This section addresses the various attributes of virtual communities. While disciplinary perspectives are muted intentionally in the proposed definition of virtual community, they take center stage in the discussion of attributes. Indeed, this author contends that the role of attributes in understanding virtual communities is where discipline-focused researchers could contribute to the generation and testing of hypotheses.
The literature suggests that five attributes could be used to characterize virtual communities: (1) Purpose, (2) Place, (3) Platform, (4) Population Interaction Structure, and (5) Profit Model. Conceptualizing virtual communities as having essential attributes gives researchers a consistent and practical way to describe virtual communities. From a managerial perspective, virtual communities with unique combinations of attributes are likely to have different critical success factors and associated outcomes for both members and organizational sponsors.
The proposed typology is based on a review of literature and on observations of design and activity in actual virtual communities. The approach used to develop the proposed typology embraces Hunt's (1991) description of an inductive grouping procedure to form categories. This grouping procedure typically results in polythetic classes of phenomena (Hunt). This means that virtual communities within a given class are likely to share common attributes, but no individual community must possess all of the attributes commonly associated with that class. The classification variables, establishment (first level) and relationship orientation (second level), distinguish one type of virtual community from all others. Thus, the attributes described in the following section could be used to describe any virtual community regardless of its type.
Attribute #1: Purpose (content of interaction)
The notion of purpose is central to the functioning of a virtual community in that communities are defined by shared purpose among community members (Gusfield, 1978). However, given the proposed definition and typology of virtual communities presented here, it is clear that virtual communities can be formed around an infinite number of shared interests. Indeed, the purpose attribute is analogous to the concept of “discourse focus” (Jones and Rafaeli, 2000, p. 218) —the subject that forms the basis of interaction in a virtual community (e.g. golfing, living with diabetes, parenting techniques).
Attribute #2: Place (extent of technology mediation of interaction)
The notion of place in virtual communities is an important but troublesome concept for researchers because of the aspatial nature of such communities (see Jones, 1997). Traditional communities often are associated with a specific, geographically bounded location. Within such a location, community-based interaction leads members to feel a sense of belongingness, shared values and understandings. Thus, the notion of community implies both something structural (e.g. a bounded location) and something socio-psychological (e.g. a sense of shared values developed through interaction with members).
Harrison and Dourish (1996) describe the structural properties of virtual communities as those that deal with the community's “space” (i.e. physical structure) and the socio-cultural properties of virtual communities as those that deal with the community's “place” (i.e. socio-cultural). They suggest that a virtual space is to a virtual place as a house is to a home that dwells within its physical boundaries. In essence, a house is a home insomuch as its members are aware that they possess the socio-cultural bonds of relationships among household members. Harrison and Dourish suggest that a house only has physical properties that could shape the development of a home. Likewise, they say that a virtual space only presents only the opportunity for a virtual place to develop (see Harrison and Dourish 1996).
Despite the clarity of the nomenclature presented by Harrison and Dourish (1996), other researchers have not fully adopted their perspective or terminology. Indeed, the notion of space and place are often conflated in the literature. For example, Blanchard (2004) suggests that a community member's perceived sense of place is influenced by perceptual cues in the virtual environment (e. g. type of access, timing of interaction and membership boundaries). She suggests that members use such cues to determine where community interaction occurs, where they “are” in the flow of conversations with other members and whether other members are present. She contends that each community member has a sense of place, even if individual diffferences in perception lead members of the same virtual community to different senses of place.
Blanchard (2004) uses a different concept of “sense of place” than that used by Harrison and Dourish (1996). Blanchard's notion of “sense of place” is one that is based on a member's psychological awareness of the location or co-presence of others in a particular location. Thus, it is more consistent with Harrison and Dourish's structural conceptualization of “space.”
Mitra and Schwartz (2001) suggest that a virtual space is comprised of both a sense of presence and location. Thus, they use the term “space” to describe Blanchard's notion of “sense of place.” Consistent with Harrison and Dourish's (1996) concept of “space”, Mitra and Schwartz suggest that technological properties can influence a community member's sense of presence. Furthermore, they suggest that the use of metaphors to the physical world (e.g. “Internet address” and “website”) enhances members' sense of location in virtual environments.
Other scholars question the appropriateness of the distinction between “space” and “place.” Some suggest that interaction in virtual communities is not confined to only virtual or only physical realities (see Bernard, 1973; Rothaermel & Sugiyama, 2001, citing Tönnies, 1967). These scholars suggest that cyberspace is conceptually embedded in physical space (Mitra & Schwartz, 2001) and that human interaction is framed by both physical and virtual space (see Harrison and Dourish's discussion on hybrid spaces). Furthermore, the fact that virtual community members use multiple modes of communication including face-to-face, telephone and mail (Blanchard, 2004) suggests that virtual and physical communities can coexist. In sum, Wilson and Peterson (2003, p. 456) suggest that “the distinction of real and imagined or virtual community is not a useful one.”
In this typology, Harrison and Dourish's (1996) socio-cultural notion of place is included, appropriately, in the definition of a virtual community. The definition put forth in this paper includes the explicit notion that interaction in a virtual community is guided by protocols or norms. A virtual community is defined, therefore, by its sense of place, as described by Harrison and Dourish.
While the distinction between “place” and “space” might prove valuable for particular types of research, for the purpose of this typology, the “place” attribute embraces the notion of “degree of virtualness” put forth by Virnoche and Marx (1997). Virnoche and Marx take a location-based approach by suggesting that communities can be categorized based on the extent to which community members share virtual space and/or physical space on an ongoing or intermittent basis. In a world where individuals maintain relationships with others in physical space, many virtual communities are composed of members who share a virtual space and, intermittently, physical space.
Virnoche and Marx (1997) define virtual extensions as real, physically based relationships that are extended into virtual space. Thus, the location attribute is conceptualized as having two levels: (1) hybrid (exists in both physical and virtual space) and (2) virtual (exists only in virtual space and never in physical space). The few researchers that have examined differences between these levels have found no significant differences in structural interactions between members of communities with the latter two levels of the location attribute (see Butler, 1999), but more research is warranted.
Attribute #3: Platform (design of interaction)
Synchronicity is an important concept related to interaction in virtual communities. Synchronicity is the degree to which a medium enables real-time interaction (Hoffman & Novak, 1996). Most researchers use the term to describe the technological design of interaction dichotomously: (1) synchronous or (2) asynchronous. For example, chat room technology supports real-time communication (i.e. synchrononous interaction) whereas email-based forums allow members to view and respond to messages at their convenience rather than in real time(i.e. asynchronous interaction) (see Preece, 2000 and Blanchard, 2004 for detailed descriptions of various technical community designs).
The fact that interaction is integral to the value of synchronous technologies gives credence to the importance of research on interactivity (see Rafaeli, 1988), particularly in electronic environments (see Zack 1993). In the context of computer-mediated communication, interactivity is conceptualized as “dependency among messages in threads” (Rafaeli and Sudweeks, 1997) as measured by thread length, depth and/or breadth (see Preece, 2001; Rafaeli & Sudweeks; Whittaker et al., 1998). Unlike the dichotomous description of synchronicity, interactivity is viewed as a continuum (Rafaeli & Sudweeks).
Synchronicity can be valuable for virtual communities provided that members actually take advantage of the synchronous technology design by interacting (Blanchard, 2004). Indeed, a highly interactive environment can enhance a member's perception of social presence, co-presence and sense of place (Blanchard, 2004). It also can facilitate the construction of social reality for members (Rafaeli & Sudweeks, 1997). However, in virtual communities, “interactivity is made possible, but not always exercised” (Rafaeli & Sudweeks, p. 2). In other words, a virtual community might have a particular design for interaction that does not necessarily result in interactivity.
Therefore, it is important to understand both the design for interaction (synchronicity) and the actual pattern of interaction (i.e., interactivity). Interactivity is addressed by the population attribute of the proposed typology (see the discussion of attribute #4 below), to the extent that not only thread length, but also the pattern of interaction among members could be a measure of interactivity. The platform attribute, therefore, focuses only on the technical design for interaction. Since the primary factor that distinguishes the capabilities of various community technologies (e.g. bulletin board, short messenger services) is synchronicity, the platform attribute is conceptualized as having three levels (1) synchronous and (2) asynchronous and (3) hybrid—where a particular community has elements of both synchronous (chat) and asynchronous communication (email forum) design.
Virtual communities as CSSNs.
Virtual communities are a form of CSSN that support strong, weak and stressful social ties among members (Garton et al., 1997; Wellman et al., 1996; Wellman et al., 1997). According to Wellman et al., strong ties emerge as a result of frequent and supportive contact among socially connected members of virtual communities. Weakly tied members also demonstrate supportive and reciprocal behavior, despite the fact that they are socially and/or physically distal. However, when communication among members becomes anti-social (e.g. flaming, spamming) these relations are described as stressful ties.
Virtual communities as small-groups or networks.
Researchers also study whether the small group or network metaphor is appropriate for empirical investigations of virtual communities. For example, Dholakia, Bagozzi and Pearo (2004) use a two-tiered typology of small groups and networks to conceptualize virtual communities. They characterize small-group-based virtual communities (e.g. Harley Owners Group) as having socially close relationships among members, high group interaction, and a focus on maintaining specific relationships within the group. Alternatively, networked-based community members are geographically and socially dispersed and focus on the functional benefits of the community such as information acquisition or problem solving (Dholakia et al., 2004). Relationships in networked-based communities are often of short duration and driven by utilitarian needs.
Furthermore, Butler (1999) found that in large communities based on email lists, interaction requires active encouragement by sponsors. Indeed, these communities are analogous to the network-based communities discussed by Dholakia et al. 2004. Accordingly, Butler suggests that high interaction is not the norm in such virtual communities and would not occur without intentional efforts on the part of a community sponsor.
In sum, prior literature suggests that there are different interaction patterns in small groups versus large networks. Small groups tend to have fixed and limited memberships, are highly interactive during sessions of limited duration and have well defined activities. Alternatively, networked social structures typically have large and variable memberships with uneven and less-active communications among members.
Virtual communities as virtual publics.
Jones and Rafaeli (2000, p. 216) contrast virtual publics with virtual communities. They suggest that virtual publics are “computer-mediated spaces, whose existence is relatively transparent and open, that allow groups of individuals to attend and contribute to a similar set of computer-mediated interpersonal interactions.” Furthermore, these virtual publics (1) might or might not be considered CSSNs, (2) can be supported by a variety of technologies, (3) can serve a variety of purposes and (4) can be owned by an organization. The concept of virtual publics is consistent with Komito's (1998) concept of virtual communities as foraging societies. In such societies, relationships serve functional or utilitarian purposes, membership is often temporary/unstable and there is less commitment and loyalty among members.
In sum, the population attribute is conceptualized as having three primary levels: (1) small group (where strong ties tend to dominate), (2) network (where weak ties are prominent and stressful ties are likely) and (3) publics (where interaction is variable and likely to include strong, weak and/or stressful ties). Some virtual communities will share attributes of both small groups and networks. As suggested by Butler (1999), a metaphor that blends small group and network attributes might best explain relationships in virtual communities, particularly organization-sponsored communities that have a large number of members. This is consistent with the concept of virtual public.
Attribute #5: Profit model (return on interaction)
The profit model attribute focuses on whether a virtual community creates tangible economic value. Although this attribute would apply most frequently to organization-sponsored communities that are commercial, it is possible that other types of communities create economic value. For example, some communities that are member-initiated could invite companies to place targeted advertisements in the community. In this example, the community would earn advertising fee revenue, while the advertisers would gain access to potential customers.
Krishnamurthy (2003) classifies virtual communities according to one of three business models: community enablers, trading/sharing communities and communities as a website feature of corporations. Community enablers host various types of communities with a variety of topical interests and often earn income via advertising and/or subscription fees (e.g. America on Line and Yahoo! groups and clubs). Trading/Sharing Communities facilitate the exchange of products or services among community members and often earn revenue via transaction fees (e.g. EBAY and Napster). Finally, firms that feature communities on their website own the community property and use it in order to generate interaction and, ultimately, revenue-generating transactions (e.g. REI.com retailer-based customer community). In sum, the profit model attribute is conceptualized as having two levels: (1) Revenue-generating (e.g. Host, Facilitator, Owner) and (2) Non-revenue generating.