(Luận văn) antecedebts of customer repurchase intention a study of online group buying in vietnam

t to UNIVERSITY OF ECONOMICS HO CHI MINH CITY ng hi International School of Business ep w n lo ad ju y th Tu Van Anh yi pl ua al n ANTECEDENTS OF CUSTOMER n va ll fu REPURCHASE INTENTION m oi A STUDY OF ONLINE GROUP-BUYING at nh z IN VIETNAM z k jm ht vb om l.c gm MASTER OF BUSINESS (Honours) n a Lu n va y te re Ho Chi Minh City – Year 2013 th t to UNIVERSITY OF ECONOMICS HO CHI MINH CITY ng hi International School of Business ep w n lo ad ju y th Tu Van Anh yi pl n ua al ANTECEDENTS OF CUSTOMER n va REPURCHASE INTENTION fu ll A STUDY OF ONLINE GROUP-BUYING oi m at nh IN VIETNAM z z k jm ht vb ID: 60340102 om l.c gm MASTER OF BUSINESS (Honours) n a Lu SUPERVISOR: Prof LE NGUYEN HAU n va y te re th Ho Chi Minh City – Year 2013 t to ng ACKNOWLEDGEMENT hi ep This thesis could not be finish without the help and support of many people who w are gratefully acknowledged here n lo ad At the very first, I would like to express my deepest gratitude to my supervisor, y th Prof Le Nguyen Hau With his guidance, I could have worked out this thesis He ju had offered me valuable suggestions and criticisms with his profound knowledge in yi pl rich research experience al ua I am grateful to express my sincere to Prof Nguyen Dinh Tho I have learned n from him a lot not only about research design, but also data analysis technique I am va n also extremely to give thankful to UEH – International School of business (ISB) ll fu supported me in all process m oi I would like to extent my sincere thanks to all my classmate and friends Their nh at kindness and supports have contributed very much in my working process Most z important, I would like to express my most sincere thanks to my father, my mother z k jm ht vb and my brother for their continuous encouragement and support om l.c gm n a Lu n va y te re th t to ng Declaration hi ep I hereby declare that this thesis, to the best of my knowledge and belief, is my w own work and effort and that is has not been submitted, either in part or whole, n lo anywhere for any award ad y th Information and ideas taken from other sources as cited as such This work has ju not been published yi pl n ua al n va ll fu Signature: Tu Van Anh oi m at nh Date: 18/02/2013 z z k jm ht vb om l.c gm n a Lu n va y te re th t to ng TABLE OF CONTENTS hi ep ABSTRACT CHAPTER INTRODUCTION w n 1.1 Research interest lo ad 1.2 Motivation of the study y th 1.3 Research objectives ju 1.4 Methodology and scope of research yi pl 1.5 Research structure ua al CHAPTER THEORETICAL BACKGROUND n 2.1 Introduction va 2.2 Theoretical background n ll fu 2.2.1 Online group-buying model oi m 2.2.2 IS repurchasing behavior model nh 2.2.2.1 Online customer retention 10 at 2.2.2.2 Online group-buying repurchase model 13 z z 2.2.3 Website quality 14 ht vb 2.3 Research model 16 jm 2.3.1 Perceived Usefulness 19 k 2.3.2 Customer satisfaction 20 gm 2.3.3 Customer trust 20 om l.c 2.3.4 Hypotheses related to website quality 21 2.4 Conclusion 23 a Lu CHAPTER RESEARCH METHOD 24 th 3.4.2 EFA for website quality scales 31 y 3.4.1 Cronbach Alpha 29 te re 3.4 Pilot test results 29 n 3.3 Instrument construction 26 va 3.2 Research design 24 n 3.1 Introduction 24 t to ng 3.5 Conclusion 32 hi ep CHAPTER ANALYSIS AND RESULTS 33 4.1 Introduction 33 w 4.2 Respondents demographic 33 n lo 4.3 Scale validation 34 ad 4.3.1 Preliminary results 34 y th 4.3.2 Confirmatory factor analysis (CFA) 38 ju yi 4.3.2.1 CFA results 38 pl 4.3.2.2 Saturated model 43 al n ua 4.4 Modified research model 45 va 4.5 Model fitness 46 n 4.6 Bootstrap 47 fu ll 4.7 Hypotheses testing 48 m oi 4.8 Conclusion 49 at nh CHAPTER CONCLUSIONS AND LIMITATIONS 50 5.1 Introduction 50 z z 5.2 Discussion and conclusion 50 vb jm ht 5.3 Managerial implications 52 5.4 Limitations and future research 54 om l.c gm APPENDICES k REFERENCES n a Lu n va y te re th t to ng LIST OF FIGURE hi ep w n lo Figure 2.1 A post-acceptance model of IS continuance 11 ad Figure 2.2 Research model 19 y th Figure 3.1 Research process 25 ju yi Figure 4.1 CFA model of online group-buying repurchase intention scale 39 pl Figure 4.2 CFA model of trust scale 40 al ua Figure 4.3 CFA model of satisfaction scale 40 n Figure 4.4 CFA model of website quality scale 41 va n Figure 4.5 Saturated model of main survey 44 fu Figure 4.6 Modified research model 46 ll m Figure 4.7 SEM result of research model (Standardized) 46 oi at nh z z k jm ht vb om l.c gm n a Lu n va y te re th t to ng LIST OF TABLE hi ep w n lo Table 2.1 Some integrated models of customer retention 12 ad Table 3.1 Sources of questionnaire items 28 y th ju Table 3.2 Cronbach alpha result of pilot test 30 yi Table 3.3 EFA result of pilot test 32 pl Table 4.1 Respondents demographic 34 al n ua Table 4.2 Cronbach alpha result 35 va Table 4.3 EFA result of main study 37 n Table 4.4 Confirmatory factor analysis of measurement model 42 fu ll Table 4.5 Relationship between constructs of research 45 m oi Table 4.6 Relationship between constructs in research model (standardized) 47 nh Table 4.7 Bootstrap estimate result with N = 1000 47 at z Table 4.8 Result of hypotheses testing 48 z k jm ht vb om l.c gm n a Lu n va y te re th t to ng LIST OF ABBREVIATION hi ep w n Average Variance Extracted lo AVE ad CFA y th CR Confirmatory Factor Analysis ju Expectation-Confirmation Model yi ECM Composite Reliability Exploratory Factor Analysis ua Electronic commerce n E-commerce Expectation-Confirmation Theory al EFA pl ECT va Electronic commerce – Service quality IS Information System ML Maximum Likelihood OGB Online group-buying SEM Structural Equation Model SPSS Statistical software package TAM Technology Acceptance Model TPB Theory of Planned Behavior TRA Theory of Reasoned Action n E-SQ ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th t to ng ABSTRACT hi ep Online group-buying has emerged as a new e-commerce model and received attention in both academics and practice Prior researches focused on investigating w n online group-buying mainly from the marketing perspective, such as the transaction lo ad process, price mechanism, and benefits (Li & Liu, 2012) Few of studies have y th investigated the relation between consumer’s acceptance and their purchasing ju behavior Hence, this study focuses in measuring relationship between cognitive yi pl factors – trust, satisfaction and perceived usefulness – and online group-buying ua al repurchase intention Simultaneous, this study also integrates website quality - a direct n and indirect variable – to measure its impact on customer repurchase behavior va To examine research model, information and data is accessed by using n ll fu questionnaire for respondents more than 18 years old and have ever purchased on oi m online group-buying websites Sample size of this quantitative research is 365 nh respondents Confirmatory factor analysis (CFA) is used to test measurement scale at and the structural equation modeling (SEM) is used as the main method for analyzing z z research model and hypotheses vb ht Results in this study show that individual user’ intention to repurchase in online k jm group-buying websites is motivated by trust, satisfaction and website quality Among gm three impact factors, website quality has the strongest direct influence, followed by trust and satisfaction Besides direct influence, website quality also has indirect l.c impact on customer repurchase intention through both of trust and satisfaction Trust om also has strong impact on customer satisfaction, thus, affect indirectly to customer n a Lu retention consider about their strategies of increasing trust and satisfaction, they should also y te re results also suggest that in order to increase customer retention should not only n va With these results, research framework can be seen fitted with data market Study technical adequacy th take consideration of increasing website quality, especially in content quality and t to ng CHAPTER hi ep INTRODUCTION This chapter is starting point of this study In this chapter, the research interest of w thesis is first introduced Subsequently, the motivation of this research is discussed n lo Research objectives are presented in the next section Then, methodology and scope of ad this research is introduced summarily Final, structure of this study is presented This y th chapter aims at specifying the purpose of this research and the research content ju yi pl 1.1 Research interest: ua al In recent decades, science and technology develops surprisingly Achievements of n these progresses create more opportunities as well as challenges for development of va economies Forms of transactions become more diversified; suppliers have more n ll fu approach ways to interact with customers Traditional markets have not caught up oi m with higher demands of consumers Buyers become more and more busy and have so nh many choices that make their buying decisions more difficult Hence, to satisfy at customers, electronic commerce appears and grows extraordinarily with widespread z z Internet usage vb Electronic commerce, commonly known as e-commerce, consists of the buying ht jm and selling of products or services over electronic systems such as the Internet and k other computer networks A form of e-commerce is online shopping which is the gm process whereby consumers directly buy goods or services from a seller in real time, om l.c without an intermediary service, over the Internet This form of shopping are attracting more buyers, suppliers and researchers because it offers a lot of benefits a Lu such as convenience, speed, time, a 24-hour opening, information on products and n reviews To nowadays, online shopping continues their rapid spread and gains many sites offer bargains on everything from meals to travel packages Customers can make th cooperate and buy goods at a discount price (Matsuo, 2009) Innovative group-buying y for applying agent technologies Group-buying is a model in which multiple buyers te re Group-buying is seen as an effective form of online shopping and a promising field n va increasing importance in the lives of a wider range of the population 3 t to ng comparison of product/service prices and choose the supplier with the lowest price hi ep (Gounaris et al., as cited in Li & Liu, 2012) when using this new model With the tremendous growth, online group-buying attracted more and more w attention of practitioners as well as researchers However, most previous online group- n lo buying studies focus mainly on the pricing mechanisms, coalition formation, benefits ad of bidder cooperation, uncertain demand, incentive mechanisms and consumer y th adoption (Fan et al., 2010) There are not many studies investigated the relationship ju yi between customers’ acceptance of online group-buying and their purchasing behavior pl In recent years, this topic begins become a hot issue, especially, continuance behavior al n ua receives more attention because this issue at an individual level has been regarded as va crucial for sustainable web-based services (Premkumar & Bhattacherjee, 2008) n Besides that, many online group-buying websites as well as others online websites fu ll are facing strong competition due to the evolution and proliferation of web-based m oi services Moreover, web-based services have low entry barriers by its nature, if one at nh service is created, then a number of comparable alternative web-based services follow, resulting in a high switching rate between those services by users z z (Vatanasombut et al., as cited in Lee & Kwon, 2011) Thus, many online group- vb jm ht buying providers are struggling to find strategies to exist in this difficult period Retaining their existing customers becomes a strategic way to ensure the company's k gm success and overall sustainability Exploring and analyzing which factors influence customer retention have significant meaning to online providers l.c om Research on continuance intention in both online shopping and online group- a Lu buying is still in its infancy Prior researches on consumer online repurchase placed more emphasis on the impact of psychological factors such as trust, satisfaction n model Very few studies have attempted to investigate the impact of product/service y te re expectation-confirmation theory and information system (IS) continuance intention n va formation, loyalty (Cheung et al., 2003) These researches primarily base on th characteristics as well as website quality on online consumer continuance 4 t to ng Website quality nowadays is assumed to have the potential to influence the future hi ep behavior of service users and have an impact on the profits of IS investment (Cronin and Taylor; Zeithaml, Berry and Parasunaman, as cited in Li, 2010) Several studies w on website quality in physical encounters have concluded that some factors of website n lo quality are responsible for customers’ perception which is likely to lead to behavioral ad intentions to purchase However, in online group-buying context, according y th mentioned above, there are a few of studies attempted to measure the direct and ju yi indirect impact of website quality to repurchase behavior In particular, research on pl taking an integrated perspective to examine the predictors of website quality and al n ua investigate the relationship between website quality and repurchase intention is still n va lacking This is a promising topic for future research in this context of e-commerce ll fu 1.2 Motivation of the study: oi m In Vietnam, group-buying model was introduced in 2010 and after a short time, nh group-buying model and daily deal sites increased very quickly With a win-win-win at model for three parties: enterprises can sell their products, group-on companies profit z z from commission and customers can get good deals, group-ons or group vouchers vb have become popular among Vietnamese students and white-collar workers due to ht jm attractive discounts and a wide range of services and products1 k To the end of 2011, there are almost 100 group-on sites in Vietnam with more gm than 6700 deals and 4.2 millions sold out vouchers2 Four leading group-buying om l.c websites include: Nhommua, Hotdeal, Cungmua and Muachung Ho Chi Minh City is the strongest competition market with about 74% transaction deals and many this new model, however, amount of transaction is still limited n a Lu followers2 Other markets such as Da Nang, Binh Duong, Can Tho… begin introduce y te re between rivals also increase very strongly The biggest challenge for group-buying n va Along with development of group-buying market, perceived risk and competition http://news.smh.com.au/breaking-news-technology/group-buying-sites-boost-ecommerce-in-asia20110502-1e3l1.html http://www.slideshare.net/TaiTran/groupon-clones-in-vietnam-2752011 th t to ng sites is method of payment and privacy In Vietnam, many people are still not familiar hi ep with online transaction With weak privacy security system, customer beliefs have trend of reducing when purchasing vouchers by money transfer or cash on delivery w method Besides, there are many partners who not follow the contract, provide bad n lo service or treat group-on customers differently Many vouchers are delivered late or ad lost, customers cannot dictate the time to receive vouchers and have to book in y th advance to use the service With many problems, online-group buying websites in ju yi Vietnam are facing many difficulties in attracting customers as well as keeping pl existing customers al n ua Base on operational experiences in vouchers transaction, many group-buying va businesses begin selecting other ways to limit their obstacles Firstly, they can n become distributors, delivery directly products to customers, instead of delivering fu ll vouchers Second, group-buying company can become e-commerce exchange They m oi can buy and manage deals by themselves They also provide some flexible solutions at nh for suppliers such as: create own websites, provide orders management system, and provide customers management system… However, according to many experts the z z key to success for this kind of business is still to ensure service quality and price to vb and technology scenes – cites that: k jm ht keep customer trust Do (2012) – who is interested in Vietnam’s start-up, social media gm An increased skepticism of the online space is low on the list of dangers though With the l.c internet penetration at 34 percent and an increasing portion of them moving into social media, om the online market is getting bigger and bigger Although companies and people may not be a Lu trusting of web content, they are learning more and more to be dependent on it n In general, Vietnam is still a promising market for online group-buying model researchers in developing way of e-commerce in Vietnam th group-buying model is really an interesting research field for practitioners and y their competitive advantages and maintain their market Simultaneously, online te re Customer retention is a necessary topic for helping group-buying businesses increase n va Companies in this market need to learn and develop more and more to succeed 6 t to ng 1.3 Research objectives: hi ep As discussion in section 1.1, information system (IS) retention has been one of the most recently explored topics in the IS research field Many theoretical perspectives, w to nowadays, have been advanced in order to understand what motivates individuals n lo to repurchase in online group-buying websites Thus, based on literature covering the ad concept of IS continuance model and circumstance of online group-buying market in y th Vietnam, this paper aims: ju yi  To propose a model predicting customers’ repurchase intention in online pl group-buying context in Vietnam al n ua  To investigate impact of website quality on customers’ repurchase intention va in online group-buying context n  To examine impact of cognitive factors (trust, satisfaction and perceived fu ll usefulness) on online group-buying repurchase intention m oi This study is necessary for development of group-buying market in Vietnam It nh also demonstrates that website quality is a noteworthy factor affect repurchase at z intention of customers in using online group-buying z ht vb 1.4 Methodology and scope of research: k jm Collecting data process of this study is designed into two stages First is a pilot gm test, second is main survey to collect data for examining research model Pilot test is quantitative research with sample 57 respondents to examine reliability and validity l.c of observed variables Main study is also quantitative research with sample size 365 om respondents a Lu Author accesses information and collect data by using questionnaire Respondents n Sample is selected by using non-probability sampling method – convenience sample th measurement scales are estimated using confirmatory factor analysis (CFA) to test y Purpose of this research is to confirm and examine conceptual model The te re Research is studied from September 2012 to December 2012 n va are more than 18 years old and have ever purchased on online group-buying websites 7 t to ng reliability and validity The structural equation modeling (SEM) is used as the main hi ep method for analyzing the research model and hypotheses w 1.5 Research structure: n This thesis is organized into five chapters It begins with introduction chapter lo ad which presents an outline of this research, including the motivation, objectives and ju y th scope of this research The next chapter describes online group buying model, literature of IS repurchase behavior model and website quality concept This chapter yi pl also describes research model and hypotheses The third chapter is research ua al methodology used to empirically test the research model The fourth chapter presents n the results of data analysis The final chapter discusses summarizes the study’s core va findings, its contributions and its limitations n ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th t to ng CHAPTER hi ep LITERATURE REVIEW AND HYPOTHESES w 2.1 Introduction: n This chapter mainly introduces theoretical background and research model of the lo ad study First, online group-buying model is introduced to clarify more clearly about ju y th advantages and disadvantages of this new model Second, literatures of IS repurchase models in online shopping and online group-buying context are discussed Next, yi pl concepts and instruments of website quality are reviewed systematically Finally, n discussed ua al research model, its constructs and relationship hypothesized among the constructs are n va fu 2.2 Theoretical background: ll 2.2.1 Online group-buying model: m oi Group-buying, also known as collective buying, introduced in 1999 by a few of nh companies After introducing time, this model have been facilitated by the internet and at z the easy, fast group coalition process brought by social networks (Xiong & Hu, as z vb cited in Erdogmus & Cicek, 2011) It is seen as a part of an innovative wave of online auctions, price-reduction models and group-buying models k jm ht market-based pricing mechanisms, includes traditional auctions, non-traditional gm There are mainly two different types of online group-buying systems (Fei et al., as l.c cited in Erdogmus & Cicek, 2011) First type of this system is structured based on a om dynamic pricing mechanism In this first type, masses of consumers are aggregated, a Lu and perform collective buying to enjoy price discounts online In the second type of n the online group-buying, the group-buying company offers a certain product or va service at a static large discount price This price required the total number of the th success These websites usually offers a large of products and services at significantly y Today, many online websites are using group-buying models and have got great te re of buyers n buyers must be greater than the predetermined limit of the minimum required number t to ng reduced prices These websites claim they can negotiate low prices with manufacturers hi ep and suppliers, and then pass these savings on to their customers (Kauffman & Wang, 2001) w However, group-buying websites, like other online shopping websites, are facing the n lo problem of obtaining market attention With group-buying models, online group-buying ad websites need not only a critical mass of consumer patronage and interest, but also a y th significant amount of transaction volume so as to be able to profitably deliver on ju yi their low price guarantee Thus, if they cannot reach a critical mass of users and sales pl volume, then it will be difficult for the group-buying business model to bring the al n ua customers of the firms that adopt it the savings they expect (Kauffman & Wang, 2001) va Moreover, Cook (as cited in Kauffman & Wang, 2001) points out that the group- n buying business model is too difficult for the general consumer to understand The fu ll author also argues that the mechanics of group-buying on the Internet also prevent m oi impulse buying, due to the lengthy periods consumers have to wait until the end of at nh auction cycles With these problems, amount of customers of online group-buying websites begin increasing slower; in some markets, there are reducing signs of z z customers as well as transactions vb jm ht In order to overcome obstacles, some online group-buying websites focus website quality and service quality For instance, Mobshop – a San Francisco, California- k gm based group-buying service provider- increased from 37.00 to 132.000 registered users in a three month period from January to April 2000 (as cited in Kauffman & l.c om Wang, 2001) Its early success attributed by its careful design, increasing technical a Lu adequacy and saving delivery time Focusing on quality is a good strategy for online group-buying websites in highly competition market at current time and in the future n succeed (Zang et al., 2011) In these two stages, second stage is an important subject th encourage them to repurchase, which is critical if the e-commerce vendor is to y concerned with the encouraging people to purchase online and the second is to te re Online buying behavior can be understood in two stages: the first stage is primarily n va 2.2.2 IS repurchasing behavior model: 10 t to ng of study because customer retention is often seen as a means to gaining competitive hi ep advantage Researchers have studied online customer retention in different contexts, such as w “online repurchase intention”, “continue to shop online”, “customer intention to n lo return”, “web site stickiness”, and “continued information systems/IT intention” ad (Wen et al., 2011) ju y th 2.2.2.1 Online customer retention: yi pl Online customer retention in recent years, become a hot issue in both the IT and ua al marketing areas Studies of this topic have been mainly divided into two streams n consisted of studies based on static-type models and process-type models (Lin & Ong, n va 2010) ll fu (1) Static-type researches are derivation from concepts such as theory of planned oi m behavior (TPB), Fishbein and Ajzen's theory of reasoned action (TRA) or nh technology acceptance model (TAM) The theory of Reasoned Action assumes at that if people view a behavior as positive (attitude), and if they believe that z z others would prefer them to perform the behavior (subjective norm), there will vb be a greater intention (motivation) to behave in that manner and they are thus ht jm more likely to so (Udo et al., 2010) TPB adds one major predictor – k perceived behavioral control – “to account for times when people have gm intention of carrying out a behavior, but the actual behavior is thwarted because om l.c they lack confident or control over behavior” (Miller, as cited in Udo et al., 2010) Along with these theories, TAM has been confirmed as the most popular a Lu parsimonious framework used to explain customers’ behavioral intention TAM n is model that explains user intention and behavior based on forward-looking or Bhattacherjee, 2001) However, many empirical studies also comparing the th and user populations (Davis et al.; Mathieson; Taylor and Todd, as cited in y acceptance behaviors across a broad range of end-user computing technologies te re usefulness and perceived ease of use as salient beliefs influencing IS n va prospective expectations about IT usage This model found perceived 11 t to ng relative effects of perceived usefulness and ease of use during pre-acceptance hi ep and post-acceptance stages of IS use report that: perceived usefulness impacts attitude substantively and consistently during both stages of IS use, and ease of w use has an inconsistent effect on attitude in the initial stages, which seems to n lo further subside and become non-significant in later stages (Davis et al.; ad Karahanna et al., as cited in Bhattacherjee, 2001) Hence, studies base on TAM y th to nowadays, perceived usefulness is usually used as direct variable influencing ju yi on customer behavioral intention pl (2) Second stream is process-type models, mainly based on expectation- al n ua confirmation theory (ECT) or expectation-confirmation model (ECM) This va stream had been developed from some limitations of TAM It also uses n individual cognitive factors for predicting IS continued use However, they fu ll based on their backward-looking or retrospective perceptions grounded in m oi actual usage experience, such as performance, disconfirmation, and satisfaction, at nh in addition to initial expectations (Premkumar, Bhattacherjee, 2008) z z vb Perceived Usefulness IS continuance Intention gm Confirmation k jm ht Satisfaction om l.c Source: Bhattacherjee, 2001 Figure 2.1: A post-acceptance model of IS continuance a Lu In this model, consumers’ intention to repurchase a product or continue service n use is determined primarily by their satisfaction with prior use of that product or consumer satisfaction, post-purchase behavior and service marketing in general th today, this stream develops widely in the consumer behavior literature to study y a loyal base of long-term consumers and confirmed in many studies Until te re Bhattacherjee, 2001) Satisfaction is viewed as the key to building and retaining n va service (Anderson and Sullivan 1993; Oliver 1980, 1993, as cited in 12 t to ng (Anderson and Sullivan; Dabholkar et al.; Oliver; Patterson et al.; Tse and hi ep Wilton, as cited in Bhattacherjee, 2001) Many researches try to add new construct, and integrate studies that combine these models and another theory or w model n lo In addition to the mainstream researches, there is a recent focus on affective ad factors According to Lee & Kwon (2011), the factors suggested classified into two y th categories: cognitive and affective Cognitive factors are those related to the mental ju yi process of knowing, including aspects such as perception, reasoning and judgment pl Representative cognitive factors are: perceived usefulness, satisfaction, trust, al n ua perceived ease of use, security, confirmation and disconfirmation, perceived risk, va perceived switching cost… In contrast, affective factors are related to specific n emotions or states of feeling Some affective factors are studied in recent years such fu ll as perceived playfulness (enjoyment), pleasure, arousal, familiarity and intimacy Lee m oi & Kwon (2011) also suggested that customer retention research has shifted its focus influencing on customer behavior at nh from cognition-oriented factors to affective factors to explore more factors z z Table 2.1: Some integrated models of customer retention vb Characteristics of the research model Research domain Based on ECM: introduced perceived playfulness as a Web-portal new factor Thong et al (2006) Based on ECM: introduce post-adoption beliefs and Internet service perceived enjoyment as new factors Limayen and Cheung Based on ECM: adds IS habit as a new factor Internet-based (2008) learning Atcharuyachanvanich et Based on ECM: adds customer loyalty as a new factor E-commerce al (2006) Min and Shenghua Based on ECM: adds perceived enjoyment as a new E- learning (2007) factor Roca et al.(2006) An integrated study that combines EDT model and TAM E- learning model, adds perceived quality and perceived usability as new factor Liao et al (2007) An integrated study that combines EDT model and the E-service theory of Planning Behavior, adds subjective norm as a new factor Chiu and Wang (2008) An integrated study that combines United theory of Web–based learning Acceptance and Use of Technology, adds subjective task value as a new factor k jm ht Study Lin et al (2005) om l.c gm n a Lu n va y te re th Source: Lee & Kwon (2011) 13 t to ng In general, researchers have integrated many other factors to create new framework hi ep that will improve the explanatory and predictive power for explaining online customer retention w n 2.2.2.2 Online group-buying repurchase model: lo ad Few studies have tested online repurchase intention in online group-buying ju y th context In 2010, Fan et al has used model of Bhattacherjee adapted expectationconfirmation theory (ECT) and integrates the technology acceptance model (TAM) to yi pl theorize a post-acceptance model of IS continuance In this study, relationship ua al between customer satisfaction, perceived usefulness and online group-buying n repurchase intention are confirmed Simultaneously, role of price expectation is also n va measured ll fu In 2012, Tien et al also research about repurchase intention in online group- oi m buying context They examine repurchase intention through relationship quality and nh expectation – confirmation theory views though relationship quality is regarded as an at important factor in the relationship marketing literature This study emphasize z z satisfaction is a particularly important foundation for a successful long-term vb relationship between customers and group-buying websites Hand in hand with ht jm satisfaction is trust of the second factor Just like previous researches, satisfaction and k trust continue being the important of two influence factors Satisfaction is a stronger gm predictor of repeat purchase intention than trust; this is consistent with the study in e- om l.c commerce (Deng et al., as cited in Tien et al., 2012) Trust has direct and significant effect on satisfaction and repeat purchase intention Moreover, the results of this a Lu research also show that trust, perceived value and perceived quality are important n antecedents of satisfaction Among these factors, perceived quality is highly th word-of mouth, customer satisfaction, promotional incentive, and customer y In the same year, Liu and Wu also study relationship between service quality, te re intention through satisfaction n va influential determinant of perceived value and indirectly affecting repeat purchase 14 t to ng satisfaction This study show that service quality has strong effect on customer hi ep satisfaction, thus, indirectly influence to customer loyalty In summary, very little studies in online group-buying context focus on customer w buying behavior, especially, customer retention While the online market is growing n lo and profitable, the competition for market share is also increasing To remain ad competitive, it is imperative for online providers to invest time and money to find out y th strategy to keep existed customers Studies of customer retention are really necessary ju yi for development of online group-buying model in the future pl al ua 2.2.3 Website quality: n Quality is not a new concept in information systems management and research va Information systems practitioners have always been aware of the need to improve the n ll fu information systems function so it can react to external and internal pressures oi m and face the critical challenges to its growth and survivability (Aladwani & Palvia, nh 2002) However, to nowadays, both the conceptualization and the measurement of at website quality have been two debated topics z z In study of Éthier et al (2006), research on the concept of website quality can be vb classified broadly into four complementary research categories (1) The first focuses ht jm on functionalities and/or content of website The dimensions identified have generally k been: functional issues, navigation, content, technical issues and contact information gm (2) The second category includes researches affected by technology acceptance model om l.c TAM, relationship between perceived ease of use and perceived usefulness can be seen as a relation of quality Information quality, system quality and service quality of a Lu websites are the essential components of website quality (3) The third category n includes studies that highlight service quality as a fundamental aspect of the overall delivery of goods and services; Santos (as cited in Udo et al., 2010) defines e-service th extent to which a website facilitates efficient and effective shopping, purchasing and y website, for instance, Zeithaml (as cited in Li, 2010) defines e-service quality as the te re exchangeable Many researchers try to make conceptualization of service quality of n va quality of a website E-service quality and website quality sometimes are use 15 t to ng quality is the overall customer perceptions, judgments and evaluations of the quality hi ep of service obtained from a virtual marketplace These concepts are used popular and developed from time to time However, this category still faces many debates (4) The w fourth category is composed of authors who believed that the principal criterion for n lo website quality was defined by customers' perceptions of quality For example, ad Huang (as cited in Éthier et al., 2006) describes website quality as whether the y th website meet and/or exceeded expectations in terms of information and enjoyment ju yi position; Wan (as cited in Éthier et al., 2006) states that the quality of a website was pl based on: information, friendliness, responsiveness and reliability al n ua Like concepts of website quality, many instruments that combine the diverse va aspects of website quality have also been proposed (see in appendix A) Some n instruments combine traditional service quality dimensions and web interface quality fu ll dimensions as the point of departure For instance, in 2000, Loiacono et al proposed m oi WEBQUAL, includes: ease of understanding, intuitive operation, informational fit-to- at nh task, tailored communication, trust, response time, visual appeal, innovativeness, emotional appeal, online completeness, relative advantage, and consistent image z z After that, Yoo & Donthu (2001) measures website quality by SITEQUAL scale with vb ordering, corporate and brand product assurance processing speed, product quality Aladwani and Palvia's (2002) gm and security, k uniqueness, equity, jm ht 12 dimension: aesthetic design, competitive value, ease of use, clarity of instrument focuses on website design and content It used four dimensions: specific om l.c content, content quality, appearance, and technical adequacy a Lu With the growth of recognition of different variability in the outcome of measuring website quality, many studies show more different dimensions in website quality, for n customers and providing services to meet the needs and expectations of customer; y te re dimensions scale of website quality, which is built on better understanding of n va examples: Madu &Madu (2002, as cited in Li & Suomi, 2009) develop a 15 and improving website quality by identifying e-service system entities and th Field et al (2004, as cited in Li & Suomi, 2009) develop process model for assessing 16 t to ng transactions between those entities and mapping key quality dimensions onto them hi ep Most recently years, Sohn & Tadisina (as cited in Li, 2010) proposed six dimensions to measure website quality, include: trust, speed of delivery, reliability, ease of use, w customized communication, website content and functionality n lo Among studies of website quality instruments, most researchers develop adapted ad website quality scales based on the modification of the SERVQUAL instrument y th SERVQUAL scale is measurement scale had been used widely to measure service ju yi quality with five dimensions: tangibles, reliability, responsiveness, assurance, and pl empathy (Iwaarden, Wiele, as cited in Li, 2010) When adapting to e-commerce al n ua context, Zeithaml (2000, as cited in Li & Suomi, 2009) proposed a 7-dimension va website quality scale Later, Panasuraman et al (2005, as cited in Liu & Wu, 2012) n developed it into seven constructs divided into two groups, includes: core e-SQ: fu ll efficiency, fulfillment, availability, privacy; recovery e-SQ: responsiveness, m oi compensation and contact To nowadays, these instruments is used popular in many at nh studies because it offers the surface dimensions of e-service quality based on customers’ experience and evaluation perspective, which are viewed also as the z z antecedents to the adoption of e-service (Rowley, as cited in Li & Suomi, 2009) vb jm ht In general, website quality concept remains underdeveloped and is a vastly undefined concept This is a complex concept which has multiple dimensions k gm Although recently, research on website quality has adopted a much broader scope on website quality compared to its past focus on usability and interactivity However, to l.c n 2.3 Research model: a Lu field for researchers om nowadays, there is no consensus on a definition on it This topic is still a promising (Li, 2010) Hence, some modifications and extensions to the original model were th order to improve its robustness and predictive ability across a wider range of contexts y of human behavior required additional theoretical refinement and empirical testing in te re online group-buying context However, the advancement of many preliminary model n va This study adapted literature reviews of customer retention in online shopping and 17 t to ng proposed in this study hi ep Firstly, online group-buying repurchase intention was designed as dependent variables Intention was considered the best immediate factor in the relationship w between attitude and behavior, and is appropriate to test consumers’ behavior (Wen et n lo al., 2011) Online repurchase intention was one of context of customer retention In ad many prior researches, online repurchase intention and IS continuance intention were y th used exchangeable In this study, repurchase intention was used instead of ju yi continuance intention because these concepts were still different (Wen et al., 2011) pl although in online environment, both continuance intention and repurchase intention al n ua are influenced by the initial use/purchase experience and sometimes used as the same va Continuance intention emphasizes the continued usage of e-commercial websites to n shop instead of the use of physical stores while online repurchase intention is a fu ll construct combining information system theory and marketing theory (Wen et al., m oi 2011) In online group-buying model, customer purchased products or service at nh through online group-buying websites, not directly from providers, so, both on customers’ IS use and purchasing behavior needed to investigate “Online group- z z buying repurchase intention” construct was acceptable with this study vb jm ht Secondly, two cognitive factors in expectation-confirmation model (ECM) – satisfaction and perceived usefulness - were chosen for this study because they k gm commanded a majority of factors found to affect customer retention in prior studies Simultaneously, it ensures that the nomological structure of the research model is l.c om consistent with the traditional belief-attitude-intention linkages in IS literature (Davis, a Lu Bagozzi & Warshaw; Venkatesh et al., Ajzen, as cited in Li, 2010) In these linkages, satisfaction is typically viewed as user attitude towards IS, which is primarily n and is a salient determinant of behavioral intention regarding IS use in TAM y te re continuance intention Perceived usefulness in this study represents as IS user beliefs n va measured by various beliefs about IS and posited to have strong saliency in predicting and online repurchases intention Trust in sellers is a vital key to maintaining th Thirdly, trust was added to the model due its role in influencing both satisfaction 18 t to ng continuity in the buyer relationship An individual level of trust may increase hi ep gradually based on positive outcomes from repeated behavior, although its important in determining repeat purchase intention may decrease over time (Chiu et al., 2012) w Especially, in Viet Nam market, when some online group-buying websites faced n lo many problems make customer trust decrease, studies of the impact of this factor on ad customer behavior become necessary for practitioners in improving and developing y th their business ju yi Finally, website quality in this proposed model was also incorporated as a factor pl leading and influencing to customer behavior through three constructs: perceived al n ua usefulness, customer satisfaction and customer trust va A review of the literature evaluation reflected that there were many instruments to n measure website quality In this study, instrument from study of Aladwani & Palvia fu ll (2002) was used due to its concept base on website functionalities and contents This m oi instrument included four constructs: content quality, appearance, specific content and at nh technical adequacy Appearance referred to the visual attractiveness of a website, included colors, fonts, and multimedia features as well as an appropriate overall z z layout Specific content referred to finding specific details about products or services, vb jm ht customer support, privacy policies and other important information (Aladwani & Palvia, 2002) Content quality dealt with attributed such as information usefulness, k gm completeness, accuracy, and so forth, and was similar to the “information quality” concept (Liao et al., 2006) Technical adequacy, according to Liao et al (2006), was l.c om also similar to the “system quality” concept It meant that appropriate technologies transactions and contribute to convenience of usage for IS users n a Lu had been adopted by the web retailer It could improve the security of online y te re presented in figure 2.2 n va With above discussions, the proposed research framework and hypotheses are th 19 t to ng hi ep Perceived Usefulness w n lo Website quality H5 ad Customer Satisfaction Content quality H2 y th H4b OGB Repurchase Intention H4d Specific content H3b ju yi pl Appearance H1a H1b H4a H3a Customer Trust al H4c n ua Technical adequacy va Figure 2.2: Research model n ll fu 2.3.1 Perceived usefulness: oi m Drawing from TAM, post-consumption expectation is represented as ex-post nh perceived usefulness in the proposed IS continuance model (Bhattacherjee, 2001) at Davis (1989) defines perceived usefulness as “the degree to which a person believes z z that using a particular system would enhance his or her job performance According vb to Burk (as cited in Al-maghrabi et al., 2010) perceived usefulness is the primary ht jm prerequisite for mass market technology acceptance, which depend on consumers’ k expectations about how technology can improve and simplify their lives (Peterson et gm al., as cited in Al-maghrabi et al., 2010) Numerous empirical investigations have l.c established strong empirical support for direct impact of perceived usefulness on om intention So, in this study, perceived usefulness is proposed factor captures the n a Lu instrumentality of IS use, and influences subsequent continuance decisions (Bhattacherjee, 2001), perceived usefulness also influences indirectly to customer th Davis et al., cited in Liao et al., 2006) and of intentions for continued use y Besides an important predictor of initial intention to use information system (Davis; te re group-buying repurchase intention n va Hypothesis 1a: Customer perceived usefulness is positively associated with online 20 t to ng intention through customer satisfaction Some studies also examine and confirm this hi ep relationship such as Bhattacherjee, (2001), Li (2010), Wen et al (2011) Thus, it is assumed that: w n Hypothesis 1b: Customer perceived usefulness is positively associated with lo ad customer satisfaction y th 2.3.2 Customer satisfaction: ju yi Satisfaction is initially defined by Locke (1976, as cited in Bhattacherjee, 2001) in the pl context of job performance as “a pleasurable or positive emotional state resulting al ua from the appraisal of one’s job.” This definition is extended by Oliver (as cited in n Bhattacherjee, 2001) to the consumption context as “the summary psychological state va n resulting when the emotion surrounding disconfirmed expectations is coupled with fu ll the consumer’s prior feelings about the consumption experience.” In many m oi researches, customer satisfaction has usually been applied to measure e-commerce at nh success or consumer repurchase behavior Oliver (as cited in Fan et al., 2010) shows that satisfaction has both direct and indirect connections with future intention through z z its impact on attitude Hence, proposed hypothesis of this study is: vb buying repurchase intention k jm ht Hypothesis 2: Customer satisfaction is positively associated with online group- gm om l.c 2.3.3 Customer trust: In general, trust is viewed as a set of specific beliefs dealing primarily with the benevolence, competency, and integrity of another party (Doney and Cannon, as cited a Lu in Chiu et al., 2012) According to TRA, trust can be viewed as a behavioral belief n that creates a positive attitude toward the transaction behavior, which is turn leads to intention and negative word of mouth communication Lack of trust prevents buyers th show that the violation of trust in e-commerce will lead to negative repurchase y will be cause by the distance and other impersonal factors Many previous researchers te re trust is more important compared to traditional business as increasing uncertainties n va transaction intentions (Pavlou and Gefen, as cited in Wen et al., 2011) The role of 21 t to ng from engaging in online shopping because they are unlikely to carry out transactions hi ep with sellers who fail to convey a sense of their trustworthiness, mainly because of fears of sellers’ opportunism (Hoffman et al., as cited in Wen et al., 2011) So, trust w plays an important role on driving repeat purchase intention Following hypothesis is n lo proposed: ad y th Hypothesis 3a: Customer trust is positively associated with online group-buying ju repurchase intention yi pl Beside directly impact on customer repurchase intention, users’ trust also influences ua al customer behavior through satisfaction In some previous researches, trust’s impact n has not high significance on customer repurchase intention (Tien et al., 2012), and in va study of Wen et al., (2011), this significant is very low However, in the other hand, n ll fu some studies confirm that trust is still a factor influencing customer repurchase oi m intention because lack of trust could be the main reason customers decide not to shop at is proposed that: nh online or why they have negative concerns related to shopping online So, this study z z Hypothesis 3b: Customer trust is positively associated with customer satisfaction jm ht vb 2.3.4 Hypotheses related to website quality: k Several elements of website quality, such as information quality, response time, and gm visual attractiveness have been verified to be related to perceived usefulness (Liao et om l.c al., 2006) Saeed and Abdinnour-Helm (as cited in Li, 2010) also assert that service quality has impact on users’ extended usage and the exploratory usage of a website via a Lu the perceived usefulness of a website Thus, website quality is expected to have an n influence on user perception of the usefulness of online group-buying context It is y th usefulness te re Hypothesis 4a: Website quality is positively associated with customer perceived n va assumed that: 22 t to ng Website quality is referred to “the consumers’ judgment about a product’s overall hi ep excellent or superiority” (Zeithaml, as cited in Tien et al., 2012) Consumer satisfaction is a transient and experience – specific attitude, which is based on a w consumer’s specific service encounter (Li, 2010) Numerous studies have been n lo conducted to clearly demonstrate the relationship between service quality and ad satisfaction Fornell (as cited in Tien et al., 2012) proposed that website quality y th positively affect overall customer satisfaction Hence, this study proposed: ju yi Hypothesis 4b: Website quality is positively associated with customer satisfaction pl ua al In study of Mcknight et al (as cited in Liao et al., 2006), authors showed that if n consumers perceive that website quality is of high quality, they are likely to have high va trusting beliefs about the web retailer’s competence, integrity, and benevolence; and n ll fu will develop a willingness to depend on the web retailer Zhou et al (as cited in Li, at nh satisfaction So, it is suggested that: oi m 2010) also found that service quality had a stronger impact on consumer trust and Hypothesis 4c: Website quality is positively associated with customer trust z z vb Service quality usually was demonstrated to be an antecedent to satisfaction and jm ht asserts a direct influence on consumer satisfaction (Anderson & Fornell; Sweeney & Soutar, as cited in Li, 2010) However, some prior marketing literature also had k gm evidence showing that service quality affects the purchasing intention Study of Liang &Lai (as cited in Liao et al., 2006) showed that a high quality website not only affects l.c om the customer’s purchase decision, but also is one of the main reasons for consumers to a Lu determine whether they will purchase online or not (Gehrke & Turban, as cited in Liao et al., 2006) Poor quality can result many loss in completion such as loss of n that: y te re leads to behavioral intention to use and reuse of customer So, hypothesis is proposed n va customers, reduction in profits, increasing costs… Therefore, website quality turn repurchase intention th Hypothesis 4d: Website quality is positively associated with online-group-buying 23 t to ng 2.4 Conclusion: hi ep This chapter presented theoretical background of IS repurchase models and website quality concepts and instruments Base on discussion of literature review, online w group-buying had emerged a new e-commerce model Customer retention of online n lo group-buying context was an interesting topic and necessary for development of this ad model in the future Hence, in this study, a research model was proposed with nine y th hypotheses Three cognitive factors – trust, satisfaction and perceived usefulness - was ju yi chosen for this study because they commanded a majority of factors found to affect pl customer retention in prior studies Website quality was also incorporated to research al n va model n ua model The next chapter will discuss about research method used to test this research ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th 24 t to ng CHAPTER hi ep RESEARCH METHOD 3.1 Introduction: w This chapter presents a detailed account of a research methodology of this study n lo First, it discusses data collection method and research process Then, measurement ad scales are presented to develop questionnaire Finally, results of pilot test are discussed y th to contribute for measurement scales This chapter aims at explicating the research ju yi approach choice and presenting the reasons for its use pl al ua 3.2 Research design: n Collecting data process of this study was designed into two stages First was a pilot va test, second was main survey This study was conducted in Ho Chi Minh City, the n ll fu principal business center in Vietnam Sample of this study is selected by using non- oi m probability sampling method – convenience sample Most of respondents of this study nh was full-time and part-time students who were studying at University of Economics at Ho Chi Minh City They also were individuals who had experience in group-buying z z and still had repurchase intention vb The pilot test of this study was conducted to assess the validity and reliability of ht jm the instrument before the questionnaire was distributed The initial data was collected k from a sample 70 randomly; however, only 57 questionnaires were collected The gm pilot study helped ensure that the final questions would be well understood and om l.c attempted to predict an appropriate sample size and improve upon the study design prior to performance of a full-scale research project Simultaneously, cronbach alpha n a Lu and exploratory factor analysis (EFA) was used to test measurement scales Main survey was also quantitative research to collect data for examining research directly and email to respondents After the data collection, total 550 responses were th were 41 free parameters), hence, for the survey, 600 questionnaires were distributed y estimated, the minimum sample size needed for testing overall model was 205 (there te re buying but with a large amount Based on rule of five observations per parameter n va model The sample for this survey was also individuals who had experience in group- 25 t to ng collected, 185 responses were eliminated because respondents indicated that they had hi ep never use online group-buying before or they had no intention to repurchase Finally, 365 responses were used as a valid data for this research In this main study, w exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were used n lo to assess measurement scales The structural equation model (SEM) had been used as ad the main method for analyzing the research model by testing the assumed causation y th among a set of dependent and independent variables Bootstrapping with N = 1000 re- ju yi samples was also used to assess the path significance pl al Literature Review n ua Draft Scale n va ll fu Cronbach Alpha EFA- Exploratory factor analysis oi m Pilot test Quantitative research (n=57) at nh Final Scale z z vb Cronbach Alpha EFA- Exploratory factor analysis (Examine reliability and validity) k jm ht gm Main survey Quantitative research (n=365) om l.c CFA - Confirmatory factor analysis (Examine discriminant and convergent validity) n a Lu n va Composite reliability and Average variance extracted (Examine reliability validity) y th Figure 3.1: Research process te re SEM – Structural Equation Model (Examine research model) 26 t to ng 3.3 Instrument construction: hi ep A paper-based questionnaire was developed to collect data to validate the constructs and theory pointed in the research framework (see in appendix F) This w questionnaire was firstly developed in English, and was translated into Vietnamese n lo later It was divided into three parts ad - The first part of the survey instrument was designed to get information about y th the respondents’ experience and online group-buying habit ju The second part of this survey instrument contained questionnaire items that yi - pl measure five constructs in the proposed model These questionnaire items were al n ua measured using a seven-point Likert scale (from 1- strongly disagree to 7- va strongly agree) These items were selected from many previous related n researches and subsequently modified to fit the online group-buying fu ll experience (Figure 3.2 listed the items for each measure and provided the oi The third part of the survey included questions regarding demographic and social economic status at nh - m sources of measures) z z In this questionnaire, five main constructs were measured in this study, included: vb jm ht online group-buying repurchase intention, satisfaction, perceived usefulness, trust and website quality Among these constructs, website quality was a multi-dimensional k specific content l.c Perceived usefulness items were adapted from Fan et al (2010) These items om - gm construct with four constructs: content quality, technical adequacy, appearance and a Lu were modified from Davis et al.'s (1989) four-item perceived usefulness scale The n first three items of this scale measured the performance productivity, and Satisfaction was measured by items modified from Wen et al (2011) and Li scale, designed to assess users' satisfaction with camcorder use According to th (2010) Three items used originally from Spreng et al.'s (1996) overall satisfaction y - te re assessed overall usefulness n va effectiveness dimensions of online group-buying usefulness, while the fourth item 27 t to ng Ajzen and Fishbein (as cited in Bhattacherjee, 2001), this scale was appropriate hi ep because affect such as satisfaction was best measured along bipolar evaluative dimensions The last item of this scale was taken from Li (2010) to confirm more w clearly customer satisfaction n Trust measured by items modified from Wen et al (2011), original from Gefen lo - ad et al (2003) and Enrique et al (2008) These items measured trustworthy base on y th Online group-buying repurchase intention was measured using five items yi pl - ju ability of protecting customer privacy and providing good service of website ua al adapted from Kim et al (2012) These items were modified from Khalifa and Liu (2007) and Zhou et al (2009) The four initial items measured respondents' n n va intention to continue purchase on the same online group-buying website The fifth item assessed respondents' overall continuance intention to control for potential ll fu Website quality was taken from Aladwani & Palvia (2002) This measurement m - oi scale mostly targeted the web users, so it measured website quality focus on nh at functionality and content of website It included 25 observed variables to measure z four constructs: content quality, specific content, appearance and technical z ht vb adequacy However, in this study, two items – finding firm general information and personalization or customization - were removed when modifying scale for k jm online group-buying context om l.c gm n a Lu n va y te re th 28 t to ng Table 3.1: Sources of questionnaire items hi ep Measure items References PU1 Using website X improves my performance in finding out group-buying Fan et al products (2010) PU2 Using website X increases my productivity in group-buying PU3 Using website X provides me with more diverse channels to enhance my effectiveness in group-buying products PU4 As for overall platform design, I find website X useful Satisfaction SAT5 I was very satisfied with my experience to website X Wen et al (SAT) SAT6 I was very pleased with my experience to website X (2011), Li SAT7 I was absolutely delighted with my experience to website X (2010) SAT8 I think I made a right decision when using website X Trust (TR) TR9 I feel safe on my transaction with website X Wen et TR10 I believe website X can protect my privacy al.(2011) TR11 I select website X, which I believe are honest TR12 I feel that website X would provide me with good service TR13 I feel that website X is trustworthy OGB INT14 I intend to continue to purchase goods from website X Kim et al repurchase INT15 I intend to acquire product information from website X (2012) intention INT16 I intend to recommend website X to people around me (INT) INT17 I intend to use website X as the priority online store for future purchases INT18 Except for any unanticipated reasons, I intend to continue to use website X Webiste CQ19 The content of website X is useful Aladwani & quality CQ20.The content of website X is complete Palvia (2002) CQ21.The content of website X is clear CQ22.The content of website X is current CQ23.The content of website X is concise CQ24.The content of website X is accurate SC25 In website X, one can find contact information SC26 In website X, one can find details about products and/or services SC27 In website X, one can find information related to customers’ policies SC28 In website X, one can find information related to customer service AP29 Website X looks attractive AP30 Website X looks organized AP31 Website X uses fonts properly AP32 Website X uses colors properly AP33 Website X uses multimedia features properly TA34 Website X looks secured for carrying out transactions TA35 Website X looks easy to navigate through TA36 Website X has adequate search facilities TA37 Website X is always up and available TA38 Website X has valid links TA39 Website X loads fast TA40 Website X has many interactive features TA41 Website X is easy to access Note: Website X is online group-buying website which respondent most regularly used Construct Perceived usefulness (PU) w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th 29 t to ng 3.4 Pilot test results: hi ep Measurement scales of pilot test were tested by quantitative research with sample n= 57 These scales were modified by using two methods: (1) cronbach alpha and (2) w exploratory factor analysis (EFA) (used only for perceived website quality scale) n lo Observed variables had cronbach alpha less than 0.6, item-total correlation under 0.3 ad (Nunally and Berstein, as cited in Nguyen & Nguyen, 2008) and factor loading less y th than 0.4 in EFA (Gerberg &Anderson, as cited in Nguyen & Nguyen, 2008) would be ju yi removed After modifying scales, remaining observed variables would be used for pl main survey ua al n 3.4.1 Cronbach Alpha: va Reliability is one of the most critical elements in assessing the quality of the n ll fu construct measures (Churchchill, as cited in Fan et al., 2010) A statically reliable oi m scale provides consistent and stable measures of a construct This study used at and main survey nh Cronbach’s coefficient alpha to evaluate the construct reliability for both pilot test z z The reliability coefficients of seven constructs in pilot test - perceived usefulness, vb repurchase intention, trust, perceived quality with four constructs (content quality, ht jm specific content, appearance, technical adequacy) - were greater than 0.6 (see in Table k 5.1) and were consider acceptable Item-total correlation of observed variables of gm these constructs also satisfied greater than 0.3.Only satisfaction construct was not om l.c accepted because Cronbach alpha was 0.34, less than threshold 0.6; item-total correlation of SAT8 was low with value 0.21; and Cronbach alpha if item deleted was a Lu very high, more than 0.9 Thus, this item needed to remove out of measurement scale n However, this item was used as a suitable scale in many prior researches of customer y te re scale again in main study n va satisfaction; hence, author did not remove this item and would test reliability of this th 30 t to ng Table 3.2: Cronbach alpha result of pilot test hi ep w n Corrected ItemTotal Correlation lo Alpha if Item Deleted ad ju y th yi pl n ua al n va ll fu oi m at nh z z 685 683 673 629 411 421 481 211 242 249 227 919* 781 733 880 800 884 920 929 901 917 900 807 552 682 735 685 808 874 840 827 840 jm ht vb 490 492 510 585 779 810 740 743 792 k 891 883 894 894 887 om 843 836 827 819 n a Lu n va 691 707 729 750 l.c gm th 910 933 907 911 923 y 859 748 872 852 783 te re Observed Scale Mean if Scale Variance if Variable Item Deleted Item Deleted Perceived usefulness (PU): alpha = 0.728 PU1 13.0000 9.571 PU2 13.4211 10.177 PU3 13.0351 9.499 PU4 12.9649 9.320 Satisfaction (SAT): alpha = 0.347*** SAT5 14.4561 79.110 SAT6 14.4211 79.891 SAT7 14.4561 78.574 SAT8 13.5088 17.076 Trust (TR): alpha = 0.930 16.7895 27.741 TR9 17.1053 29.132 TR10 17.2281 25.501 TR11 16.9474 27.301 TR12 16.8421 26.385 TR13 Repurchase intention (INT): alpha = 0.867 19.5439 23.681 INT14 19.7895 26.169 INT15 19.7018 25.892 INT16 19.7719 24.215 INT17 19.5088 25.576 INT18 Content quality (CQ): alpha = 0.909 23.0702 33.888 CQ19 23.3860 31.313 CQ20 23.0000 32.679 CQ21 23.0702 30.781 CQ22 23.1228 32.288 CQ23 Specific content (SC): alpha = 0.868 13.2456 17.260 SC25 13.2632 17.305 SC26 14.0702 16.602 SC27 14.1053 16.453 SC28 Appearance (AP): alpha = 0.932 19.1579 23.707 AP29 19.4211 23.355 AP30 18.9298 23.245 AP31 18.9298 23.566 AP32 19.3509 24.375 AP33 31 t to ng hi ep w n lo Technical adequacy (TA): alpha = 0.927 33.4211 TA34 33.1754 TA35 33.2982 TA36 33.0526 TA37 32.8421 TA38 33.3684 TA39 33.4211 TA40 32.6842 TA41 ad 796 629 826 760 795 733 760 707 913 926 911 917 913 918 916 920 ju y th 64.927 70.826 65.463 69.836 65.850 68.058 68.498 70.684 yi pl 3.4.2 EFA for website quality scale: al ua Exploratory factor analysis is usually used to determine whether the initially n suggested items were derived as a single factor or separately from those measuring va n other concepts In this study, eight constructs were suggested with measurement fu ll scales from different previous researches According to cronbach alpha results, almost m oi of measurement scales were acceptable Author wanted to use EFA to test at nh discriminant validity of measurement constructs, however, sample size of pilot test is too small, so, using EFA to test require a strictly conditions EFA of this testing was z z used as a reference for main survey Hence, EFA was used only for website quality vb jm ht construct because among measurement scales of this study, website quality was a multi-dimensional construct; its measurement scale was taken from Aladwani & k gm Palvia (2002) Moreover, when using this scale, author removed two items, so EFA was necessary to use in pilot test to assess validity and reliability of this modified om l.c measurement scale Results of EFA for website quality showed that four factors were extracted with a Lu eigen-value 0.74 and extraction sums of squared loadings more than 70% In general, n were more than 0.4 Thus, no variables of perceived quality scale were removed th examine in main study y with original scale (see in table 3.3) Reliability and validity of these scales needed to te re However, many observed variables had factor loading on different factor compare n va extracted variance was acceptable because factor loading of all observed variables 32 t to ng Table 3.3: EFA result of pilot test hi Observed Variable Factor ep w n lo ad ju y th AP32 985 AP31 903 AP29 659 AP33 TA38 597 556 SC26 AP30 497 441 305 408 409 378 292 820 736 yi CQ22 CQ19 pl al CQ23 681 ua CQ20 n 489 274 n 657 274 512 414 412 808 668 ll fu 448 m TA40 SC27 va SC25 CQ21 oi SC28 TA36 407 at nh TA34 CQ24 TA37 -.267 388 TA41 TA35 Eigen value Extracted Variance 317 510 321 499 289 758 681 233 1.157 57.566 5.591 5.029 3.222 om l.c 3.5 Conclusion: 546 429 741 gm 13.240 305 1.286 k jm ht vb -.226 315 z 254 z TA39 635 629 This chapter presented research method used to test research model, measurement a Lu scale construction and results of pilot test This study was designed into two phases: n first was a pilot test, second was main survey A paper-based questionnaire was present data analysis results of main survey th 365 respondents and used CFA and SEM as main method The next chapter will y consequently, they were used for the main survey Main survey had sample size with te re test indicated that almost items measuring the constructs were acceptable; n va developed to collect data; it was distributed directly and email to respondents Pilot 33 t to ng CHAPTER hi ep ANALYSIS AND RESULTS 4.1 Introduction: w Chapter presented research methodology to examine scale validation and n lo research model In chapter 4, results of study will be shown and analyzed with sample ad n=365 Firstly, respondents demographic are analyzed by using the SPSS – Statistical y th software package Secondly, results of scale validation are presented In this section, ju yi scale in this research was evaluated through two steps Step uses cronbach alpha pl and exploratory factor analysis (EFA) to examine reliability and validity In step 2, al n ua CFA – confirmatory factor analysis, composite reliability (CR) and the average va variance extracted (AVE) are used to examine discriminant and convergent validity n Thirdly, research model and hypotheses are examined by using SEM – structural fu ll equation model and bootstrap estimate in AMOS software package Finally, results of oi at nh 4.2 Respondents demographic: m hypotheses testing are discussed z z The collected data were analyzed using the SPSS – Statistical software package vb The results of the demographic analysis were shown in table 4.1 Initial analysis of ht jm data indicated that gender was relatively equally represented with 59.2% of k respondents were female and 40.8% male Age ranged between 18 to 40 years old, gm with 21.9% of the respondents between 18 and 22 years old, 69.6% of respondents om l.c between 22 and 30 years old, and 8.5% aged over 30 years old The academic attainment of the respondents was relative high, with 50.1% of respondents having a Lu completed that at least a university education, 47.2% of respondents having education n level under bachelor, only 2.7% having higher education Income representation was y te re million (17.3%) n va divided between those earning: less than 10 million VND (82.7%), more than 10 th 34 t to ng Table 4.1: Respondents demographic hi Demographic profile Gender Category ep w n lo Age ad ju y th Income yi pl al n ua Education Frequency n va Female Male Total 18-22 23-30 Over 30 Total Less than 10 million More than 10 million Total High school College Bachelor Master & PhD Total 302 82.7 63 17.3 365 32 140 183 10 365 100.0 8.8 38.4 50.1 2.7 100.0 ll fu 216 149 365 80 254 31 365 Percentage (%) 59.2 40.8 100.0 21.9 69.6 8.5 100.0 m oi 4.3 Scale validation: Scale in this research was evaluated through two steps CFA – confirmatory factor nh at analysis was used as main method to examine measurement scales because all of the z z latent constructs and corresponding measurements are derived from previous research vb and their reliability and validity were shown to be acceptable in prior papers ht l.c gm 4.3.1 Preliminary results: k confirm reliability and validity jm Composite reliability and the average variance extracted (AVE) was also calculated to om In pilot test, all measurement scales was tested with cronbach alpha index Almosr a Lu constructs were acceptable However, one construct had lower cronbach alpha value, n hence, main survey also use cronbach alpha index to examine reliability of all scales n va in research model y te re th 35 t to ng Table 4.2: Cronbach alpha result hi ep Observed Variable Scale Mean if Item Deleted w n lo ad Scale Variance if Corrected Item- Cronbach's Item Deleted Total Alpha if Item Correlation Deleted Perceived usefulness (PU): alpha = 0.762 PU1 12.92 10.801 538 717 PU2 13.28 11.214 505 734 PU3 12.68 9.811 627 668 PU4 12.59 10.363 571 699 Satisfaction (SAT): alpha = 0.902 SAT5 13.08 12.699 809 863 SAT6 13.06 12.626 837 853 SAT7 13.08 13.175 766 879 SAT8 13.18 13.702 713 898 Trust (TR): alpha = 0.903 16.77 23.781 700 894 TR9 16.88 23.416 706 893 TR10 16.83 22.623 835 865 TR11 16.69 23.985 772 879 TR12 16.72 23.290 783 876 TR13 Repurchase intention (INT): alpha = 0.901 18.13 26.347 797 870 INT14 18.32 28.301 604 912 INT15 18.21 26.704 786 872 INT16 18.24 26.203 792 871 INT17 18.12 26.443 800 869 INT18 Content quality (CQ): alpha = 0.903 23.08 29.098 680 894 CQ19 23.20 27.754 785 879 CQ20 23.00 27.585 784 879 CQ21 22.92 27.719 791 878 CQ22 23.01 28.016 762 882 CQ23 Specific content (SC): alpha = 0.687 13.86 22.177 572 588 SC25 13.65 15.096 344 848 SC26 14.37 21.300 630 556 SC27 14.30 21.567 625 562 SC28 Appearance (AP): alpha = 0.908 19.11 21.414 725 897 AP29 19.17 20.790 755 890 AP30 18.94 20.406 837 873 AP31 18.85 20.850 810 879 AP32 19.09 21.403 716 898 AP33 ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th 36 t to ng hi ep w n lo ad Technical adequacy (TA): alpha = 0.927 33.57 TA34 33.28 TA35 33.40 TA36 33.19 TA37 33.07 TA38 33.30 TA39 33.38 TA40 y th TA41 33.05 61.141 60.941 60.609 60.833 59.872 60.670 62.385 683 729 782 797 780 761 717 923 919 915 914 915 917 920 59.920 766 916 ju yi The resulting cronbach alpha values ranged from 0.76 to 0.93, which was above pl ua al the acceptable threshold, suggested by Nunnally and Berstein (as cited in Nguyen & Nguyen, 2008) Item-total correlation of all observed variables also satisfied greater n n va than 0.3 (Nunally and Berstein, as cited in Nguyen & Nguyen, 2008) Satisfaction fu construct in main study was accepted because cronbach alpha was 0.902, greater than ll 0.6 SAT8 was not removed out of measurement scale m oi After testing with cronbach alpha, EFA also used to test for all of items in nh measurement scales In pilot test, author only tested website quality construct, hence, at z discriminant validity of other construct was not examined Moreover, EFA can help z vb CFA testing more convenient, so, this method was necessary for main survey jm ht Results of EFA for all variables in research model were shown in appendix B k KMO of this analysis was very high 0.951, extraction sums of squared loadings = gm 58.564%, more than 50% However, compare to research model, there were six l.c factors after reduction by using principal axis factoring method with promax rotation om Factor loadings of many observed variables were less than 0.5 According Hair et al a Lu (as cited in Nguyen, 2009), factor loadings need greater than 0.5 to have practical n meaning Thus, some observed variables were removed out of measurement scales va Firstly, repurchase intention scale removed one item INT15 due to factor loading n th combine Thirdly, trust scale had five factors, from TR9 to TR12 Factor loadings of y perceived usefulness - had high factor loading in same factor, so these items had to te re was less than threshold 0.5 Secondly, all of items of two constructs - satisfaction and 37 t to ng these observed variables were above the acceptable threshold and loaded to its hi ep respective construct No items must be removed from measurement scale Table 4.3: EFA result of main study w n lo ad ju y th Factor loading yi pl n ua al 208 290 n va ll fu 710 691 672 659 637 616 598 565 -.305 -.253 oi m 297 220 324 at nh 912 794 669 662 642 z z k jm ht vb 801 728 711 698 682 gm 14.242 2.385 1.181 990 865 772 711 631 680 45.941 7.695 3.809 3.194 2.194 931 885 903 904 912 l.c 209 om n a Lu n va TA41 TA38 TA37 TA36 TA39 TA40 TA35 AP33 AP31 SAT5 PU1 PU4 PU3 PU2 SAT6 SAT7 SAT8 TR11 TR13 TR9 TR10 TR12 CQ22 CQ21 CQ19 CQ20 CQ23 INT18 INT16 INT17 INT14 Eigen value Extracted Variance Cronbach α 867 848 839 832 755 696 694 529 524 th scale were removed, so this construct was not used to measure perceived website y because of lower factor loading of all items in scales Almost of items of appearance te re Related to scales of website quality, specific content construct was removed 38 t to ng quality in this study Technical adequacy scale removed TA 34 and added two hi ep additional observed variables AP33 and AP31 Content quality scale also removed one item CQ24 w n In conclusion, there were 10 removed items after testing by EFA Results of EFA lo ad were presented in table 4.3 y th 4.3.2 Confirmatory factor analysis (CFA): ju yi This section presents result of confirmatory factor analysis (CFA) of measurement pl scale by using AMOS 16 software package When using CFA to examine, the al ua following five indices (Nguyen & Nguyen, 2008) were adopted : (1) The Chi – square n value normalized by degree of freedom (χ2/df) should be less than 3; (2) Goodness-of- va n fit (GFI) values of greater than 0.9 typically were consider good; (3) Tucker & Lewis fu ll index (TLI) value should exceed 0.9; (4) Comparative fit index (CFI) values above m oi 0.9 are usually related to model that fits well; and (5) The Root mean square of at nh approximately (RMSEA) value should be between 0.03 and 0.08 Other assessment indices are also used to examine: (1) The composite reliability; z z (2) Average variances extracted (AVE); (3) Unidimensionality; (4) Convergent vb jm ht validity; (5) Discriminant validity The composite reliability of all latent constructs exceeded recommended level of 0.7 The average variances extracted, which reflect k gm overall amount of variance in the indicators accounted for the latent construct, recommended level of 0.5 Convergent validity is the degree to which multiple l.c om attempts to measure the same concept in agreement Convergent validity was assessed based on factor loading The factor loading for all items should exceed the a Lu recommended level of 0.5 Discriminant validity is the degree to which the measures n value of correlation between constructs, smaller than 0.05 was acceptable th statistical significance and assessing the construct’s reliability and variance extracted y Each construct was evaluated separately by examining the indicator loadings for te re 4.3.2.1 CFA results: n va of different concepts are distinct Discriminant validity can be examined by using p- 39 t to ng  Online group-buying repurchase intention: hi ep Online group-buying repurchase intention was measured by original scale with five items, INT14 to INT18 When using EFA, observed variable INT15 was w removed because factor loading was lower than 0.5 Model with four items tested by n lo CFA was shown in figure 4.1 ad ju y th Chi¬-square= 6.854 ; df= ; P= 032 Chi-square/df= 3.427 GFI= 991 ; TLI= 985 ; CFI= 995 RMSEA= 082 yi e17 e19 pl 70 e20 69 al INT14 INT16 INT17 ua 83 84 e21 73 85 77 INT18 88 n n va Repurchase Intention fu ll Figure 4.1: CFA model of online group-buying repurchase intention scale m oi See figure 4.1, this model had square value normalized by degree of freedom at nh (χ2/df) 3.427 more than RMSEA value was 0.082, more than 0.08, however, it was z acceptable because according to Anderson and Gerbing (as cited in Udo et al., 2010), z vb RMSEA could be less than 0.1 Other indices of this model were good GFI, TLI and jm ht CFI values above 0.9, composite reliability was 0.912, more than 0.7; variance extracted was 0.722, more than threshold All factor loadings of this model had value k om l.c  Trust: gm higher than 0.5 In general, this model fitted well and had unidimensionality Trust scale was measured with five observed variables, TR9 to TR13 When a Lu analyzing by CFA, the Chi – square value normalized by degree of freedom (χ2/df) of n this model was 5.76, slightly more than 5; all of indices GFI, TLI and CFI exceeded y th removed Result of modified model was shown in figure 4.2 te re from 0.73 to 0.89 With this CFA result, model needed to modify, thus, TR10 was n va 0.9; RMSEA value was 0.11 more than 0.08 The factor loadings for all items ranged 40 t to ng Chi¬-square= 2.044 ; df= ; P= 360 Chi-square/df= 1.022 GFI= 997 ; TLI= 1.000 ; CFI= 1.000 RMSEA= 008 hi ep e17 e19 e20 w 52 e21 74 n lo TR9 72 TR11 TR12 86 ad 72 85 75 TR13 86 ju y th Trust Figure 4.2: CFA model of trust scale yi pl After removed TR10, the Chi – square value normalized by degree of freedom al n ua (χ2/df) of this model was 1.022, less than 3; GFI values was 0.997, greater than 0.9; TLI and CFI exceeded 0.9; RMSEA value was 0.008, satisfied condition The va n composite reliability of this model was 0.896 and variance extracted was 0.683, more fu ll than recommended level The factor loadings for all items ranged from 0.72 to 0.86 m oi In general, this model fitted well and also had unidimensionality at nh  Satisfaction: When analyzing by EFA method, perceived usefulness and satisfaction scale were z z combined However, when using CFA, regression estimates in standardized model of vb jm ht some items PU1, PU2, PU3 less than 0.5 and model indices were not acceptable Thus, observed variables PU1, PU2, PU3 were removed out of measurement scale k gm There were five items in new scale includes: PU4, SAT5, SAT6, SAT7 and SAT8 Result of modified model was presented in figure 4.3 om l.c n a Lu Chi¬-square= 12.382 ; df= ; P= 030 Chi-square/df= 2.476 GFI= 986 ; TLI= 987 ; CFI= 994 RMSEA= 064 SAT5 SAT6 e3 e5 e4 n e2 va e1 PU4 70 Figure 4.3: CFA model of satisfaction scale th Satisfaction y 91 SAT8 74 te re 89 SAT7 80 41 t to ng The Chi – square value normalized by degree of freedom (χ2/df) of this model was hi ep 2.476, less than 3; GFI values greater than 0.9; TLI was 0.987, and CFI was 0,994 exceeded 0.9; RMSEA value was 0.064 This new model fitted well The composite w reliability of this model was 0.905 and average variances extracted was 0.659, more n lo than recommended level The factor loadings for all items ranged from 0.70 to 0.91, ad more than 0.5 This model also had unidimensionality y th  Website quality: ju yi Website quality was a multi-dimensional construct, included: content quality, pl specific content, appearance and technical adequacy After analyzing validity with al n ua EFA, specific content construct and appearance construct were removed out of va measurement scale of website quality EFA results showed only two constructs - n content quality and technical adequacy - were continually tested by CFA method fu ll When tested by using CFA, some observed variable were removed such as CQ22, m oi AP31, and AP33 This modification helps to improve indices of model and increased nh at reliability validity Results were shown in figure 4.4 z Chi¬-square= 123.607 ; df= 43 ; P= 000 Chi-square/df= 2.875 GFI= 941 ; TLI= 961 ; CFI= 969 RMSEA= 072 z 76 70 n va y te re 79 75 81 n Technical adequacy a Lu 83 75 80 83 om l.c gm Content quality k TA35 TA36 TA37 TA38 TA39 TA40 TA41 84 jm e5 e6 e7 e8 e9 e10 e11 75 85 ht CQ19 CQ20 CQ21 CQ23 vb e1 e2 e3 e4 th Figure 4.4: CFA model of website quality scale 42 t to ng The Chi – square value normalized by degree of freedom (χ2/df) of this model is hi ep 2.875, less than 3; GFI was 0.941, more than 0.9, TLI was 0.961, and CFI was 0,969, both of them were more than 0.9; RMSEA value was 0.072 less than 0.08 The factor w loadings for all items ranged from 0.75 to 0.84, more than 0.5 Hence, this scale had n lo convergent validity In summary, this model was acceptable ad Table 4.4: Confirmatory factor analysis of measurement model Factor Loading ju y th Items yi pl Composite Reliability Standardized Cronbach α 0.912 0.912 0.683 0.896 0.893 0.905 0.904 n ua al Repurchase intention 0.836 INT14 0.828 INT16 0.854 INT17 0.880 INT18 Trust 0.722 TR9 0.862 TR11 0.851 TR12 0.863 TR13 Satisfaction 0.887 SAT5 0.910 SAT6 0.798 SAT7 0.744 SAT8 0.701 PU4 Website quality Content quality 0.759 CQ23 0.840 CQ21 0.852 CQ20 0.748 CQ19 Technical Adequacy 0.828 TA38 0.834 TA37 0.802 TA36 0.746 TA35 0.794 TA39 0.754 TA40 0.809 TA41 Variance extracted (AVE) 0.722 n va ll fu oi m at nh 0.659 z z jm ht vb 0.887 0.633 0.924 0.876 k 0.642 l.c gm 0.923 om n a Lu n va y te re extracted is 0.642 The composite reliability of technical adequacy scale was 0.924 th The composite reliability of content quality scale was 0.887 and average variances 43 t to ng and average variances extracted is 0.633 These results concluded that content quality hi ep and technical adequacy construct of perceived website quality scale satisfied reliability standardize Moreover, CFA results also showed that correlation r(CQ,TA) w between two constructs was 0,705 with se = 0.037 This value was different with n lo significant p-value = 0.000 Hence, content quality and technical adequacy had ad discriminant validity ju y th 4.3.2.2 Saturated model: yi pl To test discriminant validity for all constructs of research model, a saturated ua al model was generated Saturated model can be defined as one in which all parameters n relating to the constructs to one another are estimated (Anderson & Gerbing, 1988) va Saturated model of this study had 242 degree of freedom The Chi–square value of n ll fu this model was 451.258, chi-square normalized by degree of freedom (χ2/df) was oi m 1.865 with p value =0.000 GFI = 0.908, TLI = 0.964 and CFI = 0.968, all indices nh were more than 0.9; RMSEA value was 0.049, less than 0.08 The factor loadings for at all items higher than 0.5, ranged from 0.71 to 0.90 In summary, this saturated model z z was fitted with data market (see in figure 4.5) k jm ht vb om l.c gm n a Lu n va y te re th 44 t to ng e1 hi ep INT14 85 e2 INT16 83 e3 INT17 e4 INT18 Chi¬-square= 451.258 ; df= 242 ; P= 000 Chi-square/df= 1.865 GFI= 908 ; TLI= 964 ; CFI= 968 RMSEA= 049 Repurchase intention 86 86 w n 80 lo ad e5 TR9 TR11 e7 TR12 e8 TR13 ju y th e6 73 86 86 85 Trust 71 yi pl 69 al CQ19 e10 CQ20 85 84 e11 CQ21 va 75 ua e9 76 e12 CQ23 65 72 n Content quality n 77 62 ll fu 63 SAT8 e17 PU4 e18 TA35 e19 TA36 e20 TA37 e21 TA38 e22 TA39 e23 TA40 e24 TA41 71 49 75 80 om n a Lu Figure 4.5: Saturated model of main survey l.c gm Technical Adequacy k 84 83 80 75 80 jm ht vb e16 Satisfaction z SAT7 71 z e15 88 90 80 76 at SAT6 nh e14 oi SAT5 m e13 va Besides results of saturated model, to test for discriminant validity, the y th and p-value is very small, less than 0.000 (see in table 4.5) Hence, five constructs - te re presented that relationship between constructs in research model were different 1.00 n correlations between the variables need to be considered SEM analysis result 45 t to ng website quality (content quality and technical adequacy), perceived usefulness, trust hi ep and online group-buying repurchase intention - had discriminant validity Table 4.5: Relationship between constructs of research w n lo ad r 0.804 0.712 0.759 0.651 0.690 0.720 0.619 0.625 0.706 0.494 ju y th Relationship INT ↔ TR INT ↔ CQ INT ↔ PU INT ↔ TA CQ ↔ TR PU ↔ TR TA ↔ TR PU ↔ CQ TA ↔ CQ TA ↔ PU yi pl 1-r 0.196 0.288 0.241 0.349 0.310 0.280 0.381 0.375 0.294 0.506 c.r 6.280 7.814 7.052 8.760 8.160 7.687 9.243 9.153 7.909 11.088 p 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 n ua al s.e 0.031 0.037 0.034 0.040 0.038 0.036 0.041 0.041 0.037 0.046 va n 4.4 Modified research model: ll fu After examining measurement scale by using EFA and CFA, almost observed oi m variables of perceived usefulness scale were removed, except PU4 was kept with four nh items of satisfaction scale Hence, perceived usefulness construct was removed out of at research model due to unfitted measurement scale With this modification, PU4 z z became an item to measure satisfaction construct and cognitive factors in research vb model only were satisfaction and trust Website quality, according to test results, was ht k was modified with follow hypotheses: jm measured by two constructs: content quality and technical adequacy Research model gm Hypothesis 1: Customer satisfaction is positively associated with online group- om l.c buying repurchase intention Hypothesis 2a Customer trust is positively associated with online group-buying n a Lu repurchase intention Hypothesis 3a: Website quality is positively associated with customer satisfaction th repurchase intention y Hypothesis 3c: Website quality is positively associated with online group-buying te re Hypothesis 3b: Website quality is positively associated with customer trust n va Hypothesis 2b: Customer trust is positively associated with customer satisfaction 46 t to ng hi Customer Satisfaction ep H1 H3a w Website QualityH5 n H2b lo OGB Repurchase Intention ad H3c Content quality y th H2a ju Technical adequacy Customer Trust yi H3b pl ua al n Figure 4.6: Modified research model n va ll fu 4.5 Model fitness: oi m After checking the validity, hypotheses of this study was tested with the structural equation model Figure 4.7 showed the results of the structural model, including the nh at path and their standardized regression estimates z z om l.c 0.50 OGB Repurchase Intention a Lu Website Quality 0.38 n 0.79 0.78 y te re Customer Trust n 0.30 va Technical adequacy 0.29 0.28 0.89 gm Content quality k Customer Satisfaction jm ht vb Chi¬-square= 456.664 ; df= 244 ; P= 000 Chi-square/df= 1.872 GFI= 907 ; TLI= 964 ; CFI= 968 RMSEA= 049 th Figure 4.7: SEM result of research model (Standardized) 47 t to ng The observed normalized Chi squared for measurement model was 1.872 (chi hi ep squares = 456.664, df = 244, p-value = 0.000) which was smaller than recommended Other fit indices also showed good fit for the measurement model The w goodness-of-fit index (GFI) is 0.907, which exceeds the recommended cut-off level of n lo 0.8 The comparative fit index (CFI) is 0.964, Tucker & Lewis index (TLI) is 0.968, ad greater than the 0.9 recommended The root mean square error (RMSEA) is 0.049, y th exceeding the recommended cut-off level of 0.08 recommended The combination of ju yi these results suggests that the demonstrated measurement model fits the data to a pl reasonable degree ua al Table 4.6: Relationship between constructs in research model (standardized) n Estimate 0.778 0.284 0.499 0.381 0.301 0.287 n va ll fu oi m se 0.069 0.089 0.086 0.080 0.075 0.057 cr 12.03 3.103 6.515 5.437 4.760 4.470 p- value 0.000 0.000 0.000 0.000 0.000 0.000 at nh Path WQ  TR WQ  SAT TR  SAT WQ  INT TR  INT SAT INT z 4.6 Bootstrap: z Analyzing with structural equation modeling (SEM) usually request a large sample vb jm ht but it also cost many time and money (Anderson & Gerbing, as cited in Nguyen & Nguyen, 2008) Bootstrap is a suitable method to replace (Schumacker & Lomax, as k gm cited in Nguyen & Nguyen, 2008) This study used bootstrap estimate with sample N= 1000 Results were presented in table 4.7 Bias of these results were very small, thus, y th se (bs) 0.001 0.004 0.004 0.003 0.003 0.003 te re Bootstrap Estimate se se(se) bs 0.042 0.001 -0.001 0.113 0.003 0.001 0.114 0.003 0.107 0.002 -0.003 0.102 0.002 0.009 0.099 0.002 -0.006 n Mean 0.776 0.285 0.498 0.377 0.31 0.28 va WQ  TR WQ  SAT TR  SAT WQ  INT TR  INT SAT INT ML Estimate ML se 0.778 0.069 0.284 0.089 0.499 0.086 0.381 0.080 0.301 0.075 0.287 0.057 n Path a Lu Table 4.7: Bootstrap estimate result with N = 1000 om l.c estimates in this model had reliability validity 48 t to ng 4.7 Hypotheses testing: hi ep According discussion in chapter 3, there were nine proposed hypotheses However, the analytical results showed some measurement scales did not fit with data, so, in w modified research, only six hypotheses were measures Results of these hypotheses n lo testing presented that all of them were supported ad Table 4.8: Result of hypotheses testing y th ju Hypothesis yi pl n ua al H1 H2a H2b H3a H3b H3c Path Results Supported Supported Supported Supported Supported Supported n va SAT  INT TR  INT TR  SAT WQ  SAT WQ  TR WQ  INT ML estimate 0.287 0.301 0.499 0.284 0.778 0.381 Hypothesis posited that customer satisfaction was positively associated with fu ll online group-buying repurchase intention The result showed that regression estimate m oi of relationship between satisfaction and online group-buying repurchase intention was nh 0.287 with se = 0.057 This estimate had p-value = 0.000 (see in table 4.6) Thus, this at z hypothesis was supported Simultaneously, impact of satisfaction was confirmed z jm ht to customer behavior vb significantly to customer repurchase intention Increasing satisfaction would influence k Hypothesis 2a and 2b proposed that customer trust was positively associated with gm online group-buying repurchase intention and customer satisfaction Regression om l.c estimate of relationship between trust and online group-buying repurchase intention was 0.301 with se = 0.075, p-value = 0.000, while regression estimate of relationship a Lu between trust and satisfaction was 0.499 with se = 0.086, p-value = 0.000 (see in n table 4.6) These results suggested that trust had impact on both satisfaction and th were supported y on customer repurchase intention with β = 0.444, all hypotheses related to customer trust te re repurchase intention through satisfaction (βindirect,TR->INT = 0.143) Thus, trust had effect n simultaneously, this factor was also had slightly the indirect impact on customer va customer repurchase intention Trust had a significant effect on customer behavior, 49 t to ng Hypothesis assumed that website quality had both a direct effect on online hi ep group-buying repurchase intention, satisfaction and trust, as well as indirect effect on customer repurchase intention The hypothesized paths between these variables were w all positive and significant The standardized coefficient from website quality to n lo online group-buying repurchase intention was 0.381 with se = 0.080 and p-value = ad 0.000 This emphasized the important role of website quality in online group-buying y th Also, the path between website quality and satisfaction is statistically significant with ju yi standardized regression coefficient of 0.284 with se = 0.089 and p-value = 0.000 pl Implication of this result is that website quality had direct impact on customer al n ua satisfaction and slightly indirect impact on repurchase intention through satisfaction va Moreover, the result showed that regression estimate of relationship between website n quality and trust was also 0.778 with se = 0.069 and p-value = 0.000 With this result, fu ll website quality seemed to have a strongly influence on customer trust, thus, affected m oi on customer repurchase intention and customer satisfaction In conclusion, all at nh hypotheses related to website quality were supported Among three antecedents to predict online group-buying repurchase intention, website quality had the strongest z jm ht vb 4.8 Conclusion: z impact with β PQ->INT = 0.808 k This chapter presented data analysis results of measurement scales, research model gm and hypotheses Results of this study indicated that almost measurement scales needed om l.c to modify to fit with market data, research model also needed to modify with fewer constructs After analyzing scales and model, new research model was examined by a Lu using structural equation model (SEM) and results were good All hypotheses of new n research model were supported The next chapter summaries all discussion and n va conclusion of this study, as well as its implications and its limitations y te re th 50 t to ng CHAPTER hi ep CONCLUSIONS AND LIMITATIONS 5.1 Introduction: w n This section presents the contributions to and implications for research and lo ad practice First, the theoretical contribution for the research stream IS continuance are y th presented Followed is the contribution and managerial implications for practice ju Finally, the limitations are discussed and directions for future research are suggested yi pl ua al 5.2 Discussion and conclusion: The purpose of this study to examine thoroughly the relationship between website n va quality, trust, satisfaction, perceived usefulness and online group-buying repurchase n intention Hence, a model for predicting customer repurchase intention can be fu ll proposed in online group-buying context m oi Some results of this study could be summarized after analyzing research data nh First is about measurement scales and second is research model at z Measurement scales in this research are adapted from previous research and are z ht vb used to measure in Viet Nam market Initial scales include forty one observed jm variables After examine by using EFA and CFA, many items are removed Online k group-buying repurchase intention scale has to remove one item Trust scale is also gm removed one item to make model more fitted Satisfaction scale and perceived om l.c usefulness scale has high factor loading in the same factor, three observed variables of perceived usefulness scale are removed Only one item is kept to measure in a Lu satisfaction scale Among measurement scales, only website quality scale is multi- n dimensional scale with four constructs: content quality, specific content, appearance measure technical website quality Content quality scale must be also removed two th when test by CFA method With this modification, there are seven items are used to y appearance scale are removed, except two items, however, these items are removed te re this construct are not used to measure in this study Like specific content, almost of n va and technical adequacy Specific content due to low factor loading in EFA testing, so 51 t to ng item In general, measurement scale in this study need to modified a lot to be suitable hi ep with data of Viet Nam market This result can be a good reference for future research related to customer retention in online group-buying context w n Besides scale results, this study also show that individual user’ intention to lo ad repurchase in online group-buying websites is mainly motivated by trust, satisfaction y th and website quality These three factors have quite equally impact on group-buying ju repurchase intention Among them, website quality has the strongest influence, yi followed by trust and satisfaction Moreover, the results indicate that trust also pl ua al indirect impact on repurchase intention through satisfaction Trust has strong influence to satisfaction, increase trust will increase customer satisfaction when n n va purchasing Website quality is also found to be a significant motivator trust and fu satisfaction Website quality has very strong impact on trust; while influence of ll quality to customer satisfaction is not too high oi m nh With these results, this study is consistent with prior researches about IS at continuance model when find that satisfaction have a significantly effect on z z repurchase intention Influence of trust on customer behavior is also confirmed its ht vb meaning in Vietnam online group-buying market Relationship between trust and jm satisfaction often changes from one study to another (Cho, Kwon & Lee, Corritore et k al., Kim et al., Wu & Chang, Yousafzai et al., as cited in Kim et al., 2011) It remains gm unclear whether consumers are satisfied because they trust online shopping, or if they om l.c report improved trust because they are satisfied with internet shopping (Kim et al., 2011) In this study, when measuring this relationship in online group-buying context, a Lu trust is found that it has a strong impact on customer satisfaction Moreover, trust is n also a significant motivator on customer repurchase intention Hence, although this th which is examined not much in previous studies Website quality is a vastly concept y This study also confirmed relationship between website quality and other factors te re with customer attitude in this study n va effect is not consistent in prior researches, role of trust can be denied in relationship 52 t to ng and multi-dimensional construct Many researchers try to propose different hi ep measurement scales to measure this concept In this study, four constructs of website quality is used to measure customer cognition to quality of online group-buying w websites However, results of study specified that content quality and technical n lo adequacy have impact on user perception of website quality Website quality ad assessing by these construct have strongest impact on customer repurchase intention y th In previous studies, there are a few studies examine directly relationship between ju yi website quality and customer behavior They focus on impact of this factor through pl customer attitude or customer belief Hence, this study may contribute to the al n va Vietnam n ua theoretical framework for website quality in the context of online group-buying in ll fu 5.3 Managerial implications: oi m This research makes important contributions to IS continuance research It nh addresses the limitation of ECT in predicting IS continuance by introducing perceived at service quality as an additional predictor of IS continuance intention Besides that, the z z results of this study offer some important implications for practitioners who prepared vb strategic plans and implement tools to improve the performance of their online group- k jm ht buying websites as well: gm First, this study can help online group-buying companies to fully understand the crucial factors that determine the customers’ repurchase intention behavior, which l.c om will allow them to improve their managerial and IT strategies, and increases profits This result highlights important of website quality, satisfaction and trust in predicting n a Lu the repurchase intention to use online group-buying websites especially focus factors influence the feeling or experiences customer have while th Consequently, online group-buying websites need to differentiate more and more, y between websites due to highly competitive markets among top-tier websites te re existed customers Consumers today see few major differences in quality or function n va Second, group-buying websites must provide good online website quality to retain 53 t to ng engaging with the websites Technical adequacy is the most critical and equally hi ep important facet of the group-buying website quality Managers must commit to maintain system operation well and make the group-buying website easy and quick to w be used When shopping online, one of problems which customers afraid of is the loss n lo of personal data and perceived risk in security Hence, provide secure system and ad secure payment mechanism is very necessary for online group-buying Besides that, y th online group-buying website also must ensure content quality Customers usually pay ju yi attention on websites which provide more information with highly reliability and pl accurate Invest in content quality will increase quality of website and can be attract al n ua more new customers in competition market n va Third, this study indicates that trust is a predictor as well as a factor influencing on fu customer repurchase intention directly and indirectly Thus, managers who run online ll group-buying websites should pay attention to improve customers’ level of trust and m oi customer satisfaction – a mediator of trust In modern society, although companies nh at have enthusiastically used internet as a key marketing tool and sales vehicle for their z products and services (Smith, as cited in Kim et al., 2011), many people not trust z vb in e-commerce security They are reluctant to release their personal information to a jm ht website, especially in Vietnam where institutions and infrastructure conducive to trust k has not been well developed Hence, play a high priority on increasing customer trust gm becomes more necessary to motivate customer purchasing behavior in Vietnam om l.c market Fourth, this research also confirmed role of satisfaction in predicting customer a Lu repurchase intention Having satisfied customers is an antidote against IS n discontinuance for online group-buying websites Customers will discontinue using satisfaction with their provided products and services They need to improve their th online group-buying companies should devote themselves to make customers feel y websites when they feel satisfaction with it Thus, in order to retain existing customer, te re Conversely, they will repurchase products or services of online group-buying n va an online websites if they are not satisfied with it, even if it is useful or well designed 54 t to ng performance to adjust with customer expectation, as well as increasing customer trust hi ep and loyalty to online group-buying websites w 5.4 Limitations and future research: n This study has offered some valuable insight into online group-buying studies lo ad However, this study involves a number of limitations that need to be acknowledged y th First, the empirical study was conducted only in Viet Nam; especially, data are ju yi collected in Ho Chi Minh City Thus, data results mainly reflect customer behaviors pl in Vietnam Author recommended replicating the study in different nations to get al n ua international sample va Second, respondents answered these questions based on various group-on n websites rather than responding to questions about a specific websites So the fu ll business type of the website and the distinctive designs that may affect customers’ m oi experience and perceptions of online group-buying nh at Third, the online group-buying customers may have different shopping intention z They can be interested in different kind of products so they have different motivation z ht vb and perception when buying in group-buying website jm Fourth, the current study uses measurement scales from prior researches These k scales were used in many countries, but not in Viet Nam They need to be test and gm modify before using them to examine empirical study However, because of limited om l.c time, this study did not implement qualitative research to test scale Hence, many observed variables were removed out of research model, especially, perceived a Lu usefulness and satisfaction constructs did not have highly discriminant validity n factor which influenced customers’ behavior in many previous studies However, y te re online repurchase intention because website quality had been regarded as a crucial n va Finally, in this study, website quality is incorporated to cognitive process leading development of enhancements to the expanded TAM and ECT th beside this factor, there are many opportunities for future research to continue t to ng REFERENCES hi ep Aladwani, A.M, Palvia, P.C (2002) Developing and validating an instrument for w n measure user-perceived web quality Information and Management, 39 (6), lo ad 467-476 y th ju Al-maghrabi, T., Dennis, C., Halliday, S.V (2010) Adapting Tam and ECT: yi continuance intention of E-shopping in Saudi Arabia Conference paper pl al n va Systems, UAE n ua conducted at European and Mediterranean Conference on Information ll fu Anderson, J.C & Gerbing, D.W (1988) Structural equation modeling in practice: oi m A review and recommended two steps – approach Psychological Bulletin, at nh 103 (3), 411 – 423 z Bhattacherjee (2001) Understanding information systems continuance: An z vb expectation-confirmation model MIS Quarterly, 25(3), 351-370 jm ht Cao, M., Zhang, Q., Seydel, J (2005) B2C e-commerce website quality: an k gm empirical examination Industrial Management & Data Systems, 105(5), om l.c 645-661 a Lu Cheung, C.M.K., Zhu, L., Kwong, T., Chan, G.W.W., Limayem, M (2003, June 9- n 11) Online consumer behavior: A review and agenda for future research n va Conference paper conducted at 16th Bled E-commerce Conference, Slovenia y te re th t to ng Chiu, C.M., Hsu, M.H., Lai, H., Chang, C.M (2012) Re-examining the influence hi ep of trust on online repeat repurchase intention: The moderating role of habit w and its antecedents Decision Support Systems, 53, 835-845 n lo Davis, F.D (1989) Perceived usefulness, perceived ease of use and user acceptance ad ju y th of Information Technology MIS Quarterly, 13(3), 319-340 Do Anh Minh (2012, Dec 3) The impact of Vietnam’s fallen group-buying star yi pl Retrieved on http://www.techinasia.com/impact-vietnams-fallen-group- n ua al buying-star/ va n Egdogmus, I.E., Cicek, M (2011) Online group-buying: What is there for these fu ll consumers? Procedia Social and Behavioral Sciences, 24, 308-316 oi m Éthier, J., Hadaya, P., Talbotm, J., Cadieux, J (2006) B2C web site quality and nh at emotions during online shopping episodes: An empirical study Information z z & Management, 43, 627-639 ht vb jm Fan, Y.W., Chiang, M.H., Wang, J.Y., Wang, E.T.G (2010), A study in consumers’ k continuing to use online group buying platforms: The impact of price gm performance expectations Chinese Business Review, (12), ISSN 1537- om l.c 1506 a Lu Kauffman, R.J, Bin, W (2001) Bid together, buy together: On the efficacy of n n va group-buying business models in Internet- based selling In Lowry, P.B., y th business and society, Boca Raton FL: CRC Press, 99-137 te re Cherrington, J.O., Watson R.R (Ed.), Handbook of Electronic Commerce in t to ng Kim, C., Galliers, R.D., Shin, N., Ryoo, J.H., Kim, J (2012) Factors influencing hi ep internet shopping value and customer repurchase intention Electronic w Commerce Research and Applications, 11, 374-387 n lo Kim, M.J., Chung, N., Lee, C.K (2011) The effect of perceived trust on electronic ad ju y th commerce: Shopping online for tourism products and services in South Korea Tourism Management, 32, 256-265 yi pl Lee, H., Choi, S.Y., Kang, Y.S (2009) Formation of e-satisfaction and repurchase ua al intention: Moderating roles of computer self-efficacy and computer anxiety n va n Expert Systems with Applications, 36, 7848-7859 fu ll Lee, Y., Kwon, O (2011) Intimacy, familiarity and continuance intention: An oi m extended expectation-confirmation model in web-based services Electronic nh at Commerce Research and Applications, 10, 342-357 z z Li, H.X (2010) E-service continuance: An insight into online travel services in jm ht vb China PhD thesis of Turku School of Economics k Li, H.X., Liu, Y (2012, June 17-20) Predicting and explaining use intention and gm purchasing intention in online group shopping Conference paper conducted l.c om at 25th Bled eConference eDependability: Reliable and Trustworthy a Lu eStructures, eProcesses, eOperations and eServices for the future Slovenia n n va Li, H.X., Suomi, R (2009) A proposed scale for measuring e-service quality y te re International Journal of u- and e- service, Science and Technology, (1) th t to ng Liao, C., Palvia, P., and Lin, H.N (2006) The roles of habit and website quality in hi ep e-commerce International Journal of Information Management, 26, 469- w 483 n lo Lin, M.Y.C., Ong, C.S (2010) Understanding information systems continuance ad ju y th intention: A five factor model of personality perspective PACIS 2010 Proceedings, available from www.pacis-net.org/file/2010/S09-01.pdf yi pl Liu, J.J., Wu, C.Y (2012, July 3-5) The empirical study of determinants of ua al customer loyalty for the group-buying website Conference paper conducted n va n at International conference on Business and Information, Sapporo, Japan fu ll Matsuo, T (2009) A reassuring mechanism design for traders in electronic group- oi m buying Applied Artificial intelligence, 23(1), 1-15 nh at Nguyen D Tho, Nguyen T.M Trang (2008) Marketing research: Implication of z z structural equation model SEM Vietnam National University Publisher ht vb jm Nguyen Khanh Duy (2009) Practicing structural equation model SEM with AMOS k software Lecture of University of Economics Ho Chi Minh City gm Premkumar, G., Anol Bhattacherjee (2008) Explaining information technology om l.c usage: A test of competing model Omega, 36, 64-75 a Lu Simin Wang (2011, May 2) Group buying sites boost e-commerce in Asia n on http://news.smh.com.au/breaking-news-technology/group- n va Retrieved th repurchase intentions in online group-buying context: An integration of y Tien, W.P., Hsu, M.H., Chuang, L.W (2012, July 3-5) Understanding the te re buying-sites-boost-ecommerce-in-asia-20110502-1e3l1.html t to ng expectation- confirmation theory and relationship quality Conference paper hi ep conducted at International conference on Business and Information, Sapporo, w Japan n T lo Tran Tai (2012) Group-on clones in Vietnam Retrieved on ad ju y th http://www.slideshare.net/TaiTran/groupon-clones-in-vietnam-2752011 Udo, G.J., Bagchi, K.K., Kirs, P.J (2010) An assessment of customers’ e-service yi pl quality perception, satisfaction and intention International Journal of ua al Information Management, 30, 481-492 n va n Wen, C.,Victor R.PryButok & Xu, C.Y (2011) An integrated model for customer fu ll online repurchase intention Journal of computer Information Systems, 52(1), oi m 14-23 nh at Yoo, B., Donthu, N (2001) Developing a scale to measure the perceived quality of z z Internet shopping sites (SITEQUAL) Quarterly Journal of Electronic jm ht vb Commerce, 2(1), 31-47 k Zang, Y., Fang, Y., Wei, K.K., Ramsey, E., McCole, P., Chen, H (2011) gm Repurchase intention in B2C e-commerce – A relationship quality om l.c perspective Information & Management , 48, 192-200 n a Lu n va y te re th t to ng hi LIST OF APPENDICES ep w n lo ad Appendix A: Review of the main studies on the dimensions of website quality y th Appendix B: EFA results of main survey before removing items ju Appendix C: SEM results of research model (unstandardized) yi pl Appendix D: SEM results of research model (standardized) ua al Appendix E: Bootstrap distribution results n Appendix F: Questionnaire n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th t to ng Appendix A: Review of the main studies on the dimensions of website quality hi ep Authors w n Dabholkar (1996) Loiacono, Watson, Goodhue (2000) Research context E-service Key dimensions of website quality lo Web retail site ad ju y th yi pl Online retailing Web retail site n ua al Zeithaml et al (2000) Yoo & Donthu (2001) Website design, information and security nh Specific content, content quality, appearance, and technical adequacy Security, communication, reliability, responsiveness and delivery Performance, features, structure, aesthetics, reliability, service ability, security and system integrity, trust, responsiveness, service differentiation and customization, web store police, reputation, assurance and empathy Website design, security, reliability, responsiveness, accessibility and customization at z z k jm ht vb om l.c gm Website design, security, reliability, customer service n a Lu n y te re th Interaction, security, reliability, responsiveness, information, delivery, customization Ease of use, appearance, linkage, structure, content, efficiency, communication, security, reliability, incentive and customer support SERVQUAL: tangibles, reliability, responsiveness, assurance, and empathy Credibility, ease of use, security, reliability, va E-commerce service Internet oi Iwaarden & Wiele (2003) Yang et al E-service in e-commerce m Wolfinbarger & Gilly (2002,2003) Surjadaja et al (2003) Santos (2003) Online retailing service Online retailing service E-service ll Yang and Jun (2002) fu Aladwani and Palvia’s (2002) Zeithaml et al (2002) Madu & Madu (2002) n Yang (2001) E-commerce service Internet banking service Online retailing Web retail site Online service Virtual business operation va Cox & Dale (2001) Jun & Cal (2001) Website designs, reliability, delivery, ease of use, enjoyment and control WEBQUAL: ease of understanding, intuitive operation, informational fit-to-task, tailored communication, trust, response time, visual appeal, innovativeness, emotional appeal, online completeness, relative advantage, and consistent image Efficiency, reliability, fulfillment, privacy, responsiveness, compensation, and contact SITEQUAL: aesthetic design, competitive value, ease of use, clarity of ordering, corporate and brand equity, security, processing speed, product uniqueness, and product assurance quality Website appearance, communication, accessibility, credibility, understanding and availability Website design, information, ease of use, access, courtesy, responsiveness, and reliability t to ng retailing service Yang et al (2004) Field et al., (2004) Kim & Stoel (2004) Shopping sites E-service hi (2003) ep w responsiveness, convenience, communication, access, competence, courtesy, personalization, collaboration, aesthetics Competence, security, reliability, responsiveness, ease of use and product portfolio Website design, security, reliability, customer service n lo ad y th ju Yang & Fang (2004) yi Web appearance, entertainment, information, transaction capability, responsiveness and trust Responsiveness, reliability, credibility, competence, access, courtesy, communication, information and web design Website design, information, trust, responsiveness and reputation pl n ua Core e-SQ: Efficiency, fulfillment, availability, privacy Recovery e-SQ: responsiveness, compensation and contact Website design, reliability, responsiveness, trust and personalization n va Panasuraman et al (2005) al Gounaris et al (2005) Online apparel website Online securities brokerage Online retailing service E-service at z z k jm ht vb om l.c gm Online service Online financial service System quality, information quality, service quality, and attractiveness Efficiency, fulfillment, system availability, privacy, responsiveness, compensation, contact, information and graphic style Graphic quality, layout, attractiveness of selection, information, ease of use, technical quality, reliability, functional benefit and emotional benefit Website design, customer service, assurance and order management Trust, speed of delivery, reliability, ease of use, customized communication, website content and functionality nh Critobal et al (2007) Sohn & Tadisina (2008) oi Fassnacht & Koese (2006) Online retailing service E- service m Kim et al (2006) ll Online retailing service Cao et al (2005) E- service fu Lee & Lin (2005) Source: Li (2010); Li & Suomi (2009) n a Lu n va y te re th t to ng Appendix B: EFA results of main survey before removing items hi ep w n lo ad ju y th Factor loading 201 yi pl ua al 361 -.288 -.238 291 349 -.215 334 201 n -.288 -.225 n va 733 681 619 615 604 514 496 487 ll fu 283 m 400 332 410 874 766 743 706 446 210 oi at nh z k 777 44.638 6.752 3.248 2.710 2.025 1.896 th 830 y 1.111 te re 1.332 n 2.768 va 18.302 n a Lu 255 894 866 493 399 -.244 269 289 om 379 764 759 756 753 668 l.c gm 704 629 523 519 513 416 jm ht vb 205 234 z TA41 TA38 TA37 TA36 TA39 TA35 TA40 TA34 AP31 AP32 AP33 AP29 SC26 PU1 PU3 PU2 SAT5 PU4 SAT6 SAT7 SAT8 INT18 INT17 INT16 INT14 INT15 TR11 TR13 TR12 TR10 TR9 CQ24 CQ21 CQ20 CQ19 CQ22 CQ23 SC27 SC28 SC25 AP30 Eigen value Extracted Variance 872 854 852 837 793 717 673 527 522 506 496 367 247 t to ng Appendix C: SEM results of research model (unstandardized) hi ep ML w n lo ad y th yi 0.084 0.112 0.081 0.06 n ua al va 0.068 0.066 0.067 n fu 0.042 0.046 0.047 at nh z z vb Mean 1.015 0.382 0.509 0.963 0.533 0.337 0.299 1.111 1.11 0.967 1.014 0.894 0.835 0.956 1.02 1.005 0.965 1.016 0.886 0.94 0.93 0.904 0.944 0.846 0.989 0.82 k jm Bias SE-Bias 0.006 0.003 0.007 0.005 0.002 0.004 0 0.013 0.003 0.005 0.012 0.004 -0.006 0.003 0 -0.001 0.002 -0.001 0.002 -0.001 0.002 0 0.001 0.001 -0.003 0.002 -0.001 0.002 0 0.002 0.002 0.001 0.002 0.003 0.001 0 0.001 0.001 0.001 0.001 0.002 0 0.001 0.002 0.002 0.004 0.002 0.003 0.002 0.003 0.002 0.004 0.002 0.001 0.002 om l.c gm n a Lu n va 0.049 0.052 0.055 0.053 0.051 0.054 0.051 oi 0.046 0.049 0.055 m 0.048 0.049 0.047 SE-SE 0.002 0.004 0.003 0.002 0.003 0.003 0.002 0.001 0.002 0.002 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 ht C.R SE 11.221 0.103 3.187 0.16 5.812 0.122 11.375 0.098 4.777 0.156 4.016 0.113 5.063 0.108 16.481 0.062 16.721 0.075 14.545 0.075 24.389 0.035 19.602 0.051 17.806 0.052 19.678 0.053 20.862 0.049 21.157 0.043 20.853 0.042 20.585 0.041 15.953 0.057 19.16 0.039 17.992 0.056 16.274 0.063 17.829 0.056 16.494 0.055 18.092 0.052 16.122 0.049 ll WQ WQ TR WQ WQ WQ TR SAT CQ CQ CQ CQ SAT SAT SAT SAT INT INT INT INT TR TR TR TR TA TA TA TA TA TA TA SAT S.E 0.09 0.118 0.087 pl < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < - ju TR SAT SAT TA CQ INT INT INT CQ23 CQ21 CQ20 CQ19 SAT5 SAT6 SAT7 SAT8 INT14 INT16 INT17 INT18 TR13 TR12 TR11 TR9 TA38 TA37 TA36 TA35 TA39 TA40 TA41 PU4 Estimate 1.009 0.375 0.507 0.95 0.533 0.325 0.304 1.113 1.111 0.968 1.013 0.897 0.836 0.954 1.019 1.002 0.965 1.015 0.885 0.941 0.928 0.899 0.941 0.843 0.985 0.818 Bootstrap y te re th t to ng Appendix D: SEM results of research model (standardized) hi ep ML w n lo ad ju y th pl n ua al n ll fu oi at nh SE-SE 0.001 0.003 0.003 0.001 0.001 0.002 0.002 0.002 0.001 0.001 0.001 0 0.001 0.001 0.001 0.001 0 0 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 z z vb Mean Bias SE-Bias 0.776 -0.001 0.001 0.285 0.001 0.004 0.498 0.004 0.793 -0.001 0.001 0.89 0.002 0.001 0.377 -0.003 0.003 0.31 0.009 0.003 0.28 -0.006 0.003 0.766 0.001 0.001 0.837 0.001 0.849 0.001 0.749 0.001 0.879 0.001 0.904 0 0.8 -0.002 0.001 0.757 -0.001 0.001 0.853 0.001 0.827 0.001 0.856 0.001 0.864 0.001 0.001 0.851 -0.001 0.001 0.863 -0.001 0.001 0.857 0.001 0.725 -0.001 0.001 0.826 -0.001 0.001 0.835 -0.002 0.001 0.802 0.001 0.749 0.002 0.001 0.797 0.001 0.756 0.001 0.001 0.806 0.001 0.001 0.711 0.001 k jm ht om l.c gm n a Lu n va SE 0.042 0.113 0.114 0.039 0.033 0.107 0.102 0.099 0.03 0.023 0.022 0.037 0.018 0.015 0.027 0.033 0.021 0.027 0.022 0.021 0.02 0.019 0.02 0.039 0.027 0.027 0.023 0.033 0.03 0.031 0.022 0.03 m WQ WQ TR WQ WQ WQ TR SAT CQ CQ CQ CQ SAT SAT SAT SAT INT INT INT INT TR TR TR TR TA TA TA TA TA TA TA SAT Estimate 0.778 0.284 0.499 0.794 0.889 0.381 0.301 0.287 0.765 0.838 0.849 0.749 0.879 0.903 0.802 0.758 0.853 0.827 0.856 0.863 0.852 0.864 0.857 0.726 0.827 0.836 0.802 0.748 0.797 0.755 0.805 0.711 va < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < - yi TR SAT SAT TA CQ INT INT INT CQ23 CQ21 CQ20 CQ19 SAT5 SAT6 SAT7 SAT8 INT14 INT16 INT17 INT18 TR13 TR12 TR11 TR9 TA38 TA37 TA36 TA35 TA39 TA40 TA41 PU4 Bootstrap y te re th t to ng Appendix E: Bootstrap distribution results hi ep w n lo ad ju y th yi N = 1000 Mean = 805.227 S e = 2.574 pl n ua al n va ll fu 613.483 654.047 694.611 735.175 775.739 816.303 856.867 897.431 937.995 978.559 1019.123 1059.687 1100.251 1140.815 1181.379 oi m | -|* |*** |********* |************** |******************** |***************** |************** |******** |***** |*** |** |* |* | |* | at nh ML discrepancy (implied vs pop) (Default model) z z | -|* |*** |********** |**************** |******************* |***************** |*********** |******** |***** |*** |** |* |* |* |* | k jm ht vb om l.c gm n a Lu N = 1000 Mean = 564.556 S e = 876 n va y te re 504.457 517.298 530.140 542.981 555.822 568.664 581.505 594.346 607.188 620.029 632.870 645.712 658.553 671.394 684.236 th t to ng K-L overoptimism (unstabilized) (Default model) hi ep w n lo ad ju y th yi N = 1000 Mean = 224.248 S e = 11.232 pl n ua al n va ll fu -960.417 -798.364 -636.312 -474.259 -312.206 -150.154 11.899 173.952 336.004 498.057 660.110 822.163 984.215 1146.268 1308.321 | -|* |* |* |*** |******** |************ |****************** |******************** |******************** |************** |*********** |****** |*** |* |* | oi m nh at K-L overoptimism (stabilized) (Default model) z z k jm ht vb om l.c gm n a Lu N = 1000 Mean = 213.929 S e = 2.525 n va y te re 24.479 59.322 94.166 129.009 163.852 198.695 233.539 268.382 303.225 338.069 372.912 407.755 442.598 477.442 512.285 | -|* |*** |******* |********** |***************** |******************* |***************** |************* |********* |**** |*** |** |* |* |* | th t to ng Appendix F: Questionnaire hi ep Dear yours! I am student of International School of Business of University of Economics Ho w n Chi Minh City I am making a research about online group-buying model – a form lo ad of online shopping – in Vietnam market y th Hopefully you can spend your precious time to answer the questionnaire below ju The survey will take about 10 minutes The responses from this survey will be used yi in a research in online group-buying model All collected material will be used pl ua al confidentially so that individual respondents will not be tracked Thank you for taking the time to complete the survey n n va Part 1: Information about the respondents’ experience  No ll  Yes fu 1/ Do you know online group-buying model? m oi 2/ Have you ever used online group-buying model? nh  Never and have no intention to so at jm ht vb  Yes, more than times z  Yes, 1- times z  Not yet but I have intention in it 3/ List some online group-buying websites which you usually purchase k gm (maximum websites) l.c ……………………………………………………………………………………… a Lu website you use most regularly (website X)? om 4/ Among websites you listed in question 3, which online group-buying n ……………………………………………………………………………………… th disagree to 7- strongly agree) with the corresponding number: y Your level of agreement is measured by a seven-point Likert scale (from 1- strongly te re (website X), please comment your level of agreement on the following statement: n Based on your purchasing experience of website you answered in question va Part 2: t to ng hi ep w n lo 1234567- Strongly disagree Disagree Disagree some what Neutral Agree somewhat Agree Strongly agree ad ju y th Strongly disagree yi Using website X improves my performance in finding out groupbuying products Using website X increases my productivity in group-buying Using website X provides me with more diverse channels to enhance my effectiveness in group-buying products As for overall platform design, I find website X useful Strongly agree 2 3 4 5 6 7 1 1 2 2 3 3 4 4 5 5 6 6 7 7 1 1 2 2 3 3 4 4 5 5 6 6 7 7 2 2 3 3 4 4 5 5 6 6 7 7 pl n ua al n va ll fu I was very satisfied with my experience to website X I was very pleased with my experience to website X I was absolutely delighted with my experience to website X I think I made a right decision when using website X oi m at nh z I feel safe on my transaction with website X I believe website X can protect my privacy I select website X, which I believe are honest I feel that website X would provide me with good service I feel that website X is trustworthy z 1 1 k om l.c gm Strongly agree 7 7 7 th 5 5 5 y 4 4 4 te re 3 3 3 n va Strongly disagree 2 2 2 n a Lu Perceived website quality The content of website X is useful The content of website X is complete The content of website X is clear The content of website X is current The content of website X is concise The content of website X P is accurate jm ht vb I intend to continue to purchase goods from website X I intend to acquire product information from website X I intend to recommend website X to people around me I intend to use website X as the priority online store for future purchases Except for any unanticipated reasons, I intend to continue to use website X 10 t to ng hi ep In website X, one can find contact information In website X, one can find details about products and/or services In website X, one can find information related to customers’ policies In website X, one can find information related to customer service w n lo ad Website X looks attractive Website X looks organized Website X uses fonts properly Website X uses colors properly Website X uses multimedia features properly ju y th yi pl n ua al Website X looks secured for carrying out transactions Website X looks easy to navigate through Website X has adequate search facilities Website X is always up and available Website X has valid links Website X load fast Website X has many interactive features Website X is easy to access n va ll fu 3 4 5 6 7 7 1 1 2 2 3 3 4 4 5 5 6 6 7 7 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4 5 5 5 5 6 6 6 6 7 7 7 7 oi m 2 at nh Part 3: Individual Information z 1/ What is your gender?  Femail z  Male vb jm ht 2/ What is your age?  22- 30 years  30-40 years  Over 40 years k  18 – 22 years gm  10- 20 million om  Under 10 million l.c 3/ What is your annual income?  Over 20 million  College student  Master’s level n  Bachelor’s level va  Highschool n a Lu 4/ What is your education background? te re  PhD level y th 