Computer Supported Collaborative Learning CSCL assumptions l Learning

C(omputer) S(upported) C(ollaborative) L(earning) CSCL assumptions? l Learning is knowledge sharing l Learning is the co-construction of meaning l Learning is the negotiation of meaning “Leren is samen leren” Argumentative writing Educational Sciences Utrecht University 1

Utrecht Interactive Learning Group (ILG) CSCL: Focus on the process of collaboration in argumentation, writing and project-based learning l Use of (cognitive) tools: l Synchronous and a-synchronous communication l tools for planning, organizing and linearization l Text and graphics l Own environments with shared editor (TC 3 & VCRI) l Analyses of dialogues with MEPA (Erkens) l Organization: l 2 EU-projects (DUNES and SCALE: Andriessen) l 4 projects funded by Dutch NSF & 3 Ph. D-projects l 14 persons; senior researchers : Andriessen & Erkens l Educational Sciences Utrecht University 2

Utrecht Interactive Learning Group l l l l Marije van Amelsvoort Jerry Andriessen Carla van Boxtel Wouter van Diggelen Jannet van Drie Gijsbert Erkens Jos Jaspers l l l l Gellof Kanselaar Jos van der Linden Lisette Munneke Maarten Overdijk Maaike Prangsma Chiel van der Puil Arja Veerman (TNO) Educational Sciences Utrecht University 3

CSCL l Students give words to their thoughts l Students write down their thoughts l A history of the interaction (CS) l Monitoring by the teacher l Interaction paradigm l Collaboration and argumentation Educational Sciences Utrecht University 4

Central question l How do collaborating students manage to coordinate and adjust their actions to the processes of knowledge construction that occur between them? l Mutual coordination & control of interaction Educational Sciences Utrecht University 6

First phase in our research Discussion Forums Educational Sciences Utrecht University 7

Discussion Forums Educational Sciences Utrecht University 8

Students differ in amount of time Educational Sciences Utrecht University 9

Students differ in participation Educational Sciences Utrecht University 10

Second phase in our research Different tools for argumentation Educational Sciences Utrecht University 11

Research questions l How to provoke argumentation? l How to support cognitive processes of l evaluating information l multiple perspective taking l elaboration Educational Sciences Utrecht University 12

Learning environment in which we can. . . l Structure the interaction at the interface l labeling utterances l Resources : text or pictorial information l separate task and communication window l turn-taking control Educational Sciences Utrecht University 13

Electronic environments (1) Dialab (Moore, 1993) (2) Conference MOO (Jermann & Schneider, 1997) (3) CLARE (Wan & Johnson, 1994) (4) Belvédère (Suthers & Weiner, 1995) (5) CTP (Andriessen, Erkens, Overeem & Jaspers, 1996) Educational Sciences Utrecht University 14

Interface Dialab Educational Sciences Utrecht University 15

Conference MOO Educational Sciences Utrecht University 16

Interface Clare system Educational Sciences Utrecht University 17

Example “Belvédère” Educational Sciences Utrecht University 18

Belverdere system Educational Sciences Utrecht University 19

Collaborative Text Production CTP Educational Sciences Utrecht University 20

Conclusions (1) l Preconditions for fruitful argumentation: shared focus on goal l shared focus on dialogue l l Provoking argumentation l task - instruction - roles students/ tutor Educational Sciences Utrecht University 21

Conclusions (2) l Support for argumentation combination of sentence openers and free input of text l turn-taking in task window coordinates the collaborative problem-solving process l turn-taking in communication window inhibits the elaboration on arguments l use of diagrams: graphical organisation of arguments stimulates (multiple perspective taking x elaboration) l Educational Sciences Utrecht University 22

Third phase in our research Own environment TC 3 For argumentative writing Educational Sciences Utrecht University 27

How to continue. . . l structured versus unstructured interaction l graphic/ textual dialogue markers and sentence openers l combinations of educational design & structured interaction Educational Sciences Utrecht University 28

Main processes in dynamics of collaborative interaction The need to achieve a joint product in collaboration stimulates 3 processes : Activation of knowledge and skills 2. Grounding (creating a common frame of reference) 3. Negotiation or coming to agreement 1. Educational Sciences Utrecht University 29

1. Activation of knowledge and skills a. Initiating degree of participation 2. proposing topics / task strategy 1. b. Articulation explicating & verbalizing 2. organizing & structuring 1. c. Exchanging knowledge and information sharing information & resources l seeking & asking for information l stimulating task directness Educational Sciences Utrecht University 30

2. Grounding or creating a common frame of reference a. Tuning l b. Checking l c. checking exchanged info with existing knowledge Focusing l d. adapting to the partner’s level of understanding mutual control of focus and topic of discussion Co-construction l complementing knowledge from the partner maintenance of a ‘collective space’ Educational Sciences Utrecht University 31

3. Negotiation and coming to agreement a. Elaboration l b. explaining and accounts Argumentation discussion, persuasion & criticizing l comparing & evaluating l c. Coming to agreement l deciding & according ‘paradox of collaborative learning’ Educational Sciences Utrecht University 32

“But We Don’t Want To…” l l l . . . Be in disagreement. . . Argue with the tutor. . . Insult our friends. . . Loose an argument …Spend so much energy! Paradox: But we have to collaborate !!? ? Educational Sciences Utrecht University 33

COSAR project l Computer Supported Collaborative Argumentative Writing l Planning argumentative texts l generating l organizing l linearization l Students aged 16 -18; College Preparatory High school l Groupware: tools & communication Educational Sciences Utrecht University 34

Basic TC 3 environment Educational Sciences Utrecht University 35

Planning tools added Outliner Diagrammer Educational Sciences Utrecht University 36

TC 3 (Text Composer: Collaborative & Computer supported) Educational Sciences Utrecht University 37

Experiments with TC 3 l Control group: 49 pairs of students (16 -18 yr) with TC 3 basic environment l 6 experimental groups: 120 pairs with TC 3 environment and planning tools l compose argumentative texts on ‘organ donation’ & ‘cloning’ l log files of actions, chats, diagrams, text Educational Sciences Utrecht University 38

Research questions l How does discussion (chat, diagram) about writing activities relate to the argumentative quality of the resulting argumentative text? l Are number and type of argumentation episodes related to the quality of the argumentative texts? Educational Sciences Utrecht University 39

Conclusion Students writing high quality texts attend more to formulating content and discuss less non-task matters. l Discussion and sharing of knowledge result in a higher argumentative quality of the text. l Arguments about meta-cognitive matters and subject matter need to be resolved. l l Coordination of activities and agreement on a common line of reasoning is needed for a successful collaboration. Educational Sciences Utrecht University 40

MEPA (Multiple Episode Protocol Analysis) Erkens Educational Sciences Utrecht University 41

Transition diagram for the Control group Educational Sciences Utrecht University 43

Transition diagram for the Diagram condition Educational Sciences Utrecht University 44

Fourth phase Broadening the context Project-based learning Pro-ICT with VCRI = Virtual collaborative research institute EU-projects : DUNES & SCALE Educational Sciences Utrecht University 45

Project-based learning Differences with previous environments: l Synchronous and a-synchronous l A-synchronous at team and class level l => 2 students at the same time l Working together on the same or different texts l Coach for monitoring by the teacher l More Web-based than CS Educational Sciences Utrecht University 46

VCRI: Cowriter Educational Sciences Utrecht University 47

VCRI: Diagrammer Educational Sciences Utrecht University 48

VCRI: Selector Educational Sciences Utrecht University 49

VCRI: Group Forum Educational Sciences Utrecht University 50

VCRI: Planner Educational Sciences Utrecht University 51

VCRI: Reflector Educational Sciences Utrecht University 52

VCRI: Coach for teacher Educational Sciences Utrecht University 53

Trends in our research l From task to project l From a pair of students to a community l More teacher involvement in projects l More tools for coordination of collaboration and task l In EU-projects (Jerry Andriessen) l Scenario-based implementation of CSCL l Also argumentation in policy making l More ecological valid research Educational Sciences Utrecht University 54

We still think that the quality of the interaction process and dialogues is the most important aspect of learning situations Thank you Educational Sciences Utrecht University 55