EDPD5014 Chap.3 Learning Architecture and Cognitive Load
Learning Architecture and Cognitive Load
Learning architecture explains why the same task can be manageable for one learner and overwhelming for another. Working memory must coordinate unfamiliar elements, while organised knowledge in long-term memory can be retrieved as larger chunks. This chapter shows how prior knowledge and teacher management of proximal variables shape learning.
It distinguishes worthwhile thinking from avoidable search and uses modelling, guided practice, externalised directions and evidence-based fading without lowering the central learning demand.
A useful load analysis names what must be coordinated. A beginning reader may need to recall a correspondence, hold a sound sequence, blend the word, retain the sentence and monitor meaning.
Another learner retrieves several of those elements automatically and can devote attention to interpretation. The task appears identical on paper, but the active demand differs because available knowledge differs.
Prior knowledge should be checked at the level needed for the next step. Familiarity with a topic does not guarantee usable knowledge of a procedure or relation.
Ask the learner to explain a known example, perform a prerequisite component and apply it under a small change. Error patterns show which elements are available for retrieval and which still compete for working-memory space.
Direct guidance supplies explanations, procedures and strategies when learners cannot yet guide search from their own knowledge.
Modelling reveals a hidden decision; guided practice tests it with support; fading transfers responsibility. These stages are conditional rather than ideological. A model that never fades can produce prompt dependence, while unsupported inquiry can add search without building the required knowledge.
Keep the worthwhile demand visible. If the lesson goal is to compare causal explanations, the comparison should remain.
Directions can be displayed, labels integrated with a diagram and prerequisite terms reviewed so working memory is used for the relationship that matters. Finish with a changed example and an explanation. Accurate imitation of the model is encouraging, but transfer shows whether the learner can retrieve and use the method independently.
What this chapter covers
- 01
working memory
- 02
long-term knowledge
- 03
prior knowledge
- 04
direct guidance
- 05
worked examples
- 06
support fading
Worked example · free
Reduce load in a multi-step decoding task
- 1Display the procedure so memory is available for the sound relation.
- 1Activate known grapheme-phoneme correspondences before adding the cluster.
- 1Model the hidden blending decision and guide a changed example.
- 1Fade one prompt after accurate performance and explanation.
Key terms
- Working memory
- The limited active space in which unfamiliar information is held and coordinated.
- Long-term memory
- The organised knowledge structures that can be retrieved to support current thinking.
- Direct guidance
- Explicit explanation and strategy support designed to make required concepts and procedures learnable.
Learning Architecture and Cognitive Load FAQ
Does reducing cognitive load mean making work easy?
No. The aim is to reduce demands that do not serve the learning goal, such as remembering lengthy directions or searching between separated labels. The central comparison, inference or decoding relation can remain challenging. Good design protects the worthwhile thinking and removes competition that hides it.
When is direct guidance most useful?
Direct guidance is especially useful when learners cannot retrieve the knowledge needed to guide search. Explanation, modelling and worked examples make the relevant decisions visible. Support should change as knowledge develops, so guided practice and fading are part of the method rather than optional additions.
How can prior knowledge be checked quickly?
Use a brief task that samples the prerequisite relation directly, ask the learner to explain a familiar example and inspect error patterns. Avoid assuming that topic familiarity equals usable knowledge. The check should tell you which element can be retrieved and which still competes for working-memory space.
How can I explain cognitive demand without merely labelling a lesson overloaded?
Predict where demand arises. Count the unfamiliar relations that must be coordinated, identify information that disappears before it can be used, and separate demands essential to the goal from those introduced by presentation. Then propose one change while protecting the target operation. Evidence might include fewer sequence errors, a more complete explanation or successful performance after prompts are withdrawn.
Expertise changes the calculation, so the same scaffold should not be assumed necessary for every learner or forever. The design cycle is complete only when support, practice and fading are connected to observable performance on a changed example.
Assessment move
For one lesson, create two columns labelled core thinking and coordination burden. Place every demand in one column, then move directions, labels and reminders out of the core-thinking column where possible. Identify one prerequisite to activate, one modelled decision and one prompt to fade. Explain why each change preserves rather than weakens the learning goal.
Working through Learning Architecture and Cognitive Load in EDPD5014? Sia is AskSia’s AI Education tutor — ask any EDPD5014 Learning Architecture and Cognitive Load question and get a clear, step-by-step explanation grounded in how EDPD5014 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
