What Is Cognitive Load Theory? – Designing eLearning That Fits

You've probably seen it happen. A learner opens a module, faces a screen packed with text, a diagram, a narrator reading something slightly different and three clickable icons, and simply stalls. Nothing is wrong with the learner. The design has asked their brain to do too much at once. What is cognitive load theory, and how can it help you design eLearning that fits the way people actually learn?

Cognitive load theory is a research-based model of how the limits of working memory affect learning: because people can only process a few new pieces of information at a time, instruction works best when it manages the complexity of the material, removes unnecessary mental effort and helps learners build lasting knowledge in long-term memory. The theory, developed by educational psychologist John Sweller beginning in the late 1980s, describes three kinds of load. Intrinsic load comes from the material itself, extraneous load comes from poor design, and germane load is the effort that goes into real learning. For designers, the practical message is to cut extraneous load ruthlessly and manage intrinsic load carefully.

How Does Working Memory Shape Learning?

Cognitive load theory rests on a simple model of memory. Working memory is where you hold and process new information right now. It is small and short-lived. Long-term memory is where knowledge is stored, and its capacity seems very large.

The exact limit of working memory is debated. George Miller's famous 1956 paper suggested about seven items, give or take two. Later work, including a widely cited 2001 review by Nelson Cowan, suggested the real limit for new, unrelated information is closer to four chunks. Either way, the number is small.

The good news is that long-term memory changes the math. Once knowledge is organized into what psychologists call schemas (connected mental structures), an expert can treat a whole pattern as one chunk. That is why an experienced electrician glances at a panel and sees a problem, while a novice sees dozens of separate parts. Learning, in this view, is largely the process of building and automating schemas.

What Are the Three Types of Cognitive Load?

The theory traditionally names three types. Knowing them helps you decide which parts of a design to cut and which to protect.

Intrinsic Load

Intrinsic load comes from the material's inherent complexity, especially how many elements a learner must hold in mind at once. Memorizing a list of safety symbols has low interactivity, because each symbol can be learned alone. Diagnosing a hydraulic fault has high interactivity, because pressure, flow, valves and symptoms all affect each other. You can't remove intrinsic load without changing the task, but you can sequence and scaffold it.

Extraneous Load

Extraneous load is mental effort spent on things that don't help learning. Hunting for related information spread across a screen, decoding a confusing navigation scheme, reading text while hearing different narration and ignoring decorative animations all add extraneous load. This is the load designers control most directly.

Germane Load

Germane load refers to the effort learners invest in building schemas, such as comparing examples or explaining a concept to themselves. In more recent writing, Sweller and colleagues have treated it less as a separate type and more as the working-memory resources devoted to handling intrinsic load. The practical point is the same: free up capacity so learners can spend it on the material that matters.

For a fuller overview of the theory and its history, see the background on cognitive load and how researchers have refined it.

How Does Design Change the Load on a Learner?

The chart below is a simple model, not measured data. It treats working memory as a fixed budget of 100 units and shows how the same lesson might spend that budget under three designs. The intrinsic load stays roughly constant until pre-training reduces it; the design choices change everything else.

In the cluttered version, almost nothing is left for actual learning. Cleaning up the screen frees a large share. Teaching key terms and parts beforehand, a technique called pre-training, lowers the intrinsic load of the main lesson, leaving even more room.

What Design Moves Reduce Cognitive Load in eLearning?

Researchers have identified a number of reliable effects. These translate into practical design moves.

• Use worked examples for novices. Studying a solved problem step by step usually beats solving problems from scratch early on. This is the worked example effect.
• Put related information together. Place labels on a diagram rather than in a separate legend, and keep instructions next to the thing they describe. This avoids the split-attention effect.
• Don't repeat the same words in two channels. Narration that reads on-screen text word for word can hurt learning. This is the redundancy effect.
• Pair visuals with narration for complex diagrams. Explaining an animation with audio rather than on-screen text spreads load across visual and auditory channels. This is the modality effect.
• Break complex tasks into segments. Let learners control the pace between short pieces.
• Pre-teach the parts. Introduce terms and components before showing how they work together.
• Cut the decoration. Background music, stock photos and unrelated stories compete for the same limited resources.

Many of these ideas overlap with the multimedia learning principles that Richard Mayer developed from a closely related research program, which give more detailed guidance on combining words and pictures.

What Does a Lower-Load Screen Look Like?

A before-and-after example makes the ideas concrete. Imagine a screen teaching new technicians how a pump's pressure relief valve works.

The "before" version has a paragraph of text at the top, a cutaway diagram in the middle with numbered parts, a legend at the bottom matching numbers to part names, a narrator reading the paragraph aloud and a looping background animation of water. To follow it, a learner must read, listen, look up each number in the legend, find it on the diagram and ignore the moving water. Most of their working memory goes to coordination, not understanding.

The "after" version removes the paragraph and the background animation. Part names sit directly on the diagram. The narrator explains the valve's job in a few sentences while each part highlights as it is mentioned. A "next" button lets learners pace the explanation in short segments, and a previous screen introduced the part names on their own.

The content is the same. The amount of mental effort spent on the content, rather than on the screen, is much higher.

Does Cognitive Load Change as Learners Gain Expertise?

Yes, and this is one of the most important findings for workplace training. Researchers including Slava Kalyuga described the expertise reversal effect: techniques that help novices can actually hinder experienced learners.

A step-by-step worked example is a lifeline for a beginner. For someone who already knows the procedure, it is redundant information they must process and ignore, which adds load. Experienced learners often do better with problems to solve and less guidance.

For course design, this argues for flexibility. Offer pre-tests that let experienced staff skip ahead, fade guidance as learners progress and keep detailed support available on demand rather than mandatory.

Is Shorter Always Better?

Not necessarily, and I hear this misreading of the theory more than any other. Cognitive load theory is about the load at any moment, not the total length of a course. A longer module that introduces ideas one at a time, with practice between them, can carry less load than a short module that crams everything onto three screens.

The theory also doesn't say learning should feel effortless. Effort spent on making sense of the material is the goal. Some research on "desirable difficulties," such as spaced practice and retrieval, shows that effort of the right kind improves long-term retention. The aim is to remove effort that is wasted, not effort that is productive.

What Are the Signs a Course Is Overloading Learners?

You don't need a lab to spot overload. Look for these signs in reviews, pilot tests and learner data:

• Pilot learners reread or replay the same screen several times.
• Many people fail the same assessment item, especially one that combines several ideas.
• Learners say a module was "a lot" even though it was short.
• Screens mix text, narration, a diagram and interactive elements that all compete for attention.
• SMEs insist every detail stays, and nothing is marked as reference only.

When you see these signs, start by removing extraneous load. It is the cheapest fix and often the most effective.

Designing eLearning That Fits: Putting Cognitive Load Theory to Work

Cognitive load theory explains why less clutter and better sequencing lead to more learning. Working memory is limited, long-term memory is vast, and instruction succeeds when it helps people move knowledge from one to the other without wasting capacity along the way. Cut the extraneous load, manage the intrinsic load with sequencing and pre-training, and adjust support as learners gain expertise.

Pick the densest screen in a course you own and ask three questions: What on this screen doesn't help the learner? What could be taught first? What could move into a reference? Answering those three questions is cognitive load theory in practice.

← All articles