Picture yourself taking an e-learning course. The screen shows dense blocks of text next to a complicated graphic, and at the bottom a voice reads the same text aloud. After a few minutes, your mind starts to wander: your head is full, and nothing sticks. Almost everyone knows that feeling, and cognitive load theory explains exactly why it happens.
The theory rests on one core assumption: when human working memory processes new information, its capacity is limited. It can only handle a few elements of information at once before it becomes overloaded. Lasting learning happens only when content is designed so that it moves into long-term memory and is connected there into stable schemas. Schemas bundle complex information and are what make it manageable in the first place.
Traditionally, CLT distinguishes three types of cognitive load:
| Type of load | What’s behind it |
|---|---|
| Intrinsic load | Intrinsic load depends on how complex the material is relative to the learner’s prior knowledge. You can’t lower it at will, but instructional design can influence it, for example by breaking content into individual steps or activating prior knowledge. |
| Extraneous load | Extraneous load comes from poorly designed presentation and instructions, for example when learners have to jump back and forth between text, graphic, and legend, or get distracted by unnecessary images. This load does nothing for learning and should be kept to a minimum to free up capacity. |
| Germane load | Germane load is the mental effort learners invest in processing and understanding new content and in building and automating schemas. Good instructional design deliberately encourages this productive effort. |
The current debate
More recent research questions whether “germane load” should remain a category of its own or is better understood as the share of available resources that can go toward learning once extraneous load has been reduced. Many current models therefore boil it down to the distinction between intrinsic and extraneous load and treat learning-related effort as resources that learners actively devote to learning. Less distraction alone doesn’t create understanding. In practice, though, it still pays to design learning environments so that as many cognitive resources as possible are available for the actual learning process.
Design principles based on CLT
The following principles are well supported by numerous studies and are relevant to e-learning. How well they work depends on factors such as prior knowledge:
- Worked examples: Step-by-step solutions help novices in particular build schemas and keep them from feeling overwhelmed.
- Integrated information: Graphics with text labels are easier to understand than information spread across different places (split-attention effect).
- Redundancy effect: A graphic with a matching audio explanation can reduce load; identical on-screen text running alongside the narration can add to it. Text still matters, though, for technical terms, when audio isn’t available, or when learners need it for accessibility. What helps novices can become unnecessary for more experienced learners (expertise reversal effect).
- Adaptive support: Tailor the amount and type of support to learners’ prior knowledge so that experienced learners aren’t slowed down by unnecessary repetition.
Putting it into practice
At breed.design, we apply these principles consistently. Our e-learning courses are modular, follow a clear narrative arc, and focus each module on one central message, supported by targeted visual cues. A pre-check can match the level of difficulty to learners’ prior knowledge: novices get guided examples and extra support, while more experienced learners start working on their own sooner. One way we do this is with the audience feature in our authoring tool. When choosing media, we favor clear graphics and short audio explanations over long blocks of text.
