Submitted on November 26, 2007, to the On-line Interaction and Facilitation Seminar, Fall 2007, Purdue University,
Dr. Sorin A. Matei via the I Think Blog
The widespread expansion of communication technology over the past century has resulted in a marked shift toward mediated interaction. In this transition, media professionals have been faced with the task of creating and maintaining online communities. In accomplishing this task, media professionals may draw on an increasingly rich body of theoretical literature providing useful guidance for their endeavors. This white paper suggests four theoretically and empirically motivated principles for online community designers: 1) Offer the opportunity to belong and be affiliated with others; 2) Find an appropriate balance of group homogeneity and diversity; 3) Provide appropriate moderation; and 4) Provide the community with the technical ability to seek and obtain media gratifications from all four categories. For each, practical application in an online environment will be explained, possible issues for the community manager will be identified, and steps will be prescribed for evaluation of the online community’s application of the principle.
In today’s fast-paced, highly technologized culture, the social arrangements that once provided the interaction that humans crave have largely disappeared. Social changes, land use changes, and technological changes have set the stage for a transition from sociability conducted face-to-face to sociability occurring online (Wellman et al., 2003). In this transition, media professionals have been faced with the daunting task of creating and maintaining online communities, whether for e-commerce (e.g. Amazon.com), information sharing (e.g., Wikipedia) social networking (e.g., Facebook), or any of hundreds of other niche pursuits.
In accomplishing this task, media professionals may draw on an increasingly rich body of theoretical and empirical literature providing useful guidance for their endeavors. This white paper offers four theoretically and empirically motivated principles for online community designers. For each, practical application in an online environment will be explained, possible issues for the community manager will be identified, and steps will be prescribed for evaluation of the online community’s application of the principle. This paper does not attempt to provide guidelines for usability, features, or design, but instead presents broad principles which can be used to increase the effectiveness of online communities.
Principle #1: Offer the opportunity to belong and be affiliated with others
The first major building block of a successful online community is the opportunity to belong and be affiliated with others. Social psychology research has revealed that humans have an innate need to belong and be affiliated with others, which is a primary motivation in joining both online and offline groups (Ridings & Gefen, 2004, p. 4). Advances in technology over the past several decades have resulted in a marked shift in the composition of many people’s social networks. Sociologist Barry Wellman explains that prior to widespread technologized communication, people generally found their social circles in close geographic proximity to where they lived. In these “door-to-door” social networks, individuals fulfilled their need to belong and be affiliated with others with those whose “doors” were within walking distance. As travel became more common and communication technology evolved, people began widening the geographic spread of their social networks giving rise to what Wellman terms “place-to-place” networks (Wellman, 2001). In this scenario, individuals supplemented the door-to-door interaction with the ability to write letters, and later, telephone one another to keep strong social ties over larger distances. The latest wave of innovation – mobile phones and computers, has enabled the current stage of social networks: person-to-person. In the person-to-person iteration, individuals’ location is of no concern.
Through cell phones and wireless internet, people can be reached no matter what their geographic location: the locus of social interaction is now the individual, resulting in the ability to belong and be affiliated with others with little consideration of the spatial. With this new ability to form communities of shared interest, online communities have sprung up devoted to countless niche interests which geographic boundaries have long prevented. This new type of community is similar in many ways to spatially bounded communities, with participants expressing “strong interpersonal feelings of belonging, being wanted, obtaining important resources, and having a shared identity” (Wellman, 2001, p. 247).
In practice, it’s fairly simple to provide the technological ability to belong and be affiliated with others. Interfaces as archaic and “user-hostile” as PicoSpan used by The WELL were remarkably functional in providing the ability for users to communicate and develop a sense of belonging and affiliation (Rheingold, 2000). More modern examples of online communities, such as MySpace and Facebook, have hardwired cues within their infrastructure for affiliation and belonging – the “friend” list. Although many users of these and other social networking sites list dozens, if not hundreds, of “friends,” reflecting the distinct possibility that many of these relationships are merely casual acquaintances, there is wisdom in the hard-wiring of these cues for belonging. Boneva and colleagues, in a study of adolescent instant messaging (IM), found that IM users found the notifications of friends coming online or signing off provided users with a “feeling of group presence” despite the largely dyadic nature of conversations (Boneva, Quinn, Kraut, Kiesler, & Shklovski, 2006). Even on The WELL, users had the ability to see who else was logged on to the system and send private messages (Seabrook, 1997). Current communities like Facebook and MySpace also provide these indicators of group presence – even though interaction is still primarily dyadic. From these successful examples, three rules of practice can be derived:
1) The means by which individuals interact need not be “state of the art.” While it might not hurt, it certainly is not the catalyst for a successful community.
2) Provide users with cues for belonging.
3) Provide users with indicators of group presence.
For the manager of an online community, there is one major potential problem which should be anticipated and prevented. It might be tempting for a novice community engineer to assume that the ability to belong and be affiliated is, in itself, enough to foment a community. In reality, there must be some other draw in order for an online community to attract members; site infrastructure alone does not foster usage of a site (Ludford, Cosley, Frankowski, & Terveen, 2004, p. 632). Thousands of message boards, USENET news groups, IRC channels, and website forums provide the opportunity to interact, belong, and affiliate. In order to be successful, an online community must also provide access to desired information or help in achieving goals (Ridings & Gefen, 2004, p. 4). By finding a niche of useful information or specific goals sought by individuals, the community can attract new members and increase the desirability of membership for others.
Measurement and Evaluation:
Nancy Baym provides one clear measurement of whether or not an online community space is truly being used for community activities leading to affiliation and belonging. She notes that “As groups develop over time, they generate group-specific meanings. Eventually, new forms of speech, or genres, unique to that community evolve” (Baym, 1998). In order to measure this, content analysis of communication within the group would need to be done to identify any communicated meanings which are “group specific.” (For more information on content analysis techniques, see Preece, 2000, pp. 334-338). If Baym’s observations are correct, the more widespread and frequent the use of these memes are within the community, the more individual members of the community are likely to report a sense of belonging and affiliation.
In addition to this theoretically based qualitative measurement, community members could be surveyed directly and asked to respond to a questionnaire measuring levels of community involvement, belonging, and affiliation (Preece, 2000, p. 313). Statistical analysis of a survey of this kind would yield a useful estimate of the success of the online group in fostering this key attribute of community.
Principle #2: Find an appropriate balance of group homogeneity and diversity
There are a number of theoretical models that identify various factors influencing individual contributions to groups (Ludford, Cosley, Frankowski, & Terveen, 2004, p. 633). One such factor which has been empirically studied is that of group similarity and individual uniqueness. Similarity refers to the general agreement about common topics by group members while individual uniqueness refers to the unique contributions specific group members can make to the community at large ((Ludford, Cosley, Frankowski, & Terveen, 2004, p. 634). In their experimental study within the MovieLens online film rating community, Ludford and colleagues found that the most active communities were composed of dissimilar members who were given information about the unique contributions they could make to the group (Ludford, Cosley, Frankowski, & Terveen, 2004, p. 635). Ling et al. explain that in this particular experiment, those who were given information about the uniqueness of their possible contributions increased their level of effort because they felt like they were working by themselves to provide interesting information for the group. This served to reduce the social loafing of the group, defined as “the robust phenomenon that occurs when people work less hard to achieve some goal when they think they are working jointly with others than when they think they are working by themselves” (Ling et al., 2005).
The practical application of this empirical research is achieved when finding a balance between group homogeneity and difference. Groups that are too homogeneous may find that they have little to discuss – they are too similar to be interesting to one another. Even Howard Rheingold, as mentioned in John Seabrook’s Deeper, acknowledged that to some extent, arguments, or “thrashes,” were the glue that held The WELL together and kept people interested (Seabrook, 1997). When determining the informational, entertainment, or relational focus of the community, it must be restrictive enough to provide a destination worth visiting, but not so restrictive that those who stick around have too-similar ideas, opinions, and knowledge.
This is a particularly troublesome principle for online community managers to deal with. Unlike the experiments conducted by Ludford (2004), Ling (2005), and their colleagues, community managers can’t simply pick and choose community members to produce the precise level of homogeneity that will foster the most vibrant community. In order to cope with this difficulty, it may be prudent for the manager to divide the focus of the online community into sub-categories, similar to the various conferences used by The WELL (Rheingold, 2000). Doing so will attract a more diverse group to the community site as a whole, while allowing individuals the ability to choose the specific topics which interest them the most, and thus will likely be populated by more homogeneous community members. With this community design, sub-topic conversations will still benefit from “outside” community members’ interaction (increasing diversity) while maintaining a level of specificity necessary for the “regulars” in the particular sub-topic.
Measurement and Evaluation:
There are a number of ways to evaluate the homogeneity and diversity of an online community, and the method employed is largely dependent upon the nature of the community being evaluated. A movie recommendation community, such as the MovieLens platform used by Ludford (2004), Ling (2005), and their colleagues, or any other such community in which quantitative data has already been provided by users is ideal for measurement and evaluation of this principle. In most online communities, however, this is unlikely to be the case. There are a number of alternatives to strict quantitative analysis of differences between group members. One option would be to conduct interviews with individual community members to probe their perceptions of the similarities and differences between themselves and other members. Open interviews like this “often provide very rich data because respondents may mention factors not considered by the interviewer” (Preece, 2000, p. 319). These unknown factors may prove very useful as adjustments are made to the community structure in order to facilitate higher levels of interaction by members. Alternatively, content analysis could be used to determine the frequencies of various types of communication (i.e., information seeking, information providing, entertainment seeking, entertainment providing, thrash baiting, thrash response, etc.). By establishing a baseline measurement of the percentages of each type of communication identified, an evaluator could determine whether or not the community is engaging is too much (or too little) communication of each type over time, and could formulate responses to address the problem.
Principle #3: Provide appropriate moderation
Karau and Williams’ collective effort model proposes that “one’s motivation for a given effort depends on how well that effort translates into performance, what outcomes the expected level of performance is likely to lead to, and how much those outcomes are valued.” People consider these issues from both an individual and a group perspective, and also consider the predicted difference their contribution will make to the group’s performance (Cosley, Frankowski, Kiesler, Terveen, & Riedl, 2005, p. 12). The idea of oversight becomes important in the collective effort model because it strengthens the connections between individual effort and individual performance/contribution to group performance by raising the individual’s awareness that contributions are being monitored. In addition to strengthening the links between these elements of the model, oversight also eliminates the tendency people have to “work less when it is hard for others to evaluate their individual contributions” (Cosley, Frankowski, Kiesler, Terveen, & Riedl, 2005, p. 12). Empirical research based on the collective effort model has demonstrated that appropriate oversight in online groups leads to more frequent and higher quality contributions to the community. Specifically, experiment using the movie recommendation site MovieLens.org supported their hypothesis that moderation of submissions led to more submissions and less antisocial behavior in the community. Additionally, their study found that moderation by peers was as effective in raising the quality and quantity of submissions as moderation by movie experts (Cosley, Frankowski, Kiesler, Terveen, & Riedl, 2005, p. 18).
The practical application of this research is obvious: moderators make communities better by reducing antisocial behavior and increasing the quantity of contributions. For a new online community, it is probably best to start with a small number of moderators who are tasked with removing community spam, preventing flame wars, and enforcing whatever community standards have been enacted. As the community grows, it may become more practical to distribute the moderation tasks to more members of the community or perhaps even all members of the community. This gradual movement to a community-wide distributed moderation system would be similar to the path followed by Slashdot.org as the community grew too large to be moderated by a select group. Instead, any regular Slashdot member has the power to moderate posts. (Malda, 2003).
The issue of moderation is a sensitive one, and one not to be taken lightly. It is particularly important that community standards have been established and moderators do not deviate from those standards in their exercise of power. If moderation is not distributed, it is particularly important for community managers to keep close watch of moderator activity to prevent abuse. If moderators are found to be overly eager to exercise their power beyond what the community standards and guidelines allow, they may need to be stripped of that power. If distributed moderation is used, managers must take care to ensure that the algorithms determining the relative worth of individual moderating activities do not provide too much control to any individual.
Measurement and Evaluation:
It is particularly important that the scheme used for moderating community activity is evenly applied and clearly understood by the community members. For this reason, evaluation of the moderation system is especially important. If the community uses expert moderation featuring selected members who are given the power to moderate, there are a number of indicators that the system is working. First, your online community should be relatively free of spam and other postings that fail to conform to the community standards and guidelines. If spam and inappropriate postings often go unnoticed, there are probably too few moderators to adequately patrol the community spaces. Second, with an expert moderation scheme, there must be a mechanism within the community for members to report abuse. If claims of moderator abuse are relatively infrequent, the system is likely functioning properly. Moderators with frequent accusations of abuse should be monitored by the community manager, and, if necessary, be removed from duty.
If the community uses distributed moderation, abuse is much less likely to occur unless community standards and norms are still in flux. In any case, the effectiveness of the distributed moderation system can by tested quantitatively by examining the comment database and looking for trends in variables such as amount of time before comments are moderated the first time, number of moderations per comment, fairness of moderations (if a meta-moderation scheme is also used), and moderation reversals. Depending on the features of the moderation system used, the quantitative data provided by this analysis will provide the community manager with information about the strength and weakness of the moderation system as well as guidance for future improvements. (For an example of the types of quantitative evaluation that can be conducted with this type of moderation scheme, see Lampe & Resnick, 2004).
Principle #4: Provide the community with the technical ability to seek and obtain media gratifications from all four categories
Uses and Gratifications Theory (Katz, Blumler, & Gurevitch, 1974) proposes four gratification categories that people seek from media sources: information, personal identity, integration and social interaction, and entertainment. An examination of various online communities of varying levels of success has demonstrated that those communities which are most successful are communities where members routinely seek and obtain gratifications in all four of the categories specified by the theory (Yale, 2007). Today’s most widely used online communities, Facebook.com and MySpace.com, are further evidence that successful online communities provide their users with the ability to obtain multiple media gratifications from a single online space.
In applying this observation, managers should take the necessary steps to ensure that the community space provides the technical ability to allow the users to create, exchange, and interact in ways that provide users with the opportunity for sharing and receiving information, building personal identity, interacting socially, and being entertained. The exact mechanics will differ from community to community based on purposes and norms, and in some cases, the nature of the community may not lend itself to each of the gratification categories. Whatever the case, community managers should do their best to provide space for all possible gratification categories.
For a community designer and manager, it might seem practical to immediately build a community from the ground up with a contrived framework and manufactured content designed to provide media gratifications for community members. This is almost certainly a bad idea. Rather, community managers should start small and add the features and spaces for further interaction and gratification as the community grows large enough to make use of them. Community giants like Facebook and MySpace didn’t start out by providing users with an easy platform for obtaining all four gratifications. Instead, they focused on one or two and developed the platform as the communities grew and users demanded more features.
Measurement and Evaluation:
The most practical way to measure the number of gratifications sought and obtained by community members is probably through close-ended surveys using scales of agreement. Starting with the framework of media gratifications proposed by Katz, Blumler, and Gurevitch (1974), questions can be developed to measure the frequency with which members engage in community activities leading to gratifications in each of these categories. An online community can be said to be successful to the degree that high numbers of community members routinely seek and obtain media gratifications within the community space.
This white paper has provided the theoretical basis and practical application of five key principles for developing online communities. Additionally, issues which community managers must consider and designs for measurement and evaluation of a community’s implementation of each principle have been explained. This document is intended for media professionals interested in theoretical principles which can be used to foster the development of online communities.
Baym, N. (1998). The emergence of on-line community. In S. Jones (Ed.), Cybersociety 2.0 (pp. 35-68). Thousand Oaks: Sage.
Boneva, B., Quinn, A., Kraut, R., Kiesler, S.,& Shklovski, I. (2006). Teenage communication in the instant messaging era. In R. Kraut, M. Brynin, & S. Kiesler (Eds.), Computers, phones, and the Internet: Domesticating information technology. Oxford: Oxford University Press.
Cosley, D., Frankowski, D., Kiesler, S., Terveen, L., & Riedl, J. (2005). How oversight improves member-maintained communities. Conference on Human Factors in Computing Systems. Proceedings of the SIGCHI conference on Human factors in computing systems table of contents. Portland, Oregon, USA.
Katz, E., Blumler, J. G., & Gurevitch, M. (1974). Utilization of mass communication by the individual. In J. G. Blumler & E. Katz (Eds.), The Uses of Mass Communications: Current Perspectives on Gratifications Research. London: Sage.
Lampe, C & Resnik, P. (2004). Slash(dot) and burn: Distributed moderation in a large online conversation space. Proceedings of the SIGCHI conference on Human factors in computing systems, p.543-550, April 24-29, 2004, Vienna, Austria.
Ling, K., Beenen, G., Ludford, P., Wang, X., Chang, K., Li, X., Cosley, D., Frankowski, D., Terveen, L., Rashid, A. M., Resnick, P., and Kraut, R. (2005). Using social psychology to motivate contributions to online communities. JCMC, 10(4), article 10.
Ludford, P., Cosely, D., Frankowski, D., Terveen, L. (2004). Think different: Increasing online community participation using uniqueness and group dissimilarity. Conference on Human Factors in Computing Systems. Proceedings of the SIGCHI conference on Human factors in computing systems.
Malda, R. (2003, June 4). Slashdot FAQ – Comments and Moderation. Retrieved November 19, 2007, from http://slashdot.org/faq/com-mod.shtml#cm600
Preece, J. (2000). On-line communities. New York: Wiley.
Rheingold, H. L. (2000). Virtual community: Homesteading on the electronic frontier. Cambridge, MA: MIT Press.
Ridings, C. & Gefen, D. (2004). Virtual community attraction: Why people hang out online. JCMC 10(1).
Seabrook, J. (1997). Deeper: Adventures on the net. New York: Simon & Schuster.
Wellman, B. (2001). Physical place and cyber place: The rise of personalized networks. International Journal of Urban and Regional Research, 25(2), 227-252.
Wellman, B., Quan -Haase, A., Boase, J., Chen, W., Hampton, K., Isla de Diaz, I., Miyata, K. (2003). The social affordances of the Internet for networked individualism. Journal of Computer Mediated Communication, 8(3).
Yale, Robert. (2007). Welcome to I-berspace: Media Gratifications in Successful Virtual Communities. Retrieved November 19, 2007, from I Think Web site: http://matei.org/ithink/2007/10/15/i-berspace/