Marshal B Anderson, M.Ed. Workshop 2 Co-operative Project, Spring 1999

In exploring the similarities and differences between the groups it seemed, in this case at least, that the real world group reached a mature state more quickly than the on-line group. This gives rise to several lines of thought and, possibly, further investigation. In the first instance we might ask ourselves whether we actually want this stability. It may be that certain types of collaborative groups, whether on-line or off, actually require, or at least could benefit from, a level of instability. Indeed, the very nature of some activities will prevent the onset of stability if the group task is in a particularly creative area and is actively encouraging conflict for the creativity it brings - and vice-versa. Hoffman (1978 page 85) suggests that conflicts can be conducive to creative problem solving.

However the general rule that the development of cohesiveness is important to most groups, and the groups under consideration in particular, is put forward by Rohfeld and Hiemstra (1995). Of the five main facilitative approaches to 'overcome the challenges of the electronic classroom' they suggest we should:

"… build group rapport and cohesion, introduce participants to each other, match them with partners, and assign group projects."

Both of these suggestions seem fundamental to group-making and developing as well as the educational issues they address. If stable group-making is an issue within a given activity then there are a whole range of exercises available to the real-world group-maker to accelerate the process of cohesion development. These usually take the form of 'games' and they are well established. The following recommendation was put to the administrator of the on-line group:

That there could be advantages in exploring real world solutions and methods of group making and investigating a) their basic principals and b) ways of translating then into on-line activities.

This is a huge undertaking and certainly not within the scope of this study, but the point has been raised with the establishment and they have added it to the review procedure for all their on-line courses. They were directed to the work of Blanchette and Kanuka (1998) who's 'Icebreakers for Distance Education' web page shows examples of group-building activities, most of which appear to be derived at least from the principles of real-work group games.

What is also worth noting is that fact that the results of the data from the on-line group seem to contradict their feeling about the maturity of the group. Even though they perceived a slightly lower level of maturity than the real world group, the data seems to suggest that they are much less developed, certainly in terms of inclusion. It may well be the case that they felt the group was as developed as they expected it to be.

It is noted above that the on-line group especially generated a lot of 'stars'. For a group with so many isolates the identification and use of stars as leaders in the context of group making might seem attractive. However, this does also seem to fly in the face of the ideal behind collaborative learning with its emphasis on equality. There is no doubt that all groups, even pairs, tend to give rise to a leader;

"The emergence of a leadership hierarchy seems to be a universal feature of human groups." Argyle M (1969 page 230)

If it is (nearly) impossible to have a group that has no leader then the next best thing might be to have groups in which leadership is fluid - i.e. where there are changes of leader. Jaques (1984, page 11) identifies group size as one of the major factors in producing fluid leadership:

Jaques is looking here at real world groups, but his table not only gives us a clue to the solution to diminishing the effect of leadership, it also suggests a reason for so many isolates appearing in our on-line group. It's not clear whether or not sub-groups spontaneously form in a large on-line group, and this might form another study, but the stereotyping and anonymity seem to make sense in terms of the on-line environment.

This is re-enforced when we examine further the data for anti-choices in the on-line group (see appendix A). We can see four specific 'rejectees' (4, 6, 8, 27), students who, ignoring the 'don't mind' scores, are chosen by no one to work with and by several other students as people they specifically don't want to work with. There are only two (30, 25) students who have been chosen by several and rejected by none. The remaining ten chosen students are also rejected by other students.

This leads us to the conclusion that only a total of 16 out of thirty students had a strong enough identity within the group for others to have specific feelings about them, for or against.

Having said that, there is quite a high level of 'don't mind' scores; i.e. students seem to have had some idea of the identity of others, but not strong enough to produce the more definite choices we see in the real world group.

In a crowded conference with many threads being generated and replied to by 30 students it would seem difficult for any given student to have a sense of the group as a whole. When we look at the means of identifying others we see, as mentioned elsewhere in this study, that many of these are removed by the nature of asynchronous conferencing. If asked to bring members of a group to mind the visual memory can not be invoked, nor the aural; this might make it difficult to have a sense of who is contributing what to a conference. If this were the case, then we might expect only a few group members to embed themselves specifically in the memory of the whole group. We might speculate that this would happen through several mechanisms:

The above seems to suggest that a major factor in the formation of coherent groups and sub-groups is that of identity. Alexander (1991 pages 203-207) describes on-line identity as 'telepresence' defined as 'creating the illusion of the presence, and even closeness, of other people who are, nevertheless, distant in space and time, and his suggestions about how this can be achieved are included in the recommendations below. The idea that adding information beyond plain text is also suggested by Sproull and Kiesler (1991);

"The addition of sound and pictures to computer-based communication will dramatically increase social information in computer based communication." (page 53)

The following recommendations were made to the administrators of the on-line group to increase a feeling of identity:

The use of visual cues to who is contributing should be integrated into the conferencing environment. A fairly simple way of doing this would be for students to choose a particular font and colour for their postings and stick to it - bearing in mind readability for the rest of the group. More complex, but worthwhile, would be the design of graphic elements in postings or, preferably, as an integral part of the software. Small graphic elements (like monographs) could easily be designed and attached to postings, personal photographs might be made available by links from postings or as part of the conferencing system so that they automatically appear in each posting. Alexander (see above) suggest a compromise here by using a library of faces from which students choose an identity. This has attractions from the point of view of bandwidth and the repeated storage of graphics on the local and server database and would be a reasonable temporary solution.

That group tutor/facilitators bear in mind the need for increasing identity within the group. That they post entries as personal replies, using names to refer to other postings.

It is suggested above that students, to one level or another, use the raw data they see in the size of the postings of others as a cue to status, prestige and the desirability (or, more specifically, the undesirability) of other students as co-workers. It is not immediately clear how this information might be useful to the administrators of an on-line course. However, this information is not useless and it seems to tie in with the above comments on group maturity - i.e. the group sees itself, largely, as mature but the evidence suggests that some decisions made by members of the group are still based on some fairly simplistic conceptions about other members. We must, of course, accept that in some cases these conceptions will be correct, but the anti-choices scattergram really seems to suggest a set of reasoning that is superficial but still effective even after nearly a year of contact.

The following action was suggested, which is substantially an extension of that given in the section on stability:

That it shouldn't be assumed that, having gone through a series of group-making exercises, that a group is made. There seems to be some argument at least for building into the later stages of the course a continued strategy for group development. More complex exercises like those presented by Jaques (1984 pages 234-285) hold some promise of continued group development beyond the initial 'getting to know you' activities suggested above.

In connection with the caveat about conflict and creativity above, we note Hoffman (1975 page 85) cites his own research into;

"…encouraging low-status members, via the wording of their instructions, to oppose the suggestion of a high status member o change their work method. The resulting conflict produced more creative solutions than did the comparable, low conflictful solutions."

This might suggest there these conceptions about status and prestige could actually be harnessed as a method of increasing the overall creativity of the group. While that is not the immediate goal of the institution in this case, it was accepted as a possible tool for group development.