Antecedents of Virtual Community members' participation: the Triandis Interpersonal behavior Perspective

Transcription

1 Association for Information Systems AIS Electronic Library (AISeL) AMCIS 2006 Proceedings Americas Conference on Information Systems (AMCIS) December 2006 Antecedents of Virtual Community members' participation: the Triandis Interpersonal behavior Perspective Honglei Li The Chinese University of Hong Kong Vincent Lai The Chinese University of Hong Kong Follow this and additional works at: Recommended Citation Li, Honglei and Lai, Vincent, "Antecedents of Virtual Community members' participation: the Triandis Interpersonal behavior Perspective" (2006). AMCIS 2006 Proceedings This material is brought to you by the Americas Conference on Information Systems (AMCIS) at AIS Electronic Library (AISeL). It has been accepted for inclusion in AMCIS 2006 Proceedings by an authorized administrator of AIS Electronic Library (AISeL). For more information, please contact

2 : the Triandis Interpersonal behavior Perspective Honglei Li The Chinese University of Hong Kong Vincent, S. Lai The Chinese University of Hong Kong ABSTRACT Virtual communities have emerged as a new form of business model in the electronic business field. Understanding the motivations behind members participation is essential for virtual community organizers. Using the Triandis model of interpersonal behavior, this paper investigated reasons why members participate in virtual community activities. The results showed that the interpersonal behavior perspective is effective in explaining virtual community members participation. Habit was identified as having a strong effect on members participation. Keywords Virtual communities, Triandis Model, members participation, habit, interpersonal behavior. INTRODUCTION Virtual communities (VCs), taking the form of BBS (Bulletin Board Systems), newsgroups, online forums, MUD (Multiple User Dungeons), OOM (Object Oriented MUD), and others, have become an essential part of business applications. Their function is to support the social interactions of computer users in the Internet-based communication arena. Recently, various VCs have been implemented in the Internet cyberspace for the purposes of market expansion (Kozinets, 1999), customer services provision (Kim, Lee and Hiemstra, 2004), customer loyalty retention (Kim et al., 2004), brand building (McWilliam, 2000) and business transaction support (Hagel and Armstrong, 1997; Rothaermel and Sugiyama, 2001). To date, VCs have not only become an essential channel for enterprises to communicate with their customers, but have also emerged as a new business model to bring in revenue (Hagel and Armstrong, 1997). The increasingly important role VCs are playing in the business field is making their operation an essential part of business processes. According to Hagel and Armstrong (1997), acquisition of VC participants is the prerequisite for shaping and operating a VC. Only after the aggregation of members has reached a critical mass can VCs accumulate invaluable information and sufficient diversity to be able to generate revenue for the VC organizers. Not only that, VC organizers must also have good control and management of member churning defined as members who join a community for a while and never come again to sustain the continued operation of their VCs. A further point is that VCs need a predictable number of members who do not merely visit frequently, but participate actively. Consequently, issues such as how to acquire new members and keep them are becoming critical for VC organizers to manage their VCs successfully. Thus, understanding the theories and mechanisms underlying VC members participation is becoming essential for VC professionals to operate and manage their VCs effectively and efficiently. A review of past literature shows that our understanding of the reasons for VC participation is still in its infancy. The prime research findings have suggested that social-psychological reasons, including social-orientation and relationship-building, are the main causes (Bressler and Grantham, 2000). Several other reasons researchers have explored derive from the gift economy perspective (Kollock, 1999; Rheingold, 2000; Ridings, Gefen and Arinze, 2002; Wang and Fesenmaier, 2004a, 2004b; Wasko and Faraj, 2000, 2005), the social identity theory (Tajfel and Turner, 1986), self-efficacy theory (Bandura, 1986), and the trust perspective (Ridings et al., 2002). But all of the above studies focused chiefly on the members themselves and neglected to consider members participation in the context of community dynamics. When people participate in VCs, they are in fact interacting with each other. Thus, members participation in VCs is in fact a dynamic interpersonal behavior building process. The purpose of this paper is to explore VC members participation from the interpersonal behavior perspective, using the Triandis interpersonal behavior model. Specifically, the paper seeks to identify the factors that influence members behavior 4480

3 in VCs. Two types of behavior behavior to give information and behavior to obtain information were investigated. Data were collected in three large representative Chinese VCs and tested with Lisrel The results show that habit is a strong factor influencing VC members participation. LITERATURE REVIEW In this section, we review the antecedents of VC participation and the limited theoretical and empirical studies in this field. Previous studies have shown that the factors influencing VC members behavior vary with participation activities. In general, the previous literature has identified three types of participation behavior general participation, lurking and active contribution. The General Participation Behavior The general participation behavior, constructed either as the time spent in VCs or the frequency of visiting (Wang and Fesenmaier, 2003, 2004a, 2004b), or as the intention to participate in VCs (Bagozzi and Dholakia, 2002; Teo, Chan, Wei and Zhang, 2003), has been investigated in several studies. Wang and Fesenmaier (2003; 2004a; 2004b) studied VC participation in an online travel community. Their findings suggested that social needs, psychological needs, hedonic needs and membership duration have significant correlations to members participation in the VC. The general participation behavior approach treated all the behavior in VCs as the same and failed to consider the behavior characteristics in VCs. Most importantly, it failed to distinguish the lurking behavior from the active participation behavior, which is very important in VCs. Lurking Behavior A characteristic behavior observed in VCs is lurking, which is the behaviour of only viewing messages but not posting a message. People who lurk are called lurkers. Lurking behavior has been reported in a series of studies (Brazelton and Gorry, 2003; McKee, 2002; Preece, Nonnecke and Andrews, 2004). Though lurking behavior has been widely reported, it has not been investigated extensively. Only two studies have examined, and sought to explain, the motivation behind lurking (Preece et al., 2004; Ridings et al., 2002). Overall, the lurking behavior should be distinguished from the general participation behavior and the posting behavior (active participation). The lurking behavior is also very important to the community development because lurkers will potentially become the contributors who post messages (Brazelton and Gorry, 2003; Wang and Fesenmaier, 2004a, 2004b). However, few studies have paid attention to this behavior. It is worth investigating, and theories are needed to explain this behavior. Active Participation The behavior of posting messages in VCs attracts more interests from VC researchers than lurking behavior and general participation behavior. The reason is that active VC participations are contributive to VC s continued success despite the fact that this behavior is spontaneous, unrewarding and time-consuming. In general, three perspectives have been proposed to explore and explain this behavior. The first perspective is from the gift economy concept and has been studied by several researchers (Kollock, 1999; Rheingold, 2000; Wang and Fesenmaier, 2003; Wasko and Faraj, 2005). For example, Rheingold (2000) and Kollock (1999) stated that, when a member contributes to VCs, he or she actually expects a future reciprocal return from other members. Expectations of reciprocity have been tested by several studies, but no consistent results were reported. Several studies have identified expectations of reciprocity as a significant factor to explain why people post messages in VCs (Wang and Fesenmaier, 2003, 2004b; Wasko and Faraj, 2000, 2005), although in Wasko and Faraj s recent study (2005), it was not found to have any impact on active contribution. The second perspective of active VC contribution can be attributed to and explained by social identity theory (Tajfel and Turner, 1986), self-efficacy theory (Bandura, 1986), and self-presentation theory (Schlenker, 1985), which are all subtheories of self-concept theory. Based on the self-concept theory, individuals can gain satisfaction and build their ideal self through managing their social identity (Tajfel and Turner, 1986), self impressions (Schlenker, 1985), and self-efficacy (Bandura, 1986) in social groups. VCs, as the social groups, can enable members to build their social identity, manage their self-impression, and increase their self-efficacy. The third perspective of active VC participation arises from social-related constructs such as culture (McKee, 2002), trust (Ridings et al., 2002), centrality in the network and self-related expertise (Wasko and Faraj, 2005). 4481

4 Overall, the explorative studies on the antecedents of VC participation are still in the early stages. Most studies have agreed that social-psychological reasons or needs are the main causes of VC participation, but have failed to provide consistent results and theoretical framework. The antecedents of VC participation thus still need theoretical exploration of the mechanism behind members behavior. Empirical studies are also essential to test these theories. CONCEPTUAL MODEL AND RESEARCH HYPOTHESES To address the lack of theories explaining VC members participation, this study proposes that VC members behavior is explained by the Triandis model, which aims to explain people s interpersonal behavior. Figure 1 illustrates our research model and hypotheses. Based on the literature review, we propose that VC members behavior consists of two parts behavior to obtain information (BOI) and behavior to give information (BGI). BOI corresponds to the lurking behavior in the previous literature, and BGI corresponds to active participation behavior. We have not adopted the general participation behavior described in the literature because it cannot differentiate lurking behavior from active participation behavior. These two types of behavior could be explained by an interpersonal behavior model the Triandis model. To illustrate our conceptual model, we first explain why we adopt two types of VC behavior, then give an introduction to the Triandis model, and lastly propose our hypotheses. Two Types of VC Members Participation Behavior Figure 1. Conceptual Model We adopt two constructs behavior to obtain information and behavior to give information to represent the member participation behavior because of the differences between them reported by previous studies. Earlier studies have suggested that BOI & BGI differ in two ways. First, BGI is less frequent than BOI (Christie and Azzam, 2004; McKee, 2002; Preece et al., 2004). Second, the antecedents of BGI and GOI are theoretically different (Ridings et al., 2002; Wang and Fesenmaier, 2004a, 2004b; Wasko and Faraj, 2000, 2005). The Triandis Interpersonal Behavior Model and Research Hypotheses The Triandis interpersonal behavior model was proposed by Triandis in 1977 and later developed in 1980, with the objective of building a synthesized model to include centripetal variables of attitudes, values and other acquired behavior dispositions that could be used to describe different types of interpersonal behavior. The core of the Triandis model is the mechanism of behavior motivations, which can be described by two equations built on intention and habit relationships. P a = (w H H+w I I) I.F I=w S S+w A A+w C C 4482

5 The first equation represents the probability of a behavior P a as a function of the sum of habits (H) plus behavioral intentions (I), multiplied by the organism s physiological arousal (P) and by facilitating conditions (F). The second equation states that the behavioral intention is a function of social factors (S), affect (A) and perceived consequences (C). w represents weight, which ranges from 0 to 1. According to the first equation, behavior is jointly determined by habit and intentional motivations, which supplement each other. When an individual is familiar with certain behavior, his or her ability to perform this behavior will be high and his or her action toward this behavior will become the automatic behavior, which lays the ground for this behavior to become habitual. Accordingly, the intention toward the behavior is a self-conscious and self-instructional command toward the individual. The relationship between habit and behavior is like a zero-sum game. When an individual s automatic behavior is high, i.e., his habit toward this behavior is high, his intention toward this behavior is low. But even in the presence of the habit and intention, if the external environment does not permit such behavior to happen, the individual cannot perform the behavior. The second equation states that behavior intention is jointly determined by affect, perceived consequences and social factors. Affect refers to one s cognitive feeling toward the behavior, social factors refer to the internalization of the individual s subjective culture of the group of people with whom he or she interacts most often, and consequences refers to the values the behavior may bring to the individual after the person performs the behavior. Besides these two core equations, the Triandis model also suggests other relationships among the variables. For example, habit will have a positive influence on affect, implying that previous automatic experience with habit would strengthen one s emotional feeling. The Triandis model has proved useful for explaining attitude-behavior relationships, as have other models like Fishbein (1975). It can be used across disciplines, and so can be regarded as even more useful than other models like Fishbein (Sheth, 1982). The most often used relationship is attitude-behavior relationships. Also, there is a tradeoff between attitude-behavior relationship and habit-behavior relationship. If attitude-behavior relationship is strong, habit-behavior relationship is weak (Verplanken, Aarts and Knippenberg, 1994). We thus adapt the Triandis interpersonal behavior model to explain VC members behavior. The following eight hypotheses are developed based on the adapted Triandis interpersonal behavior model. H1: Habit has a positive effect on VC members behavior to obtain information. H2: Habit has a positive effect on VC members behavior to give information. H3: Behavioral intention has a positive effect on VC members behavior to obtain information. H4: Behavioral intention has a positive effect on VC members behavior to give information. H5: Habit of VC members activities has a positive effect on their affect toward participating VCs. H6: VC members affect has a positive effect on their behavioral intention to participate in VCs. H7: The perceived consequences of VC members have a positive effect on their behavioral intention to participate in VCs. H8: The social factors of VC members have a positive effect on their behavioral intention to participate in VCs. RESEARCH METHODOLOGY In this study, an online cross-sectional survey in online forums was adopted as the research method to investigate the behavior of VC members. A total of three VC companies participated in our study. These three companies were Tencent Community an adjunctive community of the largest Instant Messenger software (QQ) in China; Xilu Community an integrative commercial community; and Microsoft Chinese Community a value-added professional community of a large software company. These three communities were selected because they are of different natures and represent three of the largest VCs in China. Data Collection An online questionnaire hosted in a service provider s site ( was developed to collect data from members of the three participating VCs in China. The use of a service provider also allowed us to deal with the problems of access control, authentication and multiple responses associated with the web-based data collection approach (Stanton and Rogelberg, 2001). The success of data collection in this study rested on the management support of the three participating VCs. Therefore, the management of all participating VCs was contacted to obtain their support for our survey. 4483

6 Variable Operationalization The dependent variables of this study are behavior to give information and behavior to obtain information. Behavior to give information and behavior to obtain information were operationalized with measures developed from the actual usage behavior in information systems (Davis, 1989; Limayem and Hirt, 2003; Ridings et al., 2002; Straub, Limayem and Karahanna-Evaristo, 1995; Wang and Fesenmaier, 2004b). Most of these measures, which were derived from information technology adoption studies (such as (Davis, 1989; Limayem and Hirt, 2003) and (Straub et al., 1995)), were based on the users time and frequency spent using IT. The variables in the Triandis model were operationalized according to the initial suggestion of Triandis and previous studies. We adapted the habit construct that Limayem and Hirt (2003) developed for VC participation. The measure of behavior intention was adapted from Davis (1989), Davis et al. (1989), Moon and Kim (2001), and Teo et al. (1999). The measure of affect was adapted from Triandis (1980), who suggested the use of four pairs of semantic differential items pleasantunpleasant, enjoyable-disgusting, exciting-depressing, and joyful-hateful. Perceived consequence was measured by the construct developed by Limayem and Hirt (2003) in their study of WebBoard for supporting Internet communications. Social factors were measured using two approaches. First, we adapted some of the items from previous studies (such as (Baumann, Brown, Fontana and Cameron, 1993; Limayem and Hirt, 2003)). Second, we used open questions to solicit additional information on the social factor. This method has been used by Limayem and Hirt (2003) to develop environment-specific factors with great success. Instrument Validation To ensure the validity and reliability of the questionnaire, a four-stage survey validation was conducted. First, whenever possible, previously validated questions were used, and generally accepted online instrument construction guidelines (Ridings et al., 2002; Stanton and Rogelberg, 2001; Wang and Fesenmaier, 2003) were observed as much as possible. Second, the questionnaire was back-translated to ensure the validity of the language usage. Third, the questionnaire was pretested by one MIS professor, seven business doctoral students, and two experienced VC webmasters. Forth, a pilot test for the questionnaire was conducted on two small VCs. Data Analysis SPSS 10.0 and Lisrel 8.5 were used to analyze all data collected. Confirmatory Factor Analysis was used to check the fitness of data for our research model. Finally, Structural Equation Modeling (SEM) was conducted to validate and test the research models proposed in this study. DATA ANALYSIS RESULT In this research, a total of 1,406 responses were collected from the three participating communities. After checking for data integrity, we eliminated seventy-six responses suffering from multiple responding problems and eighteen from ineffective response problems, thus giving us a total of 1,312 effective responses. Profile of Respondents Of the 1,312 valid responses, 98% were from Mainland China, Hong Kong, Macau or Taiwan. Most respondents were male (75%) and single (78.7%). Their occupations varied from unemployed to professionals, with most of them being engineers/computer technicians and students. Regarding their ages, the respondents were predominately (67.9%) in the range As for their education level, more than half of them (72.3%) were college graduates. The profile of our respondents is very compatible to the VC profiles announced in the official site of CNNIC 1 (China internet Network Information Center), thus suggesting that our respondents were representative of VC users as a whole. Descriptive Statistics and Reliability The descriptive statistics and Cronbach s Alpha were reported in table 1. As indicated in Table 1, the Cronbach s alpha values of all our variables are above 0.80, significantly above the 0.7 level suggested for exploratory research, thus supporting the reliability of our measurements for model testing. The correlation matrix generated by Lisrel 8.54 is shown in Table

7 Construct Mean S.D. Cronbach s Alpha Behavior to Obtain Information (BOI) BOI BOI Behavior to Give Information (BGI) BGI BGI BGI BGI Habit HAT HAT HAT HAT HAT Behavioral Intention (BI) BI BI BI Affect AFF AFF AFF AFF Perceived Consequences (PC) PC PC PC PC PC PC Social Factors (SF) SF SF SF SF Table 1. Descriptive statistics and Cronbach s Alpha 4485

8 BOI BGI BI Affect Habit PC SF BOI 1.00 BGI BI Affect Habit PC SF Table 2. Correlation Matrix Analysis of Measurement and Structural Model Lisrel 8.54 was conducted to assess the measurement and structural model of Triandis for its fitness in our study. The values of our measurement and structural model, along with the recommended values, are reported in Table 3. Overall, our measurement model and structural model suggest an acceptable level of fit. Community χ 2 df TLI CFI GFI AGFI RMSEA Recommended Value Measurement Model Structural Model Table 3. Results of measurement and structural model The overall explanatory power of the Triandis model was examined using the R2 and the individual path coefficients. Dependent Variables Effect Sizes Behavior to Obtain Information 57% Behavior to Give Information 44% Behavioral Intention 52% Affect 62% Table 4. R2 of independent constructs Figure 2 shows the path coefficients with their respective significance. As shown in Figure 2, almost all the path coefficients of the model (except BI BOI and BI BGI) were found to be significant. Six hypotheses were supported for our models. The significance of the individual paths is reported in Table 5. The effect of habit on behavior to obtain information (H1), behavior to give information (H2), and affect (H5) were all supported. The behavioral intention s effect on behavior to obtain information (H3) and behavior to give information (H4) were not supported. The effect of social factors (H8), perceived consequences (H7) and affect (H6) on behavioral intention were all supported. 4486

9 Figure 2. Result of Triandis Model Testing Hypothesis Path coefficients Effect Sizes H1 Habit BOI 0.74*** H2 Habit BGI 0.64*** H5 Habit Affect 0.79*** H3 BI BOI 0.02 H4 BI BGI 0.04 H6 Affect BI 0.79*** H7 PC BI 0.23*** H8 SF BI 0.10*** Table 5. Significance and strength of individual paths ***: Significant at level DISCUSSIONS, IMPLICATIONS AND LIMITATIONS Discussions We should first notice that six out of eight relationships were significant in Triandis model, which supports our hypothesis that the Triandis model is theoretically applicable to the explanation of VC members behavior. Overall, 57% variance and 44% variance of the two VC members behaviors, behavior to obtain information and behavior to give information, are explained by Triandis model. As hypothesized, members behavior is influenced most by habit, with almost all the variance explained by habit. Contrary to our hypothesis 3 and 4, intention did not have a significant effect on VC members behavior; similar findings have occasionally been reported in past literature. This result can be explained from the original rationale of the Triandis model. Among the two antecedents of behavior, intention and habit were proposed by Triandis (1980) as two supportive and alternative constructs, i.e., when habit was weak, intentions are strong, and when intentions are weak, habits are strong. Sheth (1982) has also discussed this trade-off relationship in comments on the choice of models in Marketing, and regarded this relationship as one of the most attractive points of the Triandis model. This conclusion is also consistent with a recent study conducted by Kim et al. (2005), who reported that the intention-usage relationship for the heavier users of a website is weaker than that for lighter users. 4487

10 As for the two types of VC behavior, the effect of habit on behavior to obtain information is stronger than its effect on behavior to give information, which is consistent with our initial model conceptualization. This result is also consistent with previous studies showing that the two types of VC behavior are different (Christie and Azzam, 2004; McKee, 2002; Preece et al., 2004; Ridings et al., 2002; Wasko and Faraj, 2000, 2005). Lastly, the result also suggests that habit has a strong effect on affect and perceived consequences, social factors and affects are also antecedents of behavioral intention, which is consistent with Triandis initial proposition (1980). Implications The implication for researchers is that the Triandis interpersonal behavior model is effective in explaining VC members behavior. The successful result also suggests that the interpersonal behavior perspective is effective for exploring members reasons to participate in VCs. Future research could use other interpersonal behavior theories to explore this further. The major implication for practitioners lies in the strong effect of habit and the zero-sum relationship of habit and intention s effect on VC members behavior. The non-significance of intention s effect implied that members intentions can easily disappear and be converted into habit. Because habit has a very strong effect in explaining VC members participation, VC organizers should organize activities that attract members initial intentions and transform these intentions into habit. Limitations The major limitation of our study is the sample selection. Our samples are self-selected, which means that only members who are interested by the content of the study will be attracted; this may weaken the randomization of the sample. Thus, this study was not able to address the question of whether respondents and non-respondents in the community differ in important characteristics. The second major limitation of our study is that we used a cross-sectional survey to collect data, which cannot ensure the consistency of the result across time. REFERENCES 1. Bagozzi, R. P. and Dholakia, U. M. (2002) Intentional social action in virtual communities, Journal of Interactive Marketing, 16, 2, Bandura, A. (1986) Social foundations of thought and action: a social cognitive theory, Prentice-Hall, Englewood Cliffs, N.J. 3. Baumann, L. J., Brown, R. L., Fontana, S. A. and Cameron, L. (1993) Testing a Model of Mammography Intention, Journal of Applied Social Psychology, 23, 21, Brazelton, J. and Gorry, A. G. (2003) Creating a knowledge-sharing community: If you build it will they come?, Association for Computing Machinery. Communications of the ACM, 46, 2, Bressler, S. E. and Grantham, C. E. (2000) Communities of commerce: building internet business communities to accelerate growth, minimize risk, and increase customer loyalty, McGraw Hill, New York. 6. Christie, C. A. and Azzam, T. (2004) What's All the Talk About? Examining EVALTALK, an Evaluation Listserv, The American Journal of Evaluation, 25, 2, Davis, F. D. (1989) Perceived Usefulness, Perceived Ease Of Use, And User Acceptance of Information Technology, MIS Quarterly, 13, 3, Davis, F. D., Bagozzi, R. P. and Warshaw, P. R. (1989) User Acceptance Of Computer Technology: A Comparison Of Two, Management Science, 35, 8, Fishbein, M. and Ajzen, I. (1975) Belief, attitude, intention, and behavior: an introduction to theory and research, Addison-Wesley Pub. Co., Reading, Mass. 10. Hagel, J. and Armstrong, A. (1997) Net gain: expanding markets through virtual communities, Harvard Business School Press, Boston. 11. Kim, S. S., Malhotra, N. K. and Narasimhan, S. (2005) Two Competing Perspectives on Automatic Use: A Theoretical and Empirical Comparison, Information Systems Research, 16, 4, Kim, W. G., Lee, C. and Hiemstra, S. J. (2004) Effects of an online virtual community on customer loyalty and travel product purchases, Tourism Management, 25, 3, Kollock, P. (1999) The Economics of Online Cooperation: Gifts and Public Goods in Cyberspace, in M.A. Smith and P. Kollock (Eds) Communities in cyberspace, Routledge, London; New York, x, 323 p. 14. Kozinets, R. V. (1999) E-tribalized marketing?: the strategic implications of virtual communities of consumption, European Management Journal, 17, 3,

11 15. Limayem, M. and Hirt, S. G. (2003) Force of Habit and Information Systems Usage: Theory and Initial Validation, Journal of the Association for Information Systems, 4, Mckee, H. (2002) "YOUR VIEWS SHOWED TRUE IGNORANCE!!!": (Mis)Communication in an online interracial discussion forum, Computers and Composition, 19, 4, Mcwilliam, G. (2000) Building Stronger Brands through Online Communities, Sloan management review, 41, 3, Moon, J.-W. and Kim, Y.-G. (2001) Extending the TAM for a World-Wide-Web context, Information & Management, 38, 4, Preece, J., Nonnecke, B. and Andrews, D. (2004) The top five reasons for lurking: improving community experiences for everyone, Computers in Human Behavior, 20, 2, Rheingold, H. (2000) The virtual community: homesteading on the electronic frontier, MIT Press, Cambridge, Mass. 21. Ridings, C. M., Gefen, D. and Arinze, B. (2002) Some antecedents and effects of trust in virtual communities, The Journal of Strategic Information Systems, 11, 3-4, Rothaermel, F. T. and 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, 3, Schlenker, B. R. (1985) Identity and self-identification, in B.R. Schlenker (Eds) The Self and social life, McGraw- Hill, New York, Sheth, J. N. (1982) Some comments on Triandis the model of choice behavior in marketing, in M. Leigh (Eds) Choice models for buyer behavior, Research in Marketing, Supplement 1, JAI Press, Greenwich, Conn., Stanton, J. M. and Rogelberg, S. G. (2001) Using Internet/Intranet Web pages to collect organizational research data, Organizational Research Methods, 4, 3, Straub, D., Limayem, M. and Karahanna-Evaristo, E. (1995) Measuring system usage: Implications for IS theory testing, Management Science, 41, 8, Tajfel, H. and Turner, J. C. (1986) The social identity theory of intergroup behavior, in W.G. Austin and S. Worchel (Eds) Psychology of intergroup relations, Nelson-Hall Publishers, Chicago, Teo, H.-H., Chan, H.-C., Wei, K.-K. and Zhang, Z. (2003) Evaluating information accessibility and community adaptivity features for sustaining virtual learning communities, International Journal of Human-Computer Studies, 59, 5, Teo, T. S. H., Lim, V. K. G. and Lai, R. Y. C. (1999) Intrinsic and extrinsic motivation in Internet usage, Omega, 27, 1, Triandis, H. C. (1977) Interpersonal behavior, Brooks/Cole Pub. Co., Monterey, Calif. 31. Triandis, H. C. (1980) Values, attitudes, and Interpersonal Behavior, in University of Nebraska (Lincoln campus). Dept. of Psychology. (Eds) Nebraska Symposium on Motivation, University of Nebraska Press., [Lincoln, Neb.], Verplanken, B., Aarts, H. and Knippenberg, A. V. (1994) Attitude versus General Habit: Antecedents of Travel Model Choice, Journal of Applied Social Psychology 24, 4, Wang, Y. and Fesenmaier, D. R. (2003) Assessing Motivation of Contribution in Online Communities: An Empirical Investigation of an Online Travel Community, Electronic Markets 13, 1, Wang, Y. and Fesenmaier, D. R. (2004a) Modeling Participation in an Online Travel Community, Journal of Travel Research, 42, 3, Wang, Y. and Fesenmaier, D. R. (2004b) Towards understanding members' general participation in and active contribution to an online travel community, Tourism Management, 25, 6, Wasko, M. M. and Faraj, S. (2000) "It is what one does": why people participate and help others in electronic communities of practice, The Journal of Strategic Information Systems, 9, 2-3, Wasko, M. M. and Faraj, S. (2005) Why should I Share? Examining Social Capital and Knowledge Contribution in Electronic Networks of Practice1, MIS Quarterly, 29, 1,