What counts as transfer, and why the distinction matters
Transfer means that improvement on a trained task spills over into an untrained one. Researchers separate near transfer, which is improvement on a task that closely resembles the training, from far transfer, which is improvement on something quite different, such as an intelligence test or school grades [2][3]. Practice effects are simply getting better at the task you practiced. The question is whether working memory training produces anything beyond practice effects.
The distinction matters because working memory is strongly linked to learning. If training could raise working memory capacity in a general way, it would be a powerful educational tool. If it only teaches people to perform the training task well, its practical value is much narrower.
Training improves the trained task, but far transfer is not reliably found
In a randomized controlled trial with healthy middle-aged adults, eight weeks of adaptive n-back training produced no near or far transfer effects compared with a non-adaptive control, and no accompanying changes in brain structure or function [3]. A separate study with an active control group and multiple measures of each outcome found strong practice effects on the training task but no training-specific transfer to fluid intelligence, updating, inhibition, switching, attention, or sustained attention [5]. A meta-analytic review likewise concluded that working memory training does not reliably improve intelligence or other measures of far transfer [4].
These results point to a consistent pattern: people get better at what they practice, and sometimes at very similar tasks, but the improvement does not generalize to broad cognitive ability. The neural findings in the middle-aged adult trial reinforce this, since no structural or metabolic changes accompanied training [3].
Academic gains appear mainly when training is tied to school content
The picture is less bleak for academic outcomes when the training itself is designed around academic material. In a school-based study of 86 children aged 8 to 12, six weeks of working memory training with n-back and complex span tasks improved working memory and math performance compared with a perceptual-matching control group, with a trend toward vocabulary improvement [1]. Both groups improved in fluid intelligence and reading, which suggests those gains were not specific to working memory training [1].
A domain-specific approach argues that this is the more promising route. Rather than trying to boost a general attentional capacity, training can teach strategies that link working memory to long-term knowledge and to the specific materials being learned, such as academic tasks performed within a working memory training format [2]. This approach reports promising effects on academic outcomes, though the evidence comes from a theory paper reviewing intervention studies rather than a single definitive trial [2].
Why some studies report gains and others do not
Differences in design help explain the mixed picture. Studies with active control groups, multiple outcome measures, and larger samples tend to find no transfer beyond the trained task [5]. The middle-aged adult trial used a randomized, single-blind design with multimodal neuroimaging and also found nothing [3]. When improvements do appear in both training and control groups, that is a sign of practice effects from taking a test twice rather than an effect of training [5].
The school study suggests that trainee characteristics also matter. Within the training group, greater training performance was associated with emotional stability, conscientiousness, endurance, and teacher-reported joy of learning and social integration [1]. That does not show these traits cause better training outcomes, but it is consistent with motivation and regulation shaping how much a child gets from practice. The domain-specific view adds that training content matters: practice that resembles the academic task is more likely to show up in academic performance than practice on abstract span tasks [2].
Sources used in this answer
Effects of Working Memory Training on Children’s Memory and Academic Performance: the Role of Training Task Features and Trainee’s Characteristics
In 8–12-year-old children, six weeks of working memory training improved working memory and math performance versus a control, with a trend toward vocabulary gains, while both groups improved in fluid intelligence and reading.
The domain-specific approach of working memory training
Traditional domain-general working memory training produces reliable near transfer but not far transfer, whereas training that incorporates task-specific academic content and strategies may produce promising academic effects.
Adaptive working memory training does not produce transfer effects in cognition and neuroimaging
Eight weeks of adaptive n-back training in healthy middle-aged adults produced no near or far transfer effects and no training-related changes in brain structure, function, or metabolism.
Working memory training does not improve performance on measures of intelligence or other measures of “far transfer” evidence from a meta-analytic review
A meta-analytic review concluded that working memory training does not improve performance on measures of intelligence or other measures of far transfer.
Working memory training does not improve executive functioning or fluid intelligence
Working memory training improved the trained task but produced no transfer to fluid intelligence or specific executive functions, and both training and control groups improved on most outcomes.