Training · Cross-cuttingMoveprovenverified

Worked Examples

also known as worked-example effect, example-based learning, solved problems, model solutions

Tags: worked-examplecognitive-loadSwellerschemanoviceexample-based

A learner studies a fully worked solution — every step shown, every decision narrated — before attempting to solve similar problems independently. Cognitive load theory explains why this works. Conventional problem-solving forces the learner to manage the search for a solution and acquire the schema for the problem type at the same time. This overloads working memory. A worked example offloads the solution search, freeing cognitive capacity for schema formation. The effect is strongest for novices; once schemas are acquired, problem-solving practice becomes more effective than further examples — the expertise reversal effect. Studied and named by Sweller and Cooper (1985) and grounded in Sweller's cognitive load theory (1988).

How the learner advances

Intent. Accelerate schema acquisition in novice learners by replacing the cognitive overhead of unguided problem-solving with the study of fully elaborated solutions.

When to apply. Apply at the beginning of a new problem type, domain, or procedure where the learner has no schema to guide problem-solving. In that state, asking them to solve without guidance would produce mostly random search rather than learning. Pair with completion problems (worked example with the final steps omitted) as a transition before full independent practice. Stop using heavily worked examples once the learner has acquired the underlying schema — switch to problem-solving practice to build fluency. Do not use worked examples as the only instructional mode: they must be followed by problems for the learner to attempt, so the schema can be tested and consolidated.

Threshold — earns the next step. The learner can, given a new problem of the same type, identify the problem class, recall the sequence of moves, and apply them without consulting the worked example — and can explain why each step was taken.

Masterpiece — the artifact that proves it. A novel problem from the same class solved correctly and efficiently on the first attempt, with the learner able to articulate the schema they applied — demonstrating that they have acquired the pattern, not just memorised the specific example.

Facets

  • Container — async
  • Mode — self-directedguidedsolo
  • Reach — individual
  • Persona — learnerinstructor
  • Craft (AI Fluency) — discernmentdiligence
  • Learner — humanautonomous-agent
  • Trainer — humanautonomous-agent

Inputs

  • Fully annotated solution — A complete, step-by-step solution to a representative problem, with each step's purpose and decision explained — not just the steps performed but why each was chosen.
  • Learner with no or minimal schema — A novice or beginner for whom the problem type is genuinely new — the worked-example effect diminishes and can reverse as expertise grows.
  • Follow-on practice problems — Problems similar to the worked example for the learner to attempt after study, allowing the newly forming schema to be applied and tested.

Outputs

  • More capable learner — A learner who has formed a problem-solving schema for the example type and can apply it to similar problems with substantially fewer errors and in less time than a learner who only practised problem-solving from the start.
  • Acquired schema — The masterpiece: an internalised problem type template — a mental structure that allows the learner to recognise future instances of the problem class and know which moves to apply, without re-deriving the solution from scratch.

Steps (5)

  1. Select a representative worked example

    Choose a problem that is prototypical for the target problem class — complex enough to show all the important moves, simple enough to be comprehensible in one pass. Avoid edge cases as first examples.

  2. Annotate every step with its rationale

    For each step, make explicit: what was done, why this move rather than an alternative, and what constraint or principle dictates this choice. The annotation is the teaching — a list of steps without explanation is a recipe, not a worked example.

  3. Have the learner study the example actively

    Ask the learner to read through the full solution and then to explain each step back without looking — self-explanation — before attempting practice problems. Passive reading without self-explanation substantially reduces the learning gain.

  4. Transition to completion problems

    Provide problems where the first steps are worked and the final steps are blank for the learner to complete. This bridges study and full independent practice, reducing the jump in cognitive demand.

  5. Move to full independent practice

    Once the learner can complete similar problems with few errors, switch entirely to problem-solving practice. Continued heavy reliance on worked examples beyond this point produces the expertise reversal effect: the worked-out steps become interference rather than support.

Principles

  • Working memory is the bottleneck: any instructional design that demands simultaneous schema search and solution search will overload it for novices. Worked examples solve only one problem at a time — schema acquisition — by removing the other.
  • Self-explanation is not optional: the learning gain from a worked example comes largely from the learner constructing their own account of why each step was taken. An example studied passively yields much weaker schema acquisition.
  • The effect reverses with expertise: for learners who have already acquired the schema, worked examples become redundant and cognitively intrusive — problem-solving practice is then the superior method.

Known uses (2)

Known failure modes (3)

  • [passive-example-study]

    Anti-pattern: the learner reads the worked example without self-explaining each step, then proceeds directly to practice problems. Passive reading produces weak schema acquisition; the learner can follow the example line by line but cannot reconstruct the reasoning on a new problem.

  • [worked-examples-for-experts]

    Anti-pattern: providing fully worked examples to learners who have already acquired the schema. The expertise reversal effect means that for these learners, the redundant worked steps become cognitive interference — problem-solving practice would produce faster and deeper learning.

  • [example-without-practice]

    Anti-pattern: presenting worked examples as the complete instructional sequence, with no follow-on problems for the learner to attempt. The schema formed from study alone remains untested and fragile — it cannot be generalised without application.

Related trainings (3)

Sources (3)

Provenance

  • Ecosystem: education, professional training, autonomous-agent training
  • Added to catalog:
  • Last updated:
  • Verification status: verified