Representation-driven mixed initiative in computer supported collaborative learning in secondary education

We investigate a computer supported approach in which pairs co-con-struct a qualitative representation of the dynamics of the industrial revolution in a shared workspace. A key feature of this approach concerns the use of a meta-vocabulary for representing cause-and-effect relationships that facilitates the use of a predefined norm-representation to automatically steer the collaborative learning process. In particular, it provides focus on the set of ingredients that the learners should use. Additionally, the workspace offers each learner pair information about progress and content-related support. An evaluation study was executed in a real classroom. A workbook provided information for constructing the representation and gave advise on how to approach this task together. How-ever, most pairs took an alternative approach and divided their actions in the shared workspace in an unbalanced way. Three types of task division occurred that showed differences in the number of errors and the number of requests for support. From this result, we formulate future directions for the development of a pedagogical approach that stimulates collaborative learning with qualitative representations and the support offered by the software.