This paper considers some of the issues and options for formal online discussions in general with a specific focus on moderated discussions as exemplified at IFETS.
The IFETS discussion model is, I believe, a powerful one which has been making an important contribution to academic debate and knowledge sharing online since mid-1998. Although there has been intermittent informal discussion about the process itself, I don't believe it has come up for formal discussion over the last few years. Therefore, given the pace of developments in the nature of online communities and the technology which supports them, I think it timely to begin a conversation in IFETS about the process of formal online discussion.
I propose five areas for discussion, the first three of which are related to the discussion space whereas the last two are essentially guidelines for their conduct and recording.
The model currently used on IFETS and ITForum is that of a moderated e-mail list using simple text (no formatting, HTML or colour/shading possible) and messages are stored in list archives.
The use of simple text email messages tends to limit the length of each contribution. Long texts in this format make difficult reading in themselves and it can be difficult to perceive the context of particular contributions in ongoing discussion.
Another consequence is that it is difficult for discussion participants to weave various contributions into their message so as to replicate the kind of multi-participant discussion which can occur in a face to face round-table discussion for example. Those who have worked with collaborative workspaces like FirstClass are likely to have experience of the value of this type of visual weaving operating in text-message-based discussion.
Traditionally these limitations have been worked around by the discussion moderator attempting to set out the issues involved in discrete chunks or topics, requesting participants to use the subject header to indicate which of these they are addressing and hoping that others will likewise respect the subject header convention so as to provide a clear and easily followed thread of messages. It also falls to the moderator to attempt to pull the various separate contributions together at intervals throughout a formal discussion so as to make it easier to see the development and flow of discussion around each of the issues being addressed.
Another arguable drawback of unformatted text messages is that they are aesthetically rather basic when compared with webpages, blogs or journals.
In addition to the limitations of simple text messages outlined above, we also need to consider the problems arising out of the use of email itself as a discussion medium – spam, filters, volume and unpredictable taxonomy of messages. Many people may find themselves in agreement with Eric Hahn’s assessment at the 2004 Inbox email technology conference that “e-mail is broken” (Asarvala 2004). He reminds us that the file/folder metaphor was designed almost 30 years ago when it was anticipated that users would receive around 5 e-mail messages per day.
Is it time, then, to reconsider the options for our formal online discussions?
I would propose the following as a possible list of general features for the system of choice:
Facilitate in-depth discussion
Portability (no vendor lock-in)
No client installation (to avoid large download; problems with institutional IT staff)
Aesthetically pleasing interface
Given the limitations of basic email-based systems, what are the other options which might be considered? One possibility would be to use a collaborative workspace (e.g.Webcrossing, Groove). An alternative would be to move to some kind of web-based system.
The relative merits of Email and Collaborative Workspaces have been cogently summarised by Michael Sampson (2004) as follows:
“Email. Strengths are universal addressability, universal interoperability, and cross-organization functionality, offset by weaknesses of a disconnect between its design and current usage pattern, a deluge of spam, and a lack of immediacy, among others.”
“Collaborative Workspaces. Strengths are the provision of a set of structured tools for group work, enforced group memory, and shared content taxonomy, with weaknesses of poor cross-platform support, lack of cross-product interoperability, and poor cross-enterprise capabilities.”
These systems allow for more a organised and visually pleasing discussion space and provide good archiving. The drawbacks include the fact that they usually involve each participant having to install client software and that one is usually locked into a particular software environment (commercial or open source) which may be difficult to integrate with other applications or to migrate from at a future date. Technical support can also be an issue.
Groove is an example of a commercial workspace which has been exciting interest recently although Martin Terre Blanche’s recent comments (2004) illustrate possible drawbacks of its use for our purposes.
It is also possible that the open-source Sakai consortium may come up with interesting developments in this field in the near future.
A general objection to moving from email to web-based systems has to do with the fact that the former is a push whereas the latter is a pull process and experience suggests that push results in more participation. So any web-based solution would need to incorporate a push component to notify participants of new comments and contributions.
Such a system could involve some sort of web-based document + comments approach whereby the context of each contribution can be seen linked to the original pre-discussion document and in a thread with others addressing the same point. This was suggested by Ben Hyde and maintained as a sort of informal parallel track during the IFETS discussion on Participation in Online Learning earlier this year (30 Jan – 13 Feb 2003); the tool used in this case was d3E Ubiquitous which has been used by JIME and others to promote academic debate. Although most of the participants in our discussion preferred to stick with the simple e-mail model there were a number who opted to use the d3e tool and the result can be seen at http://ud3e.open.ac.uk/d3e_discussion.php?url=ifets.ieee.org%2Fdiscussions%2Fdiscuss_january2004.html&f=762
In this particular case the push element was provided by Ben Hyde forwarding a digest of the web-based comments to the e-mail list.
Commkit is an example of a commercial social software package which integrates web and e-mail discussion and allows community members to choose on a topic by topic basis whether to participate by web or e-mail.
Using either d3E or Commkit we continue to have to deal with the inherent flaws of e-mail. This could be resolved by using RSS rather than e-mail to notify participants of new contributions and give a brief taster of their contents. Participants would need to install an RSS aggregator on their desktop to receive this or it could be received within MS Outlook but separate from e-mail messages (Newsgator, for example, is a commercial application which runs in MS Outlook). As the potential of RSS for this and other functions becomes apparent it is likely that its applications will be considerably expanded in the open-source sphere and this could mean that web-based systems become more attractive in the near future.
A third, and less radical, option would be to continue with the message-list approach but use a more sophisticated e-mail based tool which would allow formatting and hyperlinking. The expansion of Yahoo Groups and arrival of Google Groups suggest that there may still be potential in this model. It is a question of balancing the simplicity and universality of this medium against its inherent limitations.
Incidentally, the University of Strathclyde has been working on an interesting method of visually representing the growth of discussions (Mohamed et al, 2004) . This was originally designed to work with bulletin board discussion and could conceivably be adapted for any one of the hosting options we have been considering here.
messages stored chronologically
threading of messages
view formal discussion contributions in context of pre-discussion document
various viewing options (e.g. tree-view, titles, first paragraph)
hide quoted text option (this feature and the previous one are found in the beta version of the new Google Groups e.g. http://groups-beta.google.com/group/soc.org.nonprofit )
Decisions regarding this aspect, and the following one, are very much dependent on the outcome of any decision regarding changes in hosting. I do think it worth mentioning that I have very rarely found mention of earlier IFETS contributions in the current discussions and, knowing the quality of earlier work, I would conclude that this is a result of access and consulting issues in our archiving system. The present IFETs archives (http://ifets.ieee.org/archive.html ) are not particularly user-friendly in that it is not very easy to search for particular messages or topics. The interface employed for the IT Forum archives appears to be more flexible and accessible (http://www.listserv.uga.edu/cgi-bin/wa?A0=itforum ).
The main problems are those inherent in email discussions i.e. list-enabled spamming. One measure employed to avoid inappropriate behaviour is to get list subscribers/participants sign a code of behaviour before admission. This is the policy at e-democracy.org (2004) for example . A rather charming example of a list charter is that of the rabbit lovers list, NetBunny (Nevel 1998) .
The international nature of groups such as this one does complicate things and I suspect that signing up to a code of behaviour may not help in curbing inappropriate use of addresses harvested from the list. The first reaction to this kind of behaviour will often be a private cease-and-desist message from the group moderator and if this does not work the next recourse may be contacting the perpetrator’s ISP. However if the person is based outside N. America or W. Europe this is unlikely to achieve very much and even in these areas the negative activity may not be clearly illegal (e.g. mailing unsolicited CVs or marketing of commercial products).
Spam posts to the list itself have not represented a problem as such at IFETS because all messages are monitored before publication to avoid this. This does involve, however, a rather thankless investment of time by the list coordinators as well as introducing delays in posting and ideally it would be better if the system was self regulating in this respect.
Although I am not aware of any formal guidelines for moderators here on IFETS, my observation of formal discussions over the last four years has led me to conclude that most moderators take a role somewhere between what we might call the moderator-participants model and the moderated participant-participant model. In the former the moderator, having been responsible for setting out the pre-discussion paper, tends to respond to each contribution as it comes in and is relatively centre-stage (much like role of the teacher in more traditional classrooms). In the latter the moderator keeps a low profile once the discussion is underway and concentrates on summarising contributions (more of a facilitation role).
I know where my own preferences lie on this and would be interested to hear comments or observations from others.
Within IFETS it has become (informal) practice to mention contributors to discussion by name in the moderator’s summary but there may be a case for more specific guidelines on this. My late colleague Sally Mavor (2002) set out the issues rather well in a contribution to a discussion here in August 2002 which is worth quoting in full:
“In the IFETS discussion paper, the moderator and summariser(s) are the lead authors of the paper, but the final published paper integrates the thoughts and contributions of those who participated in the discussion. In that sense, is the authorship of the paper still solely that of the moderator/summariser(s)? Or does the paper become a collaborative effort with joint authorship? Or is the paper authored by the same moderator/summarisers with clear citations of the contributions made by members?
I tend towards the latter – especially considering the amount of work involve in writing the pre-discussion paper, in moderating the discussion and in integrating the contributions of the members in a final document.
However, at the moment IFETS doesn’t appear to have a clear policy regarding how members’ contributions should be recognised/integrated in the final, published discussion paper. (I have looked but am prepared to stand corrected, if I have missed anything!) Having also looked briefly at past discussion papers, there seem to be a wide range of approaches to the final paper from no mention of discussion contributions, to meticulous naming and citing in a post-discussion summary.
The reason I bring this to light is that in one of the last discussions, I, among others, contributed to the discussion. As it happens, I coined the term ‘contextualised flexibility’ to account for the kind of understanding of context and appropriate response to that context which is important for course design. The author of the paper kindly mentioned this phrase, saying that “One contributor used the term ‘contextualised flexibility’ to describe how overall course design should be sensitive to contexts and adaptive”
Now, with my tongue firmly and friendlily- in-cheek, if I wish to use that phrase in another paper, how should I refer to it? It has now been published in the Education, Technology and Society (ETS) journal. So should I say…
“contextualised flexibility” (a contributor, cited in ETS discussion paper no. vol. etc…….. who happened to be me)? ;-)
And what about the other contributors, who in fact, contributed much more than I did and were not mentioned by name? This paper is by no means the only one that uses this approach, so it is quite legitimately following precedents and is even integrating the discussion contributions more than some previous papers. Obviously the summarisers must be also be selective and concise in their summaries, but, my question is: when the moderator/summariser(s) use the work of members’ contributions to the discussion should the names of these contributors not be mentioned in the final, published paper?
On no account would I advocate a homogenisation of approach, but I do feel that some guidelines as to appropriate ways to integrate members´ contribution to the discussion would be helpful and consistent with the collaborative nature of the discussion paper.”
I have tried to set out above what I think are the main areas where it could be useful to share our experience and thoughts on the process of formal online discussions as exemplified at IFETS and will conclude by saying that I look forward to seeing your contributions so that we can clarify our ideas and perhaps achieve some consensus regarding the following four questions:
Asaravala (2004) The changing face of e-mail, Wired News, from http://www.wired.com/news/technology/0,1282,63692,00.html
E-democracy (2004) Forum rules – Interim retrieved 1 September 2004 from http://www.e-democracy.org/center/draftrules.html
Mohamed, R., Ferguson, J. D. , Mac Cormick, A. and Elsweiler, D., Developing Bulletin Board Visualisations in Proceedings of the IADIS International Conference on Web Based Communities 2004, pp. 11-18, retrieved 1 September 2004 from http://www.cis.strath.ac.uk/~rehman/mowbc2004.pdf
Mavor (2002) Discussion paper citation practices, retrieved 1 September 2004 from http://ifets.ieee.org/past_archives/archiv_080201_270203/0995.html
Nevel, (1998) PetBunny Charter retrieved 1 September 2004 from http://www.petbunny.net/petbunny/charter.htm
Sampson (2004) Collaboration Software Clients: Email, IM, Presence, RSS & Collaborative Workspaces Should Be Integrated for Business Communication, retrieved 1 September 2004 from http://sharedspaces.typepad.com/blog/2004/08/shared_spaces_b_13.html
Terre Blanche (2004), Goodbye Groove retrieved 1 September 2004 from http://www.criticalmethods.org/collab/2004/8/news.htm#1091431419980