Usability and sociability are two characteristics of a virtual community that are critical to its ... The challenge is thus for the virtual community managers to promote ...
Encouraging Participation in Virtual Communities through Usability and Sociability Development: An Empirical Investigation Author: Lu, Xianghua; Phang, Chee Wei; Yu, Jie ProQuest document link Abstract: Usability and sociability are two characteristics of a virtual community that are critical to its success. This study aims to explore how usability and sociability of virtual communities can be developed in order to encourage members' continuous participation in these communities. A theoretical model is proposed to explain the effects of factors related to usability and sociability on members' continuous participation through the motivational beliefs of perceived usefulness, perceived enjoyment and sense of belonging. Data was collected from members of five popular leisure-oriented virtual communities in China. The results show that both perceived enjoyment and sense of belonging impact members' continuous participation intention. Among the usability-and sociability-related factors, we find that information service quality is the most critical factor that encourages members to continuously participate in virtual communities, while interaction support quality, incentive policy, and event organization also have positive effects on members' continuous participation intention via perceived enjoyment and sense of belonging. Surprisingly, leaders' involvement has no impact on members' continuous participation. Implications of the study's findings for both research and practice are discussed. [PUBLICATION ABSTRACT] Links: Linking Service Full text: Headnote Abstract Usability and sociability are two characteristics of a virtual community that are critical to its success. This study aims to explore how usability and sociability of virtual communities can be developed in order to encourage members' continuous participation in these communities. A theoretical model is proposed to explain the effects of factors related to usability and sociability on members' continuous participation through the motivational beliefs of perceived usefulness, perceived enjoyment and sense of belonging. Data was collected from members of five popular leisure-oriented virtual communities in China. The results show that both perceived enjoyment and sense of belonging impact members' continuous participation intention. Among the usability-and sociability-related factors, we find that information service quality is the most critical factor that encourages members to continuously participate in virtual communities, while interaction support quality, incentive policy, and event organization also have positive effects on members' continuous participation intention via perceived enjoyment and sense of belonging. Surprisingly, leaders' involvement has no impact on members' continuous participation. Implications of the study's findings for both research and practice are discussed. Keywords: Usability, Sociability, Virtual communities, Continuous participation ACM Categories: H.0, K.4.2 INTRODUCTION In recent years, we have witnessed the flourishing of virtual communities in both practice and academic research. Virtual communities involve a collection of people who use the computer-supported virtual space to communicate and interact openly with each other to seek some shared purposes, such as leisure, knowledge sharing, and commerce (De Souza and Preece, 2004). Among the various reasons suggested for the success of virtual communities, it is widely recognized that continuous participation of members are essential to sustain virtual communities (Jin, Cheung, Lee et al., 2007; Khalifa and Liu, 2005; Ma and Agarwal, 2007). According to iresearch, com (2008), the main issue faced by most virtual communities today is the low virtual community
stickiness, which implies that members have a high tendency to discontinue participating in the communities after some time. The challenge is thus for the virtual community managers to promote continuous participation of their members. In response to the above, there has been a sizable number of research that investigates the motivational beliefs for individuals to participate in virtual communities. Three main categories of motivational beliefs emerge from this stream of research: utilitarian, hedonic, and social, which correspond to three major functions of virtual communities. First, virtual communities serve as utilitarian information systems. Members collectively build knowledge, solve problems, and help each other by exchanging knowledge via the virtual communities (Kankanhalli, Tan and Wei, 2005; Yu, Jiang and Chan, 2007). Second, virtual communities also function as hedonic information systems whereby members participate in fun and leisure activities (Van der Heijden, 2004). Finally, virtual communities are "social aggregations that emerge from the net" (Rheingold, 1993) with mutual engagement, joint goal and shared repertoire among members, which distinguish virtual communities from other Internet applications (Wenger, 1998). With respect to these motivations, prior research shows that individuals' initial participation tends to be driven mainly by utilitarian motivations e. g., obtaining information needed (Shah, 2006). However, for longterm, sustained participation, utilitarian motivations may not be adequate as individuals become more concerned with getting hedonic enjoyment and social engagement and attachment with others, which take time to develop (Fang and Neufeld, 2009; Shah, 2006) Therefore, the investigation of what motivate continuous participation, or long-term, sustained participation (Fang and Neufeld, 2009), requires the consideration of all three types of factors i. e., utilitarian, hedonic, and social-related motivational beliefs. Despite the considerable prior literature on individual's motivational beliefs, research that focuses on how virtual communities can be developed to enhance the motivational beliefs and in turn promote continuous participation remains limited. Among the studies that investigate the latter issue, researchers have either identified critical factors leading to virtual community success (Broß, Sack and Meinel, 2007; Leimeister, Sidiras and Krcmar, 2006; Malhotra, Gosain and Hars, 1997; Williams and Cothrel, 2000), or evaluated virtual community's performance (Bughin and Hagel lii, 2000; Preece, Nonnecke and Andrews, 2004). While these studies provided some implications for the management of virtual communities, they failed to provide direct insights on how virtual communities could be developed in order to promote members' continuous participation. For this purpose, there has been an emerging stream of research that examines the technical and social aspects of developing virtual communities (MaloneyKrichmar and Preece, 2005; Phang, Kankanhalli and Sabherwal, 2009; Preece, 2001). Research along this line argues that virtual communities, as technology-mediated spaces that support interpersonal communications, can be viewed as socio-technical systems that consist of both technical and social components (Kling and Courthght, 2003; Trist and Murray, 1993). The resultant sociotechnical framework highlights usability and sociability as two corresponding characteristics of a virtual community system. Usability refers to the capability of a system to be used easily and effectively by individuals to perform intended activities (Shackel, 1991), or the quality of a system that makes it acceptable to users (Holzinger, 2005). Sociability means that the virtual community system supports "a state of being sociable", where members find it pleasant to interact with each other in attaining virtual community shared purposes through the virtual space (Phang et al., 2009; Preece, 2001). The two characteristics may in turn form the basis of motivations for individuals to continue participating in the virtual community (Phang et al., 2009; Preece, 2001). However, there remains a paucity of research that attempts to investigate factors related to the usability and sociability of virtual communities, and how they affect the motivations for individuals to continue participating in the communities. To bridge the gaps in the extant research, we seek to answer the following research question, "How do factors related to the usability and sociability in a virtual community influence members' motivations that subsequently lead to their continuous participation in the virtual community?" Correspondingly, we propose a theoretical
model of individual's continuous participation in virtual communities that is based on Preece's (2001) usabilitysociability framework and participation motivations in the context of virtual communities. The research model explains the effects of factors related to both the usability and sociability of virtual communities on members' continuous participation intention through the utilitarian, hedonic and social motivational beliefs. For managers of virtual community, this study may provide practical guidance on how they should develop the virtual community system in order to make it attractive and sustainable in the long run. The rest of the paper is organized as follows. First, the literature on Preece's usabilitysociability framework and virtual community participation motivations is reviewed. Second, specific research model and corresponding hypotheses are proposed. The research design and methodology are elaborated followed by discussion of results. Finally, both theoretical and practical implications are presented and future research directions are suggested. Theoretical background Factors characterizing usability and sociability of virtual communities The usability-sociability framework (Preece, 2001) argues that the success of a virtual community is determined by the usability and sociability characteristics of the virtual community system. Usability refers to the capability of a virtual community system to be used easily and effectively by individuals to perform intended activities (Shackel, 1991) e. g., knowledge sharing and leisure activities. Sociability is "a state of being sociable", where members find it pleasant to interact with each other in attaining shared purposes e. g., mutual support, through the virtual community space (Preece, 2001). In this study, we aim to derive a parsimonious set of factors that capture the desired elements of usability and sociability based on the relevant seminal literature. Previous studies have begun to explore what may constitute the usability and sociability perceptions of individuals toward a virtual community system. Among them, De Souza and Preece (2004) propose a comprehensive framework for analyzing and understanding the usability and sociability aspects of virtual communities, while Lazar and Preece (2003) conducted an in-depth review of the literature that delves into the two notions. According to De Souza and Preece (2004), a virtual community system with good usability should entail the following: a navigation structure that allows members to navigate and find what they want easily; support for how information is presented; and features to support different types of communities in general such as search engines, and communication tools for interacting with others such as private messaging. Based on a review of the previous literature that investigates usability, Lazar and Preece (2003) highlight the following desired usability elements: consistency (using the same terms and procedures for achieving the same functionality throughout the system); control (users are able to control what they want to do with the system); predictability (enabling users to build on their experience in using the system); navigation (easy to navigate around the system to locate the information needed); access (easy to access the community system); and information design (content comprehension and aesthetics). Essentially, the elements outlined in the previous studies revolve around two kinds of facilities provided by most virtual communities i. e., those related to information service and interaction support. Accordingly, two factors are derived for assessing the usability of a virtual community system, namely information service quality and interaction support quality, which are also consistent with Leimeister and Krcmar (2005) . Information service quality is defined in this study as the ease of navigating and obtaining information needed in a virtual community system. Interaction support quality refers to the extent to which the virtual community system can help members interact with others with shared interests, and enable members to maintain their social relationships with others. Compared to usability, sociability is a relatively new concept that describes a virtual community system. De Souza and Preece (2004) propose a number of important elements characterizing sociability, including keeping discussions on-topic; encouraging reciprocity, empathy and trust; supporting shared understanding;
development of online identities; and protecting members' privacy. These are resonated by Lazar and Preece (2003), who highlight the importance to have a clear focus for the community, as well as the presence of trust and security, and good governance in the community. Essentially, the prior literature suggests that for a virtual community to possess good sociability, it is important to encourage a high interactivity level among the members, and have a well-received governance to ensure a pleasant yet focused avenue for members' interactions. Accordingly, we propose the following factors that may entail these elements i. e., incentive policy, event organization, and leaders' involvement (see Table 1). Incentive policy covers the set of policies with regard to members' participation that revolve around the community's focus or purposes. Based on the purpose of virtual communities, well designed incentive policies can formally or informally impact the participation behavior of members as well as the entry and exit of members (De Souza and Preece, 2004). Event organization refers to the efforts to organize events both online and offline to encourage interactions and relationship building among members (Koh and Kim, 2003; Williams and Cothrel, 2000). Leaders' involvement captures another essential aspect of sociability i. e., the guidance and coordination of members' interactions within the purpose and policies of virtual communities. Virtual communities' leaders such as active members, opinion leaders or moderators, play an important role on the members' interaction through sharing expertise, acknowledging members' contribution, promoting good examples, and penalizing violations. Table 1 below provides a summary that explains how the factors pertaining to the usability and sociability of a virtual community are derived. Individuals' motivations of continuous participation in virtual communities Motivations are reasons of why people take certain actions. The previous literature on virtual community participation has highlighted the mediating role of individuals' motivations on the relationship between virtual community characteristics and participation (Lin, 2006; Yu et al., 2007). The logic is that the presence of desired characteristics of a virtual community e. g., usability and sociability, may make individuals more motivated to participate. Broadly the motivations of participating in virtual communities can be categorized based on individuals' needs in terms of the utilitarian, hedonic, and social functions of a virtual community. The utilitarian function has long been recognized as the basis of adoption of information systems or Internet applications in general (Davis, 1989; Li, Chau and Lou, 2005). Generally users expect to obtain some utilities or outcomes from using these systems. Consistent with the previous literature, we conceptualize it as perceived usefulness (Davis, 1989) in our context. In addition to utilitarian motivations i. e., usefulness, previous research highlights that for individuals to participate in virtual communities for long term, they also desire for hedonic and social benefits (Fang and Neufeld, 2009; Shah, 2006) The hedonic motivations may come in the forms of fun or enjoyment received while participating in virtual communities (Ardichvili, Page and Wentling, 2003). Hence to capture the hedonic motivational aspect of participating in virtual communities, we include perceived enjoyment in this study. Furthermore, individuals who participate for long-term may be particularly concerned with the social function of a virtual community i. e., how it may facilitate the social engagement and attachment with others. Social function is essential to virtual communities as this is what distinguishes virtual communities from other Internet applications e. g., electronic repositories. Concomitantly, the psychological perspective suggests that it is important to consider the emotional attachment to a virtual community when investigating participation (De Souza and Preece, 2004). To this end, sense of belonging has been widely conceptualized to capture such motivations for continuous participation (Jin et al., 2007; Seddon, Skinner and Postlethwaite, 2008). To recapitulate, by synthesizing the previous findings on motivations of virtual community participation, this study summarizes perceived usefulness, perceived enjoyment, and sense of belonging as the three motivational beliefs that mediate the relationship between the factors related to usability and sociability and individual's continuous participation in a virtual community.
Research Hypotheses In developing a research model of continuous participation in virtual communities, we commence with the identified set of individual's motivational beliefs that explains why members continuously participate in virtual communities, which is the dependent variable of this study. We then examine the antecedents of these motivations that are drawn from the usability and sociability perspectives of virtual communities. Our theoretical model is presented in Figure 1, and the corresponding hypotheses are elaborated in detail. Perceived usefulness Perceived usefulness has been defined as an evaluative belief that using an information technology will enhance or improve the individual's job performance (Adams, Nelson and Todd, 1992). In the context of virtual communities, the perceived usefulness to a member may transcend the organizational job boundary to entail personal growth of knowledge and exposure e. g., through exchanging views and experience with others in the virtual community. Empirical studies have provided support for the importance of perceived usefulness in influencing continuous usage of a new technology (Karahanna and Straub, 1999; Parthasarathy and Bhattacherjee, 1998). Similarly, Hagel (1999) suggested that an important element of a viable community is the ongoing provision of content that members perceive as valuable or useful. When members believe that virtual communities have what are useful to them, they are very likely to return and participate in the virtual community. Hence, we hypothesize that: H1: There is a positive relationship between members' perceived usefulness and their intention of continuous participation. Perceived enjoyment Apart from perceived usefulness, fun or enjoyment also emerge as a major determinant of technology acceptance (Davis, Bagozzi and Warshaw, 1992). Heijden (2004) included perceived enjoyment as a predictor of technology acceptance for information systems with hedonic function. Perceived enjoyment is defined as "the extent to which the activity of using the new technology is perceived to be enjoyable in its own right, apart from any consequences that may be anticipated" (Davis et al., 1992). In the context of virtual communities, prior study proposed perceived playfulness as a similar construct which is defined as the degree to which a member in the virtual community believes that participating in the virtual community would enhance the entertainment value (Moon and Kim, 2001). Empirical investigations on virtual communities reveal that perceived enjoyment in communications (Chen, 2007) and helping others (Cheung and Lee, 2007) may increase the satisfaction of new members, which in turn improves their intention to continue participating in the virtual communities. These findings correspond to the hedonic aspect of virtual communities, which suggests that members expect to enjoy their experience in virtual communities (Koh and Kim, 2003; Webster and Martocchio, 1992). Hence, we hypothesize that: H2: There is a positive relationship between members' perceived enjoyment and their intention of continuous participation. Sense of belonging Virtual communities are more than just a technology. They represent "a social aggregation that emerges from the Net when enough people carry on public discussions, with sufficient human feeling, to form webs of personal relationships in cyberspace" (Rheingold, 1993). Recognizing such nature of virtual communities, researchers have been actively trying to understand the sustainability of virtual communities from the social perspective. Their results consistently highlight sense of belonging as an important factor for members' continuous participation in virtual communities (Bagozzi and Dholakia, 2002; Dholakia, Bagozzi and Pearo, 2004). Specifically, it has been suggested that people continuously participate in a virtual community because they need a sense of belonging to a collective (Tinto, 1993; Yoo, Suh and Lee, 2002). According to McMillan and Chavis's (1986), sense of belonging is "a feeling that members have of belonging, a feeling that members matter to one another and to the group, and a shared faith that members' needs will be
met through their commitment to be together". Building on this definition, Blanchard et al. (2004) developed the virtual community version of "sense of belonging", termed as "sense of virtual community". Such a sense of community can foster the development of virtual community relationships and may therefore affect the vitality of the community. With the increase of sense of belonging, the virtual community members perceive their membership to be salient in the virtual community and acknowledge their interdependence with the virtual community. A virtuous cycle is formed and future participation is thus expected. Hence, we hypothesize that: H3: There is a positive relationship between members' sense of belonging and their intention of continuous participation. Usability-related Factors Virtual community is in essence an information system. Technological characteristics of such system are likely to influence members' continuous participation intention through the provision of high usability, specifically in terms of information service quality and interaction support quality as we have discussed previously. Information Service Quality Accessing to information needed is regarded as a major reason of why individuals join a virtual community (Phang et al., 2009). With regard to this, a major challenge for the virtual community managers is how information should be organized and presented so that the information needs of users can be better fulfilled. Information navigation (Preece, 2001), information search quality (Preece 2001; Lazar and Preece, 2003; Leimeister et al, 2006), and information content customization (Fan and Poole, 2006) are the main components of information service quality. Through these services, members are more likely to find the information that fit their own usage, and thus perceive the virtual community system to be more useful to them. Apart from perceived usefulness, information service quality is also likely to influence the perceived enjoyment derived from participating in the virtual communities. In group discussion settings, it has been found that a higher degree of content fit is positively related to higher enjoyment for discussion participants (Ho, 2007). In the context of virtual communities, members may derive their enjoyment from interacting with others and the virtual community's contents (Koh et al. 2003). Information contents are the basis of why people come to a virtual community, and the ease of getting the information desired is likely to enhance their perceived enjoyment. For instance, members of a virtual community on movie discussion will be delighted to obtain a constant stream of information about their favorite genres of movies easily, and thus perceive the participation in the community more enjoyable. From an inverted angle, an improper organization of information, lack of search engine, and poor navigation design will likely lower the enjoyment of using the virtual community system for participation (Ardichvili et al., 2003). Lastly, information service quality has been recognized as an important factor for the sustainability of a virtual community (Leimeister et al., 2006). Previous research shows that a higher quality of information service in a virtual community may lead its members to feel more gratified during participation, which in turn enhances their feeling of being part of the community i. e., a higher sense of community (Lin, 2008). Collective, we hypothesize that: H4a: The higher the perceived information service quality is, the higher the level of usefulness perceived by virtual community members. H4b: The higher the perceived information service quality is, the higher the level of enjoyment perceived by virtual community members. H4c: The higher the perceived information service quality is, the higher the level of sense of belonging perceived by virtual community members. Interaction Support Quality Virtual communities integrate feedback channels and provide collaborative mass communication features through which members can interact with each other. In this study, we investigate the technology features that support members' interaction through interaction support quality, which is defined as the extent to which the
virtual community system can help members interact with others with shared interests, and enable members to maintain their social relationships with others. Examples of these features include private messaging, functions to track other members' activities such as posting of new messages, display of online/offline status of members, and instant messaging. The interaction support quality of a virtual community may encourage interactions among members and foster strong social ties among them. According to social capital theory, the existence of strong social ties may provide various benefits to the virtual community members, including more information sharing, companionship, and emotional and material support (Wellman, Garton and Haythornthwaite, 1997; Wellman and Gulia, 1999). Through facilitating the attainment of these benefits, a virtual community system is likely to be perceived as more useful by the virtual community members. With the ease of interacting with others and the ensuing strong ties with community peers, an individual is also likely to perceive the participation experience in the virtual community to be more enjoyable (Chan and Li, 2010; Wellman and Gulia, 1999). In addition, individuals' interactions with others within a virtual community may promote their emotional integration with the virtual community peers (Leimeister et al., 2006)). Findings from the social-psychological stream of research also indicate that interpersonal relationships is associated with positive attitude and feeling of attraction towards a group (Berg and Archer, 1983; Montgomery, 1984). Hence, when a virtual community system provides features that can help bind members together, it is likely to generate a feeling of "home" and emotional attachment among the members i. e., a higher sense of belonging (Bagozzi and Dholakia, 2007). The preceding discussion leads to the following hypotheses: H5a: The higher the perceived interaction support quality is, the higher the level of usefulness perceived by virtual community members. H5b: The higher the perceived interaction support quality is, the higher the level of enjoyment perceived by virtual community members. H5c: The higher the perceived interaction support quality is, the higher the level of sense of belonging perceived by virtual community members. Sociability-related Factors Sociability of a virtual community refers to "a state of being sociable", where members find it pleasant to interact with each other in attaining virtual community's shared purposes through the community space (Phang et al. 2009; Preece, 2001). In this study, we identify three factors characterizing the sociability of a virtual community, namely incentive policy, event organization, and leaders' involvement. Incentive Policy Policies in virtual communities are defined as the "language and protocols that guide people's interactions and contribute to the development of folklore and rituals that bring a sense of history and accepted social norms" (Preece, 2001). In this study, we focus on incentive policy that guides members' participation in terms of how contributions are recognized and rewarded with appropriate incentives. Incentives in virtual communities can be both monetary and non-monetary, and tangible and intangible (Andrews, 2002; Leimeister et al., 2006). In a virtual community, incentive policy may serve as mechanisms for assessing members' reputation and providing feedback to their behavior (Tedjamulia, 2005). Such policy may manifest in the forms of ranking and rating systems, collaborative filtering, or peer-based review (Jensen, Davis and Farnham, 2002). A carefully designed incentive policy system that is well-received by the members may help achieve two objectives. First, it helps ensure that members' interactions adhere to the focus of the virtual community, and minimize irrelevant postings such as flaming (Preece, 2001). Second, members tend to agree on how their contributions are recognized and rewarded. Together these should lead members to perceive their participation in the virtual community to be more enjoyable. Moreover, previous research has suggested that in the context of online markets, effective reputation or feedback mechanisms may enhance members' trust towards the whole online market (Bolton, Lemon and Verhoef, 2004)). Likewise in the virtual communities setting, when members trust
that the virtual community's incentive policy can evaluate and reward their contributions in a reasonable and fair manner, they are more likely to identify with the virtual community and feel that they are part of the virtual community i. e, a higher sense of belonging (Blanchard and Markus, 2004). Accordingly, we hypothesize that: H6a: The better received the incentive policy is, the higher the level of enjoyment perceived by virtual community members. H6b: The better received the incentive policy is, the higher the level of sense of belonging perceived by virtual community members. Event Organization Event organization includes both online and offline events initiated by virtual community managers or ordinary members that encourage interactions among members. Williams et al. (2000) analyzed several successful online communities and find many of these communities hold online and offline events from time to time. These events provide the opportunities for members to take part in activities together and know each other better, thus deepening their mutual understanding and foster closer relationships. The organization of offline events may also serve to fill in the void of virtual community space in terms of lack of face-to-face interaction. While online interaction has been recognized as a viable means for maintaining and extending social relationships, such as in the case of virtual communities, face-to-face interaction remains valuable to build more intimate relationships due to the possibility to convey richer cues in communications (Cummings, Butler and Kraut, 2002; Rothaermel and Sugiyama, 2001). Together the online and offline interactions afforded by event organization may allow a member to understand the context and meaning of their peers' contributions in the virtual community better, which may subsequently lead them to perceive the content of interaction in the community to be more relevant. Hence, a higher level of event organization is likely to lead a member to perceive participating in the virtual community as more useful. Additionally, with the cultivation of more intimate social bonds and relationships among the members, it is likely that a member will perceive a higher level of enjoyment derived from the interaction in the virtual community (Koh and Kim, 2003). Since members join virtual communities based on shared interests and values, formation of deeper relationships may also lead to a stronger identification with the virtual community i. e., a higher sense of belonging to the virtual community (Bargh and McKenna, 2004). Based on the preceding discussion, we hypothesize: H7a: The higher level the event organization in a virtual community, the higher the level of usefulness perceived by virtual community members. H7b: The higher level the event organization in a virtual community, the higher the level of enjoyment perceived by virtual community members. H7c: The higher level the event organization in a virtual community, the higher the level of sense of belonging perceived by virtual community members. Leaders' Involvement We define leaders as individuals who can influence other members or play leading roles in virtual communities (Koh, Kim, Butler et al., 2007). In this sense, leaders may include active members, opinion leaders, and moderators. Compared with other members, leaders usually contribute to the discussion more frequently, and are more involved in socializing with others (Zhang and Watts, 2008). Therefore, they tend to be more recognizable and respected among members in the virtual communities. Specifically, leaders may contribute to virtual communities in the following ways. First, they may serve as sources of useful knowledge since they are usually seniors in their communities and know a lot about the features and histories of the communities (Zhang and Watts, 2008). Second, they can help encourage members' contributions and guide the contents of contributions to be more focused and consistent with the virtual community's purpose. When there are members who violate the virtual community's rules and norms e. g., by posting irrelevant
messages, leaders may act to exert sanctions on them. Finally, leaders can help in developing a healthy environment that promotes productive collaboration among community members (Preece, 2002; Williams and Cothrel, 2000). Therefore, the existence of active leaders may make it more likely for members to obtain the information needed from the virtual community, and thus perceive the virtual community to be more useful. Also by ensuring that the virtual community's interactions focused on its intended purpose, members may identify with the virtual community better compared to a virtual community with ambiguous objectives (Bargh and McKenna, 2004). This is in line with previous research that highlight leaders' ability to induce members to be highly committed to a virtual community's activities (Koh and Kim, 2003). Hence, a higher level of leaders' involvement may also in turn enhance members' sense of belonging to the virtual community. This leads us to the following hypotheses: H8a: The higher the perceived leaders' involvement, the higher the level of usefulness perceived by virtual community members. H8b: The higher the perceived leaders' involvement, the higher the level of sense of belonging perceived by virtual community members. Research Method Data collection In this study, we collected data to test our research model in the context of leisure-oriented communities. The choice of investigating leisure-oriented communities (Schubert and Ginsburg, 2000) was motivated by the following reasons. First, leisureoriented communities represent a form of virtual communities that can serve among the most popular demands of Internet users i. e., obtaining information related to leisure activities on the Internet (Virtanen and Malinen, 2008). The pertinence of these virtual communities thus calls for an in-depth investigation on what may contribute toward continuous participation in them for their long-term sustainability and success. Second, leisure-oriented communities may serve as an epitome of virtual communities with all the utilitarian, hedonic, and social functions. These virtual communities provide an avenue for their members to obtain information, conduct leisure activities e. g., sharing their hobbies, and building long-term relationships with each other. Having established leisure-oriented communities as the study's context, we collected data from five virtual communities in China, namely douban. com, tianya. com, mop. com, tieba. baidu. com, and bbs. sina. com. en. These communities are among the most popular virtual communities in China, all of which have more than one million members. Furthermore, all of them have been flourishing for more than five years, and have stable memberships where a substantial number of members continuously participate in the communities. The latter characteristic is desirable for our study's objective of investigating what may contribute toward continuous participation. The survey items were adapted from extant measurements with adaptation to tailor them to our research context (refer to Appendix). The items measuring the key constructs are all designed a 5point Likert scale ranging from "strongly disagree" to "strongly agree". The initial questionnaire was translated into Chinese because our subjects were recruited from Chinese virtual communities. A pilot test with 20 undergraduate students was conducted to ensure that content validity and reliability were met. Minor changes were made on the wording to mitigate misunderstanding and improve the clarity after the pilot study. The questionnaire was then translated back into English to verify the translation validity. We first published the questionnaire on an online survey website, and contacted the administrators of these virtual communities for the permission to post the link of our questionnaire in some of their subgroups. An introduction was also attached in this posting to explain our research objectives and research scope. Timely replies were provided whenever there were comments or questions related to the questionnaire after posting. The survey was conducted over a two-week period. A total of 347 responses were received. We found most of our respondents to come from two virtual communities. This was because the administrators of these two
communities provided strong support for our study, and placed our posting on a salient position of their discussion forums. In contrast, our posting was soon submerged by other postings in the other three communities, therefore leading to a lower response rate from these communities. The number of respondents from each community is listed at the end of Table 2. In order to ensure the quality of data, we excluded those respondents from the same IP address, as well as those who finished the questionnaire within one minute. Meanwhile, to ensure that subjects were not fresh members with little experience in the respective virtual community, we excluded 10 more respondents who reported that they joined the virtual community within one month. As a result, we had 329 valid responses for data analysis. Results Table 2 presents the demographic profile of our respondents. Most of the respondents are under the age of 25 (60% of them), and half of them have been members in their respective community for more than one year. 44. 1% of the respondents are males, while 55. 9% are females. The distribution of our subjects on past participation, in terms of visiting time and number of postings, is not skewed. Partial Least Squares (PLS) was utilized to assess the measurement model and structural model of this study. PLS is a powerful second generation multivariate technique for analyzing latent variable structural equation models with multiple indicators (Sirohi, McLaughlin and Wittink, 1998). It is widely accepted as a method for testing theory in early stages (Ma and Agarwal, 2007). PLS is suitable for situations when sample sizes are small, and variables are not strictly normally distributed. In this study, we run the data on smartPLS version 2. 0 software package. Measurement model Reliability, discriminant validity and convergent validity were assessed to validate the measurement model. Confirmatory factor analyses were conducted to evaluate the construct validity. Table 2 shows the results of confirmatory factor analysis. The loadings of all items on their corresponding constructs are higher than 0. 6, which indicates good convergent validity. Cronbach's Alpha and composite reliability are often used as measurement of internal consistency of items. Composite reliability takes into account the actual loadings used to construct factor scores, and is considered as a better measure of internal reliability (Chin, 1998; Fornell and Larcker, 1981). We reported both of these two indices in Table 3. While the composite reliability score of all the constructs are above the 0. 80 threshold, the Cronbach's Alphas of two constructs i. e., event organization and usefulness, are marginally below the 0. 70 threshold (0. 67 and 0. 68 respectively). While dropping certain items may improve the Cronbach's Alpha value, we decided to keep all the items since dropping some of them may affect their content validity, and that the difference with the 0. 70 threshold is marginal while their composite reliability score is satisfactory. To assess the discriminant validity of the constructs, the square root of average variance extracted (A VE) i. e., the off-diagonal elements in Table 4, should be larger than the correlations between constructs. All constructs meet this requirement. Structural model Following Chin (1998), bootstrapping method was performed in smartPLS software to test the statistical significance of path coefficients. The results are shown in Figure 2. As can be seen from Figure 2, 12 out of the 16 hypothesized paths are found significant. Together the model explains a considerable proportion of the intention of continuous participation i. e., an R2 of 62. 8%. Enjoyment and sense of belonging significantly and positively affect intention of continuous participation (i. e., H2 and H3 are supported). However, the relationship between usefulness and intention of continuous participation is not significant (i. e., H1 is not supported). For usefulness, enjoyment, and sense of belonging, their R2 yielded are 0. 422, 0. 410, and 0. 534 respectively. This shows that the hypothesized factors related to usability and sociability explain a good proportion of variances of these mediating variables. For usability-related factors, our results indicate that information service
quality positively influences all three mediating variables (i. e., H4a, H4b, H4c are supported). Interaction support quality positively influences perceived enjoyment and sense of belonging (i. e., H5b and H5c are supported, but not H5a). For sociability-related factors, incentive policy strongly influences perceived enjoyment and sense of belonging (both H6a and H6b are supported). Event organization has positive impacts on enjoyment and sense of belonging (i. e., H7b and H7c are supported but not H7a), while leaders' involvement is only positively related with usefulness (i. e., H8a is supported but not H8b). Furthermore, the results are robust after controlling for age, length of participation, and intensity of participation. None of the control variables is significant at p
