Analyzing Collaboration

There has been much research involving the topic of collaboration in relation to learning, though the ontologies of these studies have been varied in how they define and view collaboration. In Analyzing Collaboration, Noel and Reed summarize four overarching categories of perspectives on collaboration:

  • collaboration as a window
  • collaboration for distal outcomes
  • collaboration for proximal outcomes
  • collaboration as learning

and that these four perspectives differ along four axes:

  • the unit of analysis for describing processes [individual or collective]
  • the unit of analysis for describing outcomes [individual or collective]
  • the degree to which these outcomes are within the collaboration or outside the collaboration
  • the degree to which a normative stance is taken on collaboration

They argue that in researching collaboration in the learning sciences, care should be taken to first identify which of these categories the research is operating under, which will in turn “drive the choice of methods” used. Here I will summarize their main points on each of these frameworks.

Collaboration as a Window onto Thinking

This first perspective differs from the other 3 in that it doesn’t see the collaborative activity itself as important to learning. Rather, the collaboration is seen instrumentally as a way for researchers to have their subjects’ thoughts verbalized for recording and coding. This framework is more or less dismissed by the authors as it “misses the core assumption of collaborative learning – that the ways learners respond to each other matters to both the processes and outcomes of learning”.

Collaboration as a Context that Promotes (or Constrains) Distal Learning Outcomes

In this perspective, also referred to as effect-oriented, studies the effects of collaborative activities on individual learning outcomes which occur outside of the context of the collaboration. For example, critiques of the IRE (Initiate, Respond, Evaluate) pattern of discourse between teacher and student based on its ineffectiveness to elicit deep thought falls within this framework. Research from this perspective includes operationalizing both the collaborative process and the outcomes. The process must be recorded in some way to allow for analysis, usually through transcripts which are then coded. The outcomes can range from simple measures such as test scores to more complex measures like engagement.

Collaboration Coordinated with Proximal, Collective Outcomes within the Interaction Itself

This framework has gained more traction recently, and attempts to “explain how collaborative processes contribute directly to learning”. Here, emphasis is on the collaboration as a collective process with collective outcomes that happen during the scope of the process, the shared understandingdeveloped by the collective and a way to understand individual learning. Because the greater context of of the collaborative effort is important, simple transcriptions are not effective under this framework, as they tend to leave out, or “clean up”, all of the minutia of collaborative discourse such as intonation, gesturing, use of physical representations, etc.. Studies under this framework have drawn upon techniques used in conversation analysis research for producing transcripts which retain many of these types of details.

Collaboration-as-Learning: Committing to a Distributed, Endogenous Unit of Analysis

The final category is fundamentally different from the other three as it focusses completely on the collective. It is grounded in the idea that “measuring the cognitive processes of the individual … obscures a lot of what matters to the successful realization of [the] cognitive task”. This type of research aims to understand how “members themselves are jointly managing and emergently maintaining collaboration of some particular kind to which they are jointly accountable”. It sees the distributed unit (team, group, company, etc) as “historically durable”, the unit lives on while members join and leave. Further, rather than the normative outcomes found in the other perspectives, here the outcomes are endogenous, the outcomes are the collaborative effort itself. The authors theorize that this type of research has yet to really take hold because it is misaligned with how we typically think about learning: schooling, which inherently cares more about evaluating the individual rather than collective accomplishments. In non-school settings, however, this kind of research is especially useful. The authors argue that this framework has some meaningful implications inherent to it. First, that normative approaches “often smuggle[] in cultural values of dominant or majority groups” in the way transcripts are analyzed, which they argue could be avoided by taking an endogenous approach. Second, they argue that in contemporary  learning sciences research, “understanding the existing cultural knowledge and practices of youth provides pathways toward new learning”.


While the paper presents itself as a summary of these four broad categories of research in collaboration, it is very clear that the authors are pushing the collaboration-as-learning perspective. I don’t necessarily disagree with their evaluation, but it seems that they could have been more straightforward with their intent to critique and evaluate these frameworks from the outset.