Pierre Dillenbourg & David Traum (2006)
Sharing solutions: Persistence and Grounding in Multimodal Collaborative Problem Solving
The Journal of the Learning Sciences, 15(1), 121-151
Questions
P122
‘What are the cognitive effects of specific types of interaction?’
‘Under what conditions do these interactions appear?’
Emphasis on
· differences in scale( how many, how long)
· complexities of what is being shared.
Grounding ( originally a term used in psycholinguistics for communication between pairs)
‘When community members interact over months and years they develop a specific culture ( relevance to group character). This culture is what common ground is to the pair. Cultures share not only concepts but a system of values, a frame for interpreting situations, a set of stories, and a history ( KRO also a way of communicating) .
Software for collaboration (***)
P122-123 “If collaborative learning is a side effect of the process of building shared understanding then CSCL should investigate how software contributes to build shared understanding. On obvious answer is that building a common visual representation (textual or graphical) of the problem at hand contributes to the construction of shared understanding’. WSIWIS ( what you see is what I see). N.b. no individual will share exactly the same understanding of this space as another.
Collaboration as a process
Communication
Diagnosis ( leads to feedback)
Feedback (acknowledgement, repair, KRO ?? support and can only take place in the forum)
‘Misunderstanding has a different epistemic value in research on efficient communication and research in collaborative learning’ For the former it is communication break down ( KRO the social aspect of group collaboration) and in the latter a learning opportunity.
To repair misunderstandings about knowledge learners have to engage in construction activities , i.e. extra effort. ( KRO and therefore the need for a forum)
Grounding behaviour varies with the media involved. There are media- related constraints
Clark & Brenman (1991) | Dillenbourg & Traum (2006) |
Co-presence | |
Visibility object is present | |
Audibility have to elicit information about an object as it is not visible | |
cotemporality | |
simultaneity | |
sequentiality | |
reviewability | persistence |
revisability | Mutual reveisability |
Research Question
‘What is the complementarity between a whiteboard and a chat interface in constructing shared understanding?’
Hypothesis ‘W/B would be subordinated to the chat interface and that the role of the WB would be to support the grounding of the textual interactions in the MOO’ note W/B allows for deitic gestures and therefore reduces the distance of what is being represented. That the whiteboard would enable pairs to draw schemata that carry information that is difficult to carry through verbal expression’
Participants
20 pairs of post grad psychology students. Different l amounts of MOO expertise and most had no previous experience of working with each other.
Chat interface, a MOO ( a text based virtual environment, users represented by an avatar that is able to perform actions in space , including leaving and entering). Language based, sequential)
W/B graphically and spatially orientated, colour
Cost of interaction higher in both of these environments than in f-to-f.
Task – solve a murder
Crime scene - Hotel , spatial map in the MOO environment . 11 people, 1 victim several objects.
Role of grounding – what variables were measured
· Acknowledgement rate
· Medium used ( chat or W/B)
· Content categories of acknowledgements
Task knowledge
Management
Meta communication
Technical
· Measures of redundancy ( KRO good measure to think about) in terms of learning it is a measure of inefficiency.
Findings
Interaction more likely to be acknowledge if it carries an emotional load
For task knowledge there is more acknowledgement for inferences than for facts
Hypothesis about the role of the w/b was not supported . Although some pairs drew timelines, maps graphs (SNA type) W/B was mainly used for organising information – the facts and inferences and representing the state of the problem.
Conclusions
The dialogues were instrumental for grounding whiteboard information
Due to persistency of display information on W/b acted as a trigger
W/B does not ground utterances, it grounds the discussion