Near and far transfer: Is music special?
Early childhood / preschool Primary school (grades 1–6) Expertise and practice Foundations of optimal learning
Links to claims
Challenges (2)
| Claim | Basis |
|---|---|
| C30 | Reanalysis of the same data: g ≈ 0.21–0.24 after removing near-transfer controls. |
| C35 | g = 0.208 after removing near-transfer controls, pooled with passive controls; carried over from C30. |
Review state: first-pass draft, pending human review.
Source note
Syntes: Bigand & Tillmann (2022) — re-running Sala & Gobet's own data gave a small positive effect of music training (g ≈ 0.2)
Källa: Bigand, E.; Tillmann, B. (2022 (online 2021)). Near and far transfer: Is music special?. Memory & Cognition 50(2), 339–347. https://doi.org/10.3758/s13421-021-01226-6
Verified content
- [abstract] Using the authors' data file and program, they did not confirm the effect of randomisation; reanalysing gave a significant effect size, g = .234.
- [full text] Removing 21 effect sizes where the active control trained the very skill tested (e.g. phonological training tested on phonology): g = 0.208 (RVE), p < .0001; randomisation and type of control not significant moderators (p = .610 and .193).
- [full text] In those 21 comparisons, music vs near-transfer control g = −0.126 (95% CI −0.350 to 0.099, ns).
-
[full text] Excluding post-test-only studies: g = 0.243 (RVE).
-
[full text, published version, re-read 2026-10-31] The 0.208 (RVE; 95% CI 0.121 to 0.295; m = 51, k = 233; RE 0.195) is pooled over active and non-active control studies; type of control was no longer a significant moderator (p = .193) (Table 3). Sala & Gobet's own active-control estimate was 0.056 (p = .350) (Table 1). The 21 removed comparisons: 18 phonological-training controls tested on phonology, 2 reading-intervention controls tested on reading, 1 visual-art control tested on visual form (p. 343).
Limits
- [assessment] A reanalysis of the same studies, not new data; the dispute is about which comparisons count. Read in the accepted manuscript (2026-10-18) and the published version (2026-10-31). The main text gives no active-control-only estimate after the removal (it is in the online supplement, not read).
Why it is in the library
The strongest published counter to the 'music training does nothing' conclusion, from the same data; it keeps candidate K49 honest.
Added 2026-10-18 by the project (weekly source sweep). [abstract] = from the published abstract; [full text] = read in the full text; [assessment] = the project's own assessment.
Metadata as recorded
- ID
- EXP-008
- Tier
- meta
- Titel
- Near and far transfer: Is music special?
- Författare
- Bigand, E.; Tillmann, B.
- År
- 2022 (online 2021) (Memory & Cognition 50(2), 339–347)
- Källa/DOI
- https://doi.org/10.3758/s13421-021-01226-6
- URL
- https://link.springer.com/article/10.3758/s13421-021-01226-6 (accepted version on PsyArXiv: https://osf.io/gtnza/)
- Typ
- reanalysis of Sala & Gobet's (2020) data file and code: tests randomisation as moderator and removes active-control comparisons that measured near transfer for the control group
- Åldersgrupp
- foralder-tidig-barndom;grundskola-lag
- Teman
- expertis-och-ovning;optimalt-larande-grunder
- Hämtad
- 2026-10-18
- Paywall
- published version paywalled; accepted manuscript free
- Tillagd
- source sweep 2026-10-18 (kind: adjacent); links pending human review
- Verifierad
- Crossref DOI 10.3758/s13421-021-01226-6 checked 2026-10-31 (Memory & Cognition 50(2), 339–347); numbers re-read in the published version 2026-10-31
- Status
- applied 2026-10-31 (nightly attacks); C30 weakened.
The original files (PDF or abstract) are kept in the project archive and are not published here.