Method library › Memory Learning Science

Cognitive Load Analyser

strong evidence · ⏱ 4 minutes · Memory Learning Science

Analyse a learning task for cognitive load problems and recommend specific design improvements. Use when tasks overwhelm students, instructions feel complex, or materials need simplifying.

What it does

Evaluates a learning task, instruction set, or resource for cognitive load across three dimensions: intrinsic load (inherent complexity of the content), extraneous load (unnecessary difficulty caused by poor design), and germane load (productive cognitive effort directed at schema building). Produces a specific diagnosis of where load is excessive and concrete modification suggestions. AI is specifically valuable here because cognitive load analysis requires simultaneously evaluating content complexity, instructional design quality, and learner expertise level — a skill that typically requires training in instructional design that most teachers lack.

The evidence behind it

Sweller (1988, 1994) established Cognitive Load Theory (CLT) as a framework for understanding why some instructional designs fail: human working memory can hold approximately 4-7 elements simultaneously, and learning fails when the total cognitive load exceeds working memory capacity. Sweller distinguishes intrinsic load (determined by element interactivity — how many elements must be processed simultaneously), extraneous load (caused by poor instructional design), and germane load (productive effort directed at building schemas). Paas & van Merriënboer (1994) operationalised CLT for instructional design, demonstrating that reducing extraneous load consistently improves learning outcomes. Sweller et al. (2019) updated the theory to incorporate evolutionary psychology and refine the distinction between biologically primary and secondary knowledge. Critically, Kalyuga et al. (2003) identified the "expertise reversal effect" — instructional techniques that reduce load for novices (worked examples, integrated diagrams) can actually increase load for advanced learners by requiring them to process redundant information. This means cognitive load analysis must always consider learner expertise.

Sources

Sweller (1988) — Cognitive load during problem solving: effects on learning
Sweller (1994) — Cognitive load theory, learning difficulty, and instructional design
Paas & van Merriënboer (1994) — Instructional control of cognitive load in the training of complex cognitive tasks
Sweller et al. (2019) — Cognitive Architecture and Instructional Design: 20 Years Later (updated CLT)
Kalyuga et al. (2003) — The expertise reversal effect

How to use it in your lesson

For the best results with EvidenceLesson, give it:

task_description — The learning task, instruction, or resource to analyse
student_level — Age/year group and expertise level (novice/intermediate/advanced)
task_materials (optional) — Description or text of worksheets, slides, or instructions used
student_profiles (optional) — From context engine: working memory profiles, prior knowledge data
lesson_context (optional) — What comes before and after this task in the lesson

Known limitations

Cannot observe actual student behaviour. This analysis is based on task design, not on how students actually experience the task. Two students may experience the same task with very different cognitive loads depending on their prior knowledge. Teacher observation during the task remains essential — signs of overload include task abandonment, copying without understanding, and asking procedural questions ("where do I write the answer?") rather than content questions.

Intrinsic load cannot be reduced without changing the content. If the content itself is inherently complex (high element interactivity), this analysis can only reduce extraneous load and optimise sequencing — it cannot make complex content simple. For high-intrinsic-load content, the answer is often to break the content into sub-elements taught across multiple lessons, not to simplify it within one lesson.

The expertise reversal effect means recommendations are expertise-dependent. What helps a novice hinders an expert and vice versa. If the student level is inaccurate (e.g., described as "novice" but students actually have substantial prior knowledge), the modifications may be counterproductive. Teachers must calibrate based on actual student knowledge, not assumed knowledge.