The Edu-Blogger: ITI: Simulated learning companions

Rick West reported on a session at ITI which look at some aspects of the impact of learning companion (LC) on learning outcome. As the author is focussed on a solo* learning environment and trying to use AI to support the LC, it seems to me that (LC) is an emotion reflecting chatbot. Extending such AI environment into a full semantic-understanding environment is still a rough implmentation assignment.

The solution which we implement in Fablusi online role play is quite different. Since we do not focus on solo learning, we suggest a group of players to play a single role. In this way, the players are each other's LC and I do not need to implement any AI to support this. What I need is to provide a private space where the group members can share their feelings, plan their strategy and so on. One of the important elements is the emphasis on "positive inter-dependence" of the group. Obviously, if the group is large enough, we may be able to characteristic social roles emerged from the group. At this point, I have no data to support such development as most of the role play simulation I have watched are played by small groups (up to 4 players per group) and the groups were dismantled/not followed up after the brief 3-week simulation cycles.

This may be an area of research worth putting some effort in.

*Solo learning is a term I used to describe the type of learning when learner is interacting with the material alone. The other side of solo learning is group learning where a group of learners work together to achieve some learning outcomes - not necessary the same for everyone!