Opportunistic Collaboration


Zhang, J. et al. (2009). Designs for collective cognitive responsibility in knowledge-building communities. J. Learn. Sci., 18, 7-44. Get it! Free copy

Finding Summary

Many teachers like to use fixed groups on Knowledge Forum, whether assigned or chosen by students. The main reason for this is that it would help manage and track student work and limit the number of notes that students need to read and respond to. But classroom work and research has consistently found that it produces inferior results. Approaches, in which students are encouraged to interact around and address any problem that has emerged in their community to advance their collective goals, are more effective.

In the mid-1990s, researcher Jim Hewitt and teacher Jim Webb already noted the importance of this issue (1995, p. 16):

As one student explained, ‘you don’t need to see what the other groups are doing’. It became obvious … that despite repeated reminders of the importance of cross-group collaboration, most students still focused on their own group’s work. … Consequently, it was decided to take a more class-oriented approach. …. Rather than have note clusters represent group divisions, they instead represented different lines of inquiry. And instead of working in groups of 3 or 4, the entire class was instructed to work as a single research team.

More recently, teacher Richard Messina tested three designs in subsequent school years. Each year he taught Grade 4 students who used Knowledge Forum to investigate optics. In the first year he used fixed groups, in the second year interacting groups, and in the third year whole-class,  opportunistic collaboration. In the whole-class approach small and informal groups emerged to investigate specific problems, but these groups disbanded when sufficient progress had been made, and students participated in multiple groups.

Analysis of these three designs by Jianwei Zhang and colleagues showed that each design was better than its predecessor (Zhang et al., 2009, see table):

  1. Social network analysis showed that students were more connected to each other through building on, citing, rising above (synthesising) ideas contributed in their notes. 40% of the possible connections between the students were realised.

    Jianwei Zhang

    Dr. Jianwei Zhang

  2. The networks of students in the third approach were the least centralised. This indicates that no student or the teacher dominated the online discourse.
  3. Zhang et al distinguished between questions for ideas and questions on ideas, and found that questions on ideas lead to student-initiated inquiry. The fixed group design had a small number of questions on ideas (32%) than both interacting groups (56%) and the whole-class design (53%).
  4. The whole-class, dynamic collaboration led to the greatest depth of understanding and diffusion of ideas.

Zhang et al selected=

Explanation and Implications

Using a whole class design supports several knowledge-building principles:

Collective responsibility for community knowledge. It changes the goals from advancing the knowledge of a small group to a advancing that of the community; and shifts from fixed, pre-scripted  to adaptive collaboration for collective advancement.

It enhances idea diversity. The pool of ideas within a small group is typically not large enough for the best ideas to be applied to a problem and for the scrutiny, testing, and suggestions needed for idea improvement.

It enhances democratization of knowledge by improving opportunities for participants to make contributions to the discourse. A situation may not arise in a small group in which it makes sense for a student to contribute an idea; in a larger community there may be more opportunities.

The opportunistic collaborations that emerge in a whole-class design also enhance the opportunities for participants to collaborate with many different participants, which can enhance social skills and limit the negative effects of working in groups.

The major reasons why teachers like to use fixed or interacting groups in Knowledge Forum can be addressed by other methods. The most important of these are synthesis and rise-above, which slow down the growth of the database (van Aalst, 2006).