The Evidence on Teaching

Retrieval Practice as a Tool for Teaching the Control-of-Variables Strategy: A Classroom Reversal

Kranz J, Möller J, Kaufmann E, Tempel T · 2024

grade Crctindependentunreplicatednumbers spot-checked
Sample
2 experiments, grade 5/6 (n=165, 217)
Population
Grade 5-6, authentic science textbook content (control-of-variables strategy).
Design
Classroom experiments testing retrieval practice vs restudy on complex procedural science content.
Key findings
A documented classroom REVERSAL: for a complex procedural science skill, restudy beat retrieval practice on both immediate and delayed tests. Empirically confirms the element-interactivity boundary — retrieval is not a universal method; it can lose to restudy for high-load procedural content.
Genetic confound
Domain procedural-skill outcome; no g.
Replication notes
The strongest single empirical support for the van Gog & Sweller element-interactivity boundary — a real classroom REVERSAL, not just shrinkage.
DOI / URL
10.1080/00220973.2024.2392684

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Retrieval vs restudy on complex procedural (CVS) contentdirectionRESTUDY outperformed retrieval on BOTH immediate posttest AND delayed retention/transferresearcher-designedimmediate + delayed; classroomactive-alternativeuncleardomain-skill

Cited by