Intuitive vs. Rehearsal Strategies in Dual N-Back

What are the main strategies people use in dual n-back?

There are three broad strategy categories in dual n-back: verbal rehearsal (silently repeating stimuli), intuitive or non-strategic processing (responding based on felt recognition), and chunking or shortcut strategies (grouping stimuli to reduce load). Understanding the difference matters because dual n-back is designed to train a specific cognitive system, and the strategy you use determines which system gets the workout.

When you first sit down with dual n-back, the task feels overwhelming: a square flashing on a grid, a letter playing through your headphones, and a demand to remember both from two trials ago simultaneously. The brain naturally reaches for any tool that makes this more manageable.

Verbal rehearsal is by far the most common. You silently repeat the letter sequence — “C, B, C, B, B...” — and update it with each new stimulus. It feels organized and controllable. Intuitive or non-strategic processing is harder to describe: you allow stimuli to pass without explicitly labeling them, responding based on a felt sense of familiarity or pattern rather than a tracked list. Chunking and other shortcuts involve finding patterns, grouping stimuli, or reducing the task’s complexity in ways that lower the number of items you actually have to hold in mind.

What is verbal rehearsal, and why might it be a problem?

Verbal rehearsal during dual n-back likely trains a simpler cognitive system than the one the exercise targets. Silently repeating a letter string engages what cognitive psychologists call the phonological loop, a component of working memory that stores and replays verbal information through an internal “inner voice.” But dual n-back was explicitly designed to stress the central executive: the attentional control system that coordinates multiple streams of information, updates memory, and resists interference.

Chooi and Logie (2020, Memory & Cognition) tracked participants’ error patterns across multiple experiments, including 20-session and 15-session training protocols. They found that as training progressed, the absolute number of phonological-based errors decreased — but their proportion relative to other error types did not.[4] This means that while overall performance improved, participants became relatively more reliant on verbal rehearsal over time, not less. They were getting better at n-back by leaning harder on rehearsal, not by developing the executive processes the task was meant to exercise.

Susanne Jaeggi — the researcher whose 2008 PNAS study launched the modern dual n-back movement — has been explicit on this point. In a communication to a brain training forum, she noted that some participants report the best results when they “don’t use a strategy at all and just ‘let the squares/letters flow by.’” Her group’s published work includes a formal warning: participants who develop “very task-specific strategies” may boost their n-back scores while producing gains that “obviously boosts n-back performance, but may not transfer.”[1]

One small unpublished course project from Wofford College used articulatory suppression — having participants say “the, the, the” aloud to block subvocalization — during n-back trials. Blocking rehearsal this way produced no significant change in accuracy or response time in 12 participants.[2] If rehearsal were essential to n-back performance, disrupting it should have hurt scores. The finding is suggestive but should be treated as an interesting pointer, not as evidence — this was a non-peer-reviewed undergraduate lab study with a very small sample.

What does the “intuitive” approach actually mean?

The intuitive approach means responding based on a felt recognition rather than an explicitly tracked list — allowing the task to remain effortful rather than systematizing it into something more manageable. It is easier to describe in contrast to what it is not: it is not repeating letters, not constructing a mental spatial map and scanning it, and not reducing the task to a simpler decision rule.

Morrison and Chein’s 2011 review (Psychonomic Bulletin & Review) drew a distinction between “core training” (targeting domain-general mechanisms like attentional control) and “strategy training” (learning domain-specific techniques). Their conclusion was that core training “likely” produces broader transfer effects because it targets mechanisms shared across cognitive tasks.[11] The intuitive approach is, in functional terms, a form of core training: you are forcing the central executive to do the work rather than offloading it to a simpler subsystem.

Raye and colleagues (2007) provided neurological grounding by identifying distinct neural substrates for articulatory rehearsal versus attentional refreshing — the process of briefly bringing an item back to mind through directed attention rather than verbal repetition. Refreshing is associated with the dorsolateral prefrontal cortex, a region central to executive control. Rehearsal is associated with language-related circuits such as Broca’s area and premotor cortex.[13] These are genuinely different cognitive acts, even when the behavioral result looks the same.

What intuitive training feels like in practice

For beginners, the intuitive approach is usually worse at first. Without a systematic strategy, errors are frequent and the task feels chaotic. This is not a sign that the approach is wrong — it may be a sign that the central executive is actually being engaged. The discomfort of being at the edge of capacity is arguably the point. Jaeggi’s group designed the adaptive difficulty system specifically to keep users in that zone, where automatic processing is insufficient and executive control is required.

Some practitioners describe the experience as developing a new “sense” for the task over weeks — an ability to intuit n-back matches that doesn’t feel like remembering in the traditional sense. Whether this represents central executive learning or a subtler form of task-specific skill acquisition is a question the research has not yet cleanly answered.

Why is chunking counterproductive?

Chunking reduces the working memory demand of dual n-back, which is precisely what makes it counterproductive for training purposes. The entire value of the exercise depends on keeping the cognitive load high. Anything that compresses that load makes the task easier to perform but potentially less useful as a training stimulus.

Classical chunking involves grouping individual items into larger meaningful units — remembering “FBI” instead of three separate letters. In the spatial stream of dual n-back, it tends to look like noticing that recent squares formed a recognizable shape, diagonal, or motor rhythm rather than tracking each position individually. Thalmann, Souza, and Oberauer (2019) confirmed across four experiments that chunking works by retrieving a compact representation from long-term memory that replaces multiple individual representations.[17]

By reducing the number of items actively held in mind — whether through grouping, pattern recognition, or compression — you are solving the problem more cleverly. Your n-back level rises. Your d-prime improves. But the central executive is doing less work per trial, not more. You are optimizing performance on the training task while potentially undermining the mechanism through which training was supposed to generalize.

Von Bastian and Oberauer (2014, Psychological Research) formalized this distinction between WM capacity (holding more items, associated with broad transfer) and WM efficiency (using existing capacity more cleverly via strategies). Strategy gains “would be expected to transfer to untrained tasks that use the same paradigm, but not to tasks that use a different paradigm.”[18]

Does your strategy affect whether training transfers?

The honest answer is: probably yes, but the research has not established definitively how much or through what mechanism. The far-transfer question — whether dual n-back improves fluid intelligence and other unrelated cognitive abilities — is one of the most contested debates in applied cognitive psychology. Strategy use sits at the center of that debate.

What the evidence for transfer looks like

Jaeggi et al.’s 2008 PNAS study reported fluid intelligence gains after 8–19 sessions of adaptive dual n-back, with dose-dependent effects.[1] A 2015 meta-analysis by Au and colleagues found a small but statistically significant transfer effect (g ≈ 0.24).[3] More recently, Pahor et al. (2022, Nature Human Behaviour) combined three randomized controlled trials (total N = 460) and found that near transfer to untrained n-back tasks mediated far transfer to matrix reasoning.[12]

What the skeptical evidence looks like

Redick et al. (2013) found zero transfer to fluid intelligence after 20 sessions of adaptive dual n-back when an active control group was used.[14] Melby-Lervåg, Redick, and Hulme (2016) analyzed 87 publications and 145 comparisons, concluding there was no convincing evidence of far transfer with active controls.[10] Soveri et al. (2017) found strong near-transfer within n-back paradigms but only small far-transfer (g = 0.16).[15]

Where strategy enters the argument

Hogrefe et al. (2017, Journal of Cognitive Enhancement) found that participants who adopted a task-specific strategy during n-back training achieved higher training scores but lower far transfer.[7] Studer-Lüthi et al. (2012) found the same pattern with conscientious participants: more systematic, strategy-driven training correlated with stronger near transfer but weaker generalization.[16]

The difficulty is that no study has directly manipulated strategy type (rehearsal vs. intuitive) across a full multi-week training protocol and measured far transfer with active controls. The theoretical argument is well-supported; the direct experimental evidence remains limited.

The bottom line on transfer: The research is genuinely mixed, and anyone who tells you definitively that dual n-back does or does not improve fluid intelligence is overstating what the data shows. What is clearer is that strategies which reduce the executive demands of the task — rehearsal, chunking, shortcuts — are more likely to produce task-specific gains than broadly transferable ones.

What should beginners actually do?

Based on the available research, beginners should resist verbal rehearsal, accept errors as productive, and let the adaptive difficulty system keep them at the edge of their capacity. These are evidence-informed suggestions, not definitive rules.

Do this

• Resist the urge to rehearse. When you notice yourself internally repeating letters or mentally labeling each square’s position, treat that as a prompt to stop and re-engage. The moment a strategy becomes comfortable, it is likely reducing the executive demand of the task.
• Accept errors as part of the process. More mistakes, especially early on, may mean you are staying in the uncomfortable zone where executive control is required. A low accuracy rate at a genuinely difficult level is more likely to be productive than a high accuracy rate achieved through a reliable strategy.
• Let difficulty guide you. The adaptive system is designed to keep you at your performance threshold. Trust it. If the level feels manageable and systematic, you may have found a shortcut.
• Respond on instinct when unsure. When you are uncertain whether a stimulus matches, responding based on gut feeling rather than deliberate analysis may better engage the type of recognition associated with central executive processing.

Avoid this

• Don’t rehearse out loud or subvocalize letter strings. This recruits the phonological loop rather than the executive system.
• Don’t develop spatial routines. If you find yourself mentally scanning a fixed grid pattern or tracking positions through a predictable mental routine, you have likely discovered an efficiency strategy.
• Don’t treat a rising level as the primary goal. Level improvement reflects both genuine learning and strategy acquisition — and the research suggests strategy acquisition may account for the majority of it in most users.
• Don’t panic about inconsistency. Session-to-session variability in performance is normal and may reflect genuine central executive engagement rather than a steady strategy-driven climb.

Note: The transfer column reflects theoretical predictions consistent with the research literature. Direct head-to-head comparisons across full training protocols do not yet exist.

Frequently Asked Questions

Sources

1. [1] Jaeggi, S.M., Buschkuehl, M., Jonides, J., & Perrig, W.J. (2008). Improving fluid intelligence with training on working memory. PNAS, 105(19), 6829–6833. Link
2. [2] Wofford College PSY330 Lab (2009). The effect of articulatory suppression on GSR and accuracy during an n-back task [Unpublished undergraduate course experiment]. Link
3. [3] Au, J., Sheehan, E., Tsai, N., Duncan, G.J., Buschkuehl, M., & Jaeggi, S.M. (2015). Improving fluid intelligence with training on working memory: a meta-analysis. Psychonomic Bulletin & Review, 22(2), 366–377. Link
4. [4] Chooi, W.T., & Logie, R. (2020). Changes in error patterns during n-back training indicate reliance on subvocal rehearsal. Memory & Cognition, 48(8), 1484–1503. Link
5. [5] Fellman, D., Laine, M., Waris, O., Jylkkä, J., & Nyman, T.J. (2020). The role of strategy use in working memory training outcomes. Journal of Memory and Language, 110. Link
6. [6] Forsberg, A., Fellman, D., Laine, M., Johnson, W., & Logie, R.H. (2020). Strategy mediation in working memory training in younger and older adults. Quarterly Journal of Experimental Psychology, 73(11), 1745–1764. Link
7. [7] Hogrefe, K., Studer-Lüthi, B., Kodzhabashev, A., & Perrig, W.J. (2017). Mechanisms underlying n-back training: response consistency during training influences training outcome. Journal of Cognitive Enhancement, 1(4), 406–418. Link
8. [8] Laine, M., Fellman, D., Waris, O., & Nyman, T.J. (2018). The early effects of external and internal strategies on working memory updating training. Scientific Reports, 8(1), 4045. Link
9. [9] Malinovitch, T., Jakoby, H., & Ahissar, M. (2021). Training-induced improvement in working memory tasks results from switching to efficient strategies. Psychonomic Bulletin & Review, 28(2), 526–536. Link
10. [10] Melby-Lervåg, M., Redick, T.S., & Hulme, C. (2016). Working memory training does not improve performance on measures of intelligence or other measures of “far transfer.” Perspectives on Psychological Science, 11(4), 512–534. Link
11. [11] Morrison, A.B., & Chein, J.M. (2011). Does working memory training work? The promise and challenges of enhancing cognition by training working memory. Psychonomic Bulletin & Review, 18(1), 46–60. Link
12. [12] Pahor, A., Seitz, A.R., & Jaeggi, S.M. (2022). Near transfer to an unrelated n-back task mediates the effect of n-back working memory training on matrix reasoning. Nature Human Behaviour, 6(9), 1243–1256. Link
13. [13] Raye, C.L., Johnson, M.K., Mitchell, K.J., Greene, E.J., & Johnson, M.R. (2007). Refreshing: A minimal executive function. Cortex, 43, 135–145. Link
14. [14] Redick, T.S., Shipstead, Z., Harrison, T.L., et al. (2013). No evidence of intelligence improvement after working memory training. Journal of Experimental Psychology: General, 142(2), 359–379. Link
15. [15] Soveri, A., Antfolk, J., Karlsson, L., Salo, B., & Laine, M. (2017). Working memory training revisited: A multi-level meta-analysis of n-back training studies. Psychonomic Bulletin & Review, 24(4), 1077–1096. Link
16. [16] Studer-Lüthi, B., Jaeggi, S.M., Buschkuehl, M., & Perrig, W.J. (2012). Influence of neuroticism and conscientiousness on working memory training outcome. Personality and Individual Differences, 53(1), 44–49. Link
17. [17] Thalmann, M., Souza, A.S., & Oberauer, K. (2019). How does chunking help working memory? Journal of Experimental Psychology: Learning, Memory, and Cognition, 45(1), 37–55. Link
18. [18] Von Bastian, C.C., & Oberauer, K. (2014). Effects and mechanisms of working memory training: A review. Psychological Research, 78(6), 803–820. Link