Yiuno

How Humans Learn -- The Science of Building Understanding

27 min read

How Humans Learn — The Science of Building Understanding

You have been learning your entire life, but nobody taught you how learning itself works. This article gives you the operating manual: what the brain is actually doing when it learns, why some strategies work and others waste your time, and how to match your approach to what you are trying to learn.

Who this is for

You are someone who learns constantly — new tools, new domains, new skills. You may be self-taught, professionally trained, or somewhere in between. You have noticed that some things stick and others don’t, but you have never systematically studied why.

You want a mental model of learning itself — not a list of study tips, but a framework that explains which tips work, when, and for whom.

What this article is NOT

This is not about AI or machine learning. It is about how you learn. The parallels with how machines process knowledge are real and fascinating, but they emerge naturally — the focus here is on the human brain and what a century of learning science has discovered about it.

Part 1 — The illusion of learning

Before looking at how learning works, you need to understand how it fails. The most common failure is not laziness or lack of talent — it is mistaking familiarity for understanding.

You read a chapter and feel like you understand it. You highlight key passages and the material feels clear. You watch a tutorial and nod along. But when someone asks you to explain what you learned — without looking at your notes — you discover that you cannot. The feeling of learning was real. The learning was not.1

graph LR
    R[Re-reading] -->|feels productive| F[Familiarity]
    F -->|mistaken for| M[Mastery]
    M -->|tested and| FAIL[Failure]
    style FAIL fill:#e74c3c,color:#fff
    style F fill:#e8b84b,color:#fff

Cognitive psychologists Robert and Elizabeth Bjork identified the core problem: storage strength and retrieval strength are not the same thing.2 Storage strength is whether information exists somewhere in your memory. Retrieval strength is whether you can actually access it when you need it. Re-reading boosts familiarity (storage) but does almost nothing for retrieval. The information is there, but you cannot find it.

This matters because all useful learning is retrieval. You never need to recall information while looking at it. You need to recall it in a meeting, an exam, a conversation, a coding session — when the source is not in front of you. Any learning strategy that does not practice retrieval is practising the wrong skill.

The key insight

Learning that feels easy is usually shallow. Learning that feels difficult is usually deep. The struggle to retrieve, to connect, to explain — that is not a sign that you are failing to learn. It is the learning itself. The Bjorks call these productive struggles desirable difficulties.2

This insight turns conventional study habits upside down. The strategies that feel most comfortable (re-reading, highlighting, passive review) are among the least effective. The strategies that feel most uncomfortable (testing yourself, spacing your practice, mixing topics) are among the most powerful. Every paradigm of learning science converges on this point.

Part 2 — Four ways to think about learning

Learning science is not a single theory — it is a landscape of paradigms, each illuminating a different aspect of what happens when someone learns. Understanding these paradigms gives you a vocabulary for diagnosing what kind of learning a situation demands and which strategies will serve it best.

graph TD
    LS[Learning Science] --> B[Behaviorism]
    LS --> C[Cognitivism]
    LS --> CO[Constructivism]
    LS --> CN[Connectivism]
    style LS fill:#4a9ede,color:#fff

Behaviorism: learning as conditioning

Core idea: Learning is a change in observable behaviour caused by stimuli and reinforcement. You do something, the environment responds, and you adjust. What happens inside the mind is not the focus — what matters is the input-output loop.3

Key figures: Ivan Pavlov (classical conditioning), B.F. Skinner (operant conditioning), Edward Thorndike (law of effect).

How it works: A behaviour followed by a reward is strengthened. A behaviour followed by a negative consequence is weakened. Repetition with feedback shapes performance over time. The cycle is simple: stimulus, response, reinforcement.3

graph LR
    S[Stimulus] --> R[Response]
    R --> F[Feedback]
    F -->|reinforcement| S
    style F fill:#5cb85c,color:#fff

What it gets right: Behaviorism explains habit formation, skill drilling, and procedural learning with remarkable precision. When you learn to type, ride a bicycle, or play scales on a piano, you are operating in behaviourist territory. The feedback loop — do, evaluate, adjust — is exactly how motor skills and routines are acquired.3

What it misses: Behaviorism struggles with understanding, meaning, and transfer. It can explain that you learned to respond correctly but not why the response is correct. A student who memorises multiplication tables through drill has learned a behaviour; whether they understand multiplication is a separate question.

When to use behaviourist principles

When learning procedures and skills — things where correct execution matters more than deep understanding. Typing drills, coding katas, language vocabulary through flashcards, musical scales. The feedback loop is the mechanism: do it, check the result, repeat.

Cognitivism: learning as information processing

Core idea: The mind is not a black box — it is an information processor. Learning is about how information is received, organised, stored, and retrieved in memory. The architecture of memory determines what sticks and what is lost.4

Key figures: George Miller (working memory limits), Richard Atkinson and Richard Shiffrin (multi-store memory model), John Sweller (cognitive load theory), Allan Paivio (dual coding).

How it works: Information enters through sensory memory, is filtered into working memory (which can hold roughly 4 items at a time), and is encoded into long-term memory through elaboration, organisation, and repetition. Retrieval cues — the connections between stored memories — determine whether you can access what you have stored.4

graph LR
    SM[Sensory Memory] -->|attention| WM[Working Memory<br/>4 items max]
    WM -->|encoding| LTM[Long-Term Memory<br/>unlimited]
    LTM -->|retrieval| WM
    style WM fill:#e8b84b,color:#fff
    style LTM fill:#5cb85c,color:#fff

The critical bottleneck is working memory. John Sweller’s cognitive load theory identifies three types of load:5

Load typeSourceCan you reduce it?
IntrinsicThe inherent complexity of the materialOnly by sequencing (simpler first)
ExtraneousPoor instruction, confusing layout, irrelevant detailYes — cut the noise
GermaneThe mental effort of building schemas and connectionsNo — this is the good load

The goal is not to minimise all cognitive load. It is to minimise extraneous load (eliminate distractions) so that working memory is free for germane load (actual learning). Intrinsic load can be managed by breaking complex material into sequential parts.5

What it gets right: Cognitivism explains why certain strategies work. Retrieval practice works because it strengthens retrieval cues. Spaced repetition works because it forces re-encoding at the point of near-forgetting. Chunking works because it compresses multiple items into a single working memory slot.4

What it misses: Cognitivism treats the learner as an individual processor and underplays the role of social context, motivation, and the learner’s active construction of meaning.

When to use cognitivist principles

When learning facts, concepts, and systems — things where understanding and recall matter. Use cognitive load theory to design your learning environment: remove distractions, break complex topics into parts, and always practice retrieval rather than re-reading.

Constructivism: learning as building

Core idea: Knowledge is not transmitted — it is constructed by the learner. You learn by connecting new information to what you already know, and the connection changes both the new and the old. Learning is an active process of building and restructuring mental models.6

Key figures: Jean Piaget (cognitive constructivism, schemas, assimilation/accommodation), Lev Vygotsky (social constructivism, zone of proximal development), Jerome Bruner (discovery learning, spiral curriculum).

How it works: Your brain maintains a network of mental frameworks called schemas — structured patterns for how things work. When you encounter new information, one of two things happens:7

  • Assimilation: The new information fits an existing schema. You absorb it smoothly. (You learn that swans are white — this fits your “bird” schema.)
  • Accommodation: The new information contradicts an existing schema. Discomfort arises. You must restructure your mental model. (You see a black swan — your “all swans are white” schema must change.)
graph TD
    NEW[New Information] --> FIT{Fits existing schema?}
    FIT -->|yes| ASS[Assimilation<br/>absorb smoothly]
    FIT -->|no| DIS[Disequilibrium<br/>discomfort]
    DIS --> ACC[Accommodation<br/>restructure schema]
    ASS --> GR[Schema grows]
    ACC --> GR
    style DIS fill:#e8b84b,color:#fff
    style GR fill:#5cb85c,color:#fff

Vygotsky added a crucial dimension: the zone of proximal development (ZPD). This is the gap between what you can do alone and what you can do with guidance. Material below the zone is too easy (no growth). Material above it is too hard (confusion and frustration). Effective learning targets the zone — challenging enough to stretch, scaffolded enough to succeed.8

What it gets right: Constructivism explains why two people can read the same text and come away with different understandings — they have different prior schemas. It explains why the order of learning matters (you can only build on what already exists). And it explains why passive consumption (reading, watching) produces less learning than active construction (explaining, building, teaching).6

What it misses: Constructivism can undervalue direct instruction. Research shows that for true novices — people with no prior schemas in a domain — guided instruction outperforms pure discovery. The “minimal guidance” debate (Kirschner, Sweller, and Clark, 2006) demonstrated that without sufficient schemas, unguided exploration overwhelms working memory.9

When to use constructivist principles

When learning complex, interconnected concepts — things where the relationships between ideas matter as much as the ideas themselves. Build from what you know. Seek the discomfort of accommodation — the moment your existing model breaks is the moment real learning happens.

Connectivism: learning as network navigation

Core idea: In a world of abundant information, learning is less about storing knowledge in your head and more about knowing where knowledge lives, how to access it, and how to connect it across sources. Knowledge is distributed across networks — human, digital, institutional.10

Key figures: George Siemens, Stephen Downes.

How it works: The learner builds and maintains a personal network of connections — to people, tools, communities, and information sources. The network itself becomes a form of knowledge. Knowing who knows something is as valuable as knowing the thing directly. The ability to navigate, evaluate, and synthesise across a network is the core skill.10

graph TD
    L[Learner] --> P[People]
    L --> T[Tools and Systems]
    L --> C[Communities]
    L --> D[Data Sources]
    P --> K[Distributed Knowledge]
    T --> K
    C --> K
    D --> K
    style L fill:#4a9ede,color:#fff
    style K fill:#5cb85c,color:#fff

What it gets right: Connectivism describes how learning actually works in practice for many professionals. A developer who learns by reading documentation, asking questions on forums, following experts, and building projects is learning connectivistically — even if they have never heard the term. The insight that knowledge lives in networks, not just in heads, is powerful.10

What it misses: Connectivism is young and contested as a formal theory. Critics argue it describes a learning environment more than a learning mechanism. It does not explain the cognitive processes that happen when someone actually integrates new knowledge — for that, you still need cognitivism and constructivism.

When to use connectivist principles

When learning rapidly evolving domains — things where the knowledge landscape changes faster than any single source can track. Build your network intentionally. Curate your sources. Invest in knowing where to find knowledge, not just in memorising it.

The paradigms compared

ParadigmLearner is…Best for learning…Key mechanismRisk
BehaviorismA responderProcedures, skills, habitsStimulus-response-feedbackRote without understanding
CognitivismA processorFacts, concepts, systemsEncoding, storage, retrievalIgnores motivation and context
ConstructivismA builderComplex, interconnected ideasSchema constructionOverwhelms novices without scaffolding
ConnectivismA navigatorEvolving, distributed knowledgeNetwork buildingBreadth without depth

No single paradigm is sufficient

The paradigms are not competing answers to the same question — they are answers to different questions. Behaviorism explains skills. Cognitivism explains memory. Constructivism explains understanding. Connectivism explains orientation. An effective learner uses all four, choosing the paradigm that matches what they are trying to learn.

Part 3 — What are you trying to learn?

Not all knowledge is the same. The revised Bloom’s Taxonomy identifies four types of knowledge, each requiring a fundamentally different learning strategy.11 Treating all knowledge the same — studying facts the way you study concepts, or practising procedures the way you memorise definitions — is one of the most common and costly mistakes a learner can make.

graph TD
    K[Knowledge Types] --> F[Factual<br/>discrete elements]
    K --> C[Conceptual<br/>relationships and structures]
    K --> P[Procedural<br/>how to do something]
    K --> M[Metacognitive<br/>knowing how you know]
    style K fill:#4a9ede,color:#fff
    style M fill:#9b59b6,color:#fff

Factual knowledge

Discrete, verifiable pieces of information. Terminology, specific details, dates, formulas. “The HTTP status code for ‘not found’ is 404.” “The French word for bread is pain.”

How to learn it: This is behaviourism’s strength. Retrieval practice, flashcards, spaced repetition. The atom of factual knowledge is the claim — a single assertion that can be true or false. The more precisely you isolate each claim, the more efficiently you can learn and test it.11

Common mistake: Trying to understand facts before memorising them. Some facts simply need to be committed to memory first — understanding comes from having enough facts to see patterns.

Conceptual knowledge

The relationships between facts — categories, principles, theories, models, structures. “REST is an architectural style that uses stateless communication.” “Natural selection favours traits that improve reproductive fitness.”

How to learn it: This is constructivism’s strength. You cannot memorise a concept the way you memorise a fact — you must build it by connecting it to what you already know. Elaborative interrogation (“why does this work?”), analogies, and teaching others are the most effective strategies. The atom here is not a single claim but a relationship between claims.11

Common mistake: Memorising definitions without building the underlying schema. You can recite “REST uses stateless communication” without understanding what statelessness means or why it matters. The definition is not the concept — the concept is the web of connections that makes the definition meaningful.

Procedural knowledge

How to do something — methods, techniques, sequences, algorithms. “To deploy, first run tests, then build, then push.” “To solve a quadratic equation, apply the quadratic formula.”

How to learn it: This is where Kolb’s experiential learning cycle is essential (see Part 4). Procedures are learned by doing — not by reading about doing. The behaviourist feedback loop (do, evaluate, adjust) is the mechanism. Deliberate practice — repetition with focused attention on errors — is the strategy.12

Common mistake: Reading about a procedure instead of performing it. You cannot learn to code by reading about coding. You cannot learn to cook by reading recipes. Procedural knowledge lives in your hands, not your head.

Metacognitive knowledge

Knowledge about your own cognition — how you learn, what strategies work for you, where your blind spots are, when you are confused versus when you are actually learning. This is the layer that makes all other learning more effective.11

How to learn it: Reflection. After a learning session, ask: What did I learn? What confused me? What strategy did I use and did it work? Judgements of learning — deliberately assessing whether you actually know something versus merely recognise it — build metacognition over time.1

Common mistake: Never reflecting on the learning process itself. Most people have strong opinions about how they learn (“I’m a visual learner”) but have never systematically tested those beliefs. The “learning styles” myth — the idea that people learn best through their preferred modality — has been repeatedly debunked by research.13

The matching principle

Factual knowledge → retrieval practice, spaced repetition (behaviourist). Conceptual knowledge → elaboration, connection, teaching (constructivist). Procedural knowledge → deliberate practice, do-evaluate-adjust (experiential). Metacognitive knowledge → reflection, self-testing, journaling. Match the strategy to the knowledge type.

Part 4 — How to sequence learning

Once you know what you are learning, the next question is when. The order in which you encounter material determines how well you can integrate it. Two frameworks offer complementary guidance.

Kolb’s experiential learning cycle

David Kolb proposed that effective learning is a cycle of four phases:14

graph LR
    CE[1. Concrete<br/>Experience] --> RO[2. Reflective<br/>Observation]
    RO --> AC[3. Abstract<br/>Conceptualisation]
    AC --> AE[4. Active<br/>Experimentation]
    AE --> CE
    style CE fill:#4a9ede,color:#fff
  1. Concrete experience — do something. Encounter the material directly, attempt the task, engage with the problem.
  2. Reflective observation — step back and examine what happened. What worked? What didn’t? What surprised you?
  3. Abstract conceptualisation — form a theory or mental model. Extract principles from the experience. Connect to existing knowledge.
  4. Active experimentation — test your theory. Apply your model to a new situation and see if it holds.

The cycle then repeats, each iteration deepening your understanding. The entry point matters less than completing the full cycle. Some people prefer to start with theory (step 3) and then test it (step 4). Others prefer to start with experience (step 1) and then reflect (step 2). Both are valid, but skipping steps is not.14

The sequencing insight

For procedural knowledge (skills, processes), start with experience. Do first, theorise later. For conceptual knowledge (theories, models, systems), start with a minimal framework, then test it against experience. For factual knowledge, sequence matters less — but spacing matters enormously.

Bruner’s spiral curriculum

Jerome Bruner proposed that complex topics should be revisited repeatedly at increasing levels of depth — a spiral rather than a linear sequence.15 The first encounter is simple and concrete. The second adds nuance. The third adds abstraction. Each pass builds on the schemas constructed in the previous one.

graph TD
    P1[Pass 1: Concrete<br/>what is it?] --> P2[Pass 2: Relational<br/>how does it connect?]
    P2 --> P3[Pass 3: Abstract<br/>why does it work?]
    P3 --> P4[Pass 4: Transfer<br/>where else does it apply?]
    style P1 fill:#4a9ede,color:#fff
    style P4 fill:#5cb85c,color:#fff

This is why good textbooks revisit the same concepts multiple times, each time in a richer context. It is why a second reading of a difficult book always yields more than the first — your schemas have changed between readings.

The spiral principle has a practical implication for self-directed learning: do not expect to understand something fully on the first pass. Plan to return. The gap between passes is not wasted time — it is the space where your schemas consolidate.

Go deeper

The concept of knowledge-granularity explores the same question from a different angle: how finely should knowledge be broken into pieces for effective learning and retrieval?

Part 5 — Six strategies that actually work

Learning science has identified a set of strategies with unusually strong and consistent evidence behind them. These are not tricks or hacks — they are fundamental mechanisms that align with how memory actually works. The Learning Scientists (a research communication initiative founded by cognitive psychologists Yana Weinstein and Megan Sumeracki) distilled them into six core strategies.16

1. Retrieval practice

What: Test yourself instead of re-reading. Try to recall information from memory before checking the answer.

Why it works: Retrieval strengthens the neural pathways you will need to access the information later. Every successful retrieval makes the next retrieval easier. Every failed retrieval (followed by checking the answer) identifies exactly what you do not know.1

How to apply: After reading a section, close the book and write down everything you remember. Use flashcards. Explain the material to someone. The format matters less than the act of pulling the information from memory without looking.

2. Spaced repetition

What: Distribute your learning over time rather than massing it into one session. Review material at increasing intervals.

Why it works: Each time you retrieve information at the edge of forgetting, you reset and extend the forgetting curve. The Bjorks call this a desirable difficulty — it feels harder than cramming, but it produces dramatically stronger long-term retention.2

How to apply: If you study something today, review it tomorrow, then in three days, then in a week, then in two weeks. Spaced repetition software (Anki, Brainscape) automates this scheduling.

3. Interleaving

What: Mix different topics or problem types during a single study session instead of practising one type repeatedly (blocking).

Why it works: Interleaving forces you to discriminate between problem types and select the right strategy for each one. Blocking lets you coast on the same strategy without having to decide when to use it. Interleaving feels harder and produces lower performance during practice — but higher performance on later tests.16

How to apply: If you are studying three topics, alternate between them within a session. If you are practising problem-solving, mix problem types rather than doing ten of the same kind in a row.

4. Elaboration

What: Ask why and how while learning. Connect new material to things you already know. Generate explanations.

Why it works: Elaboration creates multiple retrieval cues — more connections between the new information and existing knowledge mean more paths to find it later. It is the mechanism behind constructivist learning: building schemas by linking new information to old.16

How to apply: For every new concept, ask: “Why does this work?” “How does this relate to what I already know?” “What would happen if this were different?” Teaching someone else is the ultimate form of elaboration — it forces you to organise and articulate your understanding.

5. Concrete examples

What: Illustrate abstract concepts with specific, tangible instances. Move from the abstract to the concrete.

Why it works: Concrete examples provide an anchor for the abstract principle — a schema you can assimilate the abstraction into. Once you have multiple examples, you can extract the pattern that unifies them, which is the concept itself.16

How to apply: When studying an abstract idea, find or create at least two concrete examples. Ideally, the examples should be from different contexts so that the underlying principle — not the surface features — becomes salient.

6. Dual coding

What: Combine verbal and visual representations of the same information. Words and images together, not words alone.

Why it works: Allan Paivio’s dual coding theory proposes that the brain processes verbal and visual information through separate channels. Encoding information in both creates two independent retrieval paths — if one fails, the other may succeed.17

How to apply: Draw diagrams, sketch relationships, create timelines, map processes visually. Do not just supplement text with images — actively translate between the two. A Mermaid diagram that you build yourself is worth more than a pre-made infographic you passively view.

How the strategies combine

Imagine learning a new programming framework:

  1. Concrete example — follow a tutorial and build a small app (also: procedural experience via Kolb’s cycle)
  2. Elaboration — ask “why is it designed this way? How does this compare to the framework I already know?”
  3. Dual coding — draw a diagram of the framework’s architecture
  4. Retrieval practice — close the docs and try to rebuild the diagram from memory
  5. Interleaving — alternate between framework concepts and a different topic you are also learning
  6. Spaced repetition — revisit the framework concepts at increasing intervals over the next two weeks

Each strategy addresses a different aspect of learning. Together, they ensure that the knowledge is encoded deeply, connected broadly, and retrievable reliably.

Part 6 — From novice to expert

Learning strategies are not one-size-fits-all — they must change as your expertise grows. What helps a novice can actually hinder an expert, and vice versa. Understanding this spectrum lets you calibrate your approach as you develop.

graph LR
    N[Novice<br/>needs structure] -->|practice + feedback| I[Intermediate<br/>needs connection]
    I -->|abstraction + transfer| E[Expert<br/>needs challenge]
    style N fill:#e8b84b,color:#fff
    style I fill:#4a9ede,color:#fff
    style E fill:#5cb85c,color:#fff

The novice: scaffold everything

A novice has few or no schemas in the domain. Working memory is fully consumed by the material’s intrinsic complexity, leaving no room for exploration or creativity.5 For novices:

  • Direct instruction outperforms discovery. Worked examples, step-by-step guides, and explicit rules reduce extraneous load.
  • Vygotsky’s ZPD is critical. The material must be just beyond current ability — challenging but achievable with guidance.
  • Behaviorist strategies work well. Drill, repetition, and immediate feedback build the basic schemas that later learning will build on.

The intermediate: connect everything

An intermediate has basic schemas but has not yet integrated them. They know individual concepts but struggle to see how they relate and when to apply which one. For intermediates:

  • Constructivist strategies dominate. Elaboration, comparison, and teaching others are the primary tools. The goal is to build connections between existing schemas.
  • Interleaving becomes valuable. Mixing topics forces discrimination and integration.
  • Conceptual knowledge is the focus. The learner is moving from knowing what to knowing why.

The expert: challenge everything

An expert has rich, deeply connected schemas. Their working memory is freed up because most of the material is chunked into familiar patterns. For experts:

  • The expertise reversal effect: Scaffolding that helps novices (step-by-step instructions, worked examples) actually impairs expert performance by adding extraneous load to already-automated processes.18
  • Desirable difficulties are essential. Experts need increased challenge, novel problems, and cross-domain transfer to continue growing.
  • Metacognition becomes the edge. The difference between a good expert and a great one is not more knowledge — it is better awareness of what they know, what they don’t, and how their thinking works.

Where are you?

Ask yourself: “Can I explain this topic to someone with no background?” If no, you are a novice. If yes but you struggle to apply it in unfamiliar contexts, you are intermediate. If you can both explain it and transfer it to new situations, you are approaching expertise. Match your strategy to your stage.

Part 7 — The map of learning

Everything above connects into a single framework. Here is the full picture.

graph TD
    LS[How Humans Learn] --> PAR[Learning Paradigms]
    LS --> KT[Knowledge Types]
    LS --> SEQ[Sequencing]
    LS --> STRAT[Evidence-Based Strategies]
    LS --> EXP[Novice-Expert Spectrum]

    PAR --> BEH[Behaviorism<br/>skills + habits]
    PAR --> COG[Cognitivism<br/>memory + processing]
    PAR --> CON[Constructivism<br/>meaning + schemas]
    PAR --> CNN[Connectivism<br/>networks + navigation]

    KT --> FA[Factual]
    KT --> CA[Conceptual]
    KT --> PR[Procedural]
    KT --> ME[Metacognitive]

    SEQ --> KOLB[Kolb Cycle<br/>do-reflect-theorise-test]
    SEQ --> SPIR[Spiral Curriculum<br/>revisit at depth]

    STRAT --> RP[Retrieval Practice]
    STRAT --> SR[Spaced Repetition]
    STRAT --> IL[Interleaving]
    STRAT --> EL[Elaboration]
    STRAT --> CE[Concrete Examples]
    STRAT --> DC[Dual Coding]

    style LS fill:#4a9ede,color:#fff
    style PAR fill:#9b59b6,color:#fff
    style KT fill:#e8b84b,color:#fff
    style SEQ fill:#5cb85c,color:#fff
    style STRAT fill:#e74c3c,color:#fff

The framework works like this:

  1. Identify the knowledge type — is this factual, conceptual, procedural, or metacognitive?
  2. Choose the paradigm — behaviorism for skills, cognitivism for memory, constructivism for understanding, connectivism for orientation.
  3. Sequence appropriately — experience-first for procedures, framework-first for concepts, spiral for complex topics.
  4. Apply the right strategies — retrieval practice, spacing, interleaving, elaboration, examples, dual coding.
  5. Calibrate to your level — scaffold as a novice, connect as an intermediate, challenge as an expert.

What you now understand

Concepts you've gained

  • The illusion of learning — familiarity is not mastery; only retrieval practice produces durable learning
  • Four paradigms — behaviorism (conditioning), cognitivism (processing), constructivism (building), connectivism (navigating) are complementary lenses, not competing theories
  • Four knowledge types — factual, conceptual, procedural, and metacognitive knowledge each require different strategies
  • Cognitive load theory — working memory is the bottleneck; minimise extraneous load, maximise germane load
  • Schema construction — learning is assimilating and accommodating new information into mental frameworks
  • Desirable difficulties — struggle is not failure; it is the mechanism of deep learning
  • Six evidence-based strategies — retrieval practice, spaced repetition, interleaving, elaboration, concrete examples, dual coding
  • The novice-expert spectrum — learning strategies must evolve as expertise grows; scaffolding helps novices but hinders experts

Check your understanding

Where to go next

Path A: Knowledge Engineering

Explore how the principles of learning science connect to how machines represent and retrieve knowledge. Schemas become ontologies. Claims become triples. Granularity becomes chunking. Follow knowledge-epistemology for the full picture.

Best for: Someone who wants to understand the bridge between human cognition and machine knowledge systems.

Path B: Building with AI

Apply learning science to your practice of working with AI systems. The vault structure you are reading right now — concept cards, learning paths, taxonomies — is a practical application of constructivist and cognitivist principles. Follow agentic-design to see how agentic AI systems are structured.

Best for: Someone who wants to build AI-assisted learning or knowledge systems.

Path C: Deepen the Paradigms

Explore constructivism, schema-theory, and claims-and-propositions as standalone concept cards. Each goes deeper into the research and applications than this overview can.

Best for: Someone who wants to understand a specific paradigm more thoroughly before applying it.

Sources

Further reading

Resources

Footnotes

  1. Roediger, H. L., & Karpicke, J. D. (2006). Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention. Psychological Science, 17(3), 249-255. ↩ ↩2 ↩3

  2. Bjork, R. A., & Bjork, E. L. (2020). Desirable Difficulties in Theory and Practice. Journal of Applied Research in Memory and Cognition, 9(4), 475-479. ↩ ↩2 ↩3

  3. McLeod, S. (2024). Behaviorist Approach. Simply Psychology. ↩ ↩2 ↩3

  4. Brainscape Academy. (2025). The Cognitive Science of Studying: 16 Principles for Faster Learning. Brainscape. ↩ ↩2 ↩3

  5. Sweller, J. (1988). Cognitive Load During Problem Solving: Effects on Learning. Cognitive Science, 12(2), 257-285. ↩ ↩2 ↩3

  6. Piaget, J. (1971). Psychology and Epistemology: Towards a Theory of Knowledge. Viking Press. ↩ ↩2

  7. Psychology Notes HQ. (2026). Assimilation and Accommodation in Piaget’s Theory Explained. Psychology Notes HQ. ↩

  8. Psychology Notes HQ. (2025). Zone of Proximal Development: ZPD, 3 Levels & Scaffolding. Psychology Notes HQ. ↩

  9. Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why Minimal Guidance During Instruction Does Not Work. Educational Psychologist, 41(2), 75-86. ↩

  10. Siemens, G. (2005). Connectivism: A Learning Theory for the Digital Age. International Journal of Instructional Technology and Distance Learning, 2(1). ↩ ↩2 ↩3

  11. Anderson, L. W., & Krathwohl, D. R. (Eds.). (2001). A Taxonomy for Learning, Teaching, and Assessing: A Revision of Bloom’s Taxonomy of Educational Objectives. Longman. ↩ ↩2 ↩3 ↩4

  12. Ericsson, K. A., Krampe, R. T., & Tesch-Romer, C. (1993). The Role of Deliberate Practice in the Acquisition of Expert Performance. Psychological Review, 100(3), 363-406. ↩

  13. Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008). Learning Styles: Concepts and Evidence. Psychological Science in the Public Interest, 9(3), 105-119. ↩

  14. Kolb, D. A. (1984). Experiential Learning: Experience as the Source of Learning and Development. Prentice-Hall. ↩ ↩2

  15. Bruner, J. S. (1960). The Process of Education. Harvard University Press. ↩

  16. Weinstein, Y., Sumeracki, M., & Caviglioli, O. (2019). Understanding How We Learn: A Visual Guide. Routledge. ↩ ↩2 ↩3 ↩4

  17. Paivio, A. (1986). Mental Representations: A Dual Coding Approach. Oxford University Press. ↩

  18. Kalyuga, S. (2007). Expertise Reversal Effect and Its Implications for Learner-Tailored Instruction. Educational Psychology Review, 19, 509-539. ↩

Graph View

Backlinks