Working Memory Training: What the Science Actually Says
1. What Does "Working Memory Training" Mean?
Working memory training refers to structured, repeated practice on tasks that engage working memory — typically designed to adapt in difficulty as performance changes. The central question researchers have investigated is whether this practice produces effects that extend beyond the trained task itself.
Two types of transfer are commonly distinguished:
- Near transfer: changes in performance on tasks similar to the trained one. For example, practicing backward digit span and subsequently performing differently on forward digit span or sequencing tasks.
- Far transfer: changes in performance on tasks quite unlike the trained one — for instance, training digit span and then scoring differently on tests of fluid intelligence, reading comprehension, or real-world cognitive demands.
The distinction matters because most public discussions of "working memory training benefits" implicitly assume far transfer — that training will generalize broadly to academic performance, intelligence, or daily cognitive function. Whether that assumption holds is what the scientific debate is largely about.
2. The Near Transfer Finding
Near transfer is the more consistent finding in the research literature. Participants who engage in repeated working memory practice tend to show changes in performance on the trained tasks and on closely related ones. This pattern is broadly consistent with how the brain responds to repeated cognitive demands: the specific processes engaged become more fluent through practice.
In this sense, engaging consistently with tasks such as digit span, n-back exercises, or sorting tasks does repeatedly recruit the phonological loop and central executive in ways that researchers have measured. Whether this represents a change in the underlying capacity or a change in task-specific efficiency remains a point of ongoing discussion.
3. The Far Transfer Debate
Far transfer — whether working memory training generalizes to fluid intelligence, academic performance, or clinical outcomes — has been extensively studied and remains actively debated.
Jaeggi et al. (2008) reported that adaptive n-back training over several weeks was associated with changes in scores on tests of fluid intelligence. The report attracted substantial attention and prompted a wave of follow-up studies.
Subsequent meta-analyses reached more cautious conclusions. Melby-Lervåg and Hulme (2013) reviewed 23 studies and found that while near transfer was relatively consistently observed, evidence for far transfer to non-verbal ability, verbal ability, or word decoding was weak and inconsistent across studies. Shipstead, Redick, and Engle (2012) raised methodological concerns about many working memory training studies, questioning whether the control conditions used were sufficient to rule out placebo effects and motivational differences between groups.
The debate has not been resolved. Some research reports positive far transfer effects under specific conditions; others find those effects disappear under more rigorous controls. As of 2026, there is no scientific consensus that working memory training reliably and durably generalizes to real-world cognitive performance.
On the n-back task: Much of the far transfer literature focuses on the n-back task, which differs mechanically from digit span. Findings from n-back training research do not straightforwardly transfer to conclusions about digit span or other phonological loop-focused tasks.
4. What This Site Does and Doesn't Claim
TOMOY does not claim that practicing digit span, dual-task exercises, or filtering tasks will make you smarter, improve your academic grades, or slow cognitive decline. These claims go beyond what the research consistently supports.
What this site provides is structured practice with tasks that engage working memory components and executive functions — the same tasks used in laboratory and clinical research settings. Whether, and to what degree, such practice produces transfer effects in individual users is a question the science has not definitively answered.
Practicing these tasks trains your ability to perform these tasks. Beyond that, honest uncertainty is the appropriate stance — and we think that's worth stating directly rather than obscuring.
5. How to Think About Practice
A useful framing: these tasks impose genuine cognitive demands. Holding a six-digit sequence in mind while preparing to sort it is genuinely difficult. Performing a secondary task while maintaining a digit sequence introduces real interference. The experience of working at or near your limit — the sensation that things are just barely holding together — reflects a real cognitive state with measurable correlates in the phonological loop and central executive.
Whether that state, produced repeatedly over time, leads to lasting generalized effects is what researchers continue to investigate. What is clear is that consistent, effortful engagement is the method used in working memory training research. The task difficulty should remain challenging rather than comfortable for the practice to function as the research designs intended — because when the task becomes easy, the central executive is no longer meaningfully recruited.
Further Reading
- Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008). Improving fluid intelligence with training on working memory. Proceedings of the National Academy of Sciences, 105(19), 6829–6833.
- Melby-Lervåg, M., & Hulme, C. (2013). Is working memory training effective? A meta-analytic review. Developmental Psychology, 49(2), 270–291.
- Shipstead, Z., Redick, T. S., & Engle, R. W. (2012). Is working memory training effective? Psychological Bulletin, 138(4), 628–654.