satisfaction and preference issues is becoming more evident. 1.1. A brief description of the research topic This thesis argue that there will have some factors affecting online game service satisfaction and preference. The purpose of this study is to understand an online game service model. This model contains several dimensions including experiential value (H.-E. Yang et al., 2009), the Technology Acceptance Model (TAM) (Davis, 1989), Transaction Cost Analysis (Williamson, 1987) and Service Quality (Parasuraman, Zeithaml, & Berry, 1988), the four antecedents of online game service satisfaction and preference and test the associations among the constructs in it. After surveying some online game players in Ho Chi Minh City (Vietnam), this study found that four antecedents could have significant and positive effects on online game service satisfaction and which, in turn, significantly affect online preference. Especially, service quality has the relatively higher total positive effects on both online game satisfaction and preference. Meanwhile, online game satisfaction completely mediates the effects of these antecedents on online preference. The findings imply that how to manage online game service quality better, provide more -1- acceptable transaction cost, and offer more experiential value are the key ways for effectively enhancing players’ satisfaction with the online game service in order to retain their preference to the online game service system. 1.2. Background The “Online Games Market in Vietnam” report provides an in-depth analysis of the Vietnam online games market and this report forecasts the number of users playing online games market in Vietnam to exceed 10 million by 2011, driven by rising incomes, increasing PC and Internet penetration rates, and a large population of youth that are actively seeking out entertainment content. These findings are contained in the business intelligence and consulting the report “Online Games Market in Vietnam”. There are more than 50 online games in the market; a notable achievement given that the online games market just emerged in 2004. Other notable trends include the emergence of locally developed titles, aimed specifically at Vietnamese gamers. Key findings: + Approximately 50% of the total Vietnamese population is under the age of 25. This is an age range that is known for being tech savvy, making them a high priority demographic for digital entertainment companies. There are approximately 21 million Internet users in Vietnam with an Internet penetration rate of 24%. + In our interviews with Vietnamese gamers, many were spending 60,000 to 100,000 VND ($3 - $6) per month. In one high-end Internet café we visited, a few interviewees were spending an average of 500,000 VND ($31) per month. These consumers are driving the digital entertainment and online games market with virtual item purchases. Top online games in Vietnam can attract 200,000 users. The Internet -2- cafes that the researchers visited in Vietnam were consistently crowded with users playing online games. While Vietnam has a rapidly emerging games market, critical challenges include government regulations on online games, the worldwide slowing economy, developing infrastructure, and low income levels. 1.3. Research purpose Structural equation model analyses indicate that metrics tested through each model provide a statistically significant explanation of the variation in the online gamers' satisfaction and channel preference. There are several studies found that TAM components—perceived ease of use and usefulness—are important in forming consumer attitudes and satisfaction with the online game. Ease of use also was found to be a significant determinant of satisfaction in TCA. The study tries to find empirical support for the assurance dimension of SERVQUAL as determinant in online game satisfaction. Further, the study also verified the general support for consumer satisfaction as a determinant of channel preference. Because to keep competitiveness of online game industry is hard to hard, this study has ambition that its model can provide online game corporate to select and adopt the key point what the online game corporate should choose and how to affect the key factors of online game service satisfaction and online preference in online game industry. 1.4 Research questions The literatures revealed that the immediate factor affecting consumers to retain preference to the providers is customer satisfaction (Devraj, Fan, & Kohli, 2002). Similarly, in a B2C channel satisfaction model or online shopping satisfaction model, satisfaction is considered as an important construct because it affects participants’ -3- motivation to stay with the channel and regarded as an antecedent of repurchase (Heilman, Bowman, & Wright, 2000). But what are the key factors that can make them satisfied with the products and services, which, in turn, enhance their preference, especially, in online game service environment, are still under study (Hsu & Lu, 2004; G. Kim et al., 2013; Von Ahn, 2006). The thesis helps us to answer four main questions:  What are the key factors that make the customer playing online game?  Why do the gamers play online game?  Do SERVQUAL, TAM, TCA and EXP influence the customer satisfaction in online game service?  Does gamer satisfaction influence the gamer preference in online game service? -4- Chapter Two – Literature Review 2.1. Online game Online gaming has grown from a tiny fraction of the interactive entertainment business into a major market in its own right. In this chapter, this thesis learn about some of the features and design challenges that set online gaming apart from the more traditional single-player or multiplayer local games (Adams, 2010). Online gaming is a technology rather than a genre, a mechanism for connecting players together rather than a particular pattern of gameplay (Adams, 2010; Von Ahn, 2006; H.-E. Yang et al., 2009; X. Yang, 2013). Therefore, this thesis does not look for design commonalities as the chapters on game genres did. Instead, it addresses some of the design considerations peculiar to online games no matter what genre those games belong to. It’s a huge topic, however, and there is only room in this thesis for the highlights. Do not confuse online gaming, as Von Ahn (2006) uses the term, with online gambling or online casino gaming. Online gambling is a different industry, and is not covered here. This thesis uses the term online games to refer to multiplayer distributed games in which the players’ machines are connected by a network (Dick, Wellnitz, & Wolf, 2005; Hsu & Lu, 2004; G. Kim et al., 2013; X. Yang, 2013). This is as opposed to multiplayer local games in which all the players play on one machine and look at the same screen (Adams, 2010). While online games can, in principle, include solitaire games that happen to be provided via the Internet, such as Bejeweled, the online aspect of solitaire games is incidental rather than essential to the experience. Bejeweled is simply a puzzle game. Online games do not need to be distributed over the Internet; games played over a local area network (LAN) also qualify as online games (Adams, 2010). -5- 2.2. Experiential value Previous study discovered that experiential value can be created by traditional and electronic shopping experiences (Babin, Darden, & Griffin, 1994; Hirschman & Holbrook, 1982; Irani & Hanzaee, 2011; O’Brien, 2010). Experiential value can indeed produce both utilitarian and hedonic value when shopping at old-style and electronic store (Babin et al., 1994; Fiore, Kim, & Lee, 2005; Wolfinbarger & Gilly, 2001; H.-E. Yang et al., 2009). 2.2.1. Utilitarian Value Early research on shopping value commonly emphasis on the utilitarian feature of shopping (Bloch & Bruce, 1984). Hirschman and Holbrook (1982) have identified that in traditional information processing shopping model, the shopper is a rational decision creator wanting to make best use of utility by concentrating on tangible benefits of the product. As stated in this model, acquiring has been regarded as a problem solving action in which consumer moves through a chain of logical stages. Especially, utilitarian shopper behavior is explained through task-related and rational behavior (Batra & Ahtola, 1991; Kempf, 1999). Also, Hirschman (1984) asserted that all shopping experiences involve the motivation of thoughts and intellects. Shopping experiences viewed as a manner that provides the cognitive (utilitarian) and affective (hedonic) benefits. More specifically, tangible attributes of goods and services provide contribution to cognitive process and is closely related to calculations of utilitarian value. Thus, a consumer receives utilitarian shopping value when he or she obtains the needed product, and this value rises as the consumer obtains the product more smoothly (Babin et al., 1994). Perceived utilitarian shopping value might rely on whether how much of the consumption need that prompts the shopping experience is met (Seo & Lee, 2008), -6- regularly, this means that the consumer purchases goods in a calculated and efficient manner (Hirschman & Holbrook, 1982). Hence, utilitarian purchasing behavior is more logical, rational, related to transactions (Batra & Ahtola, 1991; Seo & Lee, 2008), and related with more information gathering compared to hedonic buying manners (Bloch & Bruce, 1984). Koufaris, Kambil, and LaBarbera (2002) suggested that utilitarian value include time saving, control, better product information granted by interactivity affecting attitude responses toward a product or website. 2.2.2. Hedonic Value Bloch and Bruce (1984) defined that customers gain hedonic value as well as utilitarian value during the shopping experience. Research about shopping has extensive emphasize the shopping experience on the utilitarian aspects, which has often been described as task-related and rational and linked thoroughly to whether a product acquisition task was successfully done or not (Babin et al., 1994; Batra & Ahtola, 1991). Nevertheless, traditional product gaining explanations may not fully describe the entirety of the shopping experience (Bloch & Bruce, 1984). Due to this, over the past several years have seen rising attention in the shopping experience on the hedonic features and academics have accepted the significance of its potential entertainment and emotional value (Babin et al., 1994). Also, hedonic value is more particular and personal than utilitarian value. That is, customers were described as either ‘‘problem solvers” or in terms of customers looking for ‘‘fun, fantasy, arousal, sensory stimulation, and enjoyment” (Hirschman & Holbrook, 1982). Utilitarian value includes shopping efficiency and making the correct product choice based on logical valuation of product information, separately customers observe shopping as an adventure. Moreover, the hedonic value of the consumption experience was assumed as these forms of pleasure, hence, hedonic value was different from utilitarian value -7- (Hirschman, 1984). Thus, MacInnis and Price (1987) said that hedonic value can be assumed as the emotional paybacks to the consumer perceives through the shopping experience other than the accomplishment of the original buying intent. In a similar context, summarized characteristics of goods and services can contribute to affective elements in shopping and are closely related to hedonic value (Cottet, Lichtlé, & Plichon, 2006). Westbrook and Black (2002) recommended that shopping satisfaction contains the opportunity for social interactions with friends, family or even strangers and the sensual incentive such as escapes from routine life, and new information about future styles and fashion. Hedonic value acts both positive and negative roles in consumption regarding consumers’ benefit. The negatively extreme form of hedonic value is impulse or compulsive (uncontrollable) purchase. Especially, ROOKH (1987) believed impulse buyers buy products from a need to buy rather than a need for a product. Also, uncontrollable buyers put their values on shopping activity itself rather than a product (Faber & O'guinn, 1992). Next, hedonic value is usually expressed by the entertaining features of store surfing whether or not an acquisition happens (Bloch & Bruce, 1984). 2.3. The Technology Acceptance Model (TAM) TAM, offered by Davis (1989) in his doctoral thesis, is an information system (IS) theory that models how individuals come to accept and use a technology (“Technology acceptance model,” n.d.). That is, TAM forecasts intention to use and acceptance of information systems by individuals. It included perceived ease of use (PEOU) and perceived usefulness (PU) that drive the approach toward an IS. TAM explains why the attitude, in turn, leads to one’s intention to use an IS and how the -8- eventual acceptance of the IS technology is affected by system design structures (Davis, 1993). “Intention” in TAM represents the effect of social norms and attitudes that can be mediated by other variables (Ajzen & Fishbein, 1980). Moore and Benbasat (1991) claimed that if no other variables interfere, as is the case in many information technology applications, then “intention” can be omitted. As a result, recent TAMbased studies have omitted “intention” without any loss of information (Lederer, Maupin, Sena, & Zhuang, 2000; Straub, Limayem, & Karahanna-Evaristo, 1995; Venkatesh, 2000). Agreeing with previous studies (Karahanna & Straub, 1999; Venkatesh, 2000; H.-E. Yang et al., 2009), my thesis setting does not contain of a circumstances that other variables interfered and therefore the use of “intention” is not necessary. However, the “attitude” is believed that captured in a number of items for channel preference. Current outcomes propose that consumer satisfaction in the online atmosphere is considerably higher than traditional channels due to the ease of use in gaining information (Shankar, Smith, & Rangaswamy, 2003). Ease of use can also influence the transaction costs when the ease of use relates to information search. Venkatesh (2000) suggests that both perceived usefulness and perceived ease of use are found to directly influence behavioral intention to use IT, which leads to eliminate the need for the attitude construct from the model. Perceived usefulness was defined as “the degree to which a person believes that using a particular system would enhance his/her job performance”, and perceived ease of use is defined as “the degree to which a person believes that using a particular system would be free of physical and mental effort” (Davis, 1989). TAM is better suited to IS because it supports to understand the technology acceptance based on ease of use and usefulness (Devraj et al., 2002). Hence, based on -9- recent research in IS, this thesis apply TAM concepts to observe consumer satisfaction with the online game. 2.4. Transaction Cost Analysis (TCA) TCA (Coase, 1937; Williamson, 1987) was firstly developed to examine the appropriate governance structures or mechanisms for firms to conduct transactions. TCA belonged to the “New institutional economics” (NIE), which studied the role of governance structure in decreasing transaction cost (P. K. Rao, 2003). A transaction happens when a good or service is shifted across a technologically separable interface (Williamson, 1987). Classical economists developed a theory that presumes perfect information symmetry in the efficient market and the transaction can be executed without costs. Nonetheless, markets are often inefficient, and this results in costs to firms in their transactions with suppliers and customers (Coase, 1937). For example, because of the absence of information about supplier’s completeness, a firm draft and negotiate a contract in order to defend their interest in the transaction. In contrast, the lack of information about customers’ credit position causes the firm to search for such information (Li & 李仲文, 2008). In an inefficient market, firms have to obtain transaction costs to safeguard a favorable deal (Coase, 1937). One of the priori acceptances of the TCA background is that market governance is more efficient than hierarchical firm governance structure because of the benefits of competition, and such acceptance is often called the production cost advantages of the market (Coase, 1937). The TCA structure builds on the relationship between three observable dimensions of transaction as proposed by Williamson (1987): asset specificity, uncertainty (environmental and behavioral) and frequency, and the two main assumptions of human behavior: bounded rationality and - 10 - opportunism. The interplay between these behavioral assumptions and transaction dimensions are elaborated as follows. 2.4.1. Bounded rationality Transactions are supposed to happen under individuals’ bounded rationality, which implies that individuals are constrained by their basic cognitive capabilities (e.g. limited short term memory, slow rate of information processing) to be perfectly rational (Simon, 1955). 2.4.2. Opportunism Excluding individual’s bounded rationality, TCA structure also proposes that agents participate in a transaction possibly will act opportunistically if given the chance. That is, Williamson (1985) defines opportunism as self- interest seeking behavior concerning planned efforts to mislead and confuse exchange partners. Opportunism ascends because buyer has imperfect control seller who has personal interests in an exchange, and the contracts between the two parties take part in an exchange are vague and incomplete (E. Anderson, 2008). 2.4.3. Uncertainty The “uncertainty” is the inability to predict relevant contingencies from two sources—unpredictable changes and information asymmetry resulting from strategic nondisclosure or distortion of information by the sellers (Masten, Meehan, & Snyder, 1991; Williamson, 1985). Also, Devraj et al. (2002) suggested that uncertainty reflects the extension to which good product and price information was provided through the online channel in context of online shopping relationship. 2.4.4. Asset specificity Opportunism poses a transactional hazard to the range that a relationship is bordered by behavioral uncertainty, and the situation is getting worse when the - 11 - relationship is supported by asset specificity (Rindfleisch & Heide, 1997). Asset specificity refers to durable investments that are undertaken in support of particular transactions; and these specific investments represent sunk costs that have much lower value outside of these specific transactions (Williamson, 1985). Namely, this makes it difficult for the buyer as well as the supplier to switch. Within context of online shopping relationship, Devraj et al. (2002) measures asset specificity by degree of variety of product and store choices that is available through online shopping. 2.4.5. Buying frequency Along with the two main transactional measurements previously defined, the complete TCA outline also contains frequency as the third measurement of transaction. The frequency dimension refers strictly to buyer activity in the market. Recurrent transactions enable economies of scale in regard to transaction costs because these costs will be easier to recover for numerous transactions of a recurring kind (Williamson, 1985). Frequency refers to the recurring nature of the transactions. But frequency has received only limited attention in empirical TCA construct (Rindfleisch & Heide, 1997). In this thesis, the research application will focus upon the analysis based on uncertainty and asset specificity. 2.4.6. Transaction costs Transaction costs for retail market organizations such as online stores consist of (1) market transaction costs for searching, bargaining, and after-sale activities and (2) managerial transaction costs to run a store (Devraj et al., 2002). In addition, the market transaction costs measure the efficiency level of the interactions of buyers and sellers during a particular market setting, while the managerial transaction costs measure the process efficiency in market organizations (Devraj et al., 2002). - 12 - 2.4.6.1. Market transaction costs In the context of market transaction costs, as a potential consumer seeks to make an online acquisition, the site may offer the product image, description, price, and feedback from other customers, in an easy-to-read format (Devraj et al., 2002; H.E. Yang et al., 2009). Also, in information system usage, S. S. Kim and Malhotra (2005) found users’ perceived ease of use and usefulness of a website at an early period positively affect their perceived ease of use and usefulness at a later period, when there is no new piece of information that changes the users’ opinions. Hence, market transaction costs basically are captured with two constructs to measure the benefits to the market: perceived ease of use (PEOU) and time efficiency. PEOU, also a TAM construct, measures the effort in shopping including searching, bargaining, and after-sale monitoring. Consequently, market transaction costs are captured with time efficiency. Time efficiency is a measure of the transaction time costs. Time efficiency is a measure of the transaction time costs. The pioneering work of Becker (1965) in consumer behavior suggests that the consumer maximizes his or her utility subject to not only income constraints but also time constraints (Dellaert, Arentze, Bierlaire, Borgers, & Timmermans, 1998). By reducing information asymmetry and surprises, such as delivering wrong products and missing delivery dates, customers find online shopping easy to use and less time consuming (Devraj et al., 2002). 2.4.6.2. Managerial transaction costs In the context of the managerial transaction costs, price savings can be considered as a measure of store efficiency because as managerial costs decrease, savings could be passed on to consumers. In the finance literature, the transaction costs of financial markets generally include commission fees, bid-ask spread, and price impact costs (Berkowitz, Logue, & Noser, 1988; Devraj et al., 2002; H.-E. Yang et al., 2009). These costs are the compensation to market makers or dealers and are - 13 - considered as a measure of market efficiency. As market institutions become more efficient, the cost of trading is lowered and consumers get better prices (Devraj et al., 2002; H.-E. Yang et al., 2009). In conclusion, transaction costs include three dimensions: PEOU, time efficiency and price saving. While PEOU and time efficiency are measures of the costs between buyer and seller interactions, relative price saving is a measure of online or conventional store transaction efficiency. Thus TCA extends TAM constructs to the cost dimension of online transactions (Devraj et al., 2002). 2.5. Service Quality (SERVQUAL) Parasuraman et al. (1988) recommended the most difference between service and goods are four characteristic: intangibility, perishability, heterogeneity, and inseparability. Therefore, the focus has different between service marketing and products marketing. Because Parasuraman et al. (1988) indicate the fact that quality is an “elusive and indistinct construct” commenced on an exploratory article that has revolutionized research and given service quality a face value in a way (Ndamnsa, 2013). SERVQUAL is a mechanism used to measure quality that sprouts from this model and works with the differences in the gaps scores derived from a questionnaire. The SERVQUAL scale (questionnaire) has two sections; one designed to record client expectation and the other to measure client perception in relation to a service segment and a service firm (Parasuraman et al., 1988). The SERVQUAL model is important to managers in service firms as it enables them to appreciate the sources of problems in quality and how they can resolve or improve on these problems (Nair, Ranjith, Bose, & Shri, 2010). Moutinho and Goode (1995) assumed that the model has conquered the service quality literature as it was founded and it is a real-world means with an - 14 - affluence of value to the industry. Studies approved that the model is a reliable apparatus in measuring service quality and has appreciated theoretical involvement in this respect. The model is also observed to be a practical and a good forecaster for service quality measurement (Sureshchandar, Rajendran, & Anantharaman, 2002). However, no model can be said to be faultless and SERVQUAL model is not an exemption and thus has its boundaries as this is justified in a study by Brown, Churchill Jr, and Peter (1993), who highlighted low reliability of the scoring recorded. Teas (1993) proclaimed that the SERVQUAL model is a tricky one as the respondents might be unable to differentiate between the different types of expectations. Nevertheless the critics of the SERVQUAL model, it is still widely and continuously used in numerous sectors. This thesis will be reexamined in detail that is an examination of the different dimensions of the model will be presented in detail. Parasuraman et al. (1988) in trying to develop the model for measuring “service quality in retail banking, maintenance, phone repairs and security brokerage” firms realized that there were core differences regarding executive perceptions of service quality and the responsibilities involves when delivering services to consumers. (Parasuraman et al., 1988) defined that service quality is a global judgment, or attitude, relating to the superiority of the service, and superiority is the gap which practical service higher than consumer expectation. When expectative service level is equal to perceived service level then it has general service quality. When perceived service level is higher than expectative service level then it has better service quality. When perceived service level is lower than expectative service level then it has worse service quality (Devraj et al., 2002; H.-E. Yang et al., 2009). - 15 - Parasuraman et al. (1988) introduce ten dimensions to measure service quality and suggest that it can be used in any service model. Previous studies use factor analysis to simplify twenty-two items to five dimensions, called SERVQUAL (Service Quality) (Devraj et al., 2002; Parasuraman et al., 1988; H.-E. Yang et al., 2009), listed as follow: 2.5.1. Tangibles: Physical facilities, equipment, and appearance of personnel. 2.5.2. Reliability: Ability to perform the promised service dependably and accurately. 2.5.3. Responsiveness: Willingness to help customers and provide prompt service. 2.5.4. Assurance: Knowledge and courtesy of employees and their ability to inspire trust and confidence. 2.5.5. Empathy: Caring, individualized attention the firm provides its customers. In electronic commerce, service quality measures have been applied to assess the quality of search engines and factors associated with Web site success. However, consumers’ perceptions of online service quality remain unexplored. There are indications that electronic commerce service issues go beyond product price and may be the reason for consumers’ preference for the channel (Devraj et al., 2002; X. Yang, 2013). SERVQUAL, a widely utilized instrument in marketing research to measure customers’ expectation and perception of service, was recently adapted to measure IS service quality. This thesis uses four dimensions of SERVQUAL, which include reliability, responsiveness, assurance, and empathy, to measure the users’ cognition of SERVQUAL in online channel. - 16 - 2.6. Satisfaction and Preference: 2.6.1 Satisfaction Business objectives are not only to deliver products or services to customers, not only aimed at selling what they are offering but also to deliver the needs to customers in a satisfactorily. Companies and organizations with an in-depth understanding of how to satisfy customers and deliver satisfaction are in a better position to increase profitability than those that might be aware of customers‟ needs but are unable to deliver them to a satisfactory level (Chen et al., 2009; Moutinho & Goode, 1995; Wu, Tao, Li, Yang, & Huang, 2011). Customer satisfaction is considered as an evaluation of the after- purchase perceptions and the pre-purchase expectations (Ndamnsa, 2013). Customer satisfaction is a practical and theoretical aspect that is important for both researchers in consumers realm and marketers in general (Irani & Hanzaee, 2011). According to (R. E. Anderson & Srinivasan, 2003), customer satisfaction is an important subject for organizations that desire to create and maintain competitive advantage in today business competitive world. Customers who are satisfied will probably inform others about their satisfactory experiences and consequently participate in sharing their experience through positive word-of-mouth (Chen et al., 2009; Shankar et al., 2003). Meuter, Ostrom, Roundtree, and Bitner (2000) customer satisfaction is defined as a decision or conclusion that a customer develops after the act of acquisition and consumption of a product /service. Other studies pointed out that customer satisfaction is affected by expectations (Moutinho & Goode, 1995). A customer is term satisfied when the outcome of performance is greater than expectations and termed satisfied when expectations exceed the outcome of performance this is simply referred to as positive and negative disconfirmation - 17 - respectively (Meuter et al., 2000; Moutinho & Goode, 1995; Ndamnsa, 2013). Customer satisfaction is described as the consumer’s perceptive appraisal of a sensitive feedback in accordance to his/her observation of whether the characteristic of the acquired offering meets his/her expectation (R. E. Anderson & Srinivasan, 2003; Chen et al., 2009). Customer satisfaction is an essential tool for survival in the business environment, the prime objective of a business is to create and maintain customer satisfaction in an optimum level (Devraj et al., 2002; Shankar et al., 2003). The concept of customer satisfaction is growing in everyday with different ideas and different definitions. Researchers have looked at customers’ satisfaction in different ways. Many arguments have been made on the aspect of customer satisfaction with majority pointing out the fact that customer satisfaction is based on experience encountered with service provider and the outcome of service rendered (Parasuraman et al., 1988). Today, gamers have greater and easier access to alternative sources to purchase products and services. However, to continue to use the online game firms websites, the gamers must believe that the online game firms offer better choices than the alternatives (H.-E. Yang et al., 2009; X. Yang, 2013). In the marketing literature, consumers’ channel-choice behavior is studied in the service outputs model (Dick et al., 2005; Hsu & Lu, 2004; H.-E. Yang et al., 2009), which argues that channel systems exist and remain viable by performing duties and providing benefits to endusers. For online game firms, spatial convenience of purchasing the products at home, reducing online waiting time and delivery time, increased product choice, and customer service are critical service outputs (Dick et al., 2005; Hsu & Lu, 2004). Overall, consumers’ satisfaction is affected by both economic and noneconomic - 18 - factors (Dick et al., 2005; Hsu & Lu, 2004; G. Kim et al., 2013; H.-E. Yang et al., 2009). When consumers find online shopping convenient, time efficient, and price saving, they will be satisfied with the general effectiveness and efficiency of the electronic channel. In addition, consumers will find the purchase experience gratifying if online vendors are responsive, concerned, and reliable during the shopping process and subsequent interactions with the customers (Devraj et al., 2002; H.-E. Yang et al., 2009). Therefore, we examine game online satisfaction based on an integrated analysis using multiple constructs of technology acceptance, transaction costs, and service quality and experimental value. 2.6.2 Preference While satisfaction is an attitude construct that affects customer’s behavioral intention, channel preference is a consumer behavior choice resulting from prior experience (Devraj et al., 2002; Heilman et al., 2000), and consumer preferences vary with the purchasing experience (Heilman et al., 2000). When consumers enter a new market, they generally show little evidence of product preference. As they gather more information for a product and with increased purchasing experience, the probability of their choosing a particular product increases. Consumers’ store-choice behavior is also largely affected by their most recent purchase experience (Aaker & Jones, 1971; P. K. Rao, 2003; T. R. Rao, 1969). In the online environment, as the overall satisfaction with the online channel increases, it is likely that consumers will use the online channel again (Bhattacherjee, 2001; Devraj et al., 2002). - 19 - Chapter III - Research Methodology This chapter discusses and justifies the author’s choices for the methodological approaches employed in the research. Starting from the discussion regarding theoretical framework, hypotheses, research approach, choices for research design, data sources, research process, data collection method, data collection instrument, sampling, data analysis method, and quality criteria are presented with the reasons behind the choices. 3.1. Theoretical framework and hypotheses 3.1.1. Integrated model An increasing electronic commerce research work has been done on the antecedents and consequences of consumer online satisfaction by adopting constructs from different theoretical frameworks in order to identify the key antecedents and explain the effects of them on the consequences in the model (Chen et al., 2009; Devraj et al., 2002; E.-J. Lee, Uniremidy, & Overby, 2004; K. Lee & Joshi, 2007). However, no such integrated models are employed to investigate online game player’s satisfaction with and preference to the game service system provided. An integrated theoretical framework, including the constructs proposed by technology acceptance model (TAM), transaction cost analysis (TCA), and service quality (SERVQUAL), was developed and validated by Devraj et al. (2002) and they demonstrated that metrics derived from traditional models in marketing, economics, and psychology can be successfully applied in e-commerce to determine customer preference. On the other hand, experience value was defined as the perception which game players gained in the past process playing (Hsu & Lu, 2004). This construct is divided into two dimensions, utilitarian value and hedonic value. Utilitarian value is defined as the best choices which do by ‘‘logical evaluation” reference to games’ - 20 - efficiency and content (Babin et al., 1994; Batra & Ahtola, 1991; O’Brien, 2010). Hedonic value is defined that to find fun, fantasy, arousal, sensory stimulation, and enjoyment is the interest for the people (Babin et al., 1994; Batra & Ahtola, 1991). From previous discusses and due to the importance of experiential value of the game service, this thesis combines the constructs for experiential values with TAM, TCA and SERVQUAL in order to assess online game service more precisely and completely. 3.1.2. Backgrounds of online game satisfaction Previous studies considered that overall satisfaction is primarily a function of perceived service quality (R. E. Anderson & Srinivasan, 2003; Chen et al., 2009; Shankar et al., 2003) and service quality is strongly related to online satisfaction (Moutinho & Goode, 1995; Sureshchandar et al., 2002; H.-e. Yang & Tsai, 2007). Recent researches have included additional constructs as the antecedents of customer satisfaction in the online satisfaction and preference model (Devraj et al., 2002; H.-e. Yang & Tsai, 2007; H.-E. Yang et al., 2009). In addition to the set of constructs suggested by the technology acceptance model (TAM), the constructs used in the transaction-cost approach (TCA) (Coase, 1937; Williamson, 1975, 1985, 1987), that is, perceived ease of use, time efficiency, and price savings, the three dimensions have been used to measure different aspects of the efficiency of online transactions and explained a large portion of customer satisfaction with Internet-based services (Devraj et al., 2002). However, according to the research conducted in Vietnam and due to the low transparency of corporate governance (McGee, 2009) and the corporate tax system in Vietnam (Chan, Whalley, & Ghosh, 1999), the price saving will be dropped. Hence, TCA will consist only two dimensions which are perceived ease of use and time efficiency. - 21 - The five dimensions of SERVQUAL: tangibility, reliability, responsiveness, assurance, and empathy, for assessing service quality, have been adapted to evaluate information system (IS) service quality recently and prior studies also indicated that SERVQUAL is appropriate for measuring IS service quality (Bhattacherjee, 2001; Brown et al., 1993; Chen et al., 2009; Cottet et al., 2006; K. Lee & Joshi, 2007; Nair et al., 2010; Ndamnsa, 2013; Wu et al., 2011; Yoshida & James, 2010). For measuring customer- perceived service quality of websites, (H.-E. Yang et al., 2009) refined and validated the current SERVQUAL and IS-SERVQUAL instruments and the results indicated that the tangibility dimension is less relevant to the e-commerce service quality and completely excluded from the model. This thesis follows the conclusion and use four dimensions: reliability, responsiveness, empathy, and assurance of online channel in this thesis (Devraj et al., 2002; H.-E. Yang et al., 2009). There is insignificant research that validated the factors anteceding to online satisfaction (R. E. Anderson & Srinivasan, 2003; Chen et al., 2009; Yoshida & James, 2010). When only considering quality as the antecedent of online satisfaction, the findings indicate that the path coefficients from system quality and service quality to online satisfaction are significantly positive (H.-E. Yang et al., 2009). When integrating the metrics of TCA, TAM, and SERVQUAL in a model, the results reveal that TAM and TCA dominate the impact on satisfaction (Devraj et al., 2002). When replacing TAM with website technology (WEBST) in the integrated model, the results show that the standardized path coefficients from WEBST, TCA, and SERVQUAL to online satisfaction are all positively significant (H.-e. Yang & Tsai, 2007). There is a research that claimed that the factors empathetic perception of emotional (aka. empathy) have significant effects to hedonic value (Zillmann, Mody, - 22 - & Cantor, 1974). Thus, this thesis will suppose that the experiential value factors have relationship with service quality factors. Based on the previous review of the relationships between the online satisfaction and the constructs from experiential value, transaction cost, TAM and SERVQUAL, we suggest that the following hypotheses can be posited in an online game service context: o H1a: Online game satisfaction (SAT) will be positively affected by utilitarian value in experiential value (EXP) o H1b: Online game satisfaction (SAT) will be positively affected by hedonic value in experiential value (EXP) o H2a: Online game satisfaction (SAT) will be positively affected by time saving (TIME) in transaction cost (TCA) o H2a1: Time saving (TIME) will be positively affected by uncertainty in transaction cost (TCA) o H2a2: Time saving (TIME) will be positively affected by asset specificity in transaction cost (TCA) o H2b: Online game satisfaction (SAT) will be positively affected by perceived ease of use (EQU) in transaction cost (TCA) o H2b1: Perceived ease of use (EQU) will be positively affected by uncertainty in transaction cost (TCA) o H2b2: Perceived ease of use (EQU) will be positively affected by asset specificity in transaction cost (TCA) o H3a: Online game satisfaction (SAT) will be positively affected by assurance (ASS) in service quality (SERVQUAL) - 23 - o H3a1: Assurance (ASS) in service quality (SERVQUAL) will be positively affected by utilitarian value in experiential value (EXP) o H3a2: Assurance (ASS) in service quality (SERVQUAL) will be positively affected by hedonic value in experiential value (EXP) o H3b: Online game satisfaction (SAT) will be positively affected by empathy (EMP) in service quality (SERVQUAL) o H3b1: Empathy (EMP) in service quality (SERVQUAL) will be positively affected by utilitarian value in experiential value (EXP) o H3b2: Empathy (EMP) in service quality (SERVQUAL) will be positively affected by hedonic value in experiential value (EXP) o H3c: Online game satisfaction (SAT) will be positively affected by reliability (REL) in service quality (SERVQUAL) o H3c1: Reliability (REL) in service quality (SERVQUAL) will be positively affected by utilitarian value in experiential value (EXP) o H3c2: Reliability (REL) in service quality (SERVQUAL) will be positively affected by hedonic value in experiential value (EXP) o H3d: Online game satisfaction (SAT) will be positively affected by responsiveness (RESP) in service quality (SERVQUAL) o H3d1: Responsiveness (RESP) in service quality (SERVQUAL) will be positively affected by utilitarian value in experiential value (EXP) o H3d2: Responsiveness (RESP) in service quality (SERVQUAL) will be positively affected by hedonic value in experiential value (EXP) o H4a: There is a positive influence of perceived ease of use (EQU) on the perceived usefulness (USE) in TAM model - 24 - o H4b: There is a positive influence of perceived usefulness (USE) on the satisfaction (SAT) in the online gaming context 3.1.3. Online game satisfaction and preference In recent marketing research, the measures of perceived quality, satisfaction, and preference on behalf of customers have been used to assess firm’s productivity and its marketing performance (Devraj et al., 2002; Heilman et al., 2000). Consumer satisfaction has been the subject of much attention in the literature because of its potential influence on consumer behavioral intention and customer retention (Heilman et al., 2000). Similarly, in a B2C channel satisfaction model, customer satisfaction is considered as an important construct because it affects participants’ motivation to stay with the channel (Devraj et al., 2002; Heilman et al., 2000). Satisfaction with a product or service offered has been identified as a key determinant for preference (Devraj et al., 2002; Heilman et al., 2000). This relationship would seem to be applicable to Internet e-commerce (R. E. Anderson & Srinivasan, 2003; Shankar et al., 2003). Past studies found that online customer preference results from customer’s satisfaction with the EC channel and that the positive impact of online satisfaction on preference is evidenced in the context of electronic commerce (R. E. Anderson & Srinivasan, 2003; Chen et al., 2009; Devraj et al., 2002; Shankar et al., 2003; H.-e. Yang & Tsai, 2007; H.-E. Yang et al., 2009). From the review of the past research, it is presumable that high online game satisfaction will yield high online game intentions and preference. Therefore, the following research hypotheses will be tested: o H5: Online game satisfaction (SAT) will have a positive impact on online game preference (PREF). - 25 - Figure 1. Integrated model Utilitarian Value Hedonic Value Perceived usefulness H4a H 1a Perceived ease of use Satisfaction H2b Preference H5 a H2 Time H3a2 d1 H3 Asset specificity H2a2 H4 b H3b H2 b1 H 2a 2 H 2b 2 Uncertainty b H1 H3a H3b1 Assurance H3b2 H3a1 Emphathy H 3c 2 H 3c H3c1 Reliability H 3d 2 H3 d Responsiveness 3.1.4. Online game satisfaction as mediator Combining the constructs discussed above, this thesis proposed the online game satisfaction and preference model, presented in Fig. 1, including constructs from experiential value, TCA, and SERVQUAL with the hypotheses of this study on the paths. According to the model, experiential value, transaction cost, and service quality, the four antecedents influence on online game satisfaction, which, in turn, affects online game preference. In this model, online game satisfaction is a function of the fours antecedents operating in a situation and helps to explain the influence of the antecedents on online preference. Although, it has attracted researchers to pay attention to the formal tests of the mediation effects (R. E. Anderson & Srinivasan, 2003; Chen et al., 2009; Devraj et al., 2002; Shankar et al., 2003; H.-e. Yang & Tsai, 2007; H.-E. Yang et al., 2009), but, to my knowledge, rare research examines the mediating effects of customer satisfaction in an integrated preference model or behavioral intentions model, much less formally tests that of online game satisfaction in the online game service environment (Devraj et al., 2002; H.-E. Yang et al., 2009). Therefore the mediating effects of online game satisfaction when the mediational - 26 - model involves latent constructs will be tested formally and the following hypothesis is posited: o H6: Online game satisfaction (SAT) will mediate the effects of the antecedents (EXP, TCA, TAM and SERVQUAL) on the online game preference (PREF). 3.2. Research Approach Ali and Birley (1999) argue that the term qualitative has no clear meaning and it can be rather explained as a term, which covers various techniques. They also state that in the use of qualitative research method, researchers try to describe, decode, and translate reality through participation (Fisher & Buglear, 2010). Therefore, the main focus is on respondents and their opinions and reactions. Thus research usually begins with questions and observations of the world and then moves to more generalized and abstract ideas (Fisher & Buglear, 2010). On the other hand, quantitative research method concerns more about actual numbers, such as frequency of occurrence, test score, or even rental costs (Ali & Birley, 1999; Fisher & Buglear, 2010). This thesis is will solely be using the quantitative research approach. In order to achieve the purpose of this research, to assess satisfaction and preference in online game of consumers in Vietnam, this thesis have based this research on integrated model created by Devraj et al. (2002) and H.-E. Yang et al. (2009). This model intended to quantitatively assess consumers’ attitudes toward online game through TCA, TAM, SERVQUAL and EXP. Considering the given nature of the model, therefore, quantitative research approach would be the most suitable approach for this case. - 27 - 3.3. Research Design Research design helps a researcher to form an appropriate design for the chosen subject and the purpose of his study. It soothes the operation of the study and is to ensure the researcher to be able to collect empirical data through his study that is necessary to meet the purpose and to answer the research question (Dhawan, 2010). There are three main types of research design; exploratory design, descriptive design, and causal design (Dhawan, 2010; Jones, Wahba, & Van der Heijden, 2008), In exploratory research design, the main purpose of the study often lies in more exact problem formulation. Thus, the emphasis for this type of research is in finding ideas and insights (Dhawan, 2010). If the study employs descriptive design, it tries to describe the characteristics of the subject to study. In this type of research design, a researcher needs to have a clear definition of his subject to study, and the study aims to gather complete data to picture the subject (Dhawan, 2010; Jones et al., 2008). Lastly, if a study takes research design of hypotheses testing, it tries to see the fundamental relationships between variables in the study and to explain if one variable causes the value of another. This type of study enables the researchers to have reduced bias, increased reliability for their research, and description of causality (Jones et al., 2008). Since this study aims to observe and obtain deeper understanding of attitudes toward online satisfaction of consumers in online gaming industry, the main interest of the study is to picture the consumers’ attitudes based on their satisfaction. In addition, this thesis conducts an intensive literature review to get insights for the study from already existing studies. Thus, exploratory and descriptive research design fits the best for the purpose of the study. - 28 - 3.4. Research process Operationalization can be described as a process of defining vague concepts in order to make the concept measurable in form of variables composing of specific observation (Bryman & Bell, 2007). They also mention steps required for successful operationalization: Theoretical insights  Define key variables  Provide operational definition of key variables  Find and list potential measures for key variables  Pretest  Design data collection instrument (Bryman & Bell, 2007; Jones et al., 2008). 3.4.1. Pretesting When collecting data through questionnaires, researchers need to conduct a pretest in order to refine a questionnaire that they are going to use. By doing so, they will be able to assure that respondents will understand the questionnaire in the way that researchers intended to and there will be no problem in recording acquired data (Jones et al., 2008). In addition, it also helps researchers to have some assessment of questions’ validity and reliability of the data (Jones et al., 2008). For the research, a series of pretests were conducted before the online questionnaires were carried out. The procedure can be summarized into two steps. First of all, the author has tested the English version of questionnaire on a senior lecturer at International University in Ho Chi Minh City, Vietnam. The primary reasons for asking a senior lecturer was to make sure that questions used were appropriate, understandable, and, well reflecting their operationalization of the concepts used (Dhawan, 2010). Second, the translated versions of questionnaire were tested on randomly chosen 16 Vietnamese gamers. The main focus of the second pretest was to make sure that all questions were understandable to anybody, as it was assumed that the levels of respondents’ background knowledge of the research topic - 29 - would vary to some extent (Dhawan, 2010). On that account, the questions’ validity and reliability, especially wording and phrasing in Vietnamese, were carefully confirmed through the second pretest. 3.4.2. Questionnaire design The study requires quantitative by conducting questionnaire survey. The questionnaire consists of 2 parts: 1st part: In this part, respondents have to evaluate statements related to their satisfaction and No table of figures entries found. preference towards playing online game. In addition, their agreement/disagreement with a statement is based on a sevenpoint Likert-type scale with anchors ranging from “1: extremely disagree” to “7: extremely agree”. 2nd part: This part asks respondents about their personal information, which place that they play game and which type of Internet connectivity that they have used. . 1. Table 3.1. Coded items of measurement scale Item Perceived Coded Description of statement References EQU1 My objective for using online game (Hsu & Lu, 2004; firm website is clear and Moon & Kim, understandable. 2001; Porter & Using online game firm website does Donthu, 2006) EQU2 not require a lot of mental effort. ease of use EQU3 I believe that it is easy to do what I want to do while using online game firm website. - 30 - EQU4 Online game firm website is easy to use. USE1 USE2 I feel using online game firm website (Devraj et al., gives me greater control over my 2002; Hsu & Lu, game account. 2004; Moon & Using online game firm website Kim, 2001; improves the quality of decision Porter & Donthu, making. 2006; X. Yang, Using online game firm website is a 2013) Perceive USE3 usefulness more effective way to make purchases online game. USE 4 I find online game firm website to be useful. USE 5 I feel using online game firm website makes it easier to play game. UT1 UT 2 Utilitarian UT 3 value UT 4 Playing online game does not require (Babin et al., a lot of mental effort. 1994; Batra & It is easy for me to become skillful at Ahtola, 1991; playing on-line game. Hirschman & Learning to play an on-line game is Holbrook, 1982; easy for me. Irani & Hanzaee, It enables me to satisfy the purpose of 2011; Kempf, playing game easier. 1999; O’Brien, 2010; Zillmann et al., 1974) - 31 - HD1 HD2 Hedonic HD3 value HD4 HD5 Compared to other things I could have (Babin et al., done, the playing experience at online 1994; Batra & game was truly a joy and comfort. Ahtola, 1991; Playing online game stimulated my Hirschman & fantasy ability. Holbrook, 1982; Online gameplay (game content and Irani & Hanzaee, tasks) let me felt a sense of adventure. 2011; Kempf, I liked the enjoy socializing with 1999; O’Brien, others when I play online game. 2010; Zillmann et I enjoyed being immersed in exciting al., 1974) new online game. HD6 While playing online game, I was able to forget my problems. UNC1 UNC2 UNC3 It was easy for me to get relevant (E. quantitative information (price, taxes 2008; Berkowitz etc.) et al., 1988; Li & I believe that it was possible for me to 李仲文, 2008; P. evaluate the trailer of online game. K. The online game firm Anderson, Rao, 2003; website Rindfleisch & Uncertainty provided adequate information such as Heide, 1997; the product image, description, price, Williamson, and feedback from other customers, in 1987; H.-E. Yang an easy-to-read format. UNC4 et al., 2009) The online game firm website provides sufficient information for the - 32 - game. ASSET1 ASSET2 ASSET3 There are many online game websites (E. Anderson, where online games are available. 2008; Berkowitz There are many online game websites et al., 1988; Li & where online game can be 李仲文, 2008; P. downloaded. K. Rao, 2003; I was satisfied with the number of Rindfleisch & websites where I could download and Heide, 1997; play online game. Williamson, Online game firms give me a wider 1987; H.-E. Yang choice of different website compared et al., 2009) Asset Specificity ASSET4 to conventional game store. ASSET5 Online game firm website offer me a wider range of product choices compared to shopping at conventional game stores. TIME1 TIME2 Time TIME3 TIME4 TIME5 Online game firm website helps me (E. Anderson, accomplish tasks more quickly. 2008; Berkowitz I did not have to spend too much time et al., 1988; P. K. to complete the transaction. Rao, 2003; I did not have to spend too much Rindfleisch & effort to complete the transaction. Heide, 1997; Purchasing the service via the online Williamson, game website seems to be easy. 1987; H.-E. Yang I can save time for purchase of game et al., 2009) - 33 - online service via the online game company website. TIME6 Purchasing online game via the website seems to require little effort. TIME7 Purchasing online game via the online game company seems to be difficult. REL1 REL2 Reliability REL3 REL4 RESP1 RESP2 I believe that online game that I play (Bhattacherjee, is reliable. 2001; Brown et I believe that what I ask for is what I al., 1993; Chen et get in playing online game. al., 2009; Cottet I think that the online game service I et al., 2006; purchased from performs the service Meuter et al., right. 2000; Nair et al., I trust the online game company to 2010; Seo & Lee, deliver the product on time. 2008) I believe the online game company is (Chen et al., responsive to my needs. 2009; K. Lee & In the case of any problem, I think the Joshi, 2007; Nair online game firm will give me prompt et al., 2010; service. Parasuraman et The customer service team at the al., 1988; Wu et online game firm will address any al., 2011; concerns that I have. Yoshida & Responsiveness RESP3 James, 2010) - 34 - EMP1 EMP2 The online game firm (Ndamnsa, 2013; remembers/recognizes me as a repeat Seo & Lee, 2008; customer (after the first time). Sureshchandar et I think online game firm can address al., 2002) the specific needs of each customer. Empathy EMP3 I was satisfied with the payment options (e.g., different credit cards) at the online game firm website that I shopped. ASS1 ASS2 Assurance I felt confident about the online game (Brown et al., purchase decision. 1993; Chen et al., I feel safe in my transactions with the 2009; K. Lee & online game firm website. Joshi, 2007; Meuter et al., ASS3 SAT1 SAT2 The online game firm had answers to 2000; Yoshida & all my questions about the product. James, 2010) The gameplay (game content and (R. E. Anderson tasks) met my needs. & Srinivasan, The online game website information 2003; Chen et al., met my needs. 2009; Devraj et It was possible for me to download al., 2002; E.-J. online game that I choose easily. Lee et al., 2004; I really enjoyed myself at this online Shankar et al., game. 2003; Shieh & Overall, I was satisfied with the online Cheng, 2007; SATISFAC SAT3 -TION SAT4 SAT5 - 35 - game experience. Sureshchandar et al., 2002) PREF1 PREF2 PREFERN- PREF3 CE PREF4 I continued to play online game, not (E.-J. Lee et al., because I had to, but because I wanted 2004; K. Lee & to. Joshi, 2007; I strongly recommend online game to Meuter et al., others. 2000; Moutinho I do not intend to switch to play other & Goode, 1995; video games. Ndamnsa, 2013; I intend to increase my use of online Wu et al., 2011; game in the future H.-e. Yang & Tsai, 2007; H.-E. Yang et al., 2009) 3.4.3. Sample Selection and Data Collection Procedure For both qualitative and quantitative research method, there is no definite answer when it comes to the sample size. Rather, it is depending on a number of considerations, time, and costs (Bryman & Bell, 2007). Therefore, to make a right decision about the sample size, researchers need to take these considerations into account. When deciding the sample size, the author had looked at some of the previous research conducted by Devraj et al. (2002) and Shankar et al. (2003) because the questioner used in this research was based on the questionnaires used in previous studies. In above-mentioned studies, researchers used paper-based surveys and personal interviews to acquire the data. However, because of some limitations, beside - 36 - paper-based surveys and personal interviews, the author had decided to distribute the questionnaires via ‘Google document’. There are mainly two reasons why it was chosen for this specific study. Comparing to paper survey, first, online questionnaires are able to reduce the amount of time required for conducting questionnaires and actually collecting data. Since the time limitation was one of critical concerns that author faced, this factor was taken into consideration. Second, it enables researchers to have an access to remotely located respondents. The questionnaires were posted on Vietnamese forums and Facebook pages that are related to online games. Additionally, the questionnaires were sent out to Vietnamese students at International University and University of Agriculture and Forestry in Ho Chi Minh City with the help of university personnel. Besides, the questionnaires were sent out to several game/net stations around Thu Duc District and District 2 with an incentive to complete the tasks of the study and offset part of the purchase cost, each participant who completed the online as well as paper-based surveys was provided with a gift certificate of 20 thousand VND. The gift certificates were redeemable at only their game/net stations. In this way, the questionnaire was open to public for 20 days. In the end, 692 respondents answered the questionnaires (658 gamers and 34 non-gamers). The details of the respondents will be discussed in the later chapter. 3.4.4. Data Analysis Method The author is going to analyze the data, using statistical software, SPSS (Statistical Package for the Social Sciences). SPSS is one of the most commonly used software to conduct quantitative analyses, which is available to researchers (Bryman & Bell, 2007; Jones et al., 2008). An analysis using SPSS takes a several steps: data coding, data entry, descriptive statistics, Cronbach’s Alpha test, Exploratory Factor - 37 - component Analysis (EFA), Confirmatory Factor Analysis (CFA), Structural Equation Modeling (SEM) and Bootstrap Test. The author was to follow the steps to display the data and complete their analysis on the data they collect through the survey. - 38 - Chapter IV – Data Analysis and findings In this chapter, the result of survey is presented. The result is presented in accordance with the data analysis method presented in chapter three. Descriptive statistics reveal the general picture of the result, and Cronbach’s Alpha test, Exploratory Factor component Analysis (EFA), Confirmatory Factor Analysis (CFA), Structural Equation Modeling (SEM) and Bootstrap Test follow to show details of the survey result. The result of hypothesis follows after the presentation of the above in the last section of this chapter. 4.1. Descriptive Statistics Analyzed data based on data collecting from 658 valid samples. Results are explained in detail as below: The percentage and frequency of respondents’ gender are shown in the table 4.1. Based on the table, number of male respondents (509 people) outweighs female respondents (149 people). . 2. Table 4.1.1. Gender Gender Frequency Percent Valid Percent Cumulative Percent Valid Male 509 77.4 77.4 77.4 Female 149 22.6 22.6 100.0 Total 658 100.0 100.0 - 39 - Figure 2. Gender Besides, the majority of respondents are in range which is from 15 to 20 years old (49.5%). People under 15 years old follows behind with 23.7% and people of other age ranges which are from 20 to 30, from 30 to 40 and above 40 years old account for 12.8%, 12.3% and 1.7% respectively. . 3. Table 4.1.2. Age Age Frequency Percent Valid Percent Cumulative Percent Valid Under 15 156 23.7 23.7 23.7 From 15 -20 326 49.5 49.5 73.3 From 20 - 30 84 12.8 12.8 86.0 From 30 - 40 81 12.3 12.3 98.3 Above 40 11 1.7 1.7 100.0 - 40 - Total 658 100.0 100.0 In terms of education, high school student or less is highest proportion in number of respondents with 39.1% and is followed by college students, university students and graduate students which make up 28.6%, 7.8% and 2.3% correspondingly. . 4. Table 4.1.3. Education Education Frequency Percent Valid Percent Cumulative Percent Valid High school or less 257 39.1 39.1 39.1 Vocational school 147 22.3 22.3 61.4 Some college 188 28.6 28.6 90.0 Bachelor’s degree 51 7.8 7.8 97.7 Graduate degree 15 2.3 2.3 100.0 658 100.0 100.0 Total According to Table 4.1.4, most respondents from survey earn under 4 million VND (38.6%) and from 8 to 12 million VND (30.2%) per month. This is reasonable because majority of respondents are students who just get money from their parents and just earn money from simple part-time or full-time jobs in their student period. Respondents who earn higher income are people who are working officially in different professions or have graduate education. . 5. Table 4.1.4. Income Income Frequency Percent Valid Percent Cumulative Percent Under 4 million VND Valid From 4 - under 8 million VND 254 38.6 38.6 38.6 155 23.6 23.6 62.2 - 41 - From 8 - under 12 million VND 199 30.2 30.2 92.4 50 7.6 7.6 100.0 658 100.0 100.0 From 12 - under 16 million VND Total . 6. Table 4.1.5. Place Frequency Percent Valid Percent Cumulative Percent Home 475 72.2 72.2 72.2 School 65 9.9 9.9 82.1 Net/Game Station 84 12.8 12.8 94.8 Coffee shop 22 3.3 3.3 98.2 Others 12 1.8 1.8 100.0 658 100.0 100.0 Valid Total The surveyed players, on average, seemed to be significant with less than 1 year experience (32.5%); 1 to 2 year experience (26.9%) and more than 3 year experience (34.2). On the contrary, from 2 to 3 year experience is lightly occurred. . 7. Table 4.1.6. Experience Experience Frequency Percent Valid Percent Cumulative Percent Valid Under 1 year 214 32.5 32.5 32.5 1–2 years 177 26.9 26.9 59.4 2–3 years 42 6.4 6.4 65.8 Above 3 years 225 34.2 34.2 100.0 Total 658 100.0 100.0 In term of place of playing game and Internet connectivity, it is clearly that most gamers play online game at home (72.2%) and they use ADSL as an Internet connectivity technology considerably (59.0%). That is, ADSL currently is the most popular technology on net service in Vietnam. In addition, net station and school use - 42 - LAN infrastructure (23.3%) can be explained by the percent of gamers who spend the most their time in school or net/game stations for playing online game. It can be interpreted by the socializing demands among the gamers. So, they prefer playing online game with others in public. . 8. Table 4.1.8. Connectivity Connectivity Frequency Percent Valid Percent Cumulative Percent Valid ADSL 388 59.0 59.0 59.0 Dial-up 32 4.9 4.9 63.8 Cable modem 47 7.1 7.1 71.0 153 23.3 23.3 94.2 Leased line 23 3.5 3.5 97.7 Others 15 2.3 2.3 100.0 658 100.0 100.0 LAN Total 4.2. Reliability Test As it has been explained in chapter three, reliability of the scale in this study is measured by Cronbach’s alpha using SPSS. As a result of the reliability test, we obtained the value of Cronbach’s Alpha as the tables on the next page shows. Using the bottom line of 0.6, as it has been discussed in the methodology chapter, all constructs in the research model except “Empathy” held acceptable levels of reliability. Rounding off to two decimal places, Good for Economy and Materialism construct keeps levels of reliability which can be considered as acceptable. The result of reliability test for Empathy factor, however, shows a significantly low value of Cronbach’s alpha. As the alpha is used to test the internal reliability, this result denotes a low internal reliability of Empathy scale (Bryman & Bell, 2007). . 9. Table 4.2.1. Cronbach's Alpha of EMP Reliability Statistics Cronbach's N of Items Alpha .482 3 Item-Total Statistics - 43 - Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted EMP1 9.18 10.919 .112 .693 EMP2 8.96 7.573 .472 .058 EMP3 8.62 9.028 .364 .279 Once the reliability for the scale was assured, the summated values for each factor, except for Empathy factor, are calculated for Cronbach’s alpha, and it is summarized in the list of tables below: . 10. Table 4.2.2. Cronbach's Alpha of EQU Reliability Statistics Cronbach's N of Items Alpha .896 4 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted EQU1 14.41 24.514 .769 .866 EQU2 14.28 24.888 .805 .852 EQU3 14.09 25.489 .756 .870 EQU4 14.12 26.885 .748 .873 . 11. Table 4.2.3. Cronbach's Alpha of USE Reliability Statistics Cronbach's N of Items Alpha .877 5 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted USE1 18.77 40.082 .677 .858 USE2 18.74 37.733 .705 .851 USE3 18.92 37.639 .750 .840 USE4 19.11 37.139 .736 .844 - 44 - USE5 19.09 38.769 .671 .860 . 12. Table 4.2.4. Cronbach's Alpha of UT Reliability Statistics Cronbach's N of Items Alpha .897 4 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted UT1 14.38 25.519 .746 .878 UT2 14.26 25.497 .806 .855 UT3 14.29 26.244 .782 .864 UT4 14.12 26.857 .756 .873 . 13. Table 4.2.5. Cronbach's Alpha of HD Reliability Statistics Cronbach's N of Items Alpha .913 6 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted HD1 24.25 60.786 .752 .898 HD2 24.36 59.395 .816 .889 HD3 24.29 60.609 .781 .894 HD4 24.21 61.049 .763 .897 HD5 24.24 60.072 .788 .893 HD6 24.36 62.911 .643 .914 . 14. Table 4.2.6. Cronbach's Alpha of UNC Reliability Statistics Cronbach's N of Items Alpha .875 4 - 45 - Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted UNC1 14.42 21.517 .742 .836 UNC2 14.26 21.591 .741 .836 UNC3 14.25 22.253 .744 .835 UNC4 14.36 22.775 .700 .852 . 15. Table 4.2.7. Cronbach's Alpha of ASSET Reliability Statistics Cronbach's N of Items Alpha .884 5 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted ASSET1 19.28 35.604 .740 .855 ASSET2 19.12 37.948 .757 .850 ASSET3 19.07 38.181 .750 .852 ASSET4 19.12 39.102 .706 .862 ASSET5 19.02 39.564 .656 .873 . 16. Table 4.2.7. Cronbach's Alpha of TIME Reliability Statistics Cronbach's N of Items Alpha .932 7 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted TIME1 29.55 85.206 .765 .923 TIME2 29.65 86.787 .775 .921 TIME3 29.52 86.232 .802 .919 TIME4 29.55 88.123 .755 .923 - 46 - TIME5 29.40 84.307 .810 .918 TIME6 29.45 87.493 .766 .922 TIME7 29.65 85.544 .790 .920 . 17. Table 4.2.8. Cronbach's Alpha of REL Reliability Statistics Cronbach's N of Items Alpha .882 4 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted REL1 14.35 26.346 .664 .879 REL2 14.64 22.671 .785 .834 REL3 14.56 25.474 .776 .838 REL4 14.42 25.595 .764 .842 . 18. Table 4.2.9. Cronbach's Alpha of RESP Reliability Statistics Cronbach's N of Items Alpha .851 3 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted RESP1 9.61 11.440 .694 .817 RESP2 9.71 10.735 .740 .773 RESP3 9.72 10.964 .729 .785 . 19. Table 4.2.10. Cronbach's Alpha of ASS Reliability Statistics Cronbach's N of Items Alpha .870 3 - 47 - Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted ASS1 9.87 10.598 .760 .811 ASS2 9.71 10.679 .729 .838 ASS3 9.64 10.350 .767 .803 . 20. Table 4.2.11. Cronbach's Alpha of SAT Reliability Statistics Cronbach's N of Items Alpha .916 5 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted SAT1 18.30 44.876 .733 .908 SAT2 18.41 45.255 .803 .893 SAT3 18.38 45.514 .805 .893 SAT4 18.46 45.786 .763 .901 SAT5 18.49 43.291 .821 .889 . 21. Table 4.2.12. Cronbach's Alpha of PREF Reliability Statistics Cronbach's N of Items Alpha .894 4 Item-Total Statistics Scale Mean if Scale Variance Corrected Item- Cronbach's Item Deleted if Item Deleted Total Alpha if Item Correlation Deleted PREF1 14.19 24.713 .731 .876 PREF2 14.25 23.255 .786 .856 PREF3 14.39 22.833 .818 .843 PREF4 14.35 24.333 .729 .877 - 48 - From table 4.2.2 to 4.2.12 indicate that all scales have alpha coefficients greater than 0.6. So keep all variables are satisfactory at this time. The final result of Cronbach’s Alpha was presented from table 4.2.2 to 4.2.12 with all the corrected item-total correlation higher than 0.3 and the Cronbach’s Alpha if item deleted lower than the Cronbach’s Alpha. In conclusion, after reliability test, the measurement scales only exclude Empathy factor because this factor does not qualified. Follow the above result; all variables would be used for Exploratory Factor Analysis. 4.3. Exploratory factor analysis Many scientific studies are featured by the fact that “numerous variables are used to characterize objects” (Williams, Brown, & Onsman, 2012). Examples are studies in which questionnaires are used that consist of a lot of questions (variables), and studies in which mental ability is tested via several subtests, like verbal skills tests, logical reasoning ability tests, etcetera (Bryman & Bell, 2007). Because of these big numbers of variables that are into play, the study can become rather complicated. Besides, it could well be that some of the variables measure different aspects of a same underlying variable. For situations such as these, (exploratory) factor analysis has been invented. Factor analysis attempts to bring intercorrelated variables together under more general, underlying variables (Trọng & Ngọc, 2005). More specifically, the goal of factor analysis is to reduce the dimensionality of the original space and to give an interpretation to the new space, spanned by a reduced number of new dimensions which are supposed to underlie the old ones (Blunch, 2012), or to explain the variance in the observed variables in terms of underlying latent factors” (Hancock & Mueller, 2006). Thus, factor analysis offers not only the possibility of gaining a clear view of the data, but also the possibility of using the output in subsequent analyses (Byrne, 2009). - 49 - In EFA, The Kaiser-Meyer-Olkin (KMO) and Barlett’s test are two of important criteria. The aim of KMO measure of sampling adequacy is to test whether the partial correlations among variables are small. Bartlett's test of sphericity tests whether the correlation matrix is an identity matrix which demonstrating that the factor model is inappropriate. In Bartlett's test, the null hypothesis Ho means that there is no correlation between observed variables. Ho is rejected if significance level is lower than 0.05, which means that correlations between variables are statistically significant. Only if KMO is in range of 0.5 and 1, the factor analysis will be only conducted. Therefore, the minimum level at which factor analysis can be conducted is 0.5 (Blunch, 2012; Byrne, 2009). In addition, the Barlett’s test of sphericity must obtain the significant level which less than 0.05 (Trọng & Ngọc, 2005). In terms of factor loading, the scales with factor loadings of 0.2 or greater are considered very sufficient due to the sample size is 658 (Bryman & Bell, 2007; Trọng & Ngọc, 2005). Therefore, in this study, only variables which have factor loadings greater than 0.2 are retained. Moreover, according to Blunch (2012), total variance explained should be greater than 50%. In this thesis, all factors which qualified with reliability test will be given of the use of factor analysis. This will be done by carrying out a factor analysis on data from a study in the field of applied linguistics, using SPSS for Windows. For this to be understandable, however, it is necessary to discuss the theory behind factor analysis. When EFA factor analysis, this study used the method to extract the main components Principal Axis Factoring (PAF) with Promax rotation, and stops when the factors extracted eigenvalues greater than 1. Research models include four EFA of Experience value (EXP), Technology Acceptance Model (TAM); Transaction Cost - 50 - Analysis (TCA); Service Quality (SERVQUAL) each pair with Satisfaction and Channel Preference due to the fact that this model have twelve independent variables measure the results of work with a total of fifty seven observations and combine four previous models. There are two basic requirements to factor analysis: sample size and the strength of the relationship of the measures. The sample size of 658 is over the 300 recommended by Blunch (2012) and is sufficient. Blunch (2012) also caution that a matrix that is factorable should include correlations in excess of .20 because the sample size of 658 is huge. If none are found, reconsider use of factor analysis. Common method variance (CMB) can be a potential source of bias in survey research. One of the procedures used to test for evidence suggesting the presence, or absence of common method bias in a data set is the Harman’s one-factor test (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). A Harman's single factor test tests to see if the majority of the variance can be explained by a single factor. To do this, constrain the number of factors extracted in your EFA to be just one (rather than extracting via eigenvalues). Then examine the unrotated solution. If a single factor is obtained or if one factor accounts for a majority of the covariance in the independent and criterion variables, then the threat of common method bias is high (Devraj et al., 2002; Podsakoff et al., 2003; H.-E. Yang et al., 2009). However, factor analysis, done by combining the independent and dependent variables, did not indicate a singlefactor structure that explained significant covariance, suggesting that common method bias is not a cause for concern in our sample. - 51 - After three rounds of EFA, the factor analysis the following results: 4.3.1. EFA of Experience value (EXP) with Satisfaction and Channel Preference . 22. Table 4.3.1. EFA of EXP KMO and Bartlett's Test Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .966 Approx. Chi-Square Bartlett's Test of Sphericity 8400.363 df 105 Sig. .000 Pattern Matrix a Component 1 2 3 4 UT1 .979 UT2 .640 UT3 .536 UT4 .472 HD1 .523 HD2 .666 HD3 .835 HD4 .750 HD5 .893 . SAT1 .314 SAT3 .839 SAT5 .880 . PREF2 .283 PREF3 .295 PREF4 .640 Extraction Method: Principal Component Analysis. Rotation Method: Promax with Kaiser Normalization. a. Rotation converged in 12 iterations. 4.3.2. EFA of Technology Acceptance Model (TAM) with Satisfaction and Channel Preference. . 23. Table 4.3.2. EFA of TAM KMO and Bartlett's Test Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .953 - 52 - Approx. Chi-Square Bartlett's Test of Sphericity 6331.006 df 66 Sig. .000 Pattern Matrix a Component 1 2 3 4 EQU1 .953 EQU2 .408 EQU4 .227 USE1 .539 USE2 .644 USE3 .953 . SAT1 .875 SAT3 .786 SAT5 .528 PREF2 .756 PREF3 .846 PREF4 .806 Extraction Method: Principal Component Analysis. Rotation Method: Promax with Kaiser Normalization. a. Rotation converged in 6 iterations. 4.3.3. EFA of Transaction Cost Analysis (TCA) with Satisfaction and Channel Preference . 24. Table 4.3.3. EFA of TCA KMO and Bartlett's Test Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .969 Approx. Chi-Square Bartlett's Test of Sphericity 11112.664 df 190 Sig. .000 Pattern Matrix a Component 1 2 3 4 5 6 EQU1 .857 EQU2 .578 EQU4 .330 SAT1 .760 - 53 - SAT3 .805 SAT5 .882 PREF2 .659 PREF3 .621 PREF4 .740 UNC1 .374 UNC3 .901 UNC4 .784 ASSET3 .533 ASSET4 .237 ASSET5 .705 TIME1 .762 TIME2 .582 TIME3 .794 TIME5 .880 TIME7 .745 Extraction Method: Principal Component Analysis. Rotation Method: Promax with Kaiser Normalization. a. Rotation converged in 7 iterations. 4.3.4. EFA of Service Quality (SERVQUAL) with Satisfaction and Channel Preference . 25. Table 4.3.4. EFA of SERVQUAL KMO and Bartlett's Test Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .967 Approx. Chi-Square Bartlett's Test of Sphericity 8311.833 df 105 Sig. .000 Pattern Matrix a Component 1 2 3 4 5 SAT1 .913 SAT3 .870 SAT5 .584 PREF2 .731 PREF3 .724 PREF4 .969 REL2 .685 REL3 .847 REL4 .922 RESP1 .537 - 54 - RESP2 .711 RESP3 1.036 ASS1 .409 ASS2 .672 ASS3 .635 Extraction Method: Principal Component Analysis. Rotation Method: Promax with Kaiser Normalization. a. Rotation converged in 7 iterations. According to table 4.3.1 to table 4.3.4, there are some observations that are dropped because does not meet the criteria: KMO is in range of 0.5 and 1; the Barlett’s test of sphericity must obtain the significant level which less than 0.05; factor loading > 0.2 and if "cross-loadings" do exist (variable loads on multiple factors), then the cross-loadings should differ by more than 0.2 (Blunch, 2012; Byrne, 2009). To sum up, 35 variables or items were retained in the final measurement after conducting EFA. It was shown below: . 26. Table 4.3.4. The retained measurement scales Item Coded Description of statement References EQU1 My objective for using online game (Hsu & Lu, 2004; firm website is clear and Moon & Kim, understandable. 2001; Porter & Using online game firm website does Donthu, 2006) Perceived EQU2 ease of use not require a lot of mental effort. EQU4 Online game firm website is easy to use. USE1 Perceive usefulness USE2 I feel using online game firm website (Devraj et al., gives me greater control over my 2002; Hsu & Lu, game account. 2004; Moon & Using online game firm website Kim, 2001; improves the quality of decision Porter & Donthu, making. 2006; X. Yang, - 55 - USE3 Using online game firm website is a 2013) more effective way to make purchases online game. UT1 UT 2 Utilitarian UT 3 value UT 4 Playing online game does not require (Babin et al., a lot of mental effort. 1994; Batra & It is easy for me to become skillful at Ahtola, 1991; playing on-line game. Hirschman & Learning to play an on-line game is Holbrook, 1982; easy for me. Irani & Hanzaee, It enables me to satisfy the purpose of 2011; Kempf, playing game easier. 1999; O’Brien, 2010; Zillmann et al., 1974) HD1 HD2 Hedonic value HD3 HD4 HD5 Compared to other things I could have (Babin et al., done, the playing experience at online 1994; Batra & game was truly a joy and comfort. Ahtola, 1991; Playing online game stimulated my Hirschman & fantasy ability. Holbrook, 1982; Online gameplay (game content and Irani & Hanzaee, tasks) let me felt a sense of adventure. 2011; Kempf, I liked the enjoy socializing with 1999; O’Brien, others when I play online game. 2010; Zillmann et I enjoyed being immersed in exciting al., 1974) new online game. - 56 - UNC1 It was easy for me to get relevant (E. Anderson, quantitative information (price, taxes 2008; Berkowitz etc.) et al., 1988; Li & 李仲文, 2008; P. UNC3 The online game firm website K. Rao, 2003; provided adequate information such as Rindfleisch & Uncertainty the product image, description, price, Heide, 1997; and feedback from other customers, in Williamson, UNC4 an easy-to-read format. 1987; H.-E. Yang The online game firm website et al., 2009) provides sufficient information for the game. I was satisfied with the number of (E. Anderson, websites where I could download and 2008; Berkowitz play online game. et al., 1988; Li & Online game firms give me a wider 李仲文, 2008; P. Asset choice of different website compared K. Rao, 2003; Specificity to conventional game store. Rindfleisch & Online game firm website offer me a Heide, 1997; wider range of product choices Williamson, compared to shopping at conventional 1987; H.-E. Yang game stores. et al., 2009) Online game firm website helps me (E. Anderson, accomplish tasks more quickly. 2008; Berkowitz I did not have to spend too much time et al., 1988; P. K. ASSET3 ASSET4 ASSET5 TIME1 Time TIME2 - 57 - TIME3 TIME5 TIME7 to complete the transaction. Rao, 2003; I did not have to spend too much Rindfleisch & effort to complete the transaction. Heide, 1997; I can save time for purchase of game Williamson, online service via the online game 1987; H.-E. Yang company website. et al., 2009) Purchasing online game via the online game company seems to be difficult. REL2 I believe that what I ask for is what I (Bhattacherjee, get in playing online game. 2001; Brown et al., 1993; Chen et REL3 Reliability REL4 I think that the online game service I al., 2009; Cottet purchased from performs the service et al., 2006; right. Meuter et al., I trust the online game company to 2000; Nair et al., deliver the product on time. 2010; Seo & Lee, 2008) I believe the online game company is (Chen et al., responsive to my needs. 2009; K. Lee & Responsive- RESP2 In the case of any problem, I think the Joshi, 2007; Nair ness online game firm will give me prompt et al., 2010; service. Parasuraman et The customer service team at the al., 1988; Wu et RESP1 RESP3 - 58 - online game firm will address any al., 2011; concerns that I have. Yoshida & James, 2010) ASS1 ASS2 Assurance I felt confident about the online game (Brown et al., purchase decision. 1993; Chen et al., I feel safe in my transactions with the 2009; K. Lee & online game firm website. Joshi, 2007; Meuter et al., ASS3 SAT1 The online game firm had answers to 2000; Yoshida & all my questions about the product. James, 2010) The gameplay (game content and (R. E. Anderson tasks) met my needs. & Srinivasan, 2003; Chen et al., SAT3 It was possible for me to download 2009; Devraj et online game that I choose easily. al., 2002; E.-J. Overall, I was satisfied with the online Lee et al., 2004; game experience. Shankar et al., SATISFAC SAT5 -TION 2003; Shieh & Cheng, 2007; Sureshchandar et al., 2002) PREF2 PREFERN- I strongly recommend online game to (E.-J. Lee et al., others. 2004; K. Lee & Joshi, 2007; CE PREF3 I do not intend to switch to play other Meuter et al., - 59 - PREF4 video games. 2000; Moutinho I intend to increase my use of online & Goode, 1995; game in the future Ndamnsa, 2013; Wu et al., 2011; H.-e. Yang & Tsai, 2007; H.-E. Yang et al., 2009) 4.4. Confirmatory factor analysis Confirmatory Factor Analysis (CFA) is the next step after EFA in order to determine the factor structure of the dataset. In the CFA, the factor structure which was extracted in the EFA is confirmed. Specifically, it is used to test whether the consistence between measures of a construct and a researcher's understanding of the nature of that construct (or factor) happens. Clearly, the aim of confirmatory factor analysis is to test whether the data fits the hypothesized measurement model. In a CFA model with multiple factors, the variance/covariance structure of the factors may be further analyzed by introducing second-order factors into the model if (1) the first-order factors are substantially correlated with each other, and (2) the second-order factors may be hypothesized to account for the variation among the firstorder factors (Wang & Wang, 2012). The table below shows Goodness-of-fit indices and criteria for convergent, discriminant validity and unidimensionity based on many references: . 27. Table 4.4.1: Criteria for measurement model Indices Chi-square/df (CMIN/DF) < 5 Meanings Model fits with References (Blunch, 2012; Byrne, - 60 - when sample size >200 survey data 2009) All standardized Model meets (Bielby & Hauser, 1977; regression weights are requirement of Blunch, 2012; Bollen, greater than 0.5 convergent validity 1998; Byrne, 2009; Chi-square/df < 3 when sample size 0.9 CFI (Comparative fit index) > 0.9 RMSEA (Root Mean Squared Error of Approximation) < 0.08   All unstandardized Hancock & Mueller, 2006; regression weights have Ullman & Bentler, 2001) statistical significance (P-value 0.7  Average variance extracted (AVE) >0.5  CR > AVE All goodness-of-fit indices are Model meets (Bielby & Hauser, 1977; met requirement of Blunch, 2012; Bollen, unidimensionality 1998; Byrne, 2009) Model meets (Hancock & Mueller, requirement of 2006; Ullman & Bentler, discriminant validity 2001)   Maximum Shared Squared Variance(MSV) < AVE Average Shared Squared Variance (ASV) < AVE - 61 - The CFA model to be tested in the present application hypothesizes a priori that (a) responses to the “satisfaction” which is a second-order factor can be explained by eight first-order factors (Perceived ease of use, Perceive usefulness, Utilitarian value, Hedonic value, Time, Reliability, Responsiveness and Assurance) and one third-order factor (Preference); (b) each item has a nonzero loading on the first-order factor it was designed to measure, and zero loadings on the other two first-order factors; (c) error terms associated with each item are uncorrelated; and (d) covariation among the three first-order factors is explained fully by their regression on the second order factor. A diagrammatic representation of this model is presented in Figure 3 Figure 3. Second-order CFA Byrne (2009) noted that, given the same number of estimable parameters, fit statistics related to a model parameterized either as a first-order structure or as a second-order structure will basically be equivalent. The difference between the two - 62 - specifications is that the second-order model is a special case of the first-order model, with the added restriction that structure be imposed on the correlational pattern among the first-order factors (Bielby & Hauser, 1977; Blunch, 2012; Byrne, 2009). However, judgment as to whether or not a measuring instrument should be modeled as a firstorder or as a second-order structure ultimately rests on substantive meaningfulness as dictated by the underlying theory. Hence, this thesis will apply second-order factors in confirmatory factor analysis. Then second-order CFA was performed and results were shown in table below: . 28. Table 4.4.2 Goodness-of-fit indices of measurement model CMIN Model NPAR P CMIN/DF CMIN DF Default model 99 3451.526 847 .000 4.075 Saturated model 946 .000 0 Independence model 43 28066.064 903 .000 31.081 RMR, GFI Model RMR GFI AGFI PGFI Default model .107 .776 .716 .799 Saturated model .000 1.000 Independence model 2.027 .061 .016 .058 Baseline Comparisons NFI RFI IFI TLI Model CFI Delta1 rho1 Delta2 rho2 Default model .877 .869 .904 .898 .904 Saturated model 1.000 1.000 1.000 Independence model .000 .000 .000 .000 .000 RMSEA Model RMSEA LO 90 HI 90 PCLOSE Default model .066 .071 .000 .068 Independence model .214 .212 .216 .000 Based on table 4.4.2, almost goodness-of-fit indices of measurement model were fairly good: CMIN= 3451.526; CMIN/df= 4.075 (< 5); CFI= 0.904 (>0.9); and RMSEA= 0.68 ([...]... surveying some online game players in Ho Chi Minh City (Vietnam), this study found that four antecedents could have significant and positive effects on online game service satisfaction and which, in turn, significantly affect online preference Especially, service quality has the relatively higher total positive effects on both online game satisfaction and preference Meanwhile, online game satisfaction. .. this study is to understand an online game service model This model contains several dimensions including experiential value (H.-E Yang et al., 2009), the Technology Acceptance Model (TAM) (Davis, 1989), Transaction Cost Analysis (Williamson, 1987) and Service Quality (Parasuraman, Zeithaml, & Berry, 1988), the four antecedents of online game service satisfaction and preference and test the associations... & Wang, 2009) With online gaming being a billion dollar industry and game companies making revenue from subscription charges (Adams, 2010), the presence of gamer satisfaction and preference issues is becoming more evident 1.1 A brief description of the research topic This thesis argue that there will have some factors affecting online game service satisfaction and preference The purpose of this study. .. determinant of satisfaction in TCA The study tries to find empirical support for the assurance dimension of SERVQUAL as determinant in online game satisfaction Further, the study also verified the general support for consumer satisfaction as a determinant of channel preference Because to keep competitiveness of online game industry is hard to hard, this study has ambition that its model can provide online. .. effects of these antecedents on online preference The findings imply that how to manage online game service quality better, provide more -1- acceptable transaction cost, and offer more experiential value are the key ways for effectively enhancing players’ satisfaction with the online game service in order to retain their preference to the online game service system 1.2 Background The Online Games Market... discusses and due to the importance of experiential value of the game service, this thesis combines the constructs for experiential values with TAM, TCA and SERVQUAL in order to assess online game service more precisely and completely 3.1.2 Backgrounds of online game satisfaction Previous studies considered that overall satisfaction is primarily a function of perceived service quality (R E Anderson... provide online game corporate to select and adopt the key point what the online game corporate should choose and how to affect the key factors of online game service satisfaction and online preference in online game industry 1.4 Research questions The literatures revealed that the immediate factor affecting consumers to retain preference to the providers is customer satisfaction (Devraj, Fan, & Kohli,... B2C channel satisfaction model or online shopping satisfaction model, satisfaction is considered as an important construct because it affects participants’ -3- motivation to stay with the channel and regarded as an antecedent of repurchase (Heilman, Bowman, & Wright, 2000) But what are the key factors that can make them satisfied with the products and services, which, in turn, enhance their preference, ... model analyses indicate that metrics tested through each model provide a statistically significant explanation of the variation in the online gamers' satisfaction and channel preference There are several studies found that TAM components—perceived ease of use and usefulness—are important in forming consumer attitudes and satisfaction with the online game Ease of use also was found to be a significant... perceived service quality of websites, (H.-E Yang et al., 2009) refined and validated the current SERVQUAL and IS-SERVQUAL instruments and the results indicated that the tangibility dimension is less relevant to the e-commerce service quality and completely excluded from the model This thesis follows the conclusion and use four dimensions: reliability, responsiveness, empathy, and assurance of online channel .. .ONLINE GAME SERVICE SATISFACTION AND PREFERENCE: AN EMPIRICAL STUDY OF VIETNAMESE ONLINE GAMING INDUSTRY In Partial Fulfillment of the Requirements of the Degree of MASTER OF BUSINESS... EFA of Transaction Cost Analysis (TCA) with Satisfaction and Channel Preference 53 4.3.4 EFA of Service Quality (SERVQUAL) with Satisfaction and Channel Preference 54 4.4 Confirmatory factor analysis... behavior, online game design issues, and drivers of consumer satisfaction with and preference for the online game Customer satisfaction can be evaluated through an assessment of the quality of efficiency,