Rethinking Cognitive Load Theory for Collaborative Learning

If you’ve ever tried group work in your classroom and wondered why it sometimes falls flat, you’re not alone. Collaborative learning is often promoted as a magic bullet: students discuss, engage, and build knowledge together—what could go wrong? As it turns out, quite a bit. But don’t throw out your group work plans just yet. The issue isn’t that collaboration doesn’t work—it’s that we often misunderstand when, how, and for whom it works best. In an article that I wrote with Jimmy Zambrano, Femke Kirschner (yes), and John Sweller, we expanded Cognitive Load Theory (CLT) to Collaborative Cognitive Load Theory (CCLT) to explain this. You can see it as an update to traditional CLT, which is based on individual learning that takes group learning seriously. By combining insights from evolutionary psychology, human cognition, and instructional design, it helps us understand not just how students learn individually but also how they learn together.

Cognitive Load Theory, developed by John Sweller and introduced to the world in 1988, is grounded in how human memory works—specifically, the relationship between working memory, which is limited and temporary, and long-term memory, which is virtually unlimited and durable. It tells us that learning involves moving information from working memory into long-term memory, where it is stored as schemas. Cognitive load can be intrinsic, tied to the complexity of what is being learned (think just a list of words and their translation to another language as compared to learning to speak or write a few sentences in a language where noun, adjective, article, number, and gender are completely interrelated) , or extraneous, caused by how material is presented (think structured explicit instruction with worked-out examples compared to messy slides, unclear instructions, or discovery learning). The key to good instruction is to manage these loads so that students can focus their mental effort and resources on learning, not just surviving the task.

Traditionally, CLT focused on individual learners. But what about when students work in groups? Isn’t collaboration supposed to reduce the load by sharing the thinking? The answer is both yes and no. Collaborative Cognitive Load Theory introduces two critical concepts to help us unpack this. First, the idea of a collective working memory suggests that in a group, working memory resources are pooled. If a task is too complex for one student to manage alone, dividing the load among peers can make it more manageable. Second, the notion of transactive activities, which refers to the communication and coordination costs of working together, explaining, clarifying, negotiating, and assigning roles. These activities can either support or derail learning, depending on how well they’re managed.

Whether collaboration helps or hinders learning depends on several key variables. The complexity of the task plays a central role. Group work only makes sense when the task is complex enough to require multiple minds; that is, when the benefits of working with others exceed the costs. Simple tasks don’t justify the coordination overhead. If you can just as easily solve a problem or carry out a task alone, why would you want to work with others? Learner expertise is also important. Novices benefit more from well-designed collaboration than experts. If students already know what they’re doing, adding collaboration can introduce unnecessary distractions, a manifestation of the expertise reversal effect. Group composition matters as well. First, mixing novices and experts can be tricky. The less students know, the more scaffolding they need—not just for the content, but for how to work together. Also, teams that have worked together before tend to be more efficient than ad hoc groups that change for each task, as they don’t need to figure out how to coordinate from scratch. The members know each others, how they work, can anticipate, etcetera. And finally, assigning clear roles and providing guidance, such as collaboration scripts in online environments, helps reduce unnecessary transactive load and makes collaboration more productive.

Drawing from evolutionary educational psychology, CCLT also distinguishes between biologically primary and secondary knowledge. Biologically primary skills, like speaking or interacting socially, are skills we evolved to learn naturally. These come easily and don’t need much teaching. In contrast, biologically secondary knowledge—things like math, writing, and science—requires effortful learning and explicit instruction. Collaboration is biologically primary—we’re wired to do it. But collaborating to learn algebra, for example, is not. That’s secondary, and it demands instructional support. We need to teach students not just the subject matter but also how to collaborate effectively in that context.

One of the most important insights from CCLT is that group work comes with hidden costs. These are the transaction costs: the time and effort needed to communicate, coordinate, and keep things on track. In Cognitive Load terms, these are a form of extraneous load. If they aren’t managed, they can cancel out the benefits of collective working memory. Students who are unclear about their roles or goals, groups that spend more time on logistics than learning, and asynchronous CSCL environments with lagging communication are all examples where high transaction costs can derail collaboration. Good instructional design can mitigate these costs. For instance, collaboration scripts in online environments, clear team roles, and group awareness tools that show who’s doing what can all help reduce the cognitive burden.

So what does this mean for teachers? First, use group work strategically. Assign collaborative tasks only when the complexity is high enough to justify the effort. Designing a complex task is a skill that needs to be acquired. Second, mind the mix. Consider students’ content knowledge and collaboration skills, and support novices differently than experts. Third, support the process. Scaffolds like worked examples, scripts, and clear role assignments can help reduce unnecessary load. Fourth, train collaboration. Don’t assume students know how to work together. Teach them how to plan, communicate, and reflect as a team. And finally, evaluate and adapt. Watch how groups function and adjust groupings, tasks, and supports based on your observations.

Collaborative learning isn’t a one-size-fits-all solution. When poorly designed, it adds confusion and waste. But when rooted in what we know about cognition and instructional design, especially through the lens of Collaborative Cognitive Load Theory, it becomes a powerful tool for learning. If we want group work to succeed, we must stop romanticising it and start designing it.

Kirschner, P. A., Sweller, J., Kirschner, F., & Zambrano R., J. (2018). From Cognitive Load Theory to Collaborative Cognitive Load Theory. International Journal of Computer-Supported Collaborative Learning, 13(2), 213–233. https://doi.org/10.1007/s11412-018-9277-y