RIJKSUNIVERSITEIT GRONINGEN – FACULTY OF ECONOMICS AND BUSINESS
From Lurking to
Contributing: How
Contextual Variables
Affect Participation
A Case Study in Online Care Communities
Sytze Boskma s2036681 June 2015
Table of Contents
Table of Contents ... 2
Abstract ... 3
Introduction ... 4
Research Question ... 5
Structure of the paper ... 6
Theoretical Review ... 7
Online communities ... 7
User roles ... 9
Motives for active and passive user behaviour ... 11
Contextual Variables ... 13
Conceptual model ... 16
Methodology ... 18
Research design ... 18
Data collection and analysis ... 18
Reliability & Validity ... 21
Results ... 23
Overview of the cases ... 23
Ease of Access ... 26
Shared Purpose ... 27
Trust ... 29
Openness of the community ... 31
Conclusions and Discussion ... 33
Limitations and Future Research ... 39
References ... 40
Abstract
Introduction
Open innovation and co-creation are topics that have been extensively covered in the literature since firms and customers can communicate and interact with each other by the internet (Chesbrough, 2003). Especially the concept of open innovation communities have found a boost through the rapid development of digital communication media (Füller, Hutter, Hautz, & Matzler, 2014). The literature discusses a great amount of practices and approaches on how to deal with open innovation and co-creation (Burke, Joyce, Kim, Anand, & Kraut, 2007; Chanal & Caron-Fasan, 2010; Enkel, Gassmann, & Chesbrough, 2009; Gassmann, 2006) and by drawing from these it is evident that users can be an important source of innovation, and thereby a potential for competitive advantage for both small and large businesses. Muhdi & Boutellier (2011) state that ‘they [the users / participants in open innovation] have been identified to be the major drivers of many innovations in different industries’.
(David & Shapiro, 2008; Füller et al., 2014). Whereas it is argued that an understanding of the different roles of individuals in these groups is essential in order to successfully manage such communities (von Krogh, G., & von Hippel, 2006; Welser et al., 2011), we may not underexpose and undervalue the role users play in a community.
Previous work on user roles in online communities suggest that ‘different roles are encountered in innovation communities that relate to the quality and kind of contributions’ (Füller et al., 2014) and found that quite similar roles and structures can be observed across online communities (Arazy & Nov, 2010; Cross et al., 2006; Toral, Martínez-Torres, & Barrero, 2010; Ye & Kishida, 2003). Found was that most communities exist of a small densely knitted core of (high) active users, and that they are surrounded by a large majority of passive (lurking) users (Arazy & Nov, 2010; Cross et al., 2006; Füller et al., 2014; Toral et al., 2010; Ye & Kishida, 2003). Throughout the literature on user roles, the large inactive majority has been titled Lurkers, Passive user, Periphery users, Silent user/contributor or Tourists.
Research Question
In her work, Nijs (2014) proposes an 8 step model (Stairway to heaven) about the emergent organization to reach strategic intent. The work of Nijs (2014) offers a contribution to the organizational controllability of the complexity of managing online communities, to realize the intended strategic goals. In step four, Nijs (2014) describes the action customers (participants in an online community) take and which emphasises the important role of customer involvement. Fuller (2014) talks about Socializers (very active in communication and interaction) as the most important user role for a successful outcome of an innovation community.
As argued by Jacucci & Wagner (2007), contextual factors are ways in which creators can intervene (shaping and augmenting) with a community, by addressing ways in which the community is structured and managed. This affects the way people view, experience and think of the community and the community members and is thus associated with both group dynamics and the community design itself. This research will go in depth on the ways in which the behaviour of the passive user can possibly be changed to active behaviour by different contextual factors and therein this research will focus on contextual factors that influence the community structure and group dynamics. Thereby this research will contribute to the framework of Nijs (2014) by adding contextual factors to the model. More specifically we will research the following question:
In order to answer this research question, the following sub-questions will be addressed:
Sub-question 1: What defines an active and passive user in an online community, and how can these be operationalized?
Sub-question 2: What drives and constraints a user in becoming active in an online community?
Sub-question 3: Which contextual variables are in play in the constraints users feel? Sub-question 4: How can these contextual factors influence the behaviour of a user?
Structure of the paper
Theoretical Review
The theoretical review section will look deeper into the existing literature. First, a brief overview on the topic of online communities will be given, followed by a more in-depth explanation of the role of users. In this explanation the model of Nijs (2014), called ‘the stairway to heaven’ will be discussed, whereas it provides an interesting view on how the input of active users can create and emerge into a generative dialogue leading to strategic intent (which can be innovation for a company for example). This section will be continued with an overview of the different roles users play in communities. Whereas there is a large body of existing literature on personal motivations to participate and to contribute in online communities, and these motivations can have a link with the contextual variables, we will follow with an overview of personal motivations to contribute. However, this will not be a too elaborate analysis whereas this is covered in the existing body of literature extensively and contextual variables are central in this thesis. For an in-depth analysis of personal motivations to participate in online innovation see Bakici, Almirall, & Wareham (2011), Koh, J., Kim, Y. G., Butler, B., Bock (2006), Ståhlbröst & Bergvall-Kåreborn (2011), Youcheng & Fesenmaier (2003), Bullinger, Haller, & Möslein (2009), Chen, Marsden & Zhang (2012), Muhdi & Boutellier (2011), Ye & Kishida (2003) and Zhou (2011). Though, we will take these motivations into account when analysing contextual variables of the community itself and overview the motivations of users to lurk in a community. We will conclude the theoretical review section with an overview of contextual variables which can influence the behaviour in an online community, which will result in propositions.
Online communities
more than ever important, whereas customers are the initiators of co-creation. Accordingly, the full potential of a community can be unlocked when a generative dialogue emerges, as explained by the model of Nijs (2014) ‘Stairway to heaven’.
Stairway to heaven
for an (complex) open context. Typical for an emergent organization is that many stakeholders are thinking with the organization about creating synergies that make themselves more relevant (Nijs, 2014). The model of Nijs (2014) is graphically presented below.
Nijs (2014) states that it is essential in realizing emerging work processes that ‘an environment is constructed that engages people, nurtures trustful and respectful relationships and unleashes everyone’s creativity in a coherent way’. She argues that when there is a feeling of being part of something extraordinary, a community can attract many talented and creative people. Also she underlines the importance of a starting engine in this process. Whereas active contributors are essential in this perspective, the different roles users take in a community are discussed below.
User roles
depth approaches seem not feasible in the scope of a thesis. Therefore we will follow the approach of Gleave & Welser (2009), which dictates that the best way for identifying roles is to take the level of meaningful social action as a starting point, from which one can work downwards to key related ‘behavioural regularities’ and distinctive positions in communities, but also upwards ‘to abstract theoretical categories that allow us to tie these particular roles to general research objectives that transcend any particular study or social context’ (Gleave & Welser, 2009; Welser et al., 2011). Starting point in our research will be the meaningful social action desirable in online communities to foster co-creational behaviour. According to the literature, and well underlined in the work of Nijs (2014), (active) interaction is crucial whereas ‘organizations are no longer closed systems for which the environment is a given, but they are open systems for whom co-creation and co-evolution of and with their environment is essential for survival’ (Nijs, 2014). This comprises mainly (frequent) posting and commenting behaviour. From this point on, we can identify many user roles in literature, which takes different points of perspectives in analysing user roles. Kim (2000) for example, takes a life-cycle perspective in analysing user roles. Customers in a community are starting as ‘visitors’, which are new to the community, and progressively becoming ‘participating novices’ before becoming ‘regulars’, ‘leaders’ and in the end ‘elders’. Kozinets (1999) divides between ‘tourists’, ‘minglers’, ‘devotees’ and ‘insiders’ according to the strength of the social ties with the group (community) and the interest in consumption behaviour. Most of the literature on user roles however focuses on the contribution and participation of the users. For example Fuller, Jawecki, & Muhlbacher (2007) distinguishes between ‘Frequent Posters’ , which contribute almost on a daily basis, ‘Posters’, which contribute on a regular basis and ‘Lurkers’, which are characterised by their passive, non-contributing and observing behaviour. Table 1 provides an overview of the different user roles that can be found in the literature.
In reviewing the literature on user roles it comes apparent that many authors differentiate between a couple of roles that involve at least some interaction or contribution of the user and a single role that describes the user(s) that is passive and non-contributing, which corresponds with the notion of Muhdi & Boutellier (2011) and others including Toral et al. (2010) who state that ‘this is the case of open source communities, which exhibit a clear onion-like structure, with a central core of highly active individuals, surrounded by other layers of progressively less-active individuals’.
will therefore be defined as the users that are related to the community, but did not (yet) post or comment in the community. In the next section, motivations to be active and passive will be discussed.
Table 1: Overview Userroles in the Literature
Author(s) User role Name Characteristics
Fuller, Jawecki, & Muhlbacher, 2007 Frequent posters Posters Lurkers
(Almost) daily contribution Regular contribution
Rarely to non-contributor and passive observer Kim, 2000 Visitors Participative novices Regulars Leaders Elders
Gradually becoming more involved and more influential in the interaction and the community respectively
Kozinets, 1999 Tourists Minglers Devotees Insiders
Weak social ties towards the group / minor interest in consumption
Strong social ties / superficial interest in consumption Weak social ties / mainly interested in consumption Strong ties and high interest in consumption activities Nonnecke et al.,
2006
Developers Discussants Peripheral users
Coming up with new ideas
Joining the interaction / discussion of ideas Silent users, with no active participation Welser et al., 2011 Advisors Parents Brokers Editors Managers Vandals
Found user roles in Wikipedia, and found roles ranging from productive to unproductive (vandals) behaviour
Motives for active and passive user behaviour
But even with strong motives to participate, users can still decide to lurk for many reasons which prohibit them from posting or responding in a community. Lurking can result from motivations ranging from intentionally lurking (Preece, Nonnecke, & Andrews, 2004) to lurking out of discomfort. Intentional Lurkers, users who did not participate in the past and never intended to participate from the outset, can lurk for multiple reasons. In a study (Nonnecke & Preece, 2001) it was found that intentional Lurkers didn’t feel a specific need to post, were not motivated to post, never intended from the outset to post and didn’t want any commitment with a community (and posting / replying to a community was felt as a commitment). Whereas intentional Lurkers are really unmotivated to post and reply, effort to activate these users is often effort wasted. More interesting are the Lurkers who lurk because they do not feel comfortable to post. The first group of unintentional Lurkers are characterized by Lurkers who are unable to post or comment due to incapability’s regarding the forum of the community, time constraints (Nonnecke & Preece, 2001) and the fact that there were too many messages (Preece et al., 2004). Furthermore, Nonnecke, Andrews, Preece, & Voutour (2004); Nonnecke & Preece (2001) and Preece et al. (2004) found that Lurkers do Lurk because they ‘still needed to find out about the group’. Users in this category of Lurkers had to develop trust in the community, whereas they didn’t know the group enough (Preece et al., 2004). Most of all they wanted to get used to the communication style of the group, and find out to what extent the group is useful for their interest, questions and problems. These, together with another major group of Lurkers can be grouped as users which are Lurking for reasons relating to the group dynamics (Nonnecke & Preece, 2001; Preece et al., 2004). Motivations to lurk ranged from a desire to remain anonymous, being shy about public posting to the fear of being offended and aggressive responses on their post. Many users lurk because they have the feeling that they will not be accepted / fit in the group. When being exposed to an existing community they felt as an outsider, and interaction with the group would feel as breaking in. Along the hesitation to interfere with a tight group, the fear of rejection and being insulted was given by respondents as major reasons to lurk (Burke et al., 2007; Nonnecke et al., 2004; Nonnecke & Preece, 2001; Preece et al., 2004). In table 2 a summary of the motivations to contribute and lurk are given.
Table 2: Motivations to contribute and lurk
Reasons to be active Reasons to lurk
Fun / Enjoyment Lurking from the outset
Monetary Reward - Never intended to participate
Peer Recognition - Not wanting to make a commitment
To make use of valuable information - Not motivated to post / nothing to offer Getting answers to specific questions Not able to post
Personal Gain - Time constraints
- Too many messages Poor fit regarding group dynamics
- Desire to remain anonymous - Shy about posting
- Fear of aggressive responses - Not feeling part of the group - New members treated poorly - Fear of rejection
- Still figuring out the group
Taking into account these motivations to participate and lurk, it seems that a user of a community must feel able to post on an accessible platform, which feels well organized and clear and within a decent timeframe. Besides that, a sense of community must be conveyed to the user to take away its desire to remain anonymous, the fear of aggressive responses, rejection and other negative behaviour against new members. Users who intended to lurk from the outset, with no intention to interact in the future are noted, but further not treated in this research.
Contextual Variables
Ease of Access
Before a user will and can post in a community, he or she will need the time to do so. Furthermore the user will need to be (or get) acquaintance with the software of the community. Finally, when there are too many messages, a user cannot find where it is supposed to post or find a question, so a clear (interface) design is needed. In their work, Chen et al. (2012) propose a taxonomy of company-sponsored value of co-creation. They propose that the value of company-company-sponsored co-creation stems from the process in which creators work on a task given by the company. Where the co-creators are (accordingly) affected by different motivations (quite similar to the motivations discussed previously), the process is affected by governance, incentives and IT support. Incentives are a tool for rewarding and motivating active behaviour, and governance is a way to regulate the discussion but IT-support is the one dealing with a motivation to lurk, undermining the reason to lurk. In this thesis, the concept of ease of access encompasses both the technical access to a community and the ease of use of the forum of the community. The following propositions are proposed:
P1a: Easy access to a community for a user is a prerequisite for active behaviour of that user
Sense of Community
Besides the constraints regarding the process of posting and unmotivated users, passive behaviour stems forth out of constraints regarding the group dynamics, as identified in the previous section. Literature on group dynamics, educational science and learning environments discuss the sense of community (in a learning context), and define this along a number of dimensions (McMillan, 1996; Rovai, 2002b, 2002c). These dimensions include sense of belonging, trust, spirit (membership), interaction, (mutual) interdependence, common expectations, shared values and shared goals. Important to note is that interaction is both an impetus of, and an outcome of ‘sense of community’. Whereas in this study the main objective is to find factors stimulating passive users (who do not interact) in active behaviour, and in respect with the constraints regarding group dynamics, we will focus on three constructs that may influence the passive behaviour through a sense of community. Before entering in interaction, individuals start out by reading and scanning the messages present in the community (Maloney-Krichmar & Preece, 2002). In this way the user sees and experiences the interaction and the division of roles, which ‘serves as a powerful example of how a member should relate to and interact with the community’ (Maloney-krichmar & Preece, 2005). Future encounters with the community will most likely be governed by the norms they observe in the community. Therefore is it essential that a community communicates their low levels of aggressive behaviour through their norms which conveys trust and safety to overcome the constraint fear of aggressive responses and rejection. Quick and helpful responses on posts of newcomers conveys acceptance and openness to newcomers, to overcome the constraint of ‘newcomers are treated badly’ and the fear of rejection (Burke et al., 2007). The constraints regarding anonymity, shyness and not feeling part of a group may be tackled by communicating a shared purpose, which conveys shared goals. These three constructs will be more elaborately discussed in the following subsections. First the importance of communicating a shared purpose is discussed, followed by trust and safety. Finally we will discuss the openness of the group in relation with the passive behaviour of users.
Shared Purpose
2010 argue that identification with a group is very effective for online groups whereas people with a strong identity attachment to a community tend to have higher rates of commitment and participation. A shared purpose can be conveyed by the company to accelerate the potential effect a shared purpose between users has on the interaction in the community. Therefore the following propositions are proposed:
P2a: In an online community, having a shared purpose between the users of the community will actively foster the behaviour of passive users
P2b: In an online community, communicating a shared purpose, by the company, will actively foster the behaviour of users
Trust
Besides influencing behaviour through a shared purpose, trust is a influential factor in reducing the constraints for lurking, whereas trust is a prerequisite for truthful interaction and interpersonal communication (Posey, Lowry, Roberts, & Ellis, 2010). In this research the definition of Posey et al., (2010) is used, defining online community trust as ‘the degree to which an individual believes that those within his or her selected online community are reliable and are trustworthy with information that makes the individual vulnerable’. Trust among members in a community is important whereas trust can reduce the desire to remain anonymous, and shyness to post. Explicitly in health communities, where the information posted can be confidential, trust is an important factor (Maloney-Krichmar & Preece, 2002). Furthermore Maloney-Krichmar & Preece (2002) state that in order to succeed, an online community needs a sense of trust among its member so that they will treat each other with respect, and that ‘their problems and concerns will be heard, and that others will provide information and support for them’ (Maloney-Krichmar & Preece, 2002). Rovai, (2002a) argues (in a learning context) that trust will elicit a feeling of safety among members, which will create the feeling that other members of the community will behave in a supportive and constructive way. Finally, trust reduces the fear that other users in a community will act opportunistically or take advantages over other contributors. Trust results from the reputation of the community and its members on the one hand, and (positive) past experiences with the community (and its members) on the other hand (Maloney-krichmar & Preece, 2005). Taking into account the important role of trust in a community, and on the premise of the argument by Maloney-krimar & Preece (2005) the following proposition is proposed:
P3b: In an online community, trust between the users of the community will actively foster the behaviour of passive users.
Openness of the community
Whereas the fear of rejection and the (dis)believe that new members are treated poorly are among the top motivation for Lurkers to lurk, openness of the community is an important factor in activating passive users. The openness of a community signals the extent to which the group is open for newcomers. The openness of a community can be considered high when newcomers (first posters and prospective members) are quickly being accepted by the incumbents of the community. Responsiveness (whether a (first)poster gets a reply) signals acceptance to a group, and is argued to have an increased likelihood of future active behaviour and more commitment towards a group (Burke et al., 2007). However, silence is interpreted as rude and unfriendly, and signals rejection, which is associated with a high degree of leaving (Burke et al., 2007). L. Chen et al. (2012) found that prior feedback is important in regard with the participation (duration) of users in a community. This indicates that a display of openness of the community, for example quick and welcome responses on newcomers, may remove the constraint Lurkers feel in being active. Therefore the following proposition is proposed:
P4: In an online community, the openness of the community will actively foster the behaviour of passive users.
Conceptual model
Methodology
The following section will discuss the method of research that was used in this study, followed by an explanation of how the data was collected, so replication of this study is possible. With the same objective in mind this section will be ended with an overview of the data analysis and a discussion of the reliability and validity.
Research design
The aim of this paper was to explain how contextual factors influence the behaviour of participants in an online community, and thereby contributing to the model of Nijs (2014) by providing a more comprehensive understanding of the effect of contextual factors regarding customer involvement. Whereas this process includes knowledge generation (exploration), rather than knowledge testing, a theory development approach was chosen for this study (Aken van, Berends, & Bij van der, 2012). In order to give a valid answer to the research question and in particular to find support for the propositions provided in the theoretical review section, case studies were conducted. The case studies had partially an interpretive research design whereas content analysis was used ‘to capture behaviours and relations within a group [community]’ (Gleave & Welser, 2009). Interpretive methods have been found useful in the identification of social roles and in the understanding of the motivations of individuals (Walsham, 2006). Interpretive research methods are often found to neglect the social structure in which these behaviours occur. Therefore the content analysis and interviews, which were held in a (semi-)structured way, were complemented with network approaches. This combination of approaches is proposed by (Gleave & Welser, 2009) who stated that ‘Integrating the two approaches [inductive and network approach] by combining the pure structural approach of social network analysis with the behavioural notion of a role makes it possible to identify, based on structural signatures, the social roles that are meaningful at the interaction level’.
Data collection and analysis
The data collection had three primary sources, interviews and open-ended questionnaires combined with a network analysis, both within two cases. Secondary data consists of content analysis. First a short description of both cases will be given, followed by an explanation of the interviews. This subsection will be concluded by providing an overview of how the data regarding the different concepts were collected.
Wijmantelzorgers
The community of Wijmantelzorgers is an initiative of PUO (Real name is withhold for privacy matters) focussed on providing a virtual space in which informal caretakers can come together and share their experiences, thoughts and problems. The initiative is started to enhance customer loyalty with its members, and to give informal care takers a podium, whereas these people mostly are not that visible. The community has 66 members at this point in time (19th of May), which include some highly active ones. Topics are nowadays started by both the community managers and the members itself. Access to the forum does not require to create an account, to respond or create a topic however does require to create an account.
Fitwerkt24
The community of Fitwerkt24 is built around the concept of vitality and mindfulness, to let people think and work about their mental-fitness. The idea is to stay healthy while working through a 28 ‘mindfulnesschallenge’ which includes exercises as yoga and meditation. Besides the challenges, a community was built around the mindfulness-challenge, with the idea that people would get involved, and think / talk about the problems and achievements they encounter along the way. This was facilitated by a MOOC (massive open online course), which was designed and made available for this project. With over a 500 initial registrations for the forum, subscription was high.
Social network analysis
towards another user (represented with a line between two nodes). An inactive user (Lurker) can be identified whenever a node is not connected at all (represented by single nodes).
Interviews & Questionnaires
Interviews were held with different parties. First, interviews with community managers were held to get an in depth understanding of the action taken to improve the participation in the communities. The interviews with G. (Community manager of Wijmantelzorgers) and K. (Community manager of Fitwerkt24) were held in a semi-structured way. The full transcription can be found in the first two appendices.
Secondly, open ended questionnaires were sent to participants of the Wijmantelzorgers-community and the Fitwerkt24-community by mail, to gain valuable information regarding the constraints they have felt in the past to become active, and if / how they did overcome these constraints. For privacy reasons, the names of the participants will not be showed. The questionnaires were of a qualitative nature, and therefore treated as such, no statistical analysis was therefore conducted.
Concepts Ease of access
Shared Purpose
The concept of shared purpose can be conveyed by the company (through the community managers and the forum content) and can be present by the users themselves. The communication of a shared purpose was obtained by content analysis of the communities, and interviews with the community managers. This was complemented by asking some of the users of the communities (selected in conjunction with the community managers) to what extent they felt an alignment of goals with the other users, to capture the presence of a shared purpose. Furthermore they were asked to what extent they believed the shared purpose (if present) affected their participation. In the analysis of the content, a shared purpose will be present whether users display common goals.
Trust
Regarding the concept of trust, questions will be asked to some users of the Wijmantelzorgers-community and the Fitwerkt24-Wijmantelzorgers-community. The questions will bring insights in the extent the members trust the company, and the community members. Also it will be asked to what extent the trust in the community has hindered them and others from participating (in the past). The questions regarding trust were adopted from Chung et al. (2010). The community managers was asked what actions he did undertake to ensure a feeling of trust regarding the community, if this affected the participation and how trusted the company is.
Openness of the group
The openness of the group will be measured in different ways. First the responsiveness on first posters will be measured, whereas the responsiveness on new posters signals acceptance. This will be complemented with questions in the interview with the active participants, asking them to what extent they feel part of the group, and to what extent they believe the group is open to newcomers. In the questionnaire of the fitwerkt24 community, questions regarding participation and openness of the group will be asked.
Reliability & Validity
Reliability is achieved when the outcomes of a study are independently from the characteristics of that study. With that intent in mind, this study takes into account the following biases: researcher bias, situation bias, instrument bias and respondent bias.
research projects). This will reduce the chance that faults were made. Finally the social network analysis was conducted in conjunction with two other researchers. Situation bias was accounted for by using two different cases (two communities) in this research. Instrument bias was controlled by using multiple research instruments; interviews / questionnaires, social network analysis and content analysis. The questions for the interview were overlooked by another researcher and the questionnaires were overlooked by a third party, to account for instrument bias. Finally respondent bias was controlled by asking different parties in each case (both the community managers and some of the members of the community).
Results
In this section an overview of the results of the empirical study will be given, which are the foundation of the conclusions to the propositions and research question which will be given in the next section. As stated in the methodology section, this research was built upon two cases, in which interviews and open-ended questionnaires were held with the community managers and members of the community respectively. In each community content analysis was conducted, complemented with a social network analysis. The results section is structured as follows: first this section continues with an overview of the two cases in respect to the participation levels, followed by an outset of the results per proposition. Not every case was useful for every proposition, therefore only the useful cases are incorporated in the results per proposition.
Overview of the cases
Case 1: Wijmantelzorgers
Figure 2: Sociogram - Wijmantelzorgers 2 months after start
Figure 3: Sociogram – Wijmantelzorgers 4,5 months after the start
Red = Identified active participants two months after the start of the community
Yellow = Identified passive participant two months after the start (and may therefore be active in the second timeframe)
Figure 4: Posts per week - Wijmantelzorgers
Case 2: Fitwerkt24
Figure 5: Sociogram Fitwerkt24
Ease of Access
The first propositions captures the ease of access for users to a community and states that easy access is a prerequisite for active behaviour, but not a driver of active behaviour:
P1a: Easy access to a community for a user is a prerequisite for active behaviour of that user
P1b: Easy access to a community for a user is not a sufficient condition itself for active behaviour of that user
Case 1: Wijmantelzorgers
In the interview with G., community manager of Wijmantelzorgers, it came apparent that the ease of access of the community was a point of concern at the start of the community. When the forum was first announced, in an offline meeting, some of the community members didn’t know what a ‘forum’ was “sommige zeiden ‘forum? Wat is dat in Godsnaam?’, ‘Wat is een post?’” [“some said ‘forum? What is a forum?, What is a post?’”]. According to the community manager, this was due to their relative high age, which had the consequence that these members had less experience in digital communication tools, and were overall very busy people. It was noted by the community manager that this also had an upside; ‘when you can make these people active, you are most likely the only platform they are active on’. Therefore the average time online per session is relative high (5 minutes per session).
The community has been renovated multiple times, starting as an open source community platform with many functions and tools. The community manager states that at first the platform was too complicated, certainly for this target audience. Therefore they have stripped down the entire platform to the basics, which has, accordingly, increased the ease of use of the community. One of the other actions that were taken to increase the ease of use was, what is called, ‘de etalage’. A snapshot at the homescreen of the community that captures the ongoing discussion of that time, and which serves as a quick summary of what’s going on. The community manager was asked if he believed that after the renovation(s), people were better able to access the forum, and more interaction took place. His answer was clear: “Ja. Zeker.” [Yes, certainly].
really had to put some effort in it before I was used to this’]. Also it became evident from the questionnaires that the forum renovations did not change the level of participation of the active users.
Case 2: Fitwerkt24
In the case of Fitwerk24, the ease of access was not really a point of concern. The community manager can only recall one single individual who didn’t know how to reach the community. The target audience of this case were working people (age <35), which almost all had experience regarding digital communication tools (this is an assumption). However she noted that the process of subscription was not optimal, whereas people had to subscribe for both the mindfulnesschallenge and the community, and that this could have been a constraint. Therefore the interviewer asked to what extent she believed that when subscription had been optimal, responds on the forum was higher. The community manager replied that she doesn’t think people would interact more on their own, and stimulation of interaction would still be necessary “Niet zo maar uit zichzelf, ik denk dat je ze nog steeds moet helpen over de eerste drempel heen om iets te posten”. Finally, regarding the ease of access, it was found in the community itself that some topics had been read over a 1000 times.
Shared Purpose
Regarding a shared purpose, the following propositions were proposed:
P2a: In an online community, having a shared purpose between the users of the community will actively foster the behaviour of passive users
P2b: In an online community, communicating a shared purpose, by the company, will actively foster the behaviour of users
Data was found in both the cases by interviews with the community managers, questionnaires of the users and analysing the content of the community.
Case 1: Wijmantelzorgers
(previous identified) customers of PUO with their ideas of starting a community. Central was the intent to connect people and to enter into a dialogue with these people to show the relevance of PUO and the community for them. This resulted in very positive responses from this audience. One way in which this shared purpose is displayed is the title of the community, which is chosen in conjunction with the members, ‘Wij Mantelzorgers’. Another way of communicating a shared purpose is by sending a newsletter with relevant and ‘hot’ topics. Furthermore there is an animated movie, which can be found in the community, explaining the purpose of the community, and how the members can support each other with the intent to activate members.
Finally the community manager points out that in the Wijmantelzorgers community there are two groups of people with different goals. One group has the goal of giving informal caretakers more public recognition (give them a ‘face’). The other group consists of people that want to tell their story, be listened to, finding recognition / share emotions among each other, and sharing tips and advices. The second group is built around the concept of providing help to each other because ‘we are all caretakers’. The community manager points out that the second group is more active, and may be less interested in the goal of the first group (whereas their goal is to be listened to, and not to make a united platform), but both groups can be found active in the community. In analysing the content of the community it came apparent that both groups do share their purpose across the topics, for example “Mijn doel als mantelzorger is erkenning, herkenning en acceptatie” [my goals as an informal caretaker is to get acknowledgement, recognition and acceptation].
In the questionnaire, sent to some of the active users in de Wijmantelzorgers community it was asked to what extent they felt their goals were aligned with the other users, and how this affected their participation. One of the respondents stated that he/she didn’t have goals, because the community was not under his/her daily attention. Similarities with the interview of the community manager can be found under the respondents that did state their goals. One of the similarities is the mutuality and cohesiveness of the process, ‘the forum is driven by (1) competence, (2) autonomy and (3) the connection with others’, this was complemented by a statement that mutuality was (in his/her personal opinion) one of the drivers of this community. On the question of how the shared purpose affected their participation it was answered ‘I share my experiences with other caretakers’, and ‘this influenced my participation in a positive way’.
Case 2: Fitwerkt24
people had the same goal in mind, ‘taking more time for yourself’, but she argues that the underlying problems (where these goals stem from) were very different. Regarding the same topic she argues that this may be one of the points where they [the initiators of the community] run short, they didn’t address the problems of the people, but just offered a solution. In contrast to the Wijmantelzorgers case, Fitwerkt24 did not convey a shared purpose from the start. ‘People expected to complete a certain program about mindfulness, and that was where we invited them for’. The community manager furthermore stated ‘we communicated the challenge, not the community’, and she believes that may be a point that they must consider for the future. In the beginning a little movie was posted by the initiators of the community which conveyed a message to trigger people to be active, however, this movie did not convey a shared purpose. Mid-way the challenge, the community manager did try to activate users by notifying them that they could use tips of each other. She states that the sharing of a purpose is also something that has to rise inside a community on its own.
When analysing the content it came apparent that there was one single topic that had exceeded the number of 3 reactions. The topic was called ‘Waarom ga jij de uitdaging aan?’ [Why did you accept the challenge?], which triggered (until that point passive) users to state their personal goals of subscribing to the mindfulnesschallenge. In this topic people both stated their personal goals, but also in some cases, their underlying problems. According to the community manager this topic had the most interaction because people can make a commitment, for themselves, to do the challenge. She notes that this is just a personal assumption.
Trust
Regarding the concept of trust interviews were held with both the community managers, and active members of the communities to find an answer to the following propositions:
P3a: In an online community, trust (of the users) in the company will actively foster the behaviour of passive users.
P3b: In an online community, trust between the users of the community will actively foster the behaviour of passive users.
Case 1: Wijmantelzorgers
get those people together, and work from that point onwards. The first trust relationships were built at a face to face meeting (offline), but the community manager argues that it is also important (for trust) that users show something personal at the forum of the community. Therefore they designed for some special topics ‘Het verhaal van’ [the story of...] and ‘het dagboek van’ [the diary of...] in which users can share a glimpse of their daily live. He argues that the fundament of trust is that when one shares his/her live, others are also tended to do the same. When asked if the trust relationships that were designed activated users to participate, he could not give an indication. To measure the trust in the community, active users were asked to what extent they believed that other users would act in a trusted manner regarding the forum. All the interviewed users did acknowledge trust as a factor that is in play, but their attitudes towards trust differed. One of the users stated: ‘I feel safe around the people at this forum and do act in trust with them. For me it means trust giving is trust receiving’ indicating that he/she trusted the other users. Another indicated that safety is something that was conveyed at the meetings, but he/she still ‘generalized’ her information to make it less personal, whereas ‘healthy distrust’ was in place (accordingly). Furthermore one indicated that he/she was aware of the fact that third parties can access and read the community too, which somehow restricted her in putting too confidential information on the forum. However he/she states that the other users are felt as warm, sympathizing, competent and friendly, and adds that the same counts for the PUO webmasters.
Regarding trust in the company (PUO), the community manager stated that this has been an important theme for them. In the first two meetings the community manager has explained to the users why they are initiating this community, which resulted in questions, discussions and finally a manifest in which, together, they made clear what the goal of the community must be, and how the company plays his part in this. This did have some effect: ‘We see that people gradually get looser and better realize that the content they post can pop up in Google’. Furthermore he notes that once people had been in contact with the company, in real live, it is easier to get them active online. ‘But when people are new to the community, and did not have any contact with the company physically, it is way harder to get them active online’. Finally he states that he thinks that PUO is a strong brand, in which people have trust and a good feeling. The community manager states that by explicitly stating online that the community is initiated by PUO, people are more tended to visit the community.
their carefulness and strong will to achieve something good with this forum’. However it comes forth out of the interviews that some believe PUO to have a ‘hidden agenda’ regarding the data in the community ‘to be honest, it would surprise me if PUO didn’t do something with all the data’. But it does not seem that this hinders or affects the trust in the good intentions of PUO. Finally it was asked if trust did affect the participation in the community. One of the respondents answered that trust did not affect her participation, because he/she felt perfectly safe in the community. Also it was answered by one of the interviewed users that it affected the kind of information posted. Confidential information was not posted, and rather conveyed in a personal way, and otherwise generalized. This was in line with the other interviewed user who stated that trust had played a role, and wonders to what extent such information can be used against you.
Case 2: Fitwerkt24
In the interview with the fitwerkt24 community manager it was asked to what extent actions or measures were taken to increase the trust among members. It came apparent that none were taken, and added that in some cases it may be positive for participation, whereas anonymity can reduce the constraint of sharing something personal. Regarding the trust in PUO, she adds that PUO has a trustworthy image, and is often linked to reliability. She adds that whether this influences if people will share personal experiences depends on how one views PUO. If one considers PUO as a pension fund administrator, this will not be the case. If one considers PUO as a care provider, this could be the case. Whereas PUO was not actively linked to the community (as was the case in the Wijmantelzorgers case), she doesn’t believe this was a constraint for people to share their experiences.
Openness of the community
Regarding the openness of the community, the following proposition was proposed:
P4: In an online community, the openness of the community will actively foster the behaviour of passive users.
Analysis of the content and interviews with the community managers and some active members were the basis of this proposition.
Case 1: Wijmantelzorgers
feel the need to meet new people. Gradually it became clear that when they [the community] want to make an impact, and help other people, they had to grow. Nowadays people are getting welcomed when they post on the forum, and we see the same openness in the offline meetings’. The same was found when analysing the content of the community, first time posters where in most of the cases thanked for their post both in the form of a ‘like’ and a written responds. To enhance this, the community manager recently introduced the role of ‘verwelkomer’, which included the specific task to make people feel welcome after their first post. Finally he states that to make the community ‘tangible’, it was communicated that the forum is ‘owned’ by its members, accordingly this elicits the feeling of acceptation and ownership.
The questioned active members of the community mostly view themselves as a group that is becoming tighter, and all do feel part of the group. The questioned user that didn’t describe the group as (becoming) tight(er), also believed his/her participation is not affected. All the interviewed users are open for newcomers, and see them as added value for the community.
Case 2: Fitwerkt24
Conclusions and Discussion
The following section presents conclusions and discussions of the results of this study. Per concept the proposition(s) will be covered with a conclusion and a discussion, including theoretical and managerial implications.
Ease of Access
Proposition 1a: Easy access to a community for a user is a prerequisite for active behaviour of that user
In the first case (Wijmantelzorgers) the ease of access was sometimes problematic due to the minor experience some of the (older) users had with the digital communication tools (including the community platform). Some users did not even know what a forum was, and therefore weren’t able to post or react at all. Over time the ease of use was improved, for example by stripping down the forum to the basics. The community manager told that more interaction took place after the forum was renovated and also the users themselves indicated that the ease of use improved. This is consistent with the findings of the social network analysis which shows an increase in active users, and the upwards slope of the posts per week graph. The percentage active users / Lurkers did descend with some percentages, but this is mainly due to the fact that the community started out of a face to face meeting, resulting in a very high starting percentage of active users. In the last two months, all new members to the community did not meet face to face first, and therefore proportionately more passive users did become active in the second period. All the previous stated arguments indicate that easy access to a community for a user is indeed a prerequisite for active behaviour of that user, and therefore proposition 1a doesn’t have to be revised.
Theoretical and Managerial Implications of Proposition 1a
learned from the cases. The cases learn that the complexity of the forum must be aligned with the target audience. Whereas the forum of the WijMantelzorgers community was in the beginning too comprehensive, and the target audience very inexperienced, a match between those two was far to be found, which withhold some of the users from becoming active. A second lesson can be learned from the first case. The Wijmantelzorgers case shows that an increase in ease of use may take away the constraints for Lurkers to post for the first time, and that the ease of access can both arise from practical improvements (for example adding ‘the etalage’ for quick access or the shortcut on the homepage where one can ask questions) or from stripping down the forum to the basics. Therefore a manager must first recognize the underlying constraint regarding ease of access before taking action.
Proposition 1b: Easy access to a community for a user is not a sufficient condition itself for active behaviour of that user
Even though the first case shows an increase in active behaviour after the ease of access increased, it is too straightforward to label ease of access as a driver of active behaviour. The active users that were interviewed were all already active before the forum was renovated. The questionnaires of the active users made it evident that the level of participation of the active users was not changed by the forum renovations. The same results were present in the second case. The users of the fitwerkt24 case were all younger than 35 years old and therefore assumed to have (at least minor) experience in digital communication tools. Both the community manager of the fitwerkt24 case and ‘the numbers of times read’ for each post also indicated that ease of access was not a constraint in the fitwerkt24 case. Therefore it can be concluded that easy access to a community is not a sufficient condition itself for active behaviour of that user (P1b), whereas the interaction in the second case was almost none.
Theoretical and Managerial Implications of Proposition 1b
from the cases is that easy access alone is not enough to make a Lurker active, and therefore other actions must be taken to increase participation.
Shared Purpose
Proposition 2a: In an online community, having a shared purpose between the users of the community will actively foster the behaviour of passive users
Proposition 2b: In an online community, communicating a shared purpose, by the company, will actively foster the behaviour of users
Out of the questionnaires of the users and the content analysis, it becomes evident that a shared purpose exists under (most of) the members of the Wijmantelzorgers community. Furthermore the active members describe the forum to be driven by competence, autonomy, the connection with others, but most of all the mutuality between the people. The participation of the users was affected ‘in a positive way’ by the presence of a shared purpose and made questioned members to share his/her experiences with other caretakers. In the second case interaction was very low, with one single exception, a topic in which members were asked to state their goals, which triggered many passive users to state their goals and entertain into interaction.
In the Wijmantelzorgers case a shared purpose can be considered to be the cornerstone of the community. From out the start it was conveyed that the community was built for, and therefore owned by its members. A shared purpose was, besides in the personal contact, conveyed by the name of the community (Wij Mantelzorgers), which underlines the notion that the community is built for and around the informal caretaker. Furthermore a shared purpose was conveyed by the introductory animated movie and the (introduction) newsletters. The intentions to build the community around a shared purpose have ‘resulted in a very positive responds’ according to the community manager, which is consistent with the active core of users that can be found in the social network analysis (figure 2). It was found that two different goals were present in the community: The goal to make informal caretakers public recognizable (on a large scale), and the goal as also displayed by the shared purpose conveyed by the company. The community manager pointed out that the second group (with the similar goals as the company) is more active than the first group (the ones with the goal of public recognition). All the above indicates that P2a and P2b do not have to be revised.
The conclusions to the propositions regarding a shared purpose are in line with the existing theory. The communication of a shared purpose, by both the company and the users, reassures that others share the same problems (Maloney-krichmar & Preece, 2005), which can overcome the constraint Lurkers feel regarding ‘figuring out about the group’, as identified in the theory section. This is in line with the findings that stress the driving factor of mutuality in the Wijmantelzorgers community, and may also be the reason that the topic about goals did elicit interaction in the second case. These findings do contribute to the model of Nijs (2014) by including the importance of realizing a shared purpose to engage people into (emerging) action. This brings in some valuable lessons for managers. First, managers must acknowledge the importance of having a shared purpose between users of a community, whereas a shared purpose can be the driving force in a community and the key to customer involvement. The second lesson builds on the first: explicitly communication a shared purpose throughout the community can have a reinforcing effect on the activation effect that a shared purpose between users has.
Trust
Proposition 3a: In an online community, trust (of the users) in the company will actively foster the behaviour of passive users
Proposition 3b: In an online community, trust between the users of the community will actively foster the behaviour of passive users
of the active members made it clear that trust did restricted some users from entering in substantive interaction, and rather keeping the conversation on a general level. A user that fully trusts the other users and the company stated that trust wasn’t a factor that constraints, and therefore he/she shared freely. Other users had, even though they stated they trust the company, the idea that PUO still would ‘use’ the data, and they would like to know the future goals of the company regarding the community. Besides trust in the company and the other users, the fact that third parties can access the forum made some users withheld from posting to private information.
Theoretical and Managerial Implications of Proposition 3a and 3b
The fact that there was no information found that indicates that the trust in the company or users directly influences the behaviour of passive users, does not mean trust is not important in the consideration Lurkers make, to become active. The company was considered a trusted brand in both cases (which makes sense whereas both communities are part of the same company), and therefore it could be that the constraint to lurk (as identified in the literature section) was therefore not present. The found interaction effect of trust is added value regarding the model of Nijs (2014), whereas customer involvement is essential to reach the strategic intent. Customer involvement may be higher when people feel free to enter in to substantive interaction, rather than a safe kind of interaction. Even though no information was found that trust influences directly the behaviour of passive users in these cases, lessons can still be learned. First, trust must still be considered an important factor whereas low levels of trust may hinder people from sharing substantive information, which will keep the interaction at a ‘safe’ level. Secondly, trust between the users of a community may elicit a sense of community, which is found to increase the interaction between users. Third, the Wijmantelzorgers case showed the importance of physical meetings in building trust relationships, which are the foundation for initial community growth. Finally, the first case shows the importance of offering the users an opportunity to tell something personal. From both the interviews and questionnaires it became evident that a display of trust of one user, triggers trust between users.
Openness of the community
P4: In an online community, the openness of the community will actively foster the behaviour of passive users.
complemented by the community manager of Fitwerkt24, who states that ‘I don’t believe many people did feel part of a community in the first place’. Also the data of the first case was not sufficient to give a conclusive indication to what extent the openness of the community fosters the behaviour of the passive participants. Therefore this proposition is recommended for future research.
All the conclusions are summarized in the following table, to provide an overview of the implications of this study.
Table 3: Conclusions and Implications
Concept Theoretical Implications Managerial Implications
Ease of Access - Easy access to a community can be found a prerequisite for active behaviour, but is on itself not a sufficient condition for active behaviour
- A contribution is made to Nijs (2014) by adding the need for an easy access in constructing an environment that supports collective creativity
- The complexity of the forum must be aligned with the target audience - An increase in ease of use can take away the constraints users feel regarding posting, but easy access alone is not enough to drive active behaviour
- An increase in ease of use can be achieved by both upgrading and stripping down the forum
Shared Purpose - Having a shared purpose between the users of a community will actively fosters the behaviour of passive users, and communicating a shared purpose by the company, actively fosters the behaviour of users
- Adds to Nijs (2014) by including the importance of realizing a shared purpose to engage people into (emerging) action.
- A shared purpose between the users of a community must be stimulated whereas this leads to active behaviour - Explicitly communicating a shared purpose can enhance the activation effect a shared purpose between users has.
Trust - Trust is found to have an effect
regarding the nature of the
conversation (safe conversation vs. substantive interaction). More trust between users, the company and third parties elicits substantive interaction.
- Regarding Nijs (2014), customer involvement may be higher when people feel ‘free’ to enter in substantive interaction
- Trust is important, whereas low levels of trust may hinder people from sharing substantive information. - Physical meeting increase trust relationships, and can be a foundation for initial community growth.
Limitations and Future Research
Limitations
Whereas this study is built on the approach of theoretical development, concepts were tested in a qualitative rather than a quantitative way. This study should be replicated in a quantitative way, to measure the effects of the proposed concepts. Furthermore, the use of cross-sectional data made it not possible to account for changes over time. Ease of access for example can improve over time by familiarization and increased experience in the community and therefore longitudinal data must be collected to give insight in how the effect of the concepts may change over time. Regarding trust of the company no constraints were detected, this may have been due the fact that both cases were part of the same company, and therefore a difference in trust between companies could not be determined. Another limitation regarding the trust concept may be that the study was held in a care setting, therefore the confidentiality of the information may have overtaken the effect trust between members could have had on the behaviour of a passive user. However this is an assumption, and should therefore be accounted for in replication of this study. This may also have had limitations regarding the generalizability of the study. Finally, although questions were sent to over 600 passive users, responds was zero, which is a limitation of this study, whereas no information of the passive users and their constraints was collected (due to the short time scale of this study, setting out a new questionnaire in a new population was not feasible).
Future research
References
Aken van, J., Berends, H., & Bij van der, H. (2012). Problem solving in organizations: A methodological handbook for business and management students. Cambridge University Press. Antikainen, M., & Väätäjä, H. (2010). Rewarding in open innovation communities — how to motivate
members. International Journal of Entrepreneurship and Innovation Management, 11, 440– 456.
Arazy, O., & Nov, O. (2010). Determinants of Wikipedia Quality. Proceedings of the 2010 ACM Conference on Computer Supported Cooperative Work, 233.
Ardichvili, A., Page, V., & Wentling, T. (2003). Motivation and barriers to participation in virtual knowledge-sharing communities of practice. Journal of Knowledge Management, 7(1), 64–77. Bakici, T., Almirall, E., & Wareham, J. (2011). Motives for Participation in On-Line Open Innovation
Platforms. Danish Research Unit for Industrial Dynamics, 14(11).
Bugshan, H. (2014). Co-innovation: the role of online communities. Journal of Strategic Marketing, 23(May), 175–186.
Bullinger, A. C., Haller, J. B. a, & Möslein, K. M. (2009). Innovation mobs – unlocking the innovation potential of virtual communities. Proceedings of the Fifteenth Americas Conference on Information Systems AMCIS San Francisco, 1–8.
Burke, M., Joyce, E., Kim, T., Anand, V., & Kraut, R. (2007). Introductions and requests: Rhetorical strategies that elicit response in online communities. Proceedings of the 3rd Communities and Technologies Conference, 21–39.
Chanal, V., & Caron-Fasan, M. (2010). The Difficulties involved in Developing Business Models open to Innovation Communities: the Case of a Crowdsourcing Platform. Management, 13(4), 318– 340.
Chen, C. J., & Hung, S. W. (2010). To give or to receive? Factors influencing members’ knowledge sharing and community promotion in professional virtual communities. Information and Management, 47(4), 226–236.
Chen, L., Marsden, J. R., & Zhang, Z. (2012). Theory and Analysis of Company-Sponsored Value Co-Creation. Journal of Management Information Systems, 29(2), 141–172.
Chesbrough, H. W. (2003). Open Innovation: the New Imperative for Creating and Profiting from. Harvard Business School Press.
Chung, J. E., Park, N., Wang, H., Fulk, J., & Mclaughlin, M. (2010). Age differences in perceptions of online community participation among non-users: An extension of the Technology Acceptance Model. Computers in Human Behavior, 26, 1674–1684.
David, P. a., & Shapiro, J. S. (2008). Community-based production of open-source software: What do we know about the developers who participate? Information Economics and Policy, 20(08), 364–398.
Enkel, E., Gassmann, O., & Chesbrough, H. (2009). Open R & D and open innovation : exploring the phenomenon. R & DManagement, 39, 311–316.
Faraj, S., Jarvenpaa, S. L., & Majchrzak, a. (2011). Knowledge Collaboration in Online Communities. Organization Science, 22(5), 1224–1239.
Füller, J., Hutter, K., Hautz, J., & Matzler, K. (2014). User Roles and Contributions in Innovation-Contest Communities. Journal of Management Information Systems, 31(1), 273–308.
Fuller, J., Jawecki, G., & Muhlbacher, H. (2007). Innovation creation by online basketball communities. Journal of Business Research, 60(1), 60–71.
Gassmann, O. (2006). Opening Up the Innovation Process : Towards an Agenda, 1–14.
Gleave, E., & Welser, H. T. (2009). A conceptual and operational definition of’social role'in online community. 42nd International Conference on System Sciences, 1–11.
Hsu, M. H., Ju, T. L., Yen, C. H., & Chang, C. M. (2007). Knowledge sharing behavior in virtual communities: The relationship between trust, self-efficacy, and outcome expectations. International Journal of Human Computer Studies, 65, 153–169.
Jacucci, G., & Wagner, I. (2007). Performative roles of materiality for collective creativity. Proceedings of the 6th ACM SIGCHI Conference on Creativity & Cognition - C&C ’07, 73–83. Kim, A. J. (2000). Community building on the web: Secret strategies for successful online
communities. Addison-Wesley Longman Publishing Co., Inc.
Koh, J., Kim, Y. G., Butler, B., Bock, G. W. (2006). Encouraging participation in virtual communities. Communications of the ACM, 50(2), 69–74.
Kozinets, R. V. E. (1999). The strategic implications of virtual communities of consumption. European Management Journal, 17(3), 252–264.
Kratzer, J., Zboralski, K., & Leenders, R. T. a. J. (2009). Interaction quality within communities of practice: contextual factors of utilising different communication media. International Journal of Networking and Virtual Organisations, 6(2), 199.
Lin, M. J. J., Hung, S. W., & Chen, C. J. (2009). Fostering the determinants of knowledge sharing in professional virtual communities. Computers in Human Behavior, 25(4), 929–939.
Maloney-Krichmar, D., & Preece, J. (2002). The meaning of an online health community in the lives of its members: Roles, relationships and group dynamics. Technology and Society, 20–27. Maloney-krichmar, D., & Preece, J. (2005). A Multilevel Analysis of Sociability , Usability , and
