Influence of authority on convergence in collaborative learning

Abstract

A typical classroom is already a community of practice and learning. Teachers and students are not blank slates when they encounter CSCL systems. They already have social roles, norms and conventions that effect social interaction and communication. College students, in particular, are not novices at working in the role of student; they are skilled students with a lot of experience.A major force affecting classroom communication and social interaction is authority, which gives certain people, objects, representations or ideas the power to affect thought and behavior. Computer-supported collaborative learning requires students and teachers to change how they understand and assign authority to be effective. Students need to assume more authority, assign authority to their peers, and to value their own thoughts and ideas. However, this is not an easy transition to make.This paper describes two studies in which students' perceptions of authority and authoritative representations in the classroom led to learning problems and misunderstandings while they were participating in collaborative learning. In both cases, the students converged on either a representation or representational style that they believed was authoritative. They did not converge on a representation or style based on how well the style or representation communicated the concept, but instead based on how well the representation or style reflected the practices of their teacher.In the first study, the students tried to use one representation presented by their instructor to answer all problems, regardless of whether or not it was an appropriate representation to use. As Roschelle and Behrend argue, students participating in collaborative learning, who are given a particular representation, will converge on a shared understanding (1995). However, the study described here shows that there can be learning problems when this occurs, especially when it occurs early in the learning process.In the second study, students used an early version of a system supporting web-based collaboration called CAROUSEL (Collaborative Algorithm Representations Of Undergraduates for Self-Enhanced Learning). Students shared their representations of algorithms anonymously, and viewed and evaluated each other's representations. In this study we found that the type and style of the representations converged on an authoritative style, an explanatory style similar to representations one would find their textbook or similar to ones used during lectures.In creating CSCL systems, we recommend to take explicit steps to lessen the effect of social identity on the formation and acceptance of ideas. Also, it is important to either adapt the system to the current social activities and arrangements in the classroom or to make it explicit how the social arrangement will change and have participants accept and take ownership of that change.

Constructive and collaborative learning of algorithms

“…This study also showed that students were converging on a representational style that was consistent with the representational style used in their lectures and textbooks [6,8]. At the beginning of the study, the students used a variety of media in creating the representations, but by the end of the study, the only media used were text and static graphics.…”

“…The implicit model of learning most algorithm animations are based on seems to be a transmission model, where the media present relevant information in appropriate formats and students absorb the information through passive viewing and limited interaction. Our qualitative study of undergraduate computerscience students indicates that students regularly employ informal and collaborative meaning-building activities while learning algorithms [6]. These include studying in small groups, solving homework problems together, and explaining and helping each other with difficult concepts.…”

Dancing hamsters and marble statues

“…Indeed, in earlier studies we found that computer science undergraduate students already form a community of practice in which learners tend to study collaboratively, discussing and constructing a shared understanding of topics covered by their instructor [Hübscher-Younger & Narayanan, 2001;. Unfortunately, we also found that authority assigned to certain explanatory visualizations (typically those provided by the instructor) led students to ignore alternative explanations and fail to understand the limitations of such visualizations [Hübscher-Younger & Narayanan, 2002]. We felt that visualizations built by students themselves carry less authority.…”