Synthesis
Collaboration is not simply a treatment which has positive effects on
participants. Collaboration is a social structure in which two or more people
interact with each other and, in some circumstances, some types of interaction
occur that have a positive effect. The conclusion of this chapter could
therefore be that we should stop using the word 'collaboration' in general and
start referring only to precise categories of interactions. The work of Webb,
reported above, showed that even categories such as 'explanation' are too large
to be related to learning outcomes. We have to study and understand the
mechanisms of negotiation to a much greater depth than we have so far.
We do not claim that conversational processes are exclusive candidates for
explaining the effects observed. The 'mere presence' of a partner can, in itself,
be responsible for individual progress. Neither should we discard the role of
non-verbal communication in collaboration. However, verbal interactions
probably provide, at present, more tractable ways in which to tackle the
development of computational models of collaborative learning.
In various areas of cognitive science psychologists and computer scientists
have developed computational models together. This is not the case for
collaborative learning. We hope that this chapter will help psychologists and
researchers in machine learning to develop models of collaborative learning.
Both in psychology and in computer science, individual learning and verbal
interactions have been studied separately. The challenge is to build a model
for how the two interrelate, for how dialogue is used as a means for carrying
22
out joint problem-solving and how engaging in various interactions may
change the beliefs of the agents involved.