The Science of Learning
Understanding how the brain processes and stores information, and why this matters for instructional design.
Why Instruction Often Fails
A teacher delivers a carefully prepared lecture. Students sit attentively. The content is accurate and well-organized. Yet three weeks later, when tested on the material, most students remember almost nothing.
What went wrong?
The problem isn't the content or the teacher's knowledge. The problem is that the instruction didn't account for how learning actually works in the brain. Understanding this process is the foundation for everything that follows in this book.
The Information Processing Model
Learning requires information to move from the external world into long-term memory, where it can be retrieved and used later. This journey involves three memory systems:
Sensory Memory
All incoming information first enters sensory memory—a very brief holding area that captures everything we see, hear, and experience. Most of this information disappears within seconds unless we pay attention to it.
Instructional implication: If learners aren't attending, information never even enters the system. Gaining attention isn't just a nice opener—it's a cognitive prerequisite.
Working Memory
Information we attend to moves into working memory, where active processing happens. Here's the critical constraint: working memory has extremely limited capacity—roughly 4 items at a time—and information fades within 20-30 seconds without rehearsal.
Instructional implication: If we present too much at once, working memory overloads. Information is lost before it can be processed. This is why "covering content" often fails—dumping information faster than working memory can handle doesn't produce learning.
Long-Term Memory
For learning to persist, information must be encoded from working memory into long-term memory. Long-term memory has virtually unlimited capacity, but encoding requires active processing—connecting new information to existing knowledge, organizing it meaningfully, practicing retrieval.
Instructional implication: Learning isn't complete when students hear or see information. It's complete when they've processed it deeply enough to store it durably. This processing takes time and effort.
The Encoding Problem
The gap between "presented" and "learned" is the encoding problem. Information can sit in working memory without ever making it to long-term storage. Students may feel like they understand—they can follow the explanation in the moment—but no durable memory trace has formed.
This explains the common experience: "I understood it in class, but I couldn't do the homework."
Understanding in the moment (information active in working memory) is different from learning (information encoded in long-term memory for later retrieval).
What Enables Encoding?
Research shows that encoding is strengthened by:
- Meaningful processing — Thinking about what information means, not just what it looks like
- Connection to prior knowledge — Linking new information to existing mental structures
- Practice retrieval — Actively pulling information out of memory, not just putting it in
- Spaced repetition — Encountering information multiple times over intervals
These aren't just nice additions to instruction—they're requirements for durable learning.
Cognitive Load Theory
Cognitive load theory, developed by John Sweller, explains what happens when instruction exceeds working memory capacity.
Three Types of Load
Intrinsic load — The inherent complexity of the material. Some topics are genuinely complex; they require holding multiple elements in mind simultaneously.
Extraneous load — The unnecessary burden created by poor instructional design: confusing layouts, irrelevant information, unclear explanations.
Germane load — The productive mental effort spent on understanding and encoding. This is the "good" cognitive work.
The Design Implication
Working memory capacity is fixed. We can't expand it. But we can:
- Manage intrinsic load through sequencing and scaffolding
- Minimize extraneous load through clear, focused design
- Maximize germane load by directing attention to what matters
When extraneous load is high, there's no capacity left for learning. When instruction is well-designed, more capacity is available for actual understanding.
The Expertise Reversal Effect
What works for novices doesn't always work for experts—and vice versa.
Novice learners benefit from:
- Detailed guidance
- Worked examples
- Step-by-step scaffolding
- Explicit structure
Expert learners are hindered by these same supports. They already have schemas that organize knowledge; adding explicit structure creates redundancy and extraneous load.
Instructional implication: Design must be appropriate to learner expertise. This is why "one size fits all" instruction often fails someone—either overwhelming novices or boring experts.
What This Means for Instruction
If learning depends on:
- Attention — focusing on the right information
- Working memory processing — within capacity limits
- Encoding to long-term memory — through meaningful, active processing
- Later retrieval — practicing pulling information out
Then instruction must be designed to support each of these processes. Presenting information and hoping students "get it" leaves too much to chance.
This is where Gagné's framework becomes essential. Each of his nine events addresses a specific cognitive process required for learning. Together, they form a systematic approach to instructional design that works with—not against—how the brain learns.
Key Takeaways
- Learning requires information to move from sensory memory → working memory → long-term memory
- Working memory has severe capacity limits (≈4 items); overloading it prevents learning
- Encoding requires active processing, not passive reception
- Cognitive load must be managed: minimize extraneous, optimize germane
- What helps novices may hinder experts (expertise reversal)
- Effective instruction must support each cognitive process, not just deliver content