The Evidence on Teaching

Do "Brain-Training" Programs Work?

Simons DJ, Boot WR, Charness N, Gathercole SE, Chabris CF, Hambrick DZ, Stine-Morrow EAL · 2016

grade Breviewindependentnot-applicable
Sample
Every peer-reviewed intervention study cited on the websites of leading brain-training companies
Population
All ages; the commercial brain-training evidence base as the industry itself presents it
Design
An unusually fair adversarial design: rather than selecting studies themselves, the authors evaluate exactly the papers the industry cites as its best evidence, against a pre-specified set of best practices (active controls, blinding, multiple measures per construct, preregistration, adequate power).
Key findings
Extensive evidence that training improves performance on the TRAINED tasks; less evidence for closely related tasks; little evidence for distantly related tasks or everyday cognition. Many of the cited studies had design or analysis shortcomings precluding definitive conclusions, and NOT ONE of the cited studies conformed to all the identified best practices.
Genetic confound
Low. Randomised designs throughout; the identified problems are control-group choice, blinding and power, not heritable selection.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Trained task performancequalitative strength of evidenceextensive evidence of improvementresearcher-designedpost-trainingunclearend-of-treatmentdomain-skill
Distantly related tasks and everyday cognitionqualitative strength of evidencelittle evidencemixedpost-trainingunclearend-of-treatmentfar-transfer
Industry-cited studies meeting all best practicescountzerounknownn/anonenot-applicableunclear

Cited by