No. 07
Does Memory Training Work? What the Research Says
Will digit span practice or brain-training apps make you smarter? What large experiments and meta-analyses say about what improves and what doesn't.
If you repeat a memory task, your score on that task goes up. Almost every study agrees on that much. The debate is about the next question: do those gains carry over (transfer) to other abilities you didn't train?
Near transfer and far transfer
- Near transfer: gains carrying over to tasks similar to the one you trained (for example, letter span after practicing digit span).
- Far transfer: gains carrying over to different kinds of abilities (for example, reasoning, reading comprehension, arithmetic or school performance). This is usually what brain-training ads promise.
A large experiment: 11,430 people
Owen et al. (2010) recruited 11,430 people through a British TV program and had them do online cognitive training for six weeks. Participants improved on the tasks they trained, but the researchers found no evidence of transfer to untrained cognitive tests. There was hardly any difference from a control group that answered simple questions by searching the internet.
Meta-analyses
Melby-Lervåg and Hulme (2013) meta-analyzed 23 studies of working memory training. Training produced short-term gains on near tasks, but those effects didn't last, and there was no convincing evidence of transfer to far tasks such as verbal ability, arithmetic or nonverbal reasoning.
Some studies once drew a lot of attention. Jaeggi et al. (2008) reported that dual n-back training raised fluid intelligence scores, but a re-test with a placebo control group by Redick et al. (2013) found no evidence of improved intelligence. Reviewing brain-training research as a whole, Simons et al. (2016) concluded that gains on the trained tasks are clear, but evidence of transfer to everyday cognitive abilities is scarce.
| What improves | Evidence |
|---|---|
| --- | --- |
| Score on the trained task | Strong — found in almost every study |
| Very similar tasks | Some — but often short-lived |
| Reasoning, schoolwork, everyday memory | Weak — not consistently found in large studies and meta-analyses |
What digit span practice gives you
As SF's case in the chunking guide shows, digit span can grow a lot once you learn a strategy. But just as his letter span stayed the same, what grew was "the skill of grouping digits". The honest interpretation of rising scores on this site is that they mostly come from getting used to the task and developing strategies.
Reasons to practice anyway
- You can enjoy the task itself, or use it as a self-check to notice days when you focus well and days when you don't.
- If you are in situations where you really need to remember numbers, like phone numbers or verification codes, learning grouping strategies is immediately useful.
- For everyday memory, getting enough sleep, writing important things down (external memory) and focusing on one thing at a time have clearer evidence than training apps.
Sources: Owen et al. (2010) Nature 465:775–778 · Melby-Lervåg & Hulme (2013) Developmental Psychology 49(2):270–291 · Jaeggi et al. (2008) PNAS 105(19):6829–6833 · Redick et al. (2013) Journal of Experimental Psychology: General 142(2):359–379 · Simons et al. (2016) Psychological Science in the Public Interest 17(3):103–186