Video demonstration of experiment into computer-mediated communication, deindividuation and group decision-making

The BBC filmed a simulation of our early experiment (Lea & Spears, 1991) into group computer-mediated communication for a programme called “Computers in Conversation” which was first broadcast on BBC2 in 1995.

The programme, which discussed the emerging potential of a relatively cueless text-based computer medium for social interaction and group collaboration, was part of an Open University series on Information Technology and Society. Although the study was conducted nearly three decades ago, the analysis and theoretical model that we used to understand text-based mediated communication are still relevant today. Electronic mail is more widely used than ever, and new forms of text-based communication platforms that were not then available, such as  SMS messaging (1994) and Twitter (2006), are now ubiquitous.

I was as an academic consultant and script writer on the programme and some of the students on my third year course in computer-mediated communication acted as participants in the experiment. My original recording of the programme was accidentally thrown, but I've found a low resolution clip of the experiment itself (6:26).



“Computers in Conversation” filmed by the BBC for the Open University course THD204 “Information Technology and Society” First broadcast BBC2: 08:50, 15/5/95.

Lea, M. & Spears, R. (1991). Computer-mediated communication, deindividuation, and group decision-making. International Journal of Man-Machine Studies, 34, 283-301 (Special issue on ‘Computer supported cooperative work and groupware'). Link to Abstract and Full text.


Narrator:         Ever since the introduction of the telephone, sociologists and psychologists have tried to predict how technologically mediated communications might differ from the way we communicate with each other face-to-face. What happens to our feelings of empathy and identification with each other? Can groups of people negotiate satisfactorily and come to acceptable compromises? Will their decisions be more or less cautious? More or less likely to involve the taking of risks? For anyone trying to manage the introduction of this technology into an organization, or for systems and software designers, reliable answers to questions like these would be invaluable.

Martin Lea:     A very influential view of mediated communication is the social presence model. Social presence is defined as being like intimacy, immediacy. It's a warm environment. It's a sense of being there, so face-to-face communication is considered to be ideal in that you have this full sense of being there with the other person. Communication media can be arranged along a continuum where media that are low end in social presence such as letter writing, fax, electronic mail, text-based media, would be that the bottom of this continuum. Telephone communication, for example, which includes audio cues would be then further up this continuum and video conferencing would be further still in that it has visual communication as well.

Narrator:         It may seem reasonable to take face-to-face communications as the ideal and then rank these various media according to how closely they can simulate it. And to expect that the greater the reduction in social cues, the more unsatisfactory, and possibly more extreme, the quality of the interactions that will follow. After all, even when you sit down face-to-face, the more complex, ambiguous or controversial a task or negotiation you attempt, the greater the chances of misunderstanding. But Martin Lea believes that the whole issue is much more complicated. Different technologies do, of course, have their own characteristics. But how much does the social context, those values, relationships and expectations that we bring along with us determine the ways in which we communicate? With some colleagues from the University of Amsterdam, he designed a series of experiments in which subjects are asked to discuss an issue about which opinions tend to be fairly strong, using a text-based conferencing system. Before the discussion, each participant fills in a questionnaire to give an indication of where they stand on the issue. And they do so again afterwards. Any movements in opinion are then measured and recorded. In this experiment, group identity has been deliberately emphasized, both in the way it was introduced to the volunteers and by the insistence on each being called a group member.

Though they may not speak directly to each other, they are all together in one location. But how might the results of their discussions differ if, for example, they're encouraged to think of themselves as individual participants? What happens if they are physically isolated from each other? Vary these conditions, but keep the technology the same, and it produced some interesting results.

Experimenter (ML):       Hi. Thanks for coming along. You'll be in this room here. Just take a seat over here…

Martin Lea:     What we found was that some of the results were counter intuitive in that we found most normative behavior, most decision making and attitude change in line with the prevailing group norms in a condition where the people are physically isolated from each other. It’s somewhat paradoxical that you can isolate people and have them behaving more as a group than you would if they were together in the same room. On the other hand, if we physically isolated the people but we encourage their sense of identity as unique individuals, rather than members of the group, then we got least normative behavior. Their attitudes, their decision making moved away from the group norm.

Narrator:         Are these results really surprising? It might be much easier, and therefore very tempting, to hope that by analyzing the characteristics of a particular technology, we will be able to predict its effect. But if these sorts of experiments have anything to teach us, it's that people make things much more complicated. The mental and emotional paraphernalia that we bring to any transaction is at least as important to its outcome as the tools that we use. Perhaps the real surprise is the discovery that such a seemingly cold and cueless medium can support such a wide range of different reactions and interactions.


If you found this article useful, please share so others can find it.