Cognitive load theory usually reaches a school through a CPD session and a reading list, and it often leaves a department holding a long list of things to rebuild. Most of the changes the evidence supports are smaller than that. They sit inside lessons that already exist: what goes on a slide, how much is explained before pupils try something, how a worked example is laid out, and when a scaffold comes away.

This guide is written for heads of department and curriculum leads who know the theory in outline and want to know what is worth changing first, and how far the research lets us be confident.

What the theory claims, in brief

The Centre for Education Statistics and Evaluation (CESE) in New South Wales summarises the core idea this way: working memory can hold only about four chunks of new information at a time, while long-term memory stores knowledge as schemas that can be retrieved as a single unit. Once something is well learned, it takes up little working memory, which leaves room for new thinking. CESE also notes that the instructional implications are backed by a significant number of randomised controlled trials, and it reports Dylan Wiliam's view that the theory is "the single most important thing for teachers to know".

Sweller, van Merriënboer and Paas (2019) set out the distinction that matters most for planning. Intrinsic load comes from the complexity of the material itself. Extraneous load comes from how the material is presented, and teachers can change it. They also conclude that most of the instructional effects found in the research reflect variation in extraneous load.

Optimise load rather than minimise it

A common misreading of the theory is that the best lesson is the easiest one. The Education Endowment Foundation's 2021 review of cognitive science approaches, written by a University of Birmingham team, states the aim as optimising load, and notes that this sometimes means increasing challenge. A lesson with nothing hard in it has simply removed the learning.

Six small changes

1. Declutter the slide. CESE's practice guide describes the redundancy effect, where learning suffers when the same information is delivered in several forms that pupils do not need, such as a slide of full sentences that the teacher then reads aloud. One idea per slide and a spoken explanation that adds to the slide, rather than repeating it, is a change that takes minutes. The EEF review also warns about what it calls lethal mutations, where a well-meant adaptation of a research finding ends up working against it. Its example is decorative images added in the name of dual coding, which add load without adding meaning.

2. Keep related information in the same place. The split-attention effect arises when pupils must hold two sources in mind and mentally combine them. A diagram with a key at the bottom of the page, or a worked solution on one slide and the problem on the next, forces that effort. Put labels on the diagram itself and place each step of a solution beside the line it explains. The EEF review adds a practical warning: core textbook materials often split attention, so it is worth checking the resources a department relies on.

3. Explain in smaller steps. Rosenshine's well-known principles (2012) recommend presenting new material in small steps, with pupil practice after each step. This is one of the cheapest changes available, because it alters the order of a lesson rather than its content. A twenty-minute explanation followed by an exercise becomes three shorter explanations, each followed by a short task.

4. Use worked examples, then fade them. CESE describes the worked example effect as widely replicated: novices who study worked examples tend to do better on later tests than those who solve equivalent problems unaided. The EEF review is more cautious about how much this has been tested in ordinary classrooms. It found 22 studies, all in maths and science and mostly in secondary schools, and only 8 of them were delivered by regular class teachers. For a department, the sensible reading is that the approach is well supported for new, complex material in those subjects, and worth trialling elsewhere with careful monitoring.

5. Check what pupils already know before choosing the level of support. The expertise reversal effect means that support which helps a novice can slow down a more knowledgeable pupil. CESE advises giving pupils gradually more independent problem solving as their proficiency grows, and the EEF review calls expertise reversal one of the biggest challenges in applying the theory. Its findings on incorrect or incomplete worked examples point the same way: they were positive for advanced pupils and mixed for pupils with low prior attainment. A short retrieval starter at the beginning of a lesson tells you who needs a full worked example and who is ready for a completion task.

6. Aim scaffolds at the thinking, and plan how they come away. Across 16 studies in KS2 to KS4, the EEF review found consistent evidence that well-targeted scaffolds work. The detail matters. A cue that was motivational rather than instructional had a negative effect, and fading that was adapted to what pupils knew worked better than fading on a fixed timetable. The review also flags the risk that pupils come to rely on the scaffold. For a department, that suggests writing the fade into the scheme of work alongside the scaffold itself.

Where the evidence is thinner than the conference talks suggest

The same EEF review is candid that applied evidence is limited, and that findings from laboratory studies do not always transfer to a classroom of thirty. The authors position cognitive science as a useful guide that offers no guaranteed fixes, and that reading is the right one for a department to adopt.

Worked examples: 22 studies, all in maths and science, and few delivered by regular teachers.

Scaffolds and guidance: 16 studies with consistent findings, but with a real risk of over-reliance.

Collaborative problem-solving: 9 studies, mostly in maths, with mixed results.

The sensible response is to trial changes in your own context and check whether pupils' work improves, rather than treating any single technique as settled.

A low-cost way to start

A department meeting is enough to begin. Each teacher brings one slide or resource from the coming week, and the group asks three questions of it. What does a pupil have to hold in mind at this point? Is anything on the page not helping them learn the content? Are the pieces that need to be read together actually together? These questions are a practical suggestion rather than a research protocol, but they target the extraneous load that the evidence says is most within a teacher's control.

This is also the thinking behind how BrainStrata's adaptive practice is designed: question difficulty and support adjust to where an individual pupil currently is, rather than to a single class-wide level.

Sources and further reading

[1] Centre for Education Statistics and Evaluation (CESE), NSW Department of Education. Cognitive load theory: Research that teachers really need to understand. First published 5 September 2017; page updated 30 July 2026.

[2] CESE, NSW Department of Education. Cognitive load theory in practice (2018 practice guide). Source for the split-attention, redundancy and modality effects.

[3] Sweller, J., van Merriënboer, J.J.G. and Paas, F. (2019). Cognitive architecture and instructional design: 20 years later. Educational Psychology Review, 31(2), 261-292.

[4] Perry, T. et al. (2021). Cognitive science approaches in the classroom: a review of the evidence. Education Endowment Foundation, July 2021. See the section on managing cognitive load, pp. 24-30.

[5] Rosenshine, B. (2012). Principles of instruction: research-based strategies that all teachers should know. American Educator, 36(1), 12-19.