How to intensify the individual’s feelings of belonging to a social networking site? Contributions from community drivers and post-adoption behaviours

A final version of this paper has been published as: Sánchez‐Franco, M.J., Buitrago‐Esquinas, E.M. and Yñiguez, R. (2012), "How to intensify the individual's feelings of belonging to a social networking site? Contributions from community drivers and post‐adoption behaviours", MANAGEMENT DECISION, Vol. 50 No. 6, pp. 1137-1154. https://doi.org/10.1108/00251741211238373 How to intensify the individual’s feelings of belonging to a social networking site? Contributions from community drivers and post-adoption behaviours Author Details Author 1 Name: MANUEL J. SÁNCHEZ-FRANCO Department: BUSINESS ADMINISTRATION AND MARKETING University/Institution: UNIVERSITY OF SEVILLE Town/City: SEVILLE Country: SPAIN Author 2 Name: EVA MARÍA BUITRAGO-ESQUINAS Department: APPLIED ECONOMICS University/Institution: UNIVERSITY OF SEVILLE Town/City: SEVILLE Country: SPAIN Author 3 Name: ROCÍO YÑIGUEZ-OVANDO Department: ECONOMIC ANALYSIS AND POLITICAL ECONOMY University/Institution: UNIVERSITY OF SEVILLE Town/City: SEVILLE Country: SPAIN Corresponding author: MANUEL J. SÁNCHEZ-FRANCO Corresponding Author’s Email: [email protected] Abstract Purpose. Our research tests an interaction effects model including perceived community support, community satisfaction, and post-adoption behaviours by analysing a social networking site called Tuenti. Inclusion of routinisation as a moderator allows, in particular, the overcoming of inconsistencies of relationships among key drivers found in previous studies. Design/methodology/approach. The research uses Partial Least Squares to estimate the parameters of the interaction effects model. Findings. Overall, community satisfaction leads Tuenti members to develop community participation and feelings of belonging, and identification with other members. In particular, routinised behaviours predispose members to a higher influence of community satisfaction on community integration –reducing the main effect of perceived support. The higher community satisfaction also results in a lower influence of perceived support on active participation. Research limitations/implications. The model does not include all the relevant variables. Additional studies are required to validate these results. Likewise, the sample is not drawn randomly and thus is not necessarily representative of the population. Practical implications. This study serves to enhance the long-term viability of social networking sites. Social networking sites, on the one hand, need to focus on maintaining high levels of perceived support among less routinised members by means of, for instance, affective resources available to the individual, clear instructions on how to browse through it, or manager moderation and mentoring, among others. On the other hand, social networking sites might enhance overall assessment of psychosocial aspects of relationships among routinised members by, for instance, locating specific threads of interest (related to different topics of conversation), hedonic cues for enjoyment and entertainment purposes, and the use of experts in a particular area. Originality/value. Though previous research provides evidence of the possible effects of perceived community support and satisfaction on social behaviours, a review of the literature reveals that there are still very few published studies that analyse the interaction effects of community satisfaction and routinisation and consequently, improve the explanatory power of the theoretical framework. Keywords: Social Networking Sites, Perceived Community Support, Community Satisfaction, Routinisation, Interaction Effects Article Classification: Research paper 1. Introduction Social networking sites (SNSs) can be regarded as “web-based services that allow individuals to (1) construct a public or semi-public profile within a bounded system, (2) articulate a list of other users with whom they share a connection, and (3) view and traverse their list of connections and those made by others within the system” (Boyd and Ellison, 2007). For the purpose of this study, our research also defines an SNS as a web-based service that allows people to create/share value and content with expansive groups of friends, thereby fostering mutual support and satisfactory participation in community activities as well as an individual’s feelings of attachment to the same and, consequently, expectations of continuity (hereinafter, Perceived Community Support, PCS; cf. Herrero and Gracia, 2007; Stefanone et al., 2011; see also Sense of Virtual Community, SVC, Blanchard, 2007, 2008; McMillan and Chavis, 1986). SNSs emphasise both mutually beneficial relationships and feelings of belonging, and identification with the other members. In particular, Vasalou et al. (2010, p.725) conclude that the ultimate goal of SNS designers is “to foster a sense of true commitment in their users”. In this regard, in 1994 Grönroos (p.9) had already proposed that a relationship can be divided into two parts: “to attract the customer and to build the relationship with that customer so that the economic goals of that relationship are achieved”. Nevertheless, despite dramatic expansion to millions of people worldwide connected to SNSs (e.g., 960 million active visitors in Facebook; cf. Doubleclick Adplanner, December 1, 2011) and the growing importance of affective member-member interactions that affect the scope and success of community strategies, research on identifying the psychological (and relational) process of using SNSs is still in the early stages (cf. Kwon and Wen, 2010). As Preece et al. (2004, p.221) conclude, “many lurkers are not selfish free-riders; there are a host of other reasons why lurkers lurk”. Research on the one hand, omits essential questions as to why members perceive community support provided by one SNS over a competing SNS. On the other hand, assuming that the success of an SNS derives from the development and sustainability of its members, post-adoption analysis has, compared to the established research stream of SNS adoption and initial usage, received scant attention (SánchezFranco and Roldán, 2010). Growing research is thus essential to identify significant relationships between PCS and relationship quality. In particular, as Lin (2008) suggests, both community satisfaction and a sense of belonging and involving others in mutually beneficial activities are determinants of member loyalty in a community. Nevertheless, while not minimising the importance of other relational drivers (e.g., affective trust), the scope of this paper is limited to PCS and community satisfaction -conceptualised herein as a positive affective state resulting from the appraisal of all aspects of the community services and interaction. Mittal et al. (2005) find that efficient firms with satisfied customers outperform other firms (cf. also Rust et al., 2002). Community satisfaction is indeed considered a necessary condition for an SNS (not always profitable; cf. Bontis et al., 2007) to operate successfully and expand (Wu et al., 2010). Satisfaction has been widely explored in management literature. However, to date, the association between PCS and community satisfaction has not been explored to any great extent. There are indeed few studies that use community satisfaction as an indicator for high PCS in SNS realms. Moreover, the intangibility of SNSs implies that the development of an enduring (or habitual) relationship approach is a strategic imperative, if SNSs (as communities of relations) are to achieve long-term success. Our research analyses, in particular, the association between community satisfaction and PCS by studying the interaction effects of routinised behaviours (related to integration of the SNS into daily activities). Although research calls for the need to gather data on potential moderators (cf. Sun and Zhang, 2006), recent studies that model PCS hardly ever include interaction effects. In this regard, inclusion of routinisation as a moderator could increase explanatory power of the relational model of PCS-satisfaction framework, and would thus contribute to this area of study. In short, a number of questions regarding the impact of SNSs remain unanswered. The main objective of this research is to empirically test the interaction effects model including PCS, community satisfaction, and post-adoption behaviours (i.e., routinisation). This paper addresses two main issues. How to maintain and intensify the individual’s feelings of satisfactory attachment to an SNS and posts continues to be a problem. Furthermore, does habitual usage strengthen/weaken the relationships between community satisfaction, exchange of mutual support and feelings of belongingness to an SNS? These questions are analysed in an intensive study of an SNS called Tuenti. The methodology section provides details of the sampling process, measures, analysis methods, and results. Finally, the discussion section outlines the managerial implications and limitations. 2. Theoretical framework and hypotheses Our research presents a PCS structure with three dimensions (cf. Herrero and Gracia, 2007). Perceived support is, on the one hand, defined as feelings of being supported by an online community (Blanchard and Markus, 2004). SNSs would be a source of satisfaction for individuals and a place where they can easily forget daily problems –by providing ties that constitute social resources available to them and satisfying affective needs as a valuable way of getting to know an SNS (e.g., tools for reading, finding or browsing community information; cf. Preece et al., 2004). Community participation is, on the other hand, conceptualised as active involvement in SNS activities or social participation in order to assist other members (Sánchez-Franco and Roldán, 2010); for instance, “accomplishing specific tasks”, “deepening salient aspects of one’s self through social interactions”, “maintaining interpersonal connectivity”, and “playing or otherwise interacting with others” (cf. Dholakia et al., 2004, p.244). If there is little or no message posting in an SNS, passive participation (or lurking) could be a problem. Community integration is finally defined as members’ feelings of membership, identity, and belonging to a group (Blanchard, 2007). Community integration (or true/affective commitment) is mentioned as (a) a requirement for achieving community goals, and (b) a responsible sense of belonging that mobilises one’s social capital effectively (Wellman et al., 2001), that is to say, an attitudinal factor emphasised when members acknowledge the value of continuing relationships between their community and themselves (Hur et al., 2011). In this community context, greater levels of active participation in an SNS will help members to share knowledge and ideas related to mutual interests and, subsequently, foster their individual feelings of belonging to a group (Koh and Kim, 2004). Through active participation members will become embedded in a virtual community (cf. Kozinets, 1999) and this often leads to strong community membership. In addition, members will be more highly motivated to contribute to the community if they receive useful (and emotional) assistance in return (i.e., perceived support). The amount and depth of support that members receive from other members will be positively correlated with community participation (cf. Turner et al., 2001). Likewise, greater levels of perceived support will lead members to feel that they are being supported by a whole portion of their SNS, reducing the uncertainty of their relationship with it, increasing their integration, and therefore, diminishing its negative side (i.e., social isolation) (cf. Casaló et al., 2007). Perceived support will reduce uncertainty and reinforce social relationships with other members at an affective level. This research thus proposes the following hypotheses: H1. Community participation positively influences integration; H2. Perceived support positively influences participation; and H3. Perceived support positively influences integration. See Figure 1. Moreover, the growing use of SNSs raises the question of what encourages members to participate enduringly and maintain accumulated social capital –thus creating a sense of belonging for people who are connected (cf. Putnam, 2000). In particular, community satisfaction (or satisfaction with member-member interactions) is essential for the development of mutually beneficial relationships. In fact, community satisfaction is (a) the result of an overall assessment of psychosocial aspects of accumulative/enduring relationships among members (Casaló et al., 2010) and (b) more concerned with maintaining an ongoing system of exchange among members. That is to say, community satisfaction is not confined to one recent exchange, but relates to the evaluation of exchanges over time (cf. Anderson et al., 1994), which is especially relevant in the industries of continuously provided services, such as SNSs. Figure 1. Theoretical model Accordingly, the consideration of SNSs as social utilities that connect users emotionally to their offline social networks, creating stronger feelings of being supported by the other members will, on the one hand, be associated with higher community satisfaction. Likewise, if members are not satisfied, there will not be any incentive to accomplish specific tasks to foster interpersonal connectivity in an SNS (e.g., by creating value and content, involving others, and staying active and loyal). Community satisfaction will finally lead to desirable outcomes such as cooperation, long-term orientation, and consequently, community integration (cf. Ganesan, 1994; Lam et al., 2004). The positive experiences from community interactions can strengthen the members’ relationship with the SNS. To sum up, SNSs that elicit satisfaction are likely to be those where members want to participate and maintain a relationship in the future. Secondly, community satisfaction precedes community engagement. Thirdly, when members are satisfied with their SNS, they are likely to participate actively in community events. This research therefore proposes the following hypotheses: H4. Perceived support positively influences satisfaction; H5. Community satisfaction positively influences participation; and H6. Community satisfaction positively influences integration. See Figure 1. Despite the influence of perceived support on active participation (cf. H2), when members progressively accumulate satisfactory experiences with the SNS, the leading favourable effects of perceived support on members’ participation will decrease. Satisfied members are more likely to become accustomed to the SNS norms, language, and the identities of other members. Compared to users who have accumulated less service encounters, more (satisfied) experienced users will reduce the uncertainty of their relationship with the SNS and consider the perceived support and community resources available to them as less essential. The balance between these community drivers (related to active participation) will be tilted here towards community satisfaction. Hence, it will not be the degree of support members perceive from other members, but rather the overall evaluations based on the cumulative experience with an SNS over time that will be the keyperformance indicator of active involvement in SNS activities. Therefore, H7. Overall, community satisfaction weakens the relationship between perceived support and community participation. According to H7, following their initial interactions, community members may develop into long-time experienced members (Langerak et al., 2003). As Sánchez-Franco and Roldán (2010) suggest, the success of an SNS lies particularly in the development of a habitual source of mutual support and a sense of true belonging. Our research, therefore, manifests the appropriateness of including routinisation as a moderating driver. Routinisation is associated with habitual usage, that is to say, integration of the technology into daily routines (Cooper and Zmud, 1990; Saga and Zmud, 1994; Schwarz and Chin, 2007), and provides members with a different PCS-satisfaction framework of reference for community evaluations. Despite its importance, routinisation continues to be an understudied area in the SNS domain. Though previous research provides evidence of the possible effects of PCS and satisfaction on social behaviours, we are unaware of any research that empirically examines the relationships between PCS and satisfaction being moderated by post-adoption behaviours. Routinisation occurs when use is no longer novel and becomes part and parcel of an individual’s daily life. It does not necessarily mean that a member uses the full degree of potential offered by an SNS (cf. infusion; Sundaram et al., 2007). Users could, on the one hand, make considerable use of an SNS without taking part fully in its activities. On the other hand, users could position themselves in the true commitment phase, for example, creating value and content, remaining active and loyal, and involving others (Vasalou et al., 2010). Routinised members (1) incorporate the SNS into their daily social-pattern, (2) simplify relationships with others by generating spontaneous usage, (3) make better decisions (on average) because of their efficient behaviours, and consequently, (4) adopt higher targeted behaviours -underestimating members’ feelings of being assisted by others in terms of resources available (i.e., perceived support). Contrariwise, less routinised members opt for general exploration and tend to engage in community participation in a limited way, preferring to be readers rather than writers, and being reticent -or very seldom posting. This initial lack of social ties among members prevents novice members from actively participating in social interactions within an SNS. According to the behaviour chain model, “users are initially guided to enter a phase of discovery, for instance, during which they learn about the site” (Vasalou et al., 2010, p.720; cf. also Fogg and Eckles, 2007; Fogg and Iizawa, 2008). Integration of the SNS into daily routines will, therefore, increase the possibility of familiarity and expertise. Community integration will be influenced mainly by the routinised members’ overall evaluation based on their experiences rather than by mere expectations of what members receive from others. As mentioned earlier herein, when an individual is familiar with an SNS, “he/she knows the specific characteristics of the virtual community, the particular language used in the community and the tools employed to interact and share files” (Casaló et al. 2010, p.30; cf. also Kozinets, 2002). As Dholakia et al. (2004, p.246) suggest, “once the member has learnt and accepted the virtual community’s norms, he or she will identify with the community more” –thus social resources available to him/her to cheer them up and get them into a better mood are not of vital importance. Firstly, when members infer from their feelings that they are being supported by the online community and that it is safe, the sense of receiving assistance from their SNS will become less relevant. Secondly, efficient (or satisfactory) access to community features will, therefore, reinforce the tendency to go back to a preferred SNS and attachment to it. Overall satisfaction with daily/habitual interactions could be one type of switching cost (a) fostering feelings of membership, identity, and belonging to an SNS and (b) becoming essential if a routinised member questions the relationship (SánchezFranco, 2009). This research proposes the following hypotheses: H8. Overall, routinised behaviour weakens the relationship between perceived support and integration; and H9. Overall, routinised behaviour strengthens the relationship between community satisfaction and integration. See Figure 1. 3. Method 3.1. Participants Although increasingly younger adolescents are joining SNSs, little is known about SNS use among very early teens (Espinoza and Juvonen, 2011). The research model is thus analysed in an intensive study of an SNS called Tuenti - the most popular computermediated SNS among the Spanish college student population, and characterised as a community of relations (cf. Carballar, 2011; Hagel and Armstrong, 1997). Tuenti is also purposefully selected because of its high penetration in Spain. Recent statistics show that 9.9 million active visitors have joined the site, 99% of which are located in Spain (cf. Doubleclick Adplanner 2011). Likewise, affective and social motivations (for social searching - seeking out offline contacts, or for social browsing -seeking out new contacts) are more relevant in explaining the usage of Tuenti in lieu of more professional SNSs (e.g., LinkedIn, among others; cf. Carballar, 2011). The theoretical model and the hypotheses discussed above were validated through nonprobabilistic sampling. A survey technique was used to gather data from a sample of undergraduate students (University of Seville, Spain). Respondents consisted of 119 males (43%) and 159 females (57%). The average age was 21.04 (SD: 2.403). Most of the respondents had used online services heavily: 88.2% of the respondents had used online services (e.g., instant messaging, playing games with others, and participating on SNS) for more than 3 years. Hence, following Kwon and Wen’s (2010) research, the responses were deemed to be qualified to allow analysis of the factors affecting the actual use of SNSs. Likewise, following the suggestion by Chin and Newsted (1999), a sample size of 150-200 is needed to attain reliable coefficient values when using Partial Least Squares (PLS) analysis (cf. 3.2 Data analysis). 3.2. Measures Fourteen items were used to assess PCS -taken from Herrero and Gracia (2007), Geyskens et al. (1996) and Sánchez-Franco (2009). On the other hand, a total of three items were employed to measure community satisfaction (cf. Flavián et al., 2006; Janda et al., 2002; Severt, 2002; Smith and Barclay, 1997). The measures of PCS and community satisfaction used in the survey are provided in Table 1. The instrument for measuring the degree of routinised behaviour is operationalised by Sundaram et al. (2007); e.g. My use of Tuenti has been incorporated into my regular life schedule; My use of Tuenti is pretty much integrated as part of my normal life routine; My use of Tuenti is a normal part of my life. All scale items for assessing the construct were obtained through a seven-point Likert scale for respondents to indicate the extent to which they disagreed or agreed with the items. To ensure face validity and content validity of the questionnaire, the draft version was reviewed by experts (i.e. four e-business professors). The e-business professors assessed the suitability of the wording and format, and the extent to which measures represented all facets of constructs. 3.3. Data analysis The hypothesis testing was done using a variance-based PLS approach that places minimal restrictions on sample size and residual distribution. Furthermore, taking into account that hypotheses 7, 8 and 9 are based on interaction effects, a well-known technique had to be applied to test these moderated relationships: product-indicator approach. While other methods of structural equation modelling (SEM) – such as the covariance-based LISREL – are indeed more widespread, PLS was finally chosen because there are a number of advantages to it in terms of estimation of interaction effects (Chin et al., 2003; Gefen et al., 2000; Henseler and Chin, 2010, and Henseler and Fassot, 2010, for a detailed analysis). 4. Results 4.1. Measurement model On the one hand, individual reflective-item reliability was assessed by examining the standardised loadings of the items with their respective construct. Individual reflective-item reliabilities are over the recommended acceptable cut-off level of 0.7 (Carmines and Zeller, 1979). Moreover, we checked the significance of the loadings with a bootstrap procedure (500 sub-samples) to obtain t-statistic values. They are all significant. See Table 1. TABLE 1. INDIVIDUAL ITEM RELIABILITY-INDIVIDUAL ITEM LOADINGS Latent Dimension Loadingsa Perceived support (PS) PS1. I could find people that helped me feel better 0.812 PS2. I could find someone to listen to me when I felt down 0.863 PS3. I could find a source of satisfaction for myself 0.871 PS4. I was able to cheer up and get into a better mood 0.873 PS5. I could relax and easily forget my problems 0.791 Community integration (CI) CI1. My affective bonds with my Tuenti community are the main reason why I continue to use its service 0.886 CI2. I enjoy being a member of my Tuenti community 0.918 CI3. I have strong feelings for my Tuenti community 0.827 CI4. In general, I relate very well to the members of my Tuenti community 0.831 Community participation (CP) CP1. I participate in order to stimulate my Tuenti community 0.879 CP2. I take part actively in activities in my Tuenti community 0.783 CP3. I take part in social groups in my Tuenti community 0.767 CP4. I respond to calls to support my Tuenti community 0.779 CP5. I take part actively in socio-recreational activities in my Tuenti community 0.883 Community Satisfaction (CS) CS1. In general terms, I am satisfied with my experience in my Tuenti community 0.907 CS2, I think that I made the correct decision to use my Tuenti community 0.917 CS3. I have obtained several benefits derived from my participation in my Tuenti community 0.924 a All loadings are significant at p<0.001- (based on t(499), two-tailed test) In addition, the composite reliability (ρc) assesses construct reliability (Werts et al., 1974). Nunnally (1978) suggests 0.7 as a benchmark for a modest reliability applicable in initial stages of research. The composite reliabilities for the multiple reflective indicators are well over the recommended acceptable 0.7 level, thus demonstrating high internal consistency (cf. also Fornell and Larcker, 1981). See Figure 2. Finally, convergent and discriminant validities were assessed by stipulating that the square root of the average variance extracted (AVE) by a construct from its indicators should be at least 0.7 (i.e., AVE > 0.5) and should be greater than that same construct’s correlation with other constructs () (Barclay et al., 1995; Chin, 1998; Fornell and Larcker, 1981). All latent constructs satisfy these conditions. See Figure 2. 4.2. Structural model The bootstrap re-sampling procedure (500 sub-samples) is used to generate the standard errors and the t-values. Firstly, the research model appears to have an appropriate predictive power for endogenous constructs to exceed the required amount of 0.10 –R2 values (Falk and Miller, 1992). Secondly, the Q2 test is a measure of the predictive relevance of the dependent variables in the model proposed (Stone, 1974; Geiser, 1975). The Q2 values of 0.02, 0.15, and 0.35 signify small, medium, and large predictive relevance of certain latent variables. The results of our study confirm that the main model offers satisfactory predictive relevance. See Figure 2. All hypotheses are supported on the basis of empirical data. As indicated in the main effects model, community participation and perceived support have a significant impact on integration with path coefficients of 0.405 (p<0.001) and 0.204 (p<0.001), respectively. Perceived support also has a significant effect on participation (β=0.401; p<0.001). Furthermore, community satisfaction shows a relevant impact on integration (β=0.327; p<0.001) and participation (β=0.275; p<0.001). Finally, perceived support has a significant impact on satisfaction (β=0.499; p<0.001). See Figure 2. On the one hand, our research conducts a Goodness of Fit (GoF) index for PLS path modeling which is defined as the geometric mean of the average communality and average R2 for all endogenous constructs (Tenenhaus et al., 2005). GoF is normed between 0 and 1, where a higher value represents better path model estimations (GoFsmall=0.1; GoFmedium=0.25; GoFlarge=0.36). Our empirical research obtains a GoF value of 0.546 for the main effects model. This allows us to conclude that our model has a high (large) prediction power. On the other hand, effect sizes of single predictors are obtained by comparing the explained amount of variance when a predictor is either included in, or omitted from, the model, that is, f2 = (R2incl – R2excl)/ (1 – R2includ). The overall variance explained by the model in terms of RCI2 is 0.600 (hereinafter, community integration, CI). According to Cohen (1988), these effects can be classified as (a) small effect size, 0.02 < f2 ≤ 0.15; (b) medium effect size, 0.15 < f2 ≤ 0.35; (c) large effect size, f2 > 0.35. Our result shows that the inclusion of perceived support as a driver of CI increases the overall variance explained for CI (ΔR2=0.03, f2=0.06). Furthermore, the inclusion of community participation also increases the overall variance explained for CI (ΔR2=0.11, f2=0.265). Likewise, community satisfaction increases the overall variance explained for CI (ΔR2=0.07, f2=0.172). Finally, the interaction effects are also included, in addition to the main effects model. As in regression analysis, the predictor and moderator variables are multiplied to obtain the interaction terms (i.e., product-indicator approach). Standardising of product indicators is recommended in the presence of significant interaction terms involving any of the main effects: no direct conclusion can thus be drawn from these main effects alone (Aiken and West, 1991). Tenenhaus, M., Esposito Vinzi, V., Chatelin, Y. and Lauro, C. (2005), “PLS path modeling”, Computational Statistics and Data Analysis, Vol. 48, pp. 159–205 Turner, J.W., Grube, J. and Meyers, J. (2001), "Developing a optimal match within online communities: An exploration of CMC support communities & traditional support", Journal of Communication, Vol. 51 No 2, 231-251 Vasalou, A., Joinson, A.N. and Courvoisier, D. (2010), "Cultural differences, experience with social networks and the nature of ‘true commitment’ in Facebook", International Journal of HumanComputer Studies, Vol. 68 No 10, pp. 719-728 Verhoef, P.C, Lemon, K.L., Parasuraman, A., Roggeveen, A., Tsiros, M. and Schlesinger, L.A. (2009), "Customer experience creation: Determinants, dynamics and management strategies", Journal of Retailing, Vol. 85 No 1, pp. 31-41 Wellman, B., Haase, A. Q., Witte, J. And Hampton, K. (2001), “Does the internet increase, decrease or supplement social capital? Social networks, participation and community commitment”, American Behavioral Scientist, Vol. 45, pp. 436–455 Werts, C.E., Linn, R.L. and Jöreskog, K.G. (1974), “Interclass reliability estimates: Testing structural assumptions”, Educational and Psychological Measurement, Vol. 34, pp. 25-33 Wu J-J., Chen Y-H. and Chung Y-S. (2010), "Trust factors influencing virtual community members: a study of transaction communities", Journal of Business Research, Vol. 63 No 9 pp. 1025-1032