The Evidence on Teaching

A meta-analysis of the experimental evidence on the near- and far-transfer effects among children's executive function skills

Kassai R, Futo J, Demetrovics Z, Takacs ZK · 2019

grade Bmeta-analysisindependentunreplicated
Sample
k = 43 near-transfer comparisons; k = 17 far-transfer comparisons
Population
Children; training of working memory, inhibitory control and cognitive flexibility
Design
Asks the tightest possible transfer question: does training ONE executive function move ANOTHER executive function? This is a far shorter jump than 'working memory training raises maths achievement', which makes the null especially informative - if transfer fails at this distance, the longer jumps are hopeless.
Key findings
Significant near transfer, g+ = 0.44 (k = 43, p < .001): the interventions did train what they targeted. No convincing far transfer, g+ = 0.11 (k = 17, p = .11): training one component did not move the others. The authors argue this explains the absence of far transfer from WM training to academic skills.
Genetic confound
Low. Experimental training contrasts in children.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Trained executive-function component (near transfer)Hedges g+g+ = 0.44 (k = 43), p < .001standardizedpost-trainingbusiness-as-usualend-of-treatmentnear-transfer
Untrained executive-function components (far transfer)Hedges g+g+ = 0.11 (k = 17), p = .11 (null)standardizedpost-trainingbusiness-as-usualend-of-treatmentfar-transfer

Cited by

A meta-analysis of the experimental evidence on the near- and far-transfer effects among children's executive function skills · The Evidence on Teaching