In Schools, Quizzing Boosted Facts. For Concepts and Transfer, No Significant Gain So Far

A weekly quiz can get students reciting the definition of a provision. Can the same quizzing help them decide whether a pending lawsuit needs one?

What they did

Ruitenburg and five colleagues ran a meta-analysis, a study that statistically combines earlier experiments. They pooled 233 results from 53 studies published 2006–2024.

Every study tested practice on curricular content, against an active comparison. Here we focus on retrieval practice: recalling material from memory rather than restudying it.

What they found

Results are reported as an effect size, a standard measure of how far one group outscored another. Across 31 studies, retrieval practice beat restudying, copying or elaborating by 0.37, which the authors call small-to-medium. Without outliers, it was 0.30.

The benefit depended on four factors, including the type of knowledge. For factual knowledge, such as definitions, the effect was 0.53 and statistically significant. For conceptual, procedural and mixed knowledge, it was not significant.

Multiple-choice tests and tests repeating practised material showed significant gains. Comprehension, meaning the same idea reworded or with new numbers, and transfer to a new context did not.

The conceptual and comprehension estimates fell below zero, though not significantly. On mixed-format final tests, retrieval practice did significantly worse. Because the factors overlap, the authors could not tell their effects apart.

The authors offer one explanation: retrieval consolidates understanding in memory but does not build it.

They recommend that textbook authors, teachers and students in schools routinely use retrieval practice to enhance learning. Their suggestion to use it mainly for factual knowledge is more tentative, and they call it conjectural.

Limits and context

Every study involved pupils aged 6 to 18 in primary or secondary school, not university students. Dutch educational publisher ThiemeMeulenhoff supported the work, with no involvement in design, analysis or writing.

The boundary rests on few studies: three each for conceptual knowledge and transfer, with 10 and 9 results. The authors highlight estimates this thin as tentative.

They add that non-significant results should not be considered definitive evidence of a lack of effect. Comprehension and transfer of concepts and procedures, they note, have been investigated only sporadically.

Evidence strength: synthesis. This is a meta-analysis of classroom experiments. The average benefit pools 31 studies, but the boundary for concepts and transfer rests on three studies each.