Cognitive load theory, developed by John Sweller, identifies three load types—intrinsic, extraneous, germane—and five principles to reduce mental effort
Cognitive load is the total mental effort a person uses when processing information in working memory. Cognitive Load Theory (CLT), developed by educational psychologist John Sweller in the 1980s, explains why learning breaks down when that effort exceeds working memory's limited capacity, and it provides actionable principles for instructional designers, UX practitioners, and trainers to prevent that breakdown. Understanding how the brain manages cognitive processes during learning is the foundation for designing materials that genuinely work.
Cognitive load is the amount of mental effort invested while performing a task. It is shaped by three factors: the complexity of the material itself, the learner's existing knowledge, and the way the information is presented. Because working memory can hold and process only a limited number of elements at once, tasks or designs that demand simultaneous processing of many interacting elements quickly exhaust available capacity.
A practical illustration: a first-grade student finds it easier to solve "two packs of six candies" than "six packs of two candies." The answer is the same, but the second problem requires the child to add the number 2 six separate times rather than simply doubling 6. A student who already knows multiplication reduces that load dramatically in both cases, because stored knowledge (a schema) handles much of the processing automatically.
John Sweller, an Australian educational psychologist, introduced cognitive load theory in the 1980s, building on earlier cognitive architecture research. His model distinguishes between a limited-capacity working memory (WM), which processes new information consciously, and an essentially unlimited long-term memory (LTM), which stores automated schemas. Schemas are generalized knowledge structures that allow the brain to treat complex, familiar information as a single unit, effectively bypassing working memory's bottleneck.
Sweller and colleagues including Fred Paas and Jeroen van Merriënboer expanded the model through the 1990s and 2000s. A widely cited summary of the framework appears in Sweller, van Merriënboer, and Paas (2019). Their core insight is that most learning requires conscious, effortful processing until a schema is built and automated. Consider how a skilled adult reader instantly recognizes every handwritten variant of the letter "A," whereas a child learning to read must work to identify each one. The adult's schema handles the recognition automatically, freeing working memory for comprehension.
The New South Wales Department of Education review of cognitive load theory research describes this model as directly relevant to classroom instruction, noting that it identifies three distinct types of cognitive load that teachers and designers must balance.
Cognitive load theory identifies three types of cognitive load: intrinsic, extraneous, and germane. These three types are additive: they combine in working memory, and the total must not exceed its capacity. Effective instructional design does not always aim to minimize all load, but to balance each type in proportion to the learning objective.
| Type | Source | Design goal |
|---|---|---|
| Intrinsic | Complexity of the material and learner's prior knowledge | Match difficulty to learner expertise; sequence content to build schemas progressively |
| Extraneous | Poor instructional design, irrelevant content, or distracting elements | Minimize or eliminate; it consumes capacity without contributing to learning |
| Germane | Mental effort directed at schema formation and automation | Promote and protect; this is where learning actually happens |
Intrinsic load is determined by the inherent complexity of the material and by how many elements must be processed simultaneously (element interactivity). A learner with relevant prior knowledge stored in long-term memory will experience lower intrinsic load than a novice facing the same content, because their schemas compress many elements into one. Division carries higher intrinsic load than two-digit addition because its elements interact in more ways. Intrinsic load cannot be removed without changing the material itself, but it can be managed by sequencing instruction from simpler to more complex concepts.
Extraneous load is imposed by the way information is presented rather than by the content itself. It includes confusing layouts, redundant on-screen text that duplicates audio narration, complex jargon that adds no conceptual value, and decorative images unrelated to the task. Extraneous load consumes working memory capacity without contributing to schema formation, making it the primary target for instructional designers seeking to improve learning efficiency. Poorly structured employee training, for example, can bury key steps in walls of text, forcing learners to extract meaning before they can even begin to learn it.
Germane load is the working memory effort directed at constructing and automating schemas. This is the productive cognitive work of learning. When a learner encounters new information that connects meaningfully to existing knowledge, they invest germane load in integrating it into a schema. Well-designed instruction increases germane load by presenting material in ways that encourage active processing, such as problem-solving, comparison, or elaboration, rather than passive reading.
Researchers have derived several design principles from cognitive load theory. The five below are foundational for instructional design, e-learning, and user experience work. Each targets extraneous load specifically, freeing capacity for germane processing.
Remove material that does not directly support the learning objective. Decorative images, background music, and tangential text all add extraneous load. On a web page or in a training module, reducing visual clutter lets the learner focus on the information that matters. A cleaner interface also improves engagement: users who can find the core content quickly are more likely to act on it.
Use visual cues, such as arrows, bold text, highlights, or call-out boxes, to direct attention to key information. This principle recognizes that working memory has separate channels for visual and auditory input. Presenting an explanation as spoken audio alongside a diagram uses both channels without overloading either, whereas presenting the same explanation as on-screen text next to a diagram competes for the visual channel. Modality-aware design maximizes available capacity and increases germane load by helping learners attend to what matters.
Avoid presenting the same information in multiple formats simultaneously when each format is self-sufficient. Reading full sentences aloud while the learner can already read them on screen doubles the processing demand without adding meaning. A well-chosen combination of narration and graphic, or text and image, reduces extraneous load and allows cognitive resources to be redirected to problem-solving and schema building.
Place related text and images physically close to each other rather than separating them on the page or screen. When a diagram appears on one side of a page and its explanatory labels appear at the bottom, learners must hold information in working memory while scanning back and forth. Annotating a diagram directly, or placing a caption immediately beside the relevant part of an image, eliminates that unnecessary search and reduces extraneous load. In user interface design, this principle supports labeling interactive elements in context rather than relying on separate help documentation.
Present corresponding words and visuals at the same time rather than one after the other. When a narration describes a process and the matching diagram appears only after the narration ends, learners must hold the verbal explanation in memory until the visual arrives, increasing extraneous load. Synchronizing the two allows working memory to integrate them immediately. In practice, this also underpins the UX design technique known as chunking: breaking content into small, tightly paired units of text and visual so that each unit can be fully processed before the next appears.
Cognitive load theory is directly relevant beyond classroom settings. Employee onboarding, software training, and process documentation all impose cognitive load on users who must learn while working. When an organization introduces new software, for example, employees face high intrinsic load from unfamiliar workflows alongside extraneous load from dense manuals and context switching. Reducing that combined load through well-sequenced, in-context guidance is the core challenge of workplace learning design.
Lemon Learning's digital adoption platform addresses this by delivering step-by-step guidance directly inside software applications, reducing extraneous load by meeting users in context rather than sending them to external documentation. Well-structured learning and development programs built on a digital adoption platform apply these cognitive load principles at scale, helping employees build schemas through practice rather than passive instruction.
Cognitive load theory also informs broader instructional design strategy: sequencing content from low to high complexity, segmenting long processes into manageable steps, and aligning presentation format to learner expertise level all reflect CLT principles in action.
Understanding and applying cognitive load theory, from its origins in Sweller's working memory model to the practical principles of coherence, modality, redundancy, and contiguity, remains one of the most evidence-grounded tools available for improving how people learn and how organizations train.
Cognitive load theory was developed by John Sweller, an educational psychologist, and first published in the 1980s. Sweller, along with colleagues including Fred Paas and Jeroen van Merriënboer, later refined the model to distinguish three types of cognitive load: intrinsic, extraneous, and germane.
The three types are intrinsic load (the inherent complexity of the material based on element interactivity and learner knowledge), extraneous load (unnecessary mental effort caused by poor instructional design or irrelevant content), and germane load (the mental effort devoted to forming and automating schemas in long-term memory).
Intrinsic load is the mental effort required by the complexity of the information itself, determined by how many elements must be processed simultaneously and how they interact. It varies with the learner's prior knowledge: a novice faces higher intrinsic load than an expert for the same material.
Five evidence-based principles help reduce cognitive load: the coherence principle (remove irrelevant material), the modality principle (use visual cues to highlight key points), the redundancy principle (avoid presenting the same information in multiple redundant formats simultaneously), spatial contiguity (place related text and images close together), and temporal contiguity (present corresponding text and visuals at the same time rather than sequentially).
Learn the 10-step SOP development process to write clear, effective standard operating procedures that improve consistency, compliance, and employee
Discover why blended learning is the smarter choice for corporate training. Explore key benefits, proven models, and how to implement a flexible...
What is e-procurement software? Learn the definition, key features, types, and benefits of electronic procurement systems that help organizations buy