The Evidence on Teaching

When Problem Solving Is Superior to Studying Worked Examples

Kalyuga, S., Chandler, P., Tuovinen, J., & Sweller, J. · 2001

grade Bquasi-experimentindependentreplicated
Sample
Trade apprentices (mechanical/electrical)
Population
Trade apprentices learning technical relay-circuit content; expertise manipulated by training.
Design
The canonical within-study crossover demonstration of expertise reversal in the worked-example paradigm.
Key findings
Direct within-study crossover: worked examples help novices, but the advantage shrinks and REVERSES (problem solving wins) as expertise grows — the worked-out steps become redundant once learners have schemas.
Genetic confound
Interaction is with PRIOR KNOWLEDGE (schema), not ability/g.
Replication notes
Canonical aptitude-treatment interaction for worked examples; replicated in classroom geometry (Bokosmaty 2015).
DOI / URL
10.1037/0022-0663.93.3.579

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Novices: worked examples vs problem solvingdirectionworked examples superiorresearcher-designedimmediate; appliedactive-alternativeend-of-treatmentdomain-skill
After training (more expert): worked examples vs problem solvingcrossoverproblem solving becomes superior (expertise-reversal crossover)researcher-designedimmediate; appliedactive-alternativeend-of-treatmentdomain-skill

Cited by