Cognitive Load Theory – “the single most important thing for teachers to know”

Earlier, renowned educationalist Dylan Wiliam tweeted:

I've come to the conclusion Sweller's Cognitive Load Theory is the single most important thing for teachers to know https://t.co/MkJJLruR8g

— Dylan Wiliam (@dylanwiliam) January 26, 2017

I therefore thought that I would take the opportunity to suggest a few articles that teachers can access to find out more.

Firstly, I have written a piece for The Conversation that provides a very basic introduction to Cognitive Load Theory (CLT), demonstrating how it applies to practical teaching problems.

In his tweet, Wiliam provides a link to a paper on the history of CLT written by John Sweller which also outlines some practical issues.

I first became aware of CLT through reading the seminal paper “Why Minimal Guidance During Instruction Does Not Work: An Analysis of the Failure of Constructivist, Discovery, Problem Based, Experiential, and Inquiry-Based Teaching,” by Paul Kirschner, John Sweller and Richard Clark. This paper uses CLT as the theoretical underpinning for an argument in favour of explicit teaching methods.

One problem with the paper is that many people read the title and say something like, “but good problem-based learning is not minimally guided!” The argument actually hinges on what the authors mean by ‘minimal guidance’ rather than a reader. To my understanding, asking students to solve a problem – this is a broad definition of ‘problem’ that would include things like writing an essay – without explicitly instructing them on the specific steps required is ‘minimal guidance’.

Perhaps this is why, when the same authors came to write an article on the subject for American Educator, they chose to argue, “the case for fully guided instruction”. This piece is aimed more squarely at teachers than the original one and is perhaps a better starting point.

The original ‘minimal guidance’ piece sparked a number of academic responses (here, here and here). Kirschner, Sweller and Clark then replied to these responses and, to my view, adequately addressed all of the points raised.

CLT has not been without controversy. The concept of ‘germane’ load was introduced and led to the theory becoming potentially unfalsifiable – a major problem. Without rehearsing the full argument here, I will try to summarise: Applications of Cognitive Load Theory generally attempt to reduce cognitive load. However, we also know that too little cognitive load will also lead to little learning. The load can be too low either because something is structurally very simple – such as learning a list of names – or because a student has already stored the solution steps in long term memory and so does not have to consciously think about them. When learning something complex, like how to solve a maths problem or how to interpret a novel, it might be better to try to reduce cognitive load whereas when learning lists of names or dates, we might want to intentionally increase it.

Reducing load for complex tasks is supported by research demonstrating the worked example effect where providing worked examples for students to study is superior to asking them to solve equivalent problems. Increasing load for simple tasks is supported by research demonstrating ‘desirable difficulties’. For instance, if you want a student to learn the capital of Australia then it might be best to first ask them to guess before telling them the correct answer. This will get them thinking about what they know about Australia before you slot in the new information.

Sweller has introduced the idea of ‘element interactivity’ to describe the difference in complexity between tasks but this is also controversial.

Beware a sleight of hand from CLT’s critics. As with any developing field, there will be controversies. Pointing out that element interactivity is a controversial concept does not invalidate the rest of the theory or the conclusions about the best ways to teach complex content, especially since these methods have been replicated a number of times in quite different fields of research.