Successful collaborative learning depends upon effective interaction amongst learners. However, when learners are left on their own, they rarely engage in productive interactions such as asking each other questions, explaining and justifying their opinions, articulating their reasoning, or elaborating and reflecting upon their knowledge. Collaboration scripts have recently been presented as a promising method to trigger these activities and to provide structure and support for open learning environments. Scripts aim to foster collaborative learning in shaping the way in which learners interact with one another. They may define for each phase what task the students have to perform, the composition of the group, the way the task is distributed, the mode of interaction, and the timing of the phase. Computer-supported scripts can be designed to facilitate collaborative learning in different ways. On the one hand, they can scaffold the interaction process per se by providing prompts, sentence starters etc. On the other hand, they can set up conditions in which favourable activities and productive interaction should occur. Beyond structuring specific activities and interaction patterns, scripts may also orchestrate individual and collaborative activities as well as virtual and physical activities within the classroom over longer time segments. However, it is also evident that scripted collaboration does not happen without problems and challenges, but different groups will act differently regardless of the same instructional interventions and environments. This symposium brings together researchers who have focused on designing scripts and evaluating their impact in computer-supported collaborative learning settings. These settings vary in terms of social levels, time-scale, tools, participants etc., but common for all of them are the theoretical and methodological challenges related to the application of computer-supported scripts.

Interest in developing improved instructional methods for mathematics has increased since TIMSS and PISA. In our project, we introduce a new way to promote knowledge acquisition in mathematics: we enhanced the Cognitive Tutor Algebra I, an intelligent tutoring system for learning mathematics in high school that only has been used for individual learning so far, to a collaborative learning setting. Although the Tutor has shown to increase learning, there also are several shortcomings. For instance, learning with the Tutor places emphasis on improving students’ problem solving skills, yet a deep understanding of underlying mathematical concepts is not necessarily achieved. To reduce these shortcomings, we extended the learning environment to a dyadic setting, adding new learning opportunities such as the possibility to mutually elaborate on the learning content. A script was developed to guide students’ interaction and to ensure that students profit from these new learning opportunities. The script structured students’ collaboration in an individual and a collaborative phase and prompted students to engage in fruitful interaction. An adaptive script component provided hints when the dyad encountered difficulties in problem solving. Finally, dyads engaged in a reflection activity following each problem to improve their collaboration over subsequent interactions. The scripts’ effect on students’ collaborative learning was tested in a classroom study that took place over the course of one week. We compared the learning of dyads collaborating at the Tutor without script support to the learning of dyads scaffolded by the script. To measure different aspects of learning, students solved several post tests on the Tutor and with paper and pencil. The results of the paper and pencil test assessing improvement of students’ conceptual understanding show significantly better performances of the script condition. The Tutor log data are still being analyzed and the results will be presented at the conference.

The Cognitive Tutor Algebra is a tutor for mathematics instruction at the high school level. Its main features are immediate error feedback, the possibility to ask for a hint when encountering impasses, and knowledge tracing, i.e. the Tutor creates and updates a model of the student’s knowledge and selects new problems tailored to the student’s knowledge level. However, students do not always benefit as much as they might from learning with the Tutor. First, a deep understanding of underlying mathematical concepts is not necessarily achieved since the Tutor emphasizes learning of procedural skills (Anderson, Corbett, Koedinger, & Pelletier, 1995). Second, students do not always make good use of the learning opportunities provided by the Cognitive Tutor. The collaborative extension we introduced to the Tutor environment aims at reducing these shortcomings. As research has shown, collaborative problem solving and learning have the potential to promote deep elaboration of the learning content (Teasley, 1995) and can yield improved conceptual understanding. However, students do often not show beneficial collaborative behaviours spontaneously (Rummel & Spada, 2005). Collaboration scripts have proven effective in helping people meet the challenges encountered when learning or working collaboratively (Kollar, Fischer, & Hesse, in press), thus we developed a script to prompt fruitful interaction on the Tutor.

For the present study we focused on “systems of equations”, a content novel to participating students. Our script consisted of three components that aimed at facilitating students to capitalize from learning opportunities offered in the collaborative Tutor environment. The fixed script component structured the problem solving process in two phases. During the individual problem solving phase, each student solved a problem with one equation in the Cognitive Tutor. In the collaborative phase, students joined on a single computer to solve a more complex system of equations problem that combined the two individual equations. They received instructions from the enhanced Tutor, e.g. prompting them on collaborative skills like explanation giving. This division in individual and collaborative phases served to increase the individual’s accountability for the joined problem solving and set up the preconditions for a productive interaction. Second, the script had an adaptive script component reacting when students met impasses that resulted in Tutor actions (e.g. hints). To encourage students to take advantage of these learning opportunities, the script asked them to elaborate on the help received. Third, a metacognitive component followed each collaborative phase: Dyads engaged in a reflection of the group process to improve their interaction during the following collaboration.

To evaluate the script’s effectiveness, we conducted a classroom study comparing scripted collaboration with an unscripted collaboration condition in which students collaborated without support. This study was an initial, small scale study to establish basic effects and to test the procedure in a classroom setting. Three classes taught by the same teacher participated. The unscripted condition consisted of two classes (12 and 4 students); the scripted condition consisted of one class (13 students). The study took place during three classroom periods over the course of a week. During day 1 and 2 (learning phase), students learned how to solve system of equations problems according to their condition. On day 3 (test phase), several post tests were administered measuring different aspects of learning. Three post tests at the Cognitive Tutor assessed students’ problem solving skills. Two tests asked students to solve system of equations problems isomorphic to those during instruction, thus testing the retention of the learned skills. One of these post tests was solved individually, the second was solved collaboratively. The third test assessed acceleration of future learning: If the script succeeded in improving students’ collaboration skills, learning to solve a novel problem type (inequality problems) in a collaborative setting at the Tutor should be easier for students who have had the advantage of script support during the learning phase. Finally, a paper and pencil test evaluated students’ conceptual knowledge with problems that required applying the acquired concepts to new problem types. The test consisted of two problem sets, both composed of questions with discrete answer possibilities and open format questions asking for elaborated explanations. Problem set 1 tested for students’ understanding of the basic concepts y-intercept and slope, problem set 2 assessed students’ understanding of the main new system of equations concept: the intersection point.

The results of the paper and pencil test revealed substantial differences between conditions, i.e. the script condition yielded significantly better performances. The analysis was restricted to students who always worked collaboratively when present (9 students in the unscripted condition, 10 students in the scripted condition). The statistical data are displayed in Table 1. Particularly interesting were the results of the open format questions, demonstrating a strong effect of the script on conceptual knowledge acquisition: scripted interaction substantially improved students’ ability to articulate their mathematical thinking. Scripted students also outperformed their unscripted counterparts in answering the discrete answer questions on the newly learned concept (intersection point). The discrete answer questions on basic concepts did not reveal significant differences. Results of the Cognitive Tutor post tests will be presented at the conference.

References

Anderson, J. R., Corbett, A. T., Koedinger, K. R., & Pelletier, R. (1995). Cognitive tutors: Lessons learned. Journal of the Learning Sciences, 4(2), 167-207.

Kollar, I., Fischer, F., & Hesse, F. W. (in press). Collaboration scripts - a conceptual analysis. Educational Psychology Review.

Rummel, N., & Spada, H. (2005). Learning to Collaborate: An Instructional Approach to Promoting Collaborative Problem Solving in Computer-Mediated Settings. Journal of the Learning Sciences, 14(2), 201-241.

Teasley, S. D. (1995). The role of talk in children's peer collaborations. Developmental Psychology, 31(2), 207-220.

Table 1: Means, standard deviations and statistical values of the paper and pencil post test, assessing conceptual understanding

Several studies on collaborative argumentation are based on the assumption that to foster the quality of learners’ arguments in discussions leads also to improved individual learning. More specifically, understanding and formulating sound arguments in discussions is considered as being associated with deep cognitive processing. Deep cognitive processing, in turn should promote understanding and acquisition of the concepts discussed. This, in turn, is supposed to improve the quality of arguments in the discussion. Although plausible, empirical investigations of these assumptions are rare, mainly because it has been quite difficult to analyze individual cognitive processes during collaborative argumentation. The goals of this study are to examine these assumptions. A one-factorial design (n = 48) was used to investigate the relation between the quality of arguments (low vs. high, manipulated via a collaboration script) during online discussions of groups of three, cognitive processes (which can be studied quite well in asynchronous online discussions, because learners can be asked to think aloud during collaboration), and knowledge acquisition. Results show that the quality of arguments during discussion is positively related to deep cognitive processing. Moreover, the quality of arguments had a positive effect on the individual acquisition of argumentative knowledge. The findings of this study provide support for the assumptions that high quality collaborative argumentation is associated with deeper cognitive processing as well as with the acquisition of knowledge of the individuals participating in a discussion.

Several studies in the field of Computer-Supported Collaborative Learning (CSCL) are based on the assumption that fostering the quality of learners’ arguments in discussions should lead to improved individual learning (e.g., Andriessen, Baker, & Suthers, 2003; Jermann & Dillenbourg, 2003; Kuhn & Goh, 2005). Both, the production and the reception of high-quality arguments is assumed to be related to deep cognitive processing and thus, to knowledge acquisition of the individual learner (e.g., Baker, 2003; Kuhn, Shaw, & Felton, 1997). High quality arguments are often defined as the ones that are both explicitly supported by grounds and limited by qualifications, provided that they also conform to the standards of their domain (Toulmin, 1958). Although plausible, empirical evidence for this set of interconnected claims on argumentative knowledge construction (i.e., about the relationship between the quality of arguments, the depth of cognitive processing, and knowledge acquisition) is rare. In this paper we examine these assumptions. By means of a computer-supported collaboration script (Kollar, Fischer, & Hesse, 2006) we facilitate and thereby experimentally manipulate the quality of arguments in online discussion.

Methods

In the one-factorial experimental study with the factor quality of arguments (low vs. high, manipulated via a collaboration script that aimed to facilitate the construction of single arguments), 48 students of Educational Science at the University of Munich participated during the winter term 2003/2004 as a requirement for receiving course credit in a mandatory introductory course. The experimental session covered a standard part of the course curriculum on motivation, namely Weiner’s attribution theory (1985). Performance in the experimental session did not enter the overall course grade. The participants were randomly assigned to groups of three. The groups were then randomly assigned to one of the two experimental conditions. The group’s task was to analyze three problem cases using Weiner’s theory and to come up with a joint solution for each case. The three students in each group were distributed over three laboratory rooms and were able to communicate via an asynchronous, text-based discussion board. We analysed the online discussions regarding the quality of construction of arguments, i.e. the number of arguments containing grounds and/or qualifications. All participants were asked to think-aloud during the collaborative learning phase. These think-aloud protocols were used to measure depth of cognitive elaboration by using the duration of single inferences between problem case and given theory as an indicator. Acquisition of domain-specific knowledge and knowledge on the construction of arguments were measured reliably with individual post-tests.

Results

The collaboration script increased the quality of arguments significantly and substantially in comparison to no support by the script, i.e., the experimental manipulation was successful. The depth of the cognitive elaboration of the learning material is positively correlated with the quality of a learner’s arguments in the online discussions, i.e. the higher the quality of arguments of the learner, the deeper the cognitive elaboration of the learning material. A mediator analysis has shown that the effect of quality of the learners’ own arguments was mediated by the depth of the learners’ cognitive elaboration, but not by the quality of the arguments of the learning partners. While no effect of the quality of arguments on domain-specific knowledge acquisition was found, the quality of arguments fostered the acquisition of knowledge on argumentation significantly and substantially.

Discussion

The findings of this study provide support for crucial assumptions on argumentative knowledge construction: We found empirical evidence for the claim that high quality collaborative argumentation is associated with deeper cognitive processing (as assumed, e.g., by Baker, 2003) as well as with the acquisition of knowledge of the individuals participating in a discussion. The simultaneous increase of the quality of arguments and the depth of cognitive processing by the use of the collaboration script and the mediator analysis make it plausible that deep cognitive elaboration rather precedes than follows high-quality arguments in our CSCL scenario.

The findings of this study additionally contributed to our understanding of how the quality of arguments can be supported by the use of collaboration scripts in online discussions. This also accumulates further evidence that computer-supported collaboration scripts can be employed to foster acquisition of domain-general collaborative competencies.

References

Andriessen, J. E. B., Baker, M., & Suthers, D. (Eds.). (2003). Arguing to learn. Confronting cognitions in computer-supported collaborative learning environments. Dordrecht: Kluwer.

Baker, M. (2003). Computer-mediated argumentative interactions for the co-elaboration of scientific notions. In J. Andriessen & M. Baker & D. Suthers (Eds.), Arguing to learn: confronting cognitions in computer-supported collaborative learning environments (Vol. 1, pp. 1-25). Dordrecht: Kluwer.

Jermann, P., & Dillenbourg, P. (2003). Elaborating new arguments through a CSCL script. In P. Dillenbourg (Ed.), Learning to argue (Vol. 1, pp. 205-226). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Kollar, I., Fischer, F., & Hesse, F. W. (2006). Collaboration scripts - a conceptual analysis. Educational Psychology Review, 18(2), 159-185.

Kuhn, D., & Goh, W. W. L. (2005). Arguing on the computer. In T. Koschmann, D. Suthers & T. W. Chan (Eds.), Computer Supported Collaborative Learning 2005: The Next 10 Years (pp. 125-134). Mahwah, NJ: Lawrence Erlbaum.

Kuhn, D., Shaw, V., & Felton, M. (1997). Effects of dyadic interaction on argumentative reasoning. Cognition and Instruction, 15, 3, 287-315.

Toulmin, S. (1958). The uses of argument. Cambridge: Cambridge University Press.

Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92, 548-573.

This contribution presents and evaluates how the ManyScripts environment enables teachers to design a script, prepare a session and orchestrate the activities in real time. Currently, the environment, called ManyScripts, includes the script called 'ConceptGrid'. For this script, groups of students first have to distribute roles (corresponding to theoretical approaches) among themselves. In order to learn how to play their roles, students have to read papers that describe the theory underlying their role. In the next step, each group receives a list of concepts to be defined. Students then write definitions of the concepts that were allocated to them. In order to complete the grid, groups have to assemble these concepts and to define the relationship between grid neighbours. The key task is to write 5 lines that relate or discriminate two juxtaposed concepts. During the debriefing session, the teacher compares the grid produced by different groups and asks them to justify divergences. To use a ConceptGrid script in her course, the teacher has to decide about the group size (number of roles) and edit the contents of the script: she defines the roles, the papers to be read for each role and the sets of concepts to be defined and assembled in a grid by the student groups. When the script is running, the teacher has the possibility to change some parameters such as the group composition or deadlines up to a certain level. The ManyScripts environment enables the teacher to follow the evolution of teamwork at a high level of aggregation. This new release of the ConceptGrid is currently used and studied at EPFL and the University of St. Gallen/Switzerland. The results of the empirical studies will be reported at the symposium.

Integrated learning scripts (Dillenbourg & Jermann, to appear) do not only include group activities but also individual activities and class-wide activities. These activities occur in the classroom space and are orchestrated by the teacher. In this contribution, we will first present how the ManyScripts environment enables teachers to design a script, prepare a session and orchestrate the activities in real time. Second, we will describe the research questions and some preliminary findings. The final results of the empirical studies will be reported at the symposium.

Designing a script instance

Currently, the environment, called ManyScripts, includes the script called 'ConceptGrid'.  This script is a sub-class of the Jigsaw family: 1) Groups of students have to distribute roles among themselves. Roles correspond to theoretical approaches of the domain under study. In order to learn how to play their roles, students have to read n papers that describe the theory underlying their role. 2) Each group receives a list of concepts to be defined and distributes these concepts among its members. Students write a 5 lines definition of the concepts that were allocated to them. 3) Groups have to assemble these concepts into a grid and to define the relationship between grid neighbors. The key task is to write 5 lines that relate or discriminate two juxtaposed concepts: if Concept-A has been defined by Student-A and Concept-B by Student-B, writing the Concept-A/Concept-B link requires Student-A to explain Concept-A to Student-B and vice versa. 4) During the debriefing session, the teacher compares the grid produced by different groups and asks them to justify divergences.

To use a ConceptGrid script in her course, the teacher has to decide about the group size (number of roles) and edit the contents of the script: she defines the roles, the papers to be read for each role and the sets of concepts to be defined and assembled in a grid by the student groups. The result is what we refer to as a script instance, e.g. "ConceptGridBiology2.1".

-----------------------------

Insert Figures 1a and 1b here

-----------------------------

Orchestrating a session

When the script is running, the teacher has the possibility to change some parameters such as the group composition or deadlines up to a certain level. The ManyScripts environment enables the teacher to follow the evolution of teamwork at a high level of aggregation as in figure 1b. More importantly, the 'teacher cockpit' enables the teacher to explore the contents produced by group along different axis as in figure 2: per construct concepts grids, per group, per concept or per relation between concepts.

-----------------------------

Insert Figure 2 here

-----------------------------

 Teachers may use the cockpit for grading the groups' work and, more importantly, for preparing the debriefing phase, i.e. when the teacher discusses the group productions with the whole class. The debriefing can be prepared in different ways. The teacher may annotate with her own colour codes the different productions (Figure 3a). Hence, when she uses the cockpit during the debriefing lecture, she may easily find the definitions she wants to refer to in her comments. Alternatively, she may simply integrate the student productions within her presentations as in figure 3b.

-----------------------------

Insert Figures 3a and 3b here

-----------------------------

Method

This new release of the ConceptGrid is now being used in an EPFL course, through 4 successive iterations. It is also used in a course for educational management at the University of St. Gallen. The different sessions will be evaluated and compared using content analysis. In addition, a questionnaire will be used to capture students' reactions, and the teachers using the grid will be interviewed.

Research questions and preliminary findings

The research questions for the evaluation of the Concept Grid will be:

a)        Will students see the ConceptGrid as a valuable CSCL environment to help them structure their collaboration process? For this questions, we already interviewed the pilot users of the program (N=13) in a semi-structured interview. The prelimenary findings of the pilot group were that the ConceptGrid itself provides a helpful support for the group process (10 out of 13 users completely agreed to this question), that some features, e.g. color coding, were still missing (which have been implemented by now) and that the ConceptGrid can only be applied to complex learning content because otherwise completing the grid would be to easy.

b)        Does the ConceptGrid help the teacher to prepare, implement and orchestrate the group process? To answer this research question, the teachers currently using the ConceptGrid are interviewed. The findings up to now suggest that only a short introduction is needed to use the Grid since the tool is very easy to use. The orchestration of the group process is very elaborate since it allows functions for groups with more or fewer students than concepts.

c)        How does the group process with the ConceptGrid look like? For this research question, we follow every group process in order to see how the grid is developed. The processes are locked in files so that a comprehensive content analysis will be executed once the learning processes are completed.  

References

Dillenbourg, P. & Jerman, P. (to appear). SWISH: A model for designing CSCL scripts. In F. Fischer, H. Mandl, J. Haake & I. Kollar (Eds) Scripting Computer-Supported Collaborative Learning – Cognitive, Computational, and Educational Perspectives. Computer-Supported Collaborative Learning Series, New York: Springer.

The present study focuses on the use of a particular kind of scripting, namely the use of knowledge types as a possible way to structure university students’ discourse in asynchronous discussion groups and consequently promote their learning. More specifically, the aim of the study is to determine how requiring students to label their contributions by means of knowledge types, has an impact on the knowledge construction processes reflected in the discussion. More specifically, students were asked to label each contribution with a category reflecting one of the stages of the progressive inquiry model. The categories that were provided were “Problem”, “My Explanation”, “Scientific Explanation”, “Evaluation of the Process”, and “Summary”. This script is based on the FLE3 knowledge building environment. The results suggest that the use of knowledge types significantly affects the knowledge construction processes. More specifically, it appears that requiring students to reflect on the type of message in their contributions stimulates significantly higher levels of knowledge construction reflected in students’ messages as compared to a control group engaged in regular asynchronous discussions without requirements with regard to labelling the knowledge type reflected in one’s contributions.

Objective

The present study focuses on the use of scripts to scaffold students’ online discourse and to facilitate their knowledge construction. The concept “script”, however, encompasses a broad range of methods, techniques, and approaches. In this respect it is difficult to speak about the overall efficacy of CSCL scripts (Dillenbourg, 2002). In this study, we were interested in the impact of a particular kind of scripting - the use of knowledge types - on the knowledge construction processes reflected in asynchronous discussions. As part of the course “Instructional Sciences”, 287 first-year university students were engaged in asynchronous discussion groups. Two research conditions were distinguished. In the experimental condition, students were required to tag their contributions by means of knowledge types. In the control condition students were engaged in an identical assignment. However, no requirements were made with regard to labelling the knowledge type reflected in one’s contributions. The study was guided by the following research question: Do students, who were required to tag their discussion contributions by means of knowledge types, differ from students engaged in regular asynchronous discussions with regard to their attained level of knowledge construction?

Theoretical framework

A central topic of CSCL research is how to facilitate online discourse and come to deep learning and understanding. Different approaches are studied. One approach is to realize “computer-supported collaboration scripts”. Scripts can be implemented as a kind of guideline. More specifically, a script can be defined as a detailed and explicit didactic contract between the teacher and the group of students regarding their mode of collaboration (Dillenbourg, 2002). The scripting approach is particularly interesting to specify and sequence and eventually to allocate different learning activities to learners (Weinberger et al., 2005).

In this study we investigate a computer-supported collaboration script, which provides a controlled list of message types from which the student must select before replying or creating a message. In the experimental condition, students were required to tag their messages by means of knowledge types, based on the FLE3 knowledge building environment. This environment is designed to support the collaborative process of progressive inquiry learning. The basic idea is that students gain deeper understanding by engaging in a research-like process where they generate problems, formulate hypotheses, and search out explanatory scientific information collaboratively with other students (Chen, 2004). More specifically, in the discussions students were asked to label each contribution with a category reflecting one of the stages of the progressive inquiry model. The provided categories were “Problem”, “My Explanation”, “Scientific Explanation”, “Evaluation of the Process”, and “Summary”. In this respect, students are asked to step back and to reflect upon the ongoing discussion and on how to contribute to optimise the debate. Moreover, the labels visualise the possible predominance or absence of one or more knowledge types. This can help students to create an overview of the knowledge-building activity as it unfolds and to improve their collaboration and ability to solve open-ended problems.

Method

Sample and design

All students enrolled for the course “Instructional Sciences” participated in the study (N= 286). Students were divided into discussion groups of about 8 students, with students randomly assigned to one of the 35 groups and groups randomly assigned to a research condition. Students in the experimental condition were required to tag their contributions by means of knowledge types. The online discussion environment offered a checklist interpreting the different contribution types advancing the discussion process. For each label, students received a description of what a particular knowledge type implies in terms of a discussion contribution. The discussion assignment was the same for all groups, regardless of the condition groups were in. Taken into account that transcripts of 35 discussion groups for 4 themes represent a massive amount of data, 9 groups (N=71) were randomly selected for the analysis.

Task environment and procedure

The discussions studied in the present study were a formal part of the course. Students participated during a complete semester. Four successive discussion themes of two weeks each were dealt with.

Content analysis

Content analysis was applied to study the processes of social knowledge construction reflected in the discussions. More particularly, the content analysis scheme based on Gunawardena et al. (1997) was used. For each group, the complete communication in relation to the 4 assignments was analysed. Two trained coders coded the messages independently. Inter-rater reliability was calculated and found.

Results and conclusions

The results indicate that students in the experimental condition do not experience difficulties to select knowledge building categories when contributing to the discussion. Moreover, the students seemed quite deliberate in their selection of categories. Overall, the selection of knowledge types was more or less evenly distributed among the category set. However, “My Explanation” was chosen most frequently, while “Scientific Explanation” occurred somewhat less often. In addition, the findings reveal that requesting students to label the knowledge types in their contributions augments the presence of more relevant, varied, and well-founded posts than students in the control group. The results more specifically indicate significant differences with regard to the levels of knowledge construction reflected in students’ messages favouring the experimental condition.

References

Chen, W. (2004). "Supporting Teachers Intervention in Collaborative Knowledge Building". In E.Gaudioso & L.Talavera (Eds.), Proceedings of 16^th European Conference on Artificial Intelligence (ECAI'2004). Workshop on Artificial Intelligence in Computer Supported Collaborative Learning (pp. 1-5). Valencia, Spain.

Dillenbourg, P. (2002). Over-scripting CSCL: The risks of blending collaborative learning with instructional design. In P. A. Kirschner (Ed.), Three worlds of CSCL: Can we support CSCL? (pp. 61-91). Heerlen: Open Universiteit Nederland.

Gunawardena, C. N., Lowe, C. A., & Anderson, T. (1997). Analysis of a global online debate and the development of an interaction analysis model for examining social construction of knowledge in computer conferencing. Journal of Educational Computing Research, 17, 397-431.

Weinberger, A., Stegmann, K., and Fischer, F. (2005). In T. Koschman, T. W. Chan, and D. D. Suthers (Eds.), Proceedings of Computer supported collaborative learning 2005: The next 10 Years!. Taipei, Taiwan, 30 May - 4 June 2005, retrieved October 1, 2006, from http://css.cscl2005.org/Fullpapers.aspx.

Collaboration scripts can guide the participants in dealing with the learning task, help them to choose appropriate roles to play as well as to organise and to sequence the various activities they are supposed to engage in. This study explores how three different types of pedagogical scripts guided groups’ collaboration processes, how groups’ activity level and types of activities vary during a scripted university course. This study is a design-based study, in which the participants were the first-year teacher education students (N=30) studying the pedagogy of education for a period of three months. Three different scripts (Case, Grid and Open-problem) were employed to make learning more efficient. These three scripts formed a “macro-script” for the whole online course. Process-oriented data sources included material used and produced during the computer-based activity; log data on student activities, asynchronous web-based discussions, three different outputs of each group and a self-report questionnaire. The analysis involved two levels. First, all data were verified and students’ activity levels were categorized during each script. The second level of the analysis concentrated on how well the groups proceeded through the different steps from the perspective of collaboration process. According to the findings, scripting enhanced collaboration and ensured that all groups were able to complete the task. However, despite the scripts the group activities varied during the task and the script could not guarantee “high-level” participation by all students. The activity level of participants varied between the different scripts both in terms of the number of active participants and the degree of individual participants’ activity in each phase of the script. The results of the study can be utilized in designing collaboration scripts for computer-supported settings.

Collaboration can be promoted by structuring the interaction process in order to favour the emergence of productive interactions. One way to structure interactions is to design predefined collaboration scripts for computer-supported collaborative learning environment. Collaboration scripts can guide the participants in dealing with the learning task, help them choose appropriate roles to play as well as organise and sequence the various activities they are supposed to engage in (Kobbe et al., 2006). The aim of scripting can vary, e.g. ranging from direct facilitation of specific activities to setting up the favourable conditions prior to the collaboration phase (Dillenbourg & Jermann, 2006). This study explores how different types of scripts guide groups’ collaboration processes during the web-based course. The particular aim is to describe how groups’ activity level and types of activities varied between different scripts (phases of the course).

Method

This study is a design-based study (Cobb et al., 2003), in which the participants were the first-year teacher education students (N=30) studying the pedagogy of education for a period of three months. Three different pedagogical scripts (Case, Grid and Open-problem) varying the the way and degree of scripting were employed to make learning more efficient. In the Case script, the main idea was to solve an authentic learning problem through shared work (e.g. Brown et al., 1989) based on theoretical background knowledge and authentic class room situation. In the Grid script, students worked on interrelated topics on the basis of complementary knowledge (different theoretical background information), which they had to make visible in the table to be used in shared work (e.g. Dillenbourg & Jermann, 2006). In the Open problem script, groups created and solved their own problem based on shared background material (e.g. Dochy et al., 2003). These three scripts formed a “macro-script” for the whole online course. Process-oriented data sources included material used and produced in the computer-based activity; log data on student activities, asynchronous web-based discussions, three different outputs of each group, and a self-report questionnaire (Järvenoja et al., 2005). The analysis involved two levels. First, all data were verified and students’ activity levels were categorized during each script. The second level of the analysis concentrated on how well the groups proceeded through the different steps from the perspective of collaboration process.

Results and conclusions

During the experiment, all the groups followed the Case script, although 10 of the 30 participants skipped one or more of the steps included in the script. Scripting seemed to guide collaboration at a general level ensuring that all the groups were able to find the resources and to accomplish the task. The main problem was attributable to “free riding”, which was recognised both by some active members of the group and by some “free riders” themselves. During the Grid script, 30 students started the course, but two of them dropped out the course. All the groups followed the Grid script. The script ensured that all the groups were able to complete the task. The scripting kept the process going on in an orderly manner, without skipping any phases of the script, and there were only one “free-rider” within all the groups. The Grid script involved each participant to write down his/her views and reflections in the table. Despite this positive effect, some critical issues also occurred as some students were participating in a non-collaborative way. The Open Problem script succeeded in guiding students to proceed in different steps. At the beginning, each group used a lot of effort on discussing and defining the problem. Despite of this, collaborative activity differed quite extensively, and there were a lot of differences between the groups in setting the problem, quality of knowledge construction and the activity level.   References

According to the findings, all the groups were able to complete the task. However, the scripts could not guarantee “high-level” participation by all the students. The activity level of participants varied between the different scripts both in terms of the number of active participants and the degree of individual participants’ activity in each phase of the script. There were also variations in the amount and type of activities between different groups. In general, the engaged groups seemed to benefit from “loose” scripting and the non-engaged groups from more detailed scripting. The results of the study will be discussed in more detail in the symposium, and they can be utilized in designing collaboration scripts for computer-supported settings.

References

Cobb, P., Confrey, J., diSessa, A., Lehrer, R., & Schauble, L. (2003). Design experiments in educational research. Web version of Educational Researcher, 32(1), 9-14.

Dillenbourg, P. & Jermann, P. (2006). SWISH: A model for designing CSCL scripts. In F. Fischer, H. Mandl, J. Haake, & I. Kollar (Eds.), Scripting Computer-Supported Collaborative Learning – Cognitive, Computational, and Educational Perspectives. Computer-Supported Collaborative Learning Series. New York: Springer, in press.

Järvenoja, H., Volet, S., & Järvelä, S. (2005). Investigating self-, other- and shared-regulation in socially challenging learning situations: An instrument to assess dynamics of students' emotion and motivation regulation processes. Submitted.  

Kobbe, L., Weinberger, A., Dillenbourg, P., Harrer, A., Hämäläinen, R., Häkkinen, P., & Fischer, F. (2006). Specifying Computer-Supported Collaboration Scripts. Submitted.

.