Cognitive Load in E-Learning Design

Course Design Choices

Narration + animationvs. on-screen text + animation

Redundant on-screen textcaptions duplicating narration

Split-attention layoutlabels separated from the diagram

Signaling cuesarrows / highlighting on key elements

Content & Learner

Content complexity (intrinsic load) 50%

Pacing (chunks / second) 0.8

Working Memory (Baddeley/Sweller, ~4 slots)

Intrinsic Extraneous Germane

Total load

Encoded to LTM

Dropped (overload)

Encoding rate

—

How it works

Mayer's Cognitive Theory of Multimedia Learning (CTML) models a learner's working memory as a narrow-capacity buffer (Baddeley's model, roughly 4 ± 1 chunks) fed by separate visual and auditory channels. Total cognitive load is the sum of three components:

Load_total = Load_intrinsic + Load_extraneous + Load_germane
Load_intrinsic  = f(content complexity)              — inherent to the material
Load_extraneous = f(redundancy, split-attention, ¬modality) — caused by bad design
Load_germane    = capacity remaining, spent on schema-building

Each incoming information chunk is rendered as a particle that must cross into the working-memory ring. If Load_total exceeds the ~4-slot capacity, the newest chunk is dropped (shown falling away, unencoded) instead of reaching long-term memory (LTM):

Modality principle — narration + animation (dual-channel) is processed more efficiently than on-screen text + animation, which forces both into the visual channel and competes for the same resource.
Redundancy principle — captions that duplicate narration add pure extraneous load with no learning benefit.
Spatial contiguity / split-attention — separating a label from the diagram it describes forces the eye to search and hold both locations in memory, raising extraneous load.
Signaling principle — cues (arrows, highlighting) reduce extraneous load by directing attention straight to what matters.

The simulator streams content chunks at the chosen pace and complexity; each chunk's radius reflects its own load cost, working-memory slots fill left-to-right, and any chunk arriving when all slots (and germane headroom) are full is rejected — a direct visualization of cognitive overload reducing learning throughput.

Intrinsic chunk Extraneous chunk Encoded to LTM

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.