The N-Back Task and Working Memory Training
The dual n-back task became a viral object of study after a 2008 paper by Susanne Jaeggi and colleagues reported that just 19 days of training transferred to gains on Raven's Progressive Matrices — a marker of fluid intelligence. The original finding launched an industry of brain-training apps and a decade of replication attempts that have collectively cooled the initial enthusiasm without entirely refuting it.
The dual n-back task became a viral object of study after a 2008 paper by Susanne Jaeggi and colleagues reported that just 19 days of training transferred to gains on Raven's Progressive Matrices — a marker of fluid intelligence. The original finding launched an industry of brain-training apps and a decade of replication attempts that have collectively cooled the initial enthusiasm without entirely refuting it.
The brain-training literature is large, contentious, and frequently misrepresented in popular media. The most defensible summary, drawn from the Pashler et al. consensus statement and subsequent meta-analyses, is that targeted practice on a cognitive task reliably improves performance on that task and on closely related tasks, but transfer to general cognitive ability is small at best.
On The N-Back Task and Working Memory Training specifically, the evidence supports the following pattern. The intervention itself produces measurable, often substantial, improvement on the trained task within a few weeks. The improvement transfers reasonably well to highly similar tasks with the same underlying cognitive demand. Transfer to remote tasks (different cognitive demand, different stimulus material) is much weaker and often statistically indistinguishable from zero in well-controlled studies.
This pattern is not surprising once you understand the cognitive science. Practice strengthens the specific neural circuits engaged by the trained task. The closer another task is to the trained task — in stimulus material, response demand, and underlying cognitive operation — the more those strengthened circuits transfer. Distant tasks engage different circuits and benefit less.
For practical purposes, the implication is that brain training is worth doing for the specific abilities you care about, but you should not expect general cognitive enhancement from any single program. If you want to improve your performance on matrix reasoning, practice matrix reasoning; if you want to improve working memory for verbal material, practice verbal working memory tasks. Pick the training to match the goal.
The intervention with the broadest documented benefits across cognitive domains is regular aerobic exercise. Meta-analyses consistently show small-to-moderate improvements in attention, processing speed, and executive function across age groups, with the largest effects in older adults. Sleep, social engagement, and challenging cognitive activity are all also robustly associated with cognitive maintenance across the lifespan.