行政院國家科學委員會專題研究計畫 成果報告
計畫類別：個別型計畫
計畫編號：NSC93-2416-H-390-001
執行期間：93 年 8 月 1 日至 94 年 9 月 31 日
執行單位：國立高雄大學資訊管理學系
計畫主持人：吳建興
報告類型：完整報告
處理方式：本計畫報告可公開查詢
中華民國 94 年 10 月 31 日
網路化知識社群之研究：網路社群之內聚力及與網路行銷
策略之偶合力
行政院國家科學委員會專題研究計畫 成果報告
網路化知識社群之研究：網路
社群之內聚力及與網路行銷策略之偶合力
Knowledge net community: Cohesion of network community and its
support to Internet marketing
計畫編號：NSC93-2416-H-390-001
執行期間：93 年 8 月 1 日至 94 年 7 月 31 日
主
持
人：吳建興
國立高雄大學資訊管理學系
共同主持人：高淑珍
崑山科技大學資訊管理系
中文摘要 網路社群的快速發展為網路使用者帶來無窮 的機會與方便，例如分享購物經驗，談論共同 的興趣，線上學習、以及網路商業等等。如果 從社會化的觀點來看的話 社群內聚力是來自 社群成員的接受程度及社會行為，因此是網路 社群生命週期的一個重要的問題。本研究以實 證方法探討網路社群內聚力的決定因子及相 關的管理意函。研究結果發現：1)社群領導對 社群內聚力有負面且顯著的影響；2) 社會關 係對社群內聚力沒有顯著影響，主因是虛擬信 任對網路使用者彼此行為造成的影響；3) 溝 通工具對社群內聚力沒有顯著影響，主因是溝 通工具只能視為網路使用者參與社群的一個 先決條件，而不是直接影響提高內聚力的因 素。相關的討論與管理意函也描述在本研究 中。 AbstractIt has been seen the rapid growth of the number of Virtual Internet Community (VIC) over the Web cyberspace to bring together the Internet users who have common interests, want to circulate and share information, wish to learn something, and desire to do business. However, Little studies have discussed about cohesion of VIC from Internet users point of view. This research presents an empirical investigation to explore cohesion of VIC, where composite variables including leadership, social relationships, communication tools, and functions are examined. There were remarkable results that 1) leadership showed a negative and significant impact, 2) social relationships did not indicate significance mostly due to the virtual trust having an effect on users behavior, and 3)
communication tools did not indicate significance because it is a prerequisite to motivate users participation, rather than an immediate factor to enable the VIC cohesion. Discussions and implications are addressed in this research also.
Keywords: Virtual community, Cohesion, Internet
1. Research Background
Internet technology has demonstrated its success on the enhancement of interactions among computer users This results in rapid growth of Internet users. Consequently, Virtual Internet Community (VIC) is extensively built up in the Web cyberspace to bring together the Internet users who have common interests, want to circulate and share information, wish to learn something, and desire to do business. It has been seen that VIC is increasingly capturing the attentions of management in various domains as well as of people whose life depends highly on Internet. Therefore, potential issues of VIC become hot topics for Internet applications as well as Internet users behavior in information society, such as “what are the main determinants to start and/or run a VIC?”, “is VIC ready for supporting healthy social development”, “do communication tools and functions of VIC affect Internet users to join into?”, “does a computer companion in a VIC really enrich daily life?”, “is it better to do online marketing via VIC than via postal ?”, “is it better to go shopping via VIC than via eBay?”, and so on so forth. Information discovery to help in figuring out these questions becomes essential for VIC management.
From the social development stand viewpoint, research in literature has witnessed that the VIC has turned the way that people seek
for and build up companionships, social support and belongingness (Wellman & Gulia, 1999; Venkatesh, 2003). On the one hand, Walther (1992) mentioned that face-to-face activities could not totally affect the development of the interpersonal relationships, and therefore the textual information in the computer-mediated communication should be also considered as a kind of social information processing. Importantly, the textual information in the computer-mediated communication can help communicators to form impressions for each other and can facilitate developing social relationships. Therefore, it is thought that the VIC is not only the function embedded in a website, but indeed becomes an important social development channel. On the other hand, Cummings et al. (2001) argued that with regard to the social relationships Internet is still less effective than real world.
Basically, the cohesion comes from users social behavior. It would involve some fundamental factors and is considered to be the degree of unification that both prosperously improves and enables VIC to survive in groups for members (Mbaatyo, 2001). Forrest & Kearns (2001) indicated that in general a society that is short of cohesion would lead to social chaos and conflict, and in consequence may result in social inequality and low levels of social interaction. However, there are no suggestible and significant determinants and processes so far to manage a VIC with respect to the enhanceent of cohesion. Literature have shown that there are four major facets (see Table 1) that may have potential effects on cohesion of VIC, including leadership, social relationships, communication tools, and functions. Therefore, research problems may be raised that “should a VIC need leaderships just like traditional community?”, “is a VIC able to help in social development at the current stage?”, “should a VIC need good communication tools?”, “should a VIC be a virtual learning environment?”, and “should a VIC include both virtual and physical activities?”. By understanding so, cohesion of a VIC being built can be enhanced effectively. More precisely, understanding that “what brings Internet users to frequently connect on for doing something in the same VIC?”may be important to those who would like to institute and/or make longer a VIC. Therefore, while many studies have made efforts on issues of the fundamental concepts of formation and advantages of VIC, motivation of building a VIC, and types of VIC, the current study puts emphasis on the exploration of VIC cohesion. The remaining part of this paper is organized as follows. Literature review is presented in the second section. Section three describes the research
methodology where research model, research hypothesis, sampling plan, and results of data analysis are included. The discussions and implications with respect to the current research results are presented in the forth section. Final section concludes our current research and addresses some future research focuses.
2. A Review of Literature 2.1 VIC and Cohesion
In the Internet environment, the textual information in the computer-mediated communication can be considered as a kind of social information (Walther, 1992). It can drive communicators to form impressions on each other and facilitate developing social relationships. This results in the quick development of virtual community over Internet. Socially, VIC may naturally build up communal regulations and beliefs for members behavior (Ridings et al., 2002). By focusing on the social combinations, Howard (1993) and Carver (1999) indicated that VIC is a social phenomena that sufficient Internet users come forward onto the web to proceed with public discussions, to distribute and share information, and to build up personal relationships. Romm and Clarke (1995) regarded the VIC precisely as a group of people who communicate for anything at anytime from anywhere over the information superhighways. From the information contents stand viewpoint, Hagel and Armstrong (1997) defined the VIC as a computer-mediated spaces where community members provided information of common purposes and interests. Craig and Zimring (2000) put emphasis on the sense of community, that in a VIC it is not guaranteed by opportunities for interaction; rather, it must maintain a sufficient size out of interaction itself to keep it alive. Ho et al. (2000) indicated that a VIC should be a computer-mediated web space in which participants can proceed with multi-user engagement in a multiple as well as real-time interaction style. Lee et al. (2003) also provided a survey by looking at its scope and potential research issues in a basic detail for virtual community. From the life cycle stand view point, Palloff & Pratt(1999) indicated that VIC naturally had five stages: forming, norming, storming, performing, and adjourning. Particularly, they regarded the storming as inbuilt conflicts and adjourning as fading. Apparently, issues involved in VIC establishment, development, implementation, and termination are quite ill-defined, and in consequence the determinants with respect to the VIC cohesion would be multifaceted, but quite essential to VIC life cycle.
Cohesion is defined to be the unified degree that supports the groups or organizations
to keep alive, and in consequence affects the team works (Evans & Jarvis, 1980; Mbaatyo, 2001; Regina, 2002). Importantly, the cohesion is regarded as to gather the interpersonal attractions together, and aggregate as a whole (McGrath, 1984). It is in general believed that without sufficient degree of cohesion, a community or a society would cause social disorder, conflict, inequality, diversity of moral values, and would soon lead to fade. As suggested by Forrest & Kearns (2001) in discussion of social cohesion, social capital, and neighborhood, the domains of social cohesion include common values and civic culture, social order and social control, social solidarity, social networks and social capital, and place attachment and identity. Although entirely dealing with the real world community, instead of the VIC, these domains are important to VIC related issues. Regina (2002) have also mentioned that common values and shared values are important for a real world community as long as the cohesion is concerned. These imply accordingly that a cohesive society and community in both real world and cyberspace should have common objectives.
Generally, the lack of common value may cause members to abandon the community, and as a result will break the solidarity. The weak social solidarity will result in unequal opportunities to access to welfare of community. Moreover, the community acts to achieve the common goal and implement the tasks through social order and control. Therefore, the weak social order and control will result in ignorance of common value and thus the community members will lose their focus. Equally importantly, the tight network and association are necessary to form cohesion (Forrest and Kearns, 2001). The tight network is to frequently interact within the community and the members will share the knowledge and experiences to gain the maximum benefits (Mbaatyo, 2001). Eventually, the cohesion is to show the great belonging among members and the strong attachment to community (Reginal, 2002; Forrest & Kearns, 2001). Therefore, it can be found that common values, solidarity, equal opportunities, high degree of social relations, and belonging of place are all potential factors that may have impacts on the cohesion of VIC. 2.2 Leadership for VIC
In order to have all activities matching the common objective, the leadership of the community members is in general voluntarily enthusiastic and interacts with each member in the virtual world. Moreover, to be the leader of a VIC he/she has to reply to themembers’requests for the community rapidly. As suggested by
William & Cothrel (2000), the leadership should reply to any request in 48 hours. Therefore, it is argued that (1) the leadership should actively interact with his/her members in VIC and (2) there is a certain amount of experts who are familiar with topic contents in VIC (William & Cothrel, 2000). More importantly, forums should keep the balance proportions of professional personage and beginner in the virtual community, because by doing so the members who have some problematic demands can enable to gain knowledge from the experienced one or the person who has professional capability (William & Cothrel, 2000). It is expected that while members participate in a community, the leader who might act as an expert is to provide the necessity of knowledge faster for the specialized information. Although the leadership is important to a VIC, Sivan (2000) has found that leadership in a virtual world needs various leadership qualities, such as leadership entries, leadership of process, and knowledge of management and organization. It is believed up to now that any community, institute, or organization in the real world should need a leader (e.g. president, master) to maintain the association purposes. However, members in a VIC can communicate with each other for any subject at anytime from anywhere. This is totally different from those in the real world. Therefore, our concern in the current study is both to see whether or not the Internet users still regard this issue as a key VIC cohesion and to draw some implications for this facet.
2.3 Social relationships for VIC
Basically, a VIC featured by social behavior forms with some people who have the common interests and goals (Hagel & Armstrong, 1997). Adrews (2002) indicated that the community has to express the purpose that members are interested in. By doing so, once a member looks for building companionship on someone who has the same interests, he/she can be satisfied with the interpersonal relationship by joining or creating the relevant theme community (Hagel & Armstrong, 1997; Kozinets, 1999). Thereby, as far as the interpersonal relationships in VIC are concerned, members can span the restrictions of both time and region through the medium, and therefore can easily know someone who has the same experience in his/her life (Hagel & Armstrong, 1997). Particularly, when a VIC is rapidly to develop its cultural cohesion, the manager of VIC has to build up the relationship of sharing capability. The procedure of sharing capability may contain conflicts and clarifications in the politics of VIC (Kozinets, 1999). By this, the members of VIC could easily seek for the people who have the
common brief (Igbaria, 1999). Ridings et al. (2002) also suggested that trust was significantly related to the virtual community members’desire to share information. While this process develops the interpersonal relationships among members in community, the socialization and the sense of belonging among members should be considered as a key point (Ba, 2001; Commings et al., 2002). Equally importantly, it is necessary to continuously create the sense of belonging within the VIC, so its cohesion can be improved and strengthened relatively. Moreover, the aspect of impression demands and transaction demands are also important for the enhancement of cohesion of VIC. In the impression demand, it is generally believed that the role members want to be is mostly a fantasy player. It is almost impossible to act in reality. For example, on-line game can attract many players to come together at the same time over Internet (Hangel & Armstrong, 1997). For the transaction demands, it is also believed that members tend to accept each other if they are interested in the same product and service. Accordingly, while members exchange the information and discuss shopping experiences in virtual community, this phenomenon will raise the transaction demands within the VIC. However, as Comm et al. (1997) indicated that in a VIC members can have full freedom to describe themselves to others, which could develop “virtual identity”as they would like them to be such as age, social status, occupation, gender, interests, etc. These virtual characteristics are very possible to lead to the development of implausible relationships among members, and in subsequence to an unhealthy “society”with fairly weak cohesion.
2.4 Communication tools for VIC
The communication in VIC basically includes information exchange, cultural norms, communal relationships, and cultural cohesion. It is considered to be a driver to exchange all kinds of information within the community, thus communication tool is basically a fundamental component for developing relationship, forming common norms, and sharing information as well as knowledge. Fulk et al. (1987) and Yates & Orlikowski (1992) indicated that constraints for particular environments, interpretation for the symbols used, and contexts for societies are important composites of media richness, and therefore should be included in the communication alternative list. In addition, the facet of topic expression is also crucial for the improvement of cohesion of VIC. The topic expression, usually is ignored by the planner of VIC development, can greatly influence the discussing contents. It can also bring an explicit position that guides other potential members to
join by the contents, tools, and service in community (Williams & Cothrel, 2000). Moreover, Andrews (2002) indicated that a clear community theme is quite important for the cohesion of VIC.
Romm et al. (1997) suggested that technology be an important variable that would influence people to join in a virtual community, although only on the field of e-mail adoption. Advanced Internet technology has provided many channels nowadays for communication that members can use to post multimedia materials in VIC. These channels mainly include e-mail, discussion forum, chat, instant messages, and online meeting (Williams & Cothrel, 2000; White, 2001). Moreover, if members want to know and communicate with each other, community can develop the function of searching member directories that include the basic personal information and profiles. In addition to knowing members’ background, members can quickly find the members who can help to solve problems (Andrews, 2002). This directory is quite beneficial to communication and know-who and -what management in virtual community. More importantly, group support technology has become important in supporting communication. Bieber, et al. (2002) indicated that infrastructure ready for the interactions among large dispersed groups by using multiple media forms is quite essential for a VIC. Obviously, Internet technology has made it possible that people could only think before, but now have come to their everyday life, such as online talk via Internet, TV watching on computer, stock buying and selling via personal digital assistant (PDA). Therefore, a VIC with multiple-channel communication tools is advantageous to its cohesion.
2.5 Functions for VIC
In the study of strategic implications of virtual communities of consumption, Kozinets (1999) proposed a four-facet model to describe the interaction-based segmentation. These are relational mode, information mode, transformational mode, and recreation mode. In recreation mode, particularly, the online communication is usually selfish and in consequence may not last too long. Although the cohesion of such a VIC may be weak, it is thought that a VIC with the entertainments function embedded can keep VIC members staying in longer, and accordingly enhance their interests on not only the entertainments itself, but other functions a community can provide, such as hotline news and learning topics. In addition to the enhancement of attraction for visitors who have the same interests, the VIC should also be able to provide members with a
learning platform when the general purpose of a VIC is concerned. As suggested by Rothaermel & Sugiyama (2001) and Koh & Kim (2004), knowledge sharing activity is significantly related to the heath of a VIC. It should be aware essentially that the motivation of joining a virtual community for an Internet user is able to find useful solutions for problems (Ho et al., 2000). By this, the members of VIC can obtain the related consumption knowledge through cultural norms, specialized concepts and languages, and experts’ comments (Kozinets, 1999). Moreover, the community can arrange professional knowledge systematically and therefore its members can share the subjects and their contents (Andrews, 2002). This exchange knowledge function can enable members to learn something effectively by providing members with individualistic request (Ho et al., 2000).
It is thought that the virtual community can facilitate the interactions both in virtual space and in physical activities (Fischer, 1998; Borthick & Jones, 2000). Rothaermel & Sugiyama (2001) argued that a member’s off-line interaction is positively related to the economic transaction within a VIC. In addition to relationship that has been built in the VIC, members’interactions by the face-to-face way can also improve their relationships, such as physical meetings or conferences (Andrews, 2002). The primary function in physical activities is a kind of face-to-face communication, thus some non-language characteristics can be expressed such as the look on the face and the actions of members. As suggested by Johnson (2001) face-to-face contact should be added as a function of VIC. Moreover, it is generally believed everyone may have different interpretation and cognition about the utility words. Members of VIC who communicate with each other may not totally be satisfied with all needs and wants only via virtual world. Therefore, the real world activities can soundly enhance the on-line relationship in virtual activities of virtual community.
3. Research Methodology 3.1 Research model
The research model illustrated in Figure 1 is constructed using a review of literature and a series of in-depth interview with domain specialists. There are four research composite components as the independent variables, including the leadership for VIC, social relationships for VIC, communication tools for VIC, and functions for VIC. The dependent variable is the cohesion of VIC. The research problem is to disclose the significant connections between these variables and VIC cohesion. The sample population is selected
from people who are experienced in using Internet. It is noted that the current research did not take on only those Internet users who have to be a member of at least a VIC. In fact, it is realized that the inexperienced VIC members may not be knowledgeable to provide information and/or comments about VIC. However, for the concern of broadly gathering Internet users opinions, this current study put emphasis on both the experienced and the inexperienced with respect to VIC members. 3.2 Research hypotheses
The person responsible for the leadership of VIC can be an expert or maintainer who acts as a server to satisfy member’s needs in VIC (William & Cothrel, 2000). Therefore, to determine the importance of a leadership for VIC, this current study mainly considers (1) the leader is necessary for VIC, (2) the leader's charisma is important for VIC, (3) the leader should actively interact with his/her members in VIC, (4) there should be a certain number of experts who are familiar with topic contents in VIC, and (5) a VIC should have someone to maintain its contents. The first research hypothesis is then defined as:
H1: The leadership for VIC is significantly related to the cohesion of VIC.
It is important to develop the interpersonal relationships and the sense of belonging among members, as well as the socialization in community (Ba, 2001; Commings et al, 2002). To measure the significance of social relationships among members with respect to the cohesion of VIC, the current study uses the items as follows: (1) a VIC should be able to satisfy members’interests, (2) a VIC should be able to satisfy the wants of friendships, (3) a VIC should be able to satisfy wants of thinking in fantasy and imagination, (4) a VIC should be able to bring people together who have similar interests, (5) a VIC should be able to help in sharing everyday information with others, and (6) the members of VIC can participate in the politics and execute their missions. The following hypothesis is then formed.
H2: The social relationship is significantly related to the cohesion of VIC.
Tools for VIC members to communicate are important. Basically, there should be many communication channels by which members can post their suggestions, opinions, comments, or even complaints in virtual community, such as e-mail, discussion forum, chat, online meeting, bulletin board, instant message, MSN messenger, and Internet broadcasts (Williams & Cothrel, 2000; White, 2001). To investigate whether or not the communication tools is significant in support of cohesion of VIC, the third research
hypothesis is then defined. The measure basis are (1) a VIC should provide a multi-channel for members to express their suggestions, questions and request, such as e-mail, chat room, forum, MSN messenger, (2) a VIC should have a friendly communicational interface, and (3) a VIC should have the guideline to direct its members interacting with other members.
H3: The communication tool is significantly related to the cohesion of VIC.
It is thought that entertainment of a VIC is important to its cohesion. More precisely, the entertainments function may have a certain amount of impact on the attraction of VIC members for members acting in the community more frequently (Kozinets, 1999). In addition, learning and sharing knowledge in VIC have been seen two important functions for a VIC (Rothaermel & Sugiyama, 2001; Johnson, 2001). By these two functions, members can obtain and/or search solutions for their problems, and therefore may have impacts on the cohesion of VIC (Ho et al, 2000; Andrews, 2002). Furthermore, it is generally believed that the interactions made by members of VIC in real world can also facilitate expressing the non-language or text-only characteristics among members, and therefore would improve the degree of solidarity for a virtual community (Andrews, 2002). The final hypothesis is then formed. The measure basis is centered on (1) a VIC should provide the function of game playing, (2) a VIC should have the activities of lottery, (3) a VIC should have the activities of accumulating bonus points, (4) a VIC should provide members with professional knowledge to members for various topics, (5) a VIC should provide the information to stimulate members’ motivation and interests for topics, (6) a VIC should have a variety of current information based on the members’preferences, (7) a VIC should have some periodic seminars or activities referenced to VIC’s theme, (8) a VIC should held physical meetings or activities to increase members’relationships, and (9) a VIC should interact with other physical communities and organizations.
H4: The VIC function is significantly related to the cohesion of VIC.
In regard to the measure of the cohesion of VIC, this current study focuses on the concept that a computer-mediated space is an organization that can provide members with a common value, mental and mutual benefits, problem solutions, a forum where members like to participate frequently (Evans & Jarvis, 1980; McGrath, 1984; Mbaatyo, 2001; Forrest & Kearns, 2001; Regina, 2002). Therefore, this research uses five items as follows to measure this dependent variable of cohesion of VIC: (1)
the members have common value within VIC, (2) there are group advantages in VIC, (3) the members in VIC are equal and help each other, (4) the members in VIC interact with others frequently, and (5) the members in VIC both have commitments with other members and willingly join related activities.
3.3 Sampling plan
The sample population covered Internet users containing both the experienced VIC members and the inexperienced ones. According to an Internet users investigation report for the year of 2004 by MIC (Market Intelligence Center, Taiwan), there were about 72.1 % users whose ages were between 20 and 34 (Hong, 2004, http://mic.iii.org.tw/intelligence/). The current study therefore took on the convenient sampling method to collect data from universities, colleges, and high-tech industries. Although not very accurately, participants in these groups generally aged about in this range. With regard to participants’ experience on Internet community, this current study did not set this constraint to only experienced VIC members. In fact, it is well believed that an experienced Internet user can provide more useful information. As suggested by Rothaermel and Sugiyama (2001), experience was positively related to the economic transaction in a VIC. In this case, however, it could not be denied that some information from inexperienced participants may be lost. By a discussion with domain experts in depth, therefore, this current study at this particular time considered both experienced VIC members and inexperienced Internet users as the information providers and indicated the percentage of both groups of participants later in this section. The number of samples was derived from the conditions that 1-α = 0.975, sampling error = 7%, minimum response number = 196, and expected response rate = 70%. It should be noted that because this study took on the convenient sampling method to collect data, the expected respond rate was higher than average, and accordingly was set to be 70%. Therefore we sent out 280 questionnaires. In addition, in order to evenly collect data from different areas, 70 questionnaires were sent to participants in each group of local high-tech industries, colleges, research institutes, and universities. However, this current study did not intend to go into the comparison manner for these groups.
Concerning the design of questionnaire, a 7-digit rating scale (from 1 to 7) using bi-polar descriptors for each question was constructed. Each item was measured by a 7-point Likert scale where 1 represents “strongly disagree”and 7 “strongly agree.” Moreover, to have the
designed questionnaires more appropriate, many in-depth interviews of domain specialists in the area of VIC were conducted. Before the questionnaires were sent out to the participants, a pre-test was conducted via a limited number of individuals to develop the validity, readability and reliability of the questionnaires. Based on their comments, the final version of the questionnaire was determined, containing a total of 28 items, of them 23 were for independent composite variables and 5 for the dependent.
The research took on both e-mail and post-mail to send questionnaires to participants. A numerical checkout list was used for the entire sample to check off the list when a respondent was returned. After four weeks, in order to increase the response rate, e-mail and telephone follow-up were made to those who had not responded yet. Of the 280 questionnaires that were sent, a total of 243 were returned, which showed an 86.79% returned rate. Of the 243 responses, 61 (87.14%) were from local industries, 65 (92.86%) from colleges, 50 (71.43%) from research institutes, and 67 (95.71%) from universities, indicating the response rate for each group was not unacceptable. However, of the returned questionnaires, 9 were invalid and could not be used in the statistical analysis. It is noted that a returned questionnaire was regarded as invalid if one or more than one question was not answered. Moreover, it was found that of the total returned questionnaires, 82.4% was from experienced VIC members while 17.6% from the inexperienced.
3.4 Data analysis and results
The research first conducted the factor analysis to reveal the final research variables and the reliability of each variable were confirmed statistically. The results were listed in Table 2, including composites, factors, items, loadings, Eigonvalues, item to total correlation, and Cronbach’sα. This indicated that the composite variables were grouped into seven factors. They were leadership, members’interactions, common interests, communication tools, entertainments, learning, and physical activity. It should be noted that an item grouped in the composite “social relationships”was eliminated to have the factor “members’interactions”reliable. The value of Cronbach’sαfor each factor was confirmed to be reliable, which were all above 0.65 (Nually, 1978). The research then moved to the confirmation of model fitness and hypothesis tests. The data analysis tool, Lisrel, was employed to derive their results. The path diagram of Lisrel model was illustrated in Figure 2. In the Lisrel model, all variables are divided into manifest variables that could be observed
directly and latent variables that could not. All latent variables could be further divided into exogenous variables (latent independent variables) that come from variables outside the model and endogenous variables (latent dependent variables) that depend on other variables inside the model. An exogenous latent variable was presented byξ(e.g. leadership) and endogenous latent variable by η(e.g. cohesion of VIC). The exogenous manifest variable was presented by X used to be the measurement of exogenous latent variable and endogenous manifest variable by Y used to be the measurement of endogenous latent variable.
To ensure whether or not the model is fitful, the value of RMR (root mean square residual) had to be less than 0.05, GFI (goodness of fit index) greater than 0.9, and AGFI (adjusted goodness of fit index) greater than 0.9. Table 3 showed that the RMR was 0.007, GFI 0.999, and AGFI 0.987, respectively, indicating that the fit statistics were acceptable. This was according to the suggestions for this statistical method by both Churchill (1979) for evaluation of psychometric properties in terms of measurement scale for constructs and Lehmann et al. (1999) for convergence and discrimination validation. It was found that the path relation that “Members’interactions (λx2) → Physical
activity (λx7)”was advised by the utilized tool to
improve the model fitness. This implies that members’interaction has a partial compensation and a significant effect on the physical activity because the estimate is 0.91. Results of hypothesis tests were also shown in Table 3. It was found that leadership, entertainments, learning, and physical activity were significantly related to the cohesion of VIC. Particularly, the leadership showed a negative and significant impact on the cohesion of VIC.
4. Discussions and Implications
From the data analysis results in Table 3, the research findings indicated acceptance for four determinants that were leadership, entertainments, learning, and physical activity. Discussions and implications contain two parts. The first part is for the determinants that showed significant to the cohesion of VIC while the second for those that did not.
4.1 For leadership
Firstly, the leadership shows significant and negative effects on the cohesion of VIC. This implies that members are not likely to welcome too much involvement of the leadership of a VIC. From the literature, most research suggested that VIC needs leadership to enhance its cohesion by directing subjects, linking experts for information and knowledge
sharing, and solving technical problems. Our research result is not according to this expectation that a VIC needs leadership. A possible reason could be that leadership can attract an Internet user to join into and enjoy the VIC for a short period, but may not last too long, and in consequence has no positive and significant effect on the cohesion of the VIC. Moreover, the virtual community over the Internet is growingly changing the needs of leadership. As suggested by Sivan (2000), the current leadership theory cannot sufficiently describe leadership in VIC, because the lack of leadership experience in virtual world. Importantly, according to our current examination, leadership should realize that once an Internet user has become a member of VIC he/she is most likely to have and wish full freedom to express opinions, to interact with members, and to share information without any interference or disturbance. On the one hand, our suggestion for this therefore is that the leadership of a VIC could take much care of the VIC service spectrum at the starting stage and gradually let the VIC develop in its own way. On the other hand, when more and more Internet users have obtained a basic amount of experience in VIC implementation, leadership precedents with respect to VIC organization and management would be more useful. Furthermore, although it was found that mean of the measure item that “a VIC should have someone to maintain contents”for this variable was relatively high, the remaining items showed that leadership of a VIC was relatively low to its cohesion. We also suggested that VIC initiators and/or runners should not at the wheel of the VIC, instead they should focus mainly on the maintenance of contents. Particularly, based on the investigation result for “functions”variable, they should put efforts on the functions development based on the members discussion topics such as learning environment establishment for subjects, physical activities, and collaboration with other physical communities.
4.2 For functions
Secondly, the variable of functions for VIC showed a significant influence on cohesion of VIC. This implies that Internet users are likely to accept their VIC having a multiple functions for members. In other words, the entertainments, learning, and physical activity do in fact give significant effects on cohesion of VIC. Apparently, Internet users are becoming increasingly function-sensitive as the various types of VIC become more and more numerous, and thus are broadly to make use of all VICs available, particularly for learning and
entertainments. To enhance cohesion of VIC, it is our suggestion that a VIC runner should keep tracing all changes of members’and Internet users’needs and wants with respect to learning knowledge, entertainment, and physical events. Concerning physical activities, it is well believed that real world activities cannot totally be replaced soon by virtual cyberspace supported by Internet technology. This is exactly according to the suggestion by White (2001) that “by combining the interactive tools available online with some of the targeted offline information delivery vectors, you can meet your audience’s needs.”Therefore, physical activities raised for the cohesion of VIC are necessary. In addition, this has been also confirmed by our participants in this current research and gives a positive support to the suggestion by Rothaermel & Sugiyama (2001) that “site organizers should build mechanisms into their sites that facilitate off-site contact.”
4.3 For social relationships
For the variables that showed insignificant, it was found, surprisingly, that social relationships and communication tools so far were not strongly associated with the cohesion of VIC. The finding for social relationships indicated that the concept from some literature that VIC could be a type of social development is not convinced at this particular time. However, this finding is exactly according to the argument by Cummings et al. (2002) that “Internet is less effective than other means of forming and sustaining strong social relationships”, but in contrast to the assertions by most literatures. When looking back to the measure basis defined for this composite and the results shown in Table 2, it was found that members’interactions showed significant effects on social relationships, but common interests. For both, we have proceeded with some personnel contacts of our participants by e-mail to further disclose possible reasons and implications. Our findings for the former, on the one hand, were that although participants interact frequently with others via VIC, they are very careful for communications and information providing in VIC since ethics in the cyberspace over Internet is quite different from those in the real world. For example, one of the participants we contacted indicated that he only shared general experiences to a real estate VIC about how to negotiate and bargain, but never say which department he experienced, because doing so may cause serious problems. However, if any one of whom he already knows asks for it in personal talk he will say without too much thinking. In a virtual world, everything could be unreal. Privacy and freedom in the Internet
society is important, but still ill-defined so far. Relationships among members that were established under such unreal environment may very possibly lead to artificialities, and in consequence will lead to a bad cohesive community. This is closely according to the suggestion by Comm et al. (1997) that virtual identity may end in negative organizational results such as demoralization, dishonesty, and disobedience of rules and regulations.
Moreover, Internet ethical problems (e.g. social boundaries and social expectations) are likely to be worse and worse, and some of them were due to the “virtual trust”in the virtual world. This “virtual trust”will strongly have a great effect on users behavior, and then on members’interactions, and finally on social relationship. For example, talking to people in public is absolutely different from that to one or two friends, particularly when they have no idea about who they are, what they are doing, and what they will react. In this current study, our participants did not give a significant support for that the “virtual social relationship” has a significant impact on cohesion of VIC. After all, they do not accept that there is no significant difference between “virtual social relationship” in VIC and social relationships in real world. This is just according to the finding by Ridings et al. (2002) that trust was a crucial aspect in VIC. On the other hand, if only to consider the part that by interactions VIC members can easily build “virtual friendships”, looking for comments and/or solutions for problems, and share political advocacy, the participants in fact gave a positive support on that VIC is a new world people can go into virtually and do something different from that in real world. In support of this, as illustrated in Figure 1 and Table 3, it was found that that “members’ interactions (λx2) were significantly related to
the physical activities (λx7)”. Our suggestion
therefore is that VIC runners (or organizers) should not entirely deny the effect of members’ interactions on the enhancement of cohesion of VIC, instead, they should raise for Internet users some physical activities. To do so, the “virtual trust” among members would be gradually shifted to “real trust”.
For the later, it is generally believed that common interests could be an important issue for building social relationships in the real world. However, we found from the research results that it was not significantly related to social relationships. Again, further contact of participants indicated that although participants can easily find people who have the common interests they still stay in the stage of expressing opinions for interested topics. They did not agree with that talking for some common interests in
virtual world is so called building social relationships, even the relationship is just “virtual”. This implies that common interests may not be able to make stronger the social relationships in Internet community, and in consequence having no significant effect on the cohesion of VIC. For example, one of them pointed out that after a series of discussions as well as comments expression for stereo systems, he found he preferred listening to real machines and having talks with his friends personally, although he can indeed obtain much valuable information via the VIC. He did not want to just talk online, instead he did prefer real feelings of any interesting entity. This just tallies with the finding of the variable of functions stated before. Our suggestion therefore is that VIC runners should realize that VIC does have the capability of gathering people with common interests, but may have no significant effect on its cohesion in accordance with social relationship. Raising physical activities may be somehow a way to strengthen the cohesion of VIC.
4.4 For communication tools
For the variable of communication tools, surprisingly, it was found that Internet users were most likely to regard it as an indirect factor affecting the cohesion of VIC. Although some studies argued that media richness do not have significant amount of effectiveness on task performance (e.g. Suh, 1999), but others do (e.g. Workman et al., 2003), we found from further contacts of participants that multiple communication tools have been regarded as a “default”requirement for Internet applications. Basically, we cannot deny that the role communication tools play in cyberspace is very important. However, when it comes to the cohesion, a VIC without basic good medias to use Internet users will not look for what it has, how interesting it is, and whether or not to join it, no less than they interact with other members very frequently to increase its cohesion at the later stage. As indicated by Romm et al. (1997) that the use of any Internet technique is in fact a matter of social rather than personal preference. It depends highly on whether or not the technique is perceived as appropriate by the VIC of potential Internet users. For example, the technique for a new service“Blog”is introduced on Internet recently, providing Internet users (e.g. product designers and/or idea generators) with a function, just like a daily, to instantaneously write down what comes to their mind and what they are thinking, and to regularly record what they have done for himself and/or other members reuse. Once this communication tool becomes popular and is regarded as a “default” requirement of VIC, then the focus can go to the
phase of how to increase cohesion. Moreover, Burke et al. (2001) also indicated that social information processing theory has indicated that community users will acclimatize their communication sooner or later, regardless of the media used to support the community. This implies that there must be something else that Internet users are likely to think more important. This could be how to design a VIC closely linking to their everyday life. It is our suggestion that VIC management should put more efforts on the connection of real world life and their VIC, instead of just having the VIC embedded with some convenient tools, although they can stimulate users participation willingness.
5. Conclusion
In this paper, we have drawn attention to the importance of VIC, highlighted the potential determinants on cohesion of VIC, and described the research model containing research hypotheses and sampling plan, and data analysis results. Factors analysis was used to group the composite variables and Lisrel to confirm the model fitness and to test the defined hypotheses. The objective of the current study was to statistically confirm the impact of leadership, social relationships, communication tools, and functions on the cohesion of VIC. Results showed that the leadership had negative and significant impact, the social relationships was not significantly related to VIC cohesion mostly due to the virtual trust having an effect on users behavior, and the communication tools did not indicate significance because it is a prerequisite to motivate users participation, rather than an immediate factor to enable the VIC cohesion. Furthermore, it is believed that within this swiftly changing Internet era, something is true at present, but may very possibly be wrong soon. In spite of our current study that has drawn out some information about the cohesion of VIC so far, it is believed that there still are a number of variables that vary in their impact on cohesion of VIC as well as Internet users behavior. Generally, it is not simple to collect potential VIC related variables, because of the change of perspectives after a while within Internet society. This needs an in-depth review of literature and great contribution of domain experts and Internet application professionals. To support this, VIC management has to continuously search for the most advantageous connections that can both meet their members’expectations and benefit their members at the time the innovative Internet technology is obtainable, particularly from both human beings and social development stand viewpoint.
It has been seen that Internet technology is increasingly changing the way social networking
is performing. Johnson (2001) indicated that social networking basically is the practice of enlarging the number of one’s social communications by making connections through individuals or organizations, and therefore is a need of social development in which many variables involve such as social structure, social capital, and neighborhood. Web sites created for social networking can be a type of VIC that has become increasingly popular in the cyberspace because geographical blockades and time zones are broken. However, as our research results indicated, VIC at present is still at the stage of “an Internet application”by which users can communicate more efficiently. From socialization stand viewpoint, it indeed has impact on our real world society, but not significant, because Internet users are most unlikely to accept it having momentous relations to social development. Although researchers in Internet technology innovations are also broadly developing new works for real world, and surely will have more or less influence on users behavior and everyday life, we believe that characteristics of human beings is not likely to be completely replaced soon by these works. For example, the feeling of personal presence for an opera is absolutely different from that of being in front of a computer. More importantly, although the movement of Internet technology will continue endlessly, our research results suggested that at present Internet users are most likely to accept that VIC is mainly to enrich our life and to ease our works. In Internet age, after all, innovative Internet technology is produced to benefit human, instead of to change human beings value that machine will never be able to replace.
References
Andrews, D. C. (2002), “Audience-specific online community design”, Communications of the ACM, 45(4), 64-68.
Ba, S. (2001),“Establishing online trust through a community responsibility system”, Decision Support Systems, 31, 323-336. Bagozzi, R. P. & Dholakia, U. M. (2002),
“Intention social action in virtual communities”, Journal of Interactive Marketing, 12(2), 2-21.
Bieber, M., Engelbart, D., Furuta, R., Hiltz, S.R., Noll, J., Preece, J., Stohr, E.A., Turoff, M., & Van de Walle, B. (2002), “Toward virtual community knowledge evolution, Journal of Management Information Systems, 18(4), 11-35.
Borthick, A. & Jones, R. (2000), “The motivation for collaborative discovery
learning online and its application in an information systems assurance course, Issues in Accounting Education, 15(2), 181-210. Burke, K., Aytes, K. & Chidambaram, L. (2001),
“Media effects on the development of cohesion and process satisfaction in computer-supported workgroups”, Information Technology & People, 14(2), 122-141.
Catterall, M. & Maclaran, P. (2002), “Researching consumers in virtual worlds: a cyberspace odyssey”, Journal of Consumer Behavior, 1(3), 228-237.
Churchill, G.A. (1979), “A paradigm for developing better measures for marketing constructs”, Journal of Marketing Research, 16(1), 64-73.
Cummings, J. N., Butler, B. & Kraut, R. (2002), “The quality of online social relationships”, Communications of The ACM, 45(7), 103-108.
Evans, N. J. & Jarvis, P. A. (1980), “Group cohesion: a review and re-evaluation”, Small Group Behavior, 11(4), 80-89.
Fischer, M. (1998) “Using lotus notes learning space for staff development in public schools”, Journal of Interactive Learning Research, 9(3/4), 221-234.
Forrest, R. & Kearns, Ade. (2001), “Social cohesion, social capital and the neighborhood”, Urban Studies, 38(12), 2125-2143.
Fulk, J., Steifield, C.W., Schmitz, J. & Power, J.G. (1987), “A social information processing model of media use in organizations”, Communication Research, 14(5), 529-552. Hagel, J. & Armstrong, A. (1997), “Net gain:
expanding markets through virtual communities”, Mass: Harvard Business School Press.
Ho, J., Schraefel, M.C. & Chignell, M. (2000), “Toward an evaluation methodology for the development of research-oriented virtual communities”, Proceedings of the 9th international workshops on enabling technologies: infrastructure for collaborative enterprises (WET ICE’00), Gaithersburg, Maryland, 112-117.
Hong, T.H. (2004), “An analysis of Internet users behavior for the year 2004”, Market Intelligence Center, Taiwan, http://mic.iii.org.tw/intelligence, Date of access: March, 2005.
Igbaria, M. (1999), “The driving forces in the virtual society”,Communication of the ACM, 42(12), 64-70.
Johnson, C.M. (2001), “A survey of current research on online communities of practice”, Internet and Higher Eduation, 4, 45-60. Koh, J. & Kim, Y.G. (2004), “Knowledge
sharing in virtual communities: an e-business perspective”, Expert Systems with Applications, 26, 155-166.
Kozinets, R. V. (1999),“E-Tribalized marketing?: the strategic implications of virtual communities of consumption”, European Management Journal, 17(3), 256-264.
Lai, V. S., & Mahapatra, R. K., (1997), “Exploring the research in information technology implementation”, Information & Management, 32, 187-201.
Lee, F. S. L., Vogel, D. & Limayem, M. (2003), “Virtual community informatics: a review and research agenda”,Journalof Information Technology Theory And Application, 5(1), 47-61.
Lehmann, D.R., Gupta, S. & Steckel, J.H. (1999), Marketing Research, Addison & Wesley: New York.
Mbaatyo, A. (2001), “Business cohesion: managing schools of business in the age of knowledge”, S. A. M. Advanced Management Journal, 66(2), 29-44.
McGrath, J. E. (1984), “Groups: interaction and performance”, Prentice-Hall, Englewood Cliffs, NJ.
Nunally, J. (1978), Psychometric Theory, McGraw-Hill: New York.
egina, B. S. (2002), “Considering social cohesion in quality of life assessments: concept and measurement”, Social Indicators Research, 58, 403-428.
Ridings, C.M., Gefen, D. & Arinze, B. (2002), “Some antecedents and effects of trust in virtual communities”, Journal of Strategic Information Systems, 11, 271-295.
Rothaermel, F.T. & Sugiyama, S. (2001), “Virtual Internet communities and commercial success: individual and community-level, theory grounded in the atypical case of TimeZone.com”, Journal of Management, 27, 297-312.
Romm, C., Pliskin, N. & Clarke, R. (1997), “Virtualcommunities and society: toward an integrative three-phase model”,International Journal of Information Management, 17(4), 261-270.
Suh, Kil Soo (1999) “Impact of communication medium on task performance and satisfaction: an examination of media-richness theory”, Information & Management, 35(5) 295-312.
Turner, J. W., Grube, J. A. & Meyers, J.
(
2001)
,
“De
ve
l
opi
ng an optimal
match within
online communities: an exploration of CMC support communities and traditional support”, Journal of Communication, 51(2), 231-251.Venkatesh, M. 2003. The community network lifecycle: a framework for research and action, The Information Society, 19:
339-347.
Walther, J. B. (1992), “A longitudinal experiment on relational tone in computer-mediated and face to face interaction, proceeding of the twenty-fifth hawaii international conference on system sciences”, IEEE Computer Society Press, Los Alamitos, Ca, 4, 220-231.
Wellman, B., & Gulia, M. (1999). “Net-surfers don’t ride alone: virtual communities as communities”, In B. Wellman (Ed.), Networks In The Global Village, pp. 331-336. Boulder, CO: Westview Press.
White, N. (2001), “The tools of online connection”,
http://www.fullcirc.com/community/connectt ools.htm, Date of access: March, 2005. Williams, R. L. & Cothrel, J. (2000), “Four
smart way to run online communities”,MIT Sloan Management Review, 41(4), 81-91. Workman, M., Kahnweiler, W. & Bommer, W.
(2003) “The effects of cognitive style and media richness on commitment to telework and virtual teams”, Journal of Vocational Behavior, 63(2), 199-219.
Yates, J. & Orlikowski, W.J. (1992), “Genres of organizational communication: an approach to studying communication and media”, The Academy of Management review, 17(2), 299-326.
Table 1: Potential determinants for cohesion of VIC
Facets Literature Main concerns
Leadership for VIC
Romm, et al. 1997; William & Cothrel, 2000; Sivan, 2000; Johnson, 2001
1. VIC would need a leader to define norms, a code of conduct, and resolve conflicts.
2. VIC would consider the characteristics of leaders that may have impact on users’ willingness to join.
3. VIC would need a variety of leadership quality for the leadership.
Social
relationships for VIC
Hagel & Armstrong, 1997; Romm, et al. 1997; Kozinets, 1999; Igbaria, 1999; Williams & Cothrel, 2000; Ba, 2001; Rothaermel & Sugiyama, 2001; Forrest & Kearns, 2001; Johnson, 2001; Andrews, 2002; Commings et al, 2002; Venkatesh (2003)
1.
VIC would help in developing interpersonal relationships.2.
VIC would help in gathering users having common interests.3.
VIC would change society via recharacterization of personal relationships.Communication Tools for VIC
Fulk et al., 1987; Romm, et al. 1997; Williams & Cothrel, 2000; Johnson, 2001; Andrews, 2002; Bieber et al., 2002; Venkatesh (2003)
1. VIC would need multiple-channel medias to express opinions and to communicate with other members.
2. VIC should meet media richness requirements.
3. VIC would need an attractive theme for participants
Functions for VIC
Kozinets, 1999; Ho. et al, 2000; Rothaermel & Sugiyama, 2001; Johnson, 2001; Andrews, 2002
1. VIC would help in learning something by participating in virtual activities.
2. VIC would need face-to-face contents. 3. VIC would need various entertainments.
Table 2: Factor analysis results Composites Factors Items Loading Eigonvalue
Item to total correlation Cronbach’s α Leadership Leadership (1) (2) (3) (4) (5) 0.80 0.71 0.87 0.79 0.80 3.15 0.6731 0.5709 0.7622 0.6505 0.6600 0.8507 Members interactions (1) (2) (3) 0.71 0.78 0.73 3.16 0.4100 0.5500 0.4200 0.6546 Social relationship Common interests (1) (2) 0.90 0.82 1.26 0.5834 0.5834 0.7369 Comm. Tools Comm. Tools (1) (2) (3) 0.84 0.90 0.85 2.30 0.7120 0.7547 0.6771 0.8463 Entertainments (1) (2) (3) 0.85 0.94 0.89 2.39 0.6806 0.8474 0.7442 0.8725 Learning (1) (2) (3) 0.89 0.88 0.85 2.30 0.7451 0.7215 0.6768 0.8463 Functions Physical activity (1) (2) (3) 0.90 0.92 0.90 2.48 0.7724 0.8237 0.7849 0.8948 Cohesion of VIC Cohesion of VIC (1) (2) (3) (4) (5) 0.70 0.81 0.79 0.81 0.75 3.00 0.550 0.690 0.650 0.660 0.600 0.8300
Cohesion
of VIC
Communication tools for VIC
Functions for VIC
Figure 1: Research model
Leadership for VIC
Social relationships for VIC
H1
H2
H3
H4
Table 3: Results and parameter estimate from Lisrel model Model Estimates
Parameters Estimate p value
Leadership for VIC (ξ1) Leadership of VIC (λx1) 1
-Members’interactions (λx2) 0.32 ***
Social relationships (ξ2)
Common interests (λx3) 0.83
-Comm. tools for VIC (ξ3) Comm. tools for VIC (λx4) 1
-Entertainments for VIC (ξ4) Entertainments for VIC (λx5) 1
-Learning function for VIC (ξ5) Learning function for VIC (λx6) 1
-Physical activity for VIC (ξ6) Physical activity for VIC (λx7) 0.91
-Cohesion of VIC (η) Cohesion of VIC (λy) 1
-Leadership for VIC → cohesion ofVIC(γ1) -0.18 0.03*
Social relationships for VIC → Cohesion of VIC (γ2) 0.28 0.12
Entertainments function for VIC → Cohesion of VIC (γ3) 0.22 ***
Comm. tools for VIC → Cohesion of VIC (γ4) 0.09 0.35
Learning function for VIC → Cohesion of VIC (γ5) 0.21 0.02*
Physical activity for VIC → Cohesion of VIC (γ6) 0.26 *** Members’interactions (λx2) → Physical activity (λx7) 0.33 ***
RMR=0.007<0.05; GFI=0.999>0.9; AGFI=0.987>0.9 - -*: p<0.05, *-*: p<0.01, **-*: p<0.001
Note: The path relation that “Members’interactions (λx2) → Physical activity (λx7)”was advised by the
utilized tool to improve the model fitness.
Cohesion of
VIC (η)
Physical activity for VIC (ξ6) Learning for VIC (ξ5) Entertainments for VIC (ξ4) Comm. tools for VIC (ξ3) Social relationships for VIC (ξ2) Leadership for VIC (ξ1) Members’ interactions (λx2) Common interests (λx3) Entertainments for VIC (λx5) Comm. tools for VIC (λx4) Learning for VIC (λx6) Leadership for VIC (λx1) Physical activity for VIC (λx7)Cohesion of
VIC (
λy)
Figure 2: Path diagram of Lisrel model
-0.18* 0.28 0.22*** 0.09 0.20* 0.26*** 1.00 1.00 0.32*** 0.83 0.91 1.00 1.00 1.00 0.33***
