The total mental effort required to process information at any given moment. Based on John Sweller’s Cognitive Load Theory, which proposes that working memory has a fixed capacity — and that learning fails when that capacity is exceeded.
Why it matters#
Working memory can hold roughly four items at once. When a course demands more than that — through complexity, poor organisation, unnecessary media, or confusing navigation — the learner’s cognitive resources are spent managing the experience rather than acquiring the skill. Understanding cognitive load tells you where that demand comes from and which kinds you can actually reduce.
The three types#
| Type | Source | Designer’s control |
|---|---|---|
| Intrinsic | The inherent complexity of the content itself | Limited — you can sequence and scaffold, but can’t remove complexity that belongs to the subject |
| Extraneous | How the content is presented — layout, media, navigation, redundancy | High — this is where most design improvements happen |
| Germane | Mental effort spent building and consolidating new knowledge schemas | Encouraged — this is productive load; the goal is to free capacity for it |
Total cognitive load = intrinsic + extraneous + germane. When the total exceeds working memory capacity, learning stops.
Key facts#
Extraneous load is the one to eliminate. Intrinsic load belongs to the subject. Germane load is the point. Extraneous load — caused by poor design — is pure waste. Every design decision should be checked against whether it adds extraneous load without adding meaning.
The expertise reversal effect changes what’s appropriate. Content that reduces load for a novice — step-by-step guidance, worked examples, explicit scaffolding — increases load for an expert, who already has the schema and must now suppress it to follow instructions they don’t need. Design for the actual learner, not a generic one. See learner persona.
Split-attention is a common source of extraneous load. When related information is physically or temporally separated — a diagram on one screen, its explanation on another — the learner must hold both in working memory simultaneously to integrate them. Place related elements together. Multimedia learning principles address this directly.
Worked examples reduce intrinsic load during initial learning. Showing a complete, solved example before asking learners to attempt a problem reduces the mental effort of figuring out the task itself, freeing capacity to understand the underlying principle.
Redundancy increases load, not clarity. Presenting the same information twice in different forms — narration that reads the on-screen text aloud, or a diagram captioned with a full paragraph that restates it — forces the learner to reconcile two representations. Cut the duplicate.
Chunking is the primary tool for managing intrinsic load. Breaking complex content into smaller, sequenced pieces lets learners build partial schemas before integrating them. Microlearning applies this at the module level.
When to use it#
- When reviewing a course that learners report as overwhelming or confusing — diagnose whether the issue is intrinsic, extraneous, or both
- When choosing between two ways to present the same content — pick the one with lower extraneous load
- When designing for novice learners — use worked examples, scaffolding, and progressive disclosure to manage intrinsic load
- When designing for experienced learners — remove scaffolding that creates redundancy for them