No evidence of intelligence improvement after working memory training: a randomized, placebo-controlled study
Redick TS, Shipstead Z, Harrison TL, Hicks KL, Fried DE, Hambrick DZ, Kane MJ, Engle RW · 2013
grade Breplicationindependentfailed
Sample
Three arms: adaptive dual n-back training, adaptive visual-search active placebo, and no-contact control; 20 sessions
Population
Young adults (out of the 4-18 age band)
Design
The direct, well-powered, placebo-controlled test of the Jaeggi et al. (2008) claim, with the two design features that claim lacked: an ACTIVE placebo arm and MULTIPLE measures per construct rather than a single task. Fluid intelligence, multitasking, working-memory capacity, crystallised intelligence and perceptual speed all measured at pretest, midtest and posttest.
Key findings
Participants improved on both the dual n-back and the visual-search tasks with practice, and despite high statistical power there was NO positive transfer to any cognitive ability test. Under this archive's replication rule, this failed direct replication caps the Jaeggi claim regardless of its original effect size.
Genetic confound
Low - randomised with an active placebo.
DOI / URL
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Fluid intelligence, multitasking, WM capacity, crystallised intelligence, perceptual speed | transfer vs active placebo and no-contact control | no positive transfer to any measure | standardized | post-training | active-alternative | end-of-treatment | g |
| Trained dual n-back task | pre-post change | improved with practice | researcher-designed | 20 sessions | none | end-of-treatment | domain-skill |
Cited by
- Does working-memory or brain training transfer to intelligence and achievement?no effectconf: highgc: low