Education-line Home Page
Papers from the 28th Annual SCUTREA Conference
Research, Teaching and Learning: making connections in the education of adults

Learning in virtual space: potential and pitfalls in electronic communication

Mike Davis & Kate Denning, Centre for Adult and Higher Education, University of Manchester, UK


The purpose of this paper is to explore aspects of the use of computer-mediated-communication (CMC) during the teaching of a postgraduate course offered as part of two programmes: two modules in the M.Ed. in Training and Development offered by the Centre for Adult Education at the University of Manchester; and a taught unit in the Masters/Doctoral Programme in Adult Education and Human Resource Development in the Department of Adult Education, University of Georgia (UGA), USA. This collaboration had its origins in earlier exchange arrangements between the two institutions. Early attempts to establish electronic links between members of these two institutions are reported in Davis, (1997) and Davis & Holt (1998).

The course, called Reflective Technologies in Work-based Learning, comprised 15 participants. In addition to the two course tutors, there was another T-group trainer and a technical facilitator. One of the participants was a researcher working on the project. The face-to-face week was divided into two main elements: a T-group laboratory (see, for example, Miller, 1989), and an action science laboratory, during which latter time participants were introduced to the theory and practice of action science and underwent their first experience of presenting a case study. In addition, technical training and support was provided by the technical facilitator in order to familiarize the students with the software that was going to be used in the computer mediated component.

Action science works through the interrogation of case studies to discover where the gaps and inconsistencies lie in conversations: between what we think and what we actually say, between espoused theories and theories-in-use (Argyris, Putnam & Smith, 1985) A case study is a transcript (as near as can be remembered) of a conversation which is felt represents a problem in communication. The context is given then the page is divided in half. In the right hand column the actual words used are written along with any non verbal cues which seem relevant. The left hand column contains unspoken thoughts and feelings (see, for example, Brooks & Watkins, 1994)


The computer mediated component of the course, which is the focus of this paper, was conducted over a five week period, using the groups that had been formed in the initial face-to-face meetings . Each student took it in turns to present their second action science case and have it interrogated during the period Monday to Friday of each week. An hour each Saturday was dedicated to a synchronous conference, conducted in real time. In all, fifteen case studies were presented over the five week period and this data are currently undergoing analysis using the qualitative software package, NUD*IST and will be the subject of further papers. The focus of this paper is the extent of student interaction and the possible impact of the tutors' roles.

Behind the screen

The most significant potential area of failure for any virtual environment is non-attendance, or put less strongly, the lack of productive participation, by which we mean reading but not writing. By making participation a compulsory element of the course for the purpose of gaining the qualification, this problem could, to a certain extent, be circumnavigated. However, this may not be sufficient. Riel and Levin (1990) give a list of what they consider necessary to get a good, successful network to function effectively: to enable participants to recognise a presence behind their computer screen. Their criteria were as follows:

The Reflective Technologies cohort met these, to a greater or letter extent, as follows:

What is interesting from our point of view is that we had three sub-groups on line who fulfilled the criteria 2 to 5 to the same degree, and yet they were very different in their levels of participation and 'atmosphere'. This might suggest at first glance that the last two functions are particularly important.

From an analysis of the data it is possible to speculate about some of the reasons for the relative success and failure of two of the groups and it might be possible to draw some conclusions as to how to manage groups more effectively. Before we do that, however, we need to examine some of the characteristics of success and failure.

Differences in the groups

Our first analyses were mainly quantitative and at the very least give some insight into the level of activity in the three groups though not its quality. For the purposes of this paper, we are looking at Groups 1 and 3, the most and least successful, respectively.

Members of Group 1 interacted with the server on 6866 occasions, compared to Group 3 who managed only 2500. Individuals' interactions ranged from 949 - 1715 in Group 1 and 205 - 893 in Group 3. The mean for the groups was 1373 and 500 respectively. What these numbers describe is participants' engagement with the software through interaction with the server that takes them around the various screens in an on-line case study. While this is not necessarily a passive activity, it does not demonstrate the more active role of contributing to the discussion through writing an intervention. Examination of the number of interventions also revealed that Group 1 was the more active, with 497 entries, compared to 166 from Group 3. Even the least active Group 1 participant at 68 made more entries than the most active from Group 3 (47). Means were 99 and 32 respectively.

As is evident from these figures, there is a substantial difference between the number of times participants addressed the server and the number of interventions they made. Interestingly, however, the ratio of interactions to interventions is very similar across the conference as a whole, ranging from an average in Group 1 of 14:1 to that in Group 3 of 15:1.

Group 1 individually and collectively interacted most frequently with the server. In both Groups 2 and 3, however, there were participants who were more active than the norm for their group. There is a marked difference between the groups in the way in which they functioned, however. Not only did Group 1 interact more frequently, but there was a shared culture of a relationship between interaction and intervention not demonstrated by the other two groups, something which we will examine later.

This form of quantitative analysis does, of course, reveal little or nothing about the richness of activity in the groups, but as we will see, there is a relationship. As well as the tentative conclusions we can draw from this quantitative analysis it is also possible to consider how differently the groups experienced the groups by a brief look at a few interventions which show that a virtual classroom has an atmosphere of its own despite the lack of physical presence.

Jagruti (Group 3): Could we have a synchronous session this week? I for one would like to have more interaction w/in the group. To me, our group feels somewhat lifeless.

Marianna (Group 3):Karen, this IS a great exchange. [...] Do you mind if we continue and then come back to this?

Marlene (Group 1): I've got to say that I have and I'm enjoying working with you. You are a very reflective group. And I will miss this. Today I was thinking that during this five weeks I have felt helpful, happy, useless, frustrated, annoyed and I have laughed. I have been very excited and engaged with the task. In my undergraduate career I have never been through so many emotions as I have been on-line.

And this piece of dialogue:

Marlene (Group 1): Be careful not to get focused on the other person in the case.

Molly: Hi, Marlene! I'm not sure what you mean by focusing on the other person. Do you mean Jeannie, or X (in the case) and why are you suggesting this? I'd really appreciate it to know your views on this. It took my an hour to put all this together and I don't know if the result is what I intended.

In speculating about the differences in levels of engagement and activity between the groups, it is possible to identify a number of characteristics that were features of Group 1 behaviour. Group 3 almost invariably demonstrated the opposite characteristics.


All three groups had strong, articulate members but not all were able to harness their differences in a productive way. Riel and Levin advise that there should be some form of leadership in CMC. One or more people need to take responsibility for monitoring and facilitating the group interactions. Perkins and Newman (1996) define such a person as a virtuoso, 'a highly skilled practitioner of e-discourse' (Perkins and Newman, 1996: p156), someone not normally set apart, but from within the group who 'serves as guide, gentle teacher and exemplar' (Perkins and Newman, 1996: p163). We can identify a virtuoso in Group 1 who flourished within the active environment of her group. One member of Group 3 similarly showed a similar disposition but the setting did not enable her to fulfill her potential.

As with face-to-face interaction, norms were set early in the conference. Group 1's problem-solving approach to the challenges of the medium resulted in agreed synchronous meetings, a lurk room which validated reading without commenting, and a chat room. The chat room became indicative of a whole atmosphere of involvement and concern for each others' lives and a knowledge of each other evidenced by this quote:

Moira: Ah, Bernard, you guessed that I am a morning person! This morning, however, I took my husband out to breakfast, so I am a little later than usual.

In contrast, the behaviour demonstrated by the first case writer for Group 3 may have contributed towards a norm of lower levels of participation. Indeed, he only made two interventions into his own case which clearly frustrated the rest of his group although their challenge was very mild, for example:

Jagruti: It is a little challenging to offer a ladder of inference without some feedback from you, Ritesh i

Following the lack of activity of the first week, Group 3 found it hard to pick up momentum. A slightly tougher challenge finally came in the fourth week.

Mayuri: The exchange among your three - Kate, Louise and Jagruti - caused me to wonder if Ritesh is/has been lurking during the case studies. I've wanted to know what you are doing since the middle of your case, Ritesh. If you are tuning in with us, as Kate mentioned, but just haven't had anything much to say, I'd just like to know. Are you there?

There was no response from Ritesh although he did make a few more interventions unrelated to this process initiative. In a f2f environment it is possible to know if someone has heard what you have said, although their interpretation cannot necessarily be predicted. One of the difficulties of the medium is that there is no way of knowing if he even read the challenge. The other difficulty is that sending messages to someone who is not responding is bound to fail. It may, at that point, become necessary to use other channels of communication such as telephone or email, to find out why the individual is inactive.

As in f2f there is probably an optimum group size for this kind of work. Group size for this course were set at 5 but with 1 or 2 members inactive to a greater or lesser extent, effective membership was reduced to 3 or 4 which may be insufficient to generate ideas and challenge complacency. In a group of this size, it is not safe to be dangerous.

In order to encourage participation it may be thought axiomatic that the tutor has a particular role to help facilitate the student processes. Tutor interventions can take many forms from purely social, to directly challenging, to providing content or knowledge. Because e-discourse is a written form of communication all activity is powerful. This is particularly true of comments made by tutors. Comments can be read many times over and interpretations changed and flippant remarks take on a more weighty aspect than intended, hence the perceived need for emoticons (visual symbols representing some non-verbal cues).

Although the tutors intervened approximately the same number of times into the two groups, the ratios differ because Group 3 contributed so much less.

The tutors made 31 interventions in Group 1 and 30 in Group 2, but the ratio of activity was so much greater because of the unwillingness of Group 3 to enter into the conference on their own behalf; and, we think it is true to say, because Group 1 actually needed us less. Thus the ratio for Group 1 was 17:1 and for Group 3 was 7:1.

For both groups this gives a mean of approximately 6 interventions from the tutors per case study. On the face of it then it would appear that the tutors were fair about the number of interventions they made. Further examination of the data, however reveals an inequity in the distribution of tutor interventions. This difference is particularly marked at the end of week 2 when the tutors had intervened 10 times into Group 1 and only 5 times into Group 3. It is possible that this could have given the group the feeling that the tutors were giving up on them. Silence is very powerful and Hardy (199x) suggests that silence from a tutor can be interpreted as lack of interest rather than non direction. This would have compounded the factors already mentioned such as incongruent personalities and early norms of low participation and hasty (but inappropriate and uncritical) acceptance of feedback.

The question remains about how much the tutors could or indeed should have intervened in the cases. Action science is a normative process and it could be argued that it needs to be modeled and explained. The extent to which the tutors did this, as well as demonstrated other behaviors, is examined in another paper (Davis, Watkins & Milton, 1998). That these two groups were different in composition is clear, but their biographical details and their performance and demeanor during the face to face laboratory do not obviously lead us to conclude that the less engaged group were less capable of doing the work, less emotionally or intellectually robust, or less committed to the course.

In a f2f environment to intervene more in a group which is struggling. It is hard to believe that more tutor involvement could have improved what was going on in Group 1, the situation with Group 3 is less clear. When an analysis of the transcripts showed that Mike Davis had not intervened at all into the second week of Group 3 he was shocked. The reason for his surprise was that he felt as if he had done a lot because he had visited their case study so often. Both tutors visited Group 3 more than Group 1 and both report that they were bothered at the time by the lack of activity in Group 3. These frequent visits gave the tutors an impression of active involvement which was not borne out by the printed texts. This is a potential danger and facilitators need to be aware that whilst they may feel involved when they are lurking, there is no evidence for the rest of the community.

It is hard to know what precisely the tutors could have done to encourage a higher degree of interaction. It is possible that they acceded to the early norms and were unable to counteract the effect of the atmospheres in the two groups. In the first week Group 1 was already dynamic and looking for ways to make the technology work for them. On entering the conference the overriding feeling was one of commitment and energy. Group 3 by contrast was far less exciting to enter and it was possible to go in and find nothing new had been added. To some extent, the energy level in Group 3 infected the tutors as much as the group and it would be hard to know what could have been added to facilitate the group more effectively.


The evidence from this analysis so far indicates that as with groups in face to face environments, some will be more active than others. What the evidence tells us, however, is that there are different kinds of activity: for example, that to lurk in cyberspace does not mean that you are not taking part. Indeed students commented favorably about the opportunity to enter a conference, read and then leave, only to return afterwards with there more reflective observations. This is not something that is available in the clamor of the face to face seminar. This also has something to tell us about the role of the tutor: there are no easy answers. Group 1 could be quite direct in resisting our interventions, while Group 3 could not be persuaded to act against what they saw as their best interests: in this case, by remaining silent.

The data from this course is very rich and we have only begun to investigate the more qualitative features. Nevertheless, we see that issues of participation are not necessarily what we would expect and that the approach to facilitation is not obvious. Perhaps the richer data will give us a clearer insight into the complexities of the virtual classroom and enable us to avoid the most serious pitfalls of electronic communication.


Argyris, C, Putnam, R & Smith R (1985) Action Science. San Francisco: Jossey-Bass, Inc

Berge, Z (1995) 'Facilitating computer conferencing: recommendations from the field' in Educational Technology, 35 (1) (1995) 22-30

Brooks, A & Watkins, K (Eds.) (1994) The Emerging Power of Action Inquiry Technologies. San Francisco: Jossey-Bass

Davis, M (1997) 'Fragmented by technologies: a community in cyberspace', Interpersonal Communication and Technology Journal, vol. 5 no 1/2 June, 1997

Davis, M & Holt, M (1998) 'having new ways to miss the point', Innovative Higher Education 22, 4, 311-327

Davis, M, Watkins, K & Milton, J (1998) 'Reflective Technologies: an analysis of teaching/learning practices in computer mediated communication'. Paper given at Networked Lifelong Learning Conference, University of Sheffield, April, 1998

Hardy, V. (199x) 'Introducing Computer-Mediated Communications into Participative Management Education: the impact on the tutor's role', ETTI, 29/4

Miller, N (1989): Personal Experience, Adult Learning and Social Research. University of Manchester: unpublished Ph.D. thesis

Perkins, J & Newman, K (1996) 'Two archetypes in e-discourse: lurkers and virtuosos' in International Journal of Educational Telecommunications 2 (2/3) 155-170

Riel, M & Levin, J (1980) 'Building electronic communities: success and failure in computer networking' in Instructional Science, 19, 145-169

i This is an example of an emoticon - a visual expression of a smile.

This document was added to the Education-line database 01 July 1998