The Evidence on Teaching

Training of Executive Functions in Children: A Meta-Analysis of Cognitive Training Interventions

Birtwistle E, Chernikova O, Wunsch M, Niklas F · 2025

grade Cmeta-analysisindependentmixed
Sample
N = 57 studies
Population
Children; cognitive training of executive functions
Design
The strongest recent positive claim and therefore the steelman this topic has to survive. Graded C because it is a random-effects pool of published articles without the control-type moderation and publication-bias correction that Sala et al. (2019) and Sala & Gobet (2020) show are decisive; the reported total effect size is printed as 'g = 23', evidently 0.23. Its finding that NON-ADAPTIVE training produced larger effects than adaptive training is difficult to reconcile with any capacity-training mechanism and is more consistent with artefact.
Key findings
Total effect size g = 0.23, larger for younger children. Reports benefits for both near- and far-transfer tasks, with largest near-transfer gains for working memory and claimed gains in numeracy and literacy. Computer training effective; non-adaptive training associated with GREATER effects; school an effective context. Concludes in favour of starting cognitive training early.
Genetic confound
Low for the individual randomised contrasts; the vulnerability is publication bias and passive controls, not heredity.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Overall cognitive outcomesHedges gg = 0.23 (reported as 'g = 23')mixedpost-trainingbusiness-as-usualend-of-treatmentnear-transfer
Far-transfer tasks including numeracy and literacypooled effectreported as benefiting - not corrected for control type or publication biasmixedpost-trainingbusiness-as-usualend-of-treatmentfar-transfer

Cited by