Online Community and Design Feature Overview

This paper will consider the failure of the Faq-o-Matic archive designed to supplement the Gtube Mailing List community (see http://www.gtube.org)∞. The Gtube Mailing List is an active email-based community in existence since 1996 with around 500 current members that use (or care for someone who uses) a medical device called a G-tube. The community is not affiliated with any outside organization or medical community. It is fully maintained and supported by its members.

Several years ago the person who started the Gtube Mailing List set up a website using Faq-o-Matic software to create an online archive of answers to common questions and helpful tips. In addition, registered users of the Gtube Mailing List are able to read and search online archives that include all messages sent to the List for the past several years. The relationship between the List and the Faq-O-Matic is described on the site:

"If you have a question to ask about tube feeding, the mailing list is the place to put it; whereas here at the Faq-O-Matic we are interested in any answers/tips you may have!"

Although the Faq-O-Matic has been online for quite a while now, there have been almost no content contributed to it, despite the fact that the website is well known to the List member community. The remainder of this paper will relate relevant social science theories and research findings on organizational memory that may explain this failure.

Failure due to Lack of Benefits or Critical Mass?

It is tempting to assume that the Faq system failed because it does not serve a unique or worthwhile purpose. For example, it could be argued that the Mailing List and the archive of past messages serve the same purpose as the Faq. Alternatively there may be no benefit of a Faq. Although these questions are empirical in nature, there is reason to believe that a well-designed Faq could be helpful to both the community members and non-members seeking information on Gtubes. The Faq?s unique role is to provide a summary of definitive information on a particular subject. The Mailing List and its archive provide a mass of undistilled information that can be difficult to weed through (Brewer2000) and understand, especially for newcomers. The Faq could potentially serve as a boundary object for people who only want to be marginally involved in the community or as a way of peripherally participating in the community in order to learn it?s shared repertoire (Wenger1998). However, despite the important and unique role the Faq could play, it has not done so.

One simple explanation for its failure could be that it does not have sufficient critical mass to attract new people (Markus1987). However, this is an unlikely explanation because the Gtube Mailing list is so successful and the primary way of joining that list is to learn about it from the Faq site itself. Several more plausible factors that likely contribute to the failure of the Faq are explained in the remainder of this paper.

Usability and Sociability Problems

Some possible reasons for the lack of content contributed to the Faq-O-Matic relate to what Preece calls usability and sociability issues (Preece2003). The software clearly has some usability problems as indicated by the numerous posts by the name of ?New Item? and the appearance of duplicate posts. The Faq is primarily a ?pull? technology, which is different than the ?push? technology, email, that Mailing List members are used to. Furthermore, in order to post a message to the Faq, a user must first create an account, even if they already have an account for the Gtube Mailing List. Although many users are willing to create an account to join the community, it is clear that not many are willing to create an account in order to be able to post messages to the Faq. These and other usability problems raise the cost of participation in the Faq.

Serious sociability problems with the Faq system also exist. Grudin points out that many groupware systems fail because of a disparity in work and benefit (1994), and Atwood describes how this can happen in organizational memory systems (2002). The Faq is a nice example, where the people doing the work of posting ?answers and tips? to the Faq will likely not benefit (or see the benefits to others) from participating in the Faq system, where they are discouraged from asking their own questions or receiving feedback on their posts. Furthermore, because the technology does not support switching between the Faq and the Mailing List, more work must be done to get messages from one system to the other. Atwood describes how organizational memory systems should capture information with minimal overhead, which clearly is not occurring in this situation (2002).

Social Loafing Concerns

Social psychology research on social loafing also provides some clues as to why people may not be contributing to the Faq. Research has shown that social loafing is higher when work is done for strangers and when feedback is not provided (Karau1993). There are other conditions in which social loafing is higher, but these two are especially pertinent to this situation. In this context, people who post to the Faq are writing primarily for strangers on the Faq, whereas they are writing for friends or at least acquaintances when writing for the Mailing List. Furthermore, the Faq discourages provision of feedback from the users of the posts. This is not the case in the mailing list where it is common to receive ?thank you? replies or ?that wasn?t really what I was looking for? replies.

Lessons from Knowledge Management Research
Theorists on organizational memory point out that many knowledge management systems have failed because they assume knowledge is an object that exists independent of human action and perception, which can be codified and removed from the minds of people (see Wasko2000 for a summary of this and competing views). Researchers have pointed out that knowledge is not ?magically reusable? but rather it serves as a boundary object that is decontextualized and recontextualized in order to serve the needs of both the creator and the reader but lacking the full context of either. To properly serve the reader of the memory, the creator must properly project the consequences of the memory?s later use, which can be a difficult process, but is commonly performed in everyday situations (Ackerman & Halverson Forthcoming). For this reason, designers of organizational memory systems have recommended including both document and social context for messages that will be reused (Atwood 2002).

The description of the Faq by its creator (see above) clearly treats knowledge as an object that can be removed from its historical and social context (e.g, he even recommends providing answers but not questions on the Faq). The Mailing List, on the other hand, supports a dialogue between participants that often captures these contextual clues. Despite the shortcomings of the Mailing List archive of all email messages (see Brewer2000 for a description of problems in a similar context), it is likely more useful than even a populated Faq would be because it includes the historical and social context of advice and an indication of what situations it may or may not apply to.

References

Ackerman, M.S. and Halverson, C. (Forthcoming) "Organizational Memory: Processes, Boundary Objects, and Trajectories." Computer Supported Cooperative Work: The Journal of Collaborative Computing.

Atwood, M.E. (2002) ?Organizational Memory Systems: Challenges for Information Technology.? Proceedings of the 35th Hawaii International Conference on System Sciences 2002.

Brewer, R.S. (2000) ?Improving Problem-Oriented Mailing List Archives with MCS.? Proceedings of the 2000 International Conference on Software Engineering, Limerick, Ireland, June, 2000.

Karau, S. & Williams, K. (1993) ?Social loafing: A meta-analytic review and theoretical integration. Journal of Personality and Social Psychology, 65(4), 681-706.

Markus, L. (1987) ?Towards a ?critical mass? theory of interactive media: Universal access, independence, and diffusion.? Communication Research, 14, pp. 491-511.

Preece, J. and Maloney-Krichmar, D. (2003) ?Online Communities.? In J. Jacko and A. Sears, A. (Eds.) Handbook of Human-Computer Interaction, Lawrence Erlbaum Associates Inc. Publishers. Mahwah: NJ. 596-620.

Wasko, M. M. and Faraj, S. (2000). ??It is What One Does?: why people participate and help others in electronic communities of practice. Journal of Strategic Information Systems, 9, 155-173.

Wenger, Etienne (1998) Communities of Practice.