Training-induced improvement in working memory tasks results from switching to efficient strategies
Open accessPublished version of a preprint
Abstract
It is debated whether training with a working memory (WM) task, particularly n -back, can improve general WM and reasoning skills. Most training studies found substantial improvement in the trained task, with little to no transfer to untrained tasks. We hypothesized that training does not increase WM capacity, but instead provides opportunities to develop an efficient task-specific strategy. We derived a strategy for the task that optimizes WM resources and taught it to participants. In two sessions, 14 participants who were taught this strategy performed as well as fourteen participants who trained for 40 sessions without strategy instructions. To understand the mechanisms underlying the no-instruction group’s improvement, participants answered questionnaires during their training period. Their replies indicate that successful learners discovered the same strategy and their improvement was associated with this discovery. We conclude that n -back training allows the discovery of strategies that enable better performance with the same WM resources.
AI snapshot
Written from the paper’s full text.
TL;DR
Training-induced improvement in spatial n-back is explained by discovering and using an efficient 1-update strategy rather than by increasing working memory capacity. Teaching this strategy produced in three sessions the performance level that uninstructed participants reached only after 25–40 sessions.
Key findings
- The strategy-instruction group (N = 14) improved faster than the no-instruction group (N = 14): Session 2 mean n = 3.25 vs. 2.56 (p = .002, Cohen’s d = 1.29); Session 3 mean n = 4.3 vs. 2.8 (p < .00001, Cohen’s d = 2.36).
- Initial performance did not differ between groups: strategy-instruction mean n = 2.65 vs. no-instruction mean n = 2.43 (p = .38, Cohen’s d = .4).
- The strategy-instruction group’s third-session performance (mean n = 4.3) did not significantly differ from the no-instruction group’s 40th-session performance (mean n = 3.96; p = .46, Cohen’s d = .35).
- In the no-instruction group, 6/14 participants explicitly reported using the efficient 1-update strategy; 4 discovered it within the first training week and 2 later (third and fourth weeks). Improvement differed significantly between those who reported the strategy and those who did not (p = .00067, Mann–Whitney U test), with no overlap between subgroups.
Methods
- Two groups were compared: a strategy-instruction group (N = 14) that received three sessions—one naïve session and two subsequent sessions with an 8-minute strategy video and comprehension check—and a no-instruction group (N = 14) that trained for 40 sessions without explicit strategy instructions, using previously published data.
- Both groups performed the same adaptive spatial n-back task: red circles appeared in one of eight positions every 2 s; blocks comprised n + 20 stimuli with six targets; difficulty increased when accuracy was ≥85% and decreased when ≤65%, with lures added at three difficulty levels.
- Analyses used mean n per session, t-tests, repeated-measures ANOVA, Mann–Whitney U tests, and individual learning curves aligned to the week participants first reported the efficient strategy; four independent reviewers blind to performance classified weekly self-reports.
Limitations
- Stated in the paper: self-reports of 8/14 no-instruction participants were too vague to determine or rule out any specific strategy; one no-instruction participant was excluded as an extreme outlier, though all reported results remained significant when including her; and what differentiates people who spontaneously discover efficient strategies from those who do not is beyond the study’s scope.
- Evident from the text: samples were small (14 per group), the strategy-instruction group was compared with a previously published no-instruction group rather than a concurrently randomized control, and transfer to untrained tasks was not tested in the new strategy-instruction group.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
0 of 59 cited sources checked · 0 supported · 0 flagged
- Findingown data
Initial performance without instructions (performance during the first session, measured by mean n per session) did not differ between groups.
“Initial performance without instructions (performance during the first session, measured by mean n per session) did not differ between groups (strategy-instruction group: mean n = 2.65, SD = .37, 95% confidence interval (CI) [2.46, 2.84]; no-instruction group: mean n = 2.43, SD = .44, 95% CI [2.2, 2.66]; p = .38, in a two-tailed, two-sample unequal variance t test, Cohen’s d = .4).”
Quote in the paperSubstantial claim - Findingown data
Across the first three sessions, there was a significant main effect of session on n-back performance (F(2) = 62.99, p < .0001, ηp2 = .83), indicating a general improvement as participants completed more sessions.
“A repeated-measures two-way analysis of variance (ANOVA) for the three first sessions (2 groups × 3 sessions( showed a significant main effect of session, F(2) = 62.99, p < .0001, ηp2 = .83, indicating a general improvement as participants completed more sessions;”
Quote in the paperNot a substantial claim - Findingown data
In the third session, performance was very different between the groups, with the strategy-instruction group performing significantly higher than the no-instruction group.
“In the third session, performance was very different between the groups (strategy-instruction: mean n = 4.3, SD = .66, 95% CI [3.95, 4.65]; no-instruction: mean n = 2.8, SD = .6, 95% CI [2.49. 3.11]; p < .00001, Cohen’s d = 2.36).”
Quote in the paperSubstantial claim - Findingown data
In two sessions, 14 participants who were taught this strategy performed as well as fourteen participants who trained for 40 sessions without strategy instructions.
“In two sessions, 14 participants who were taught this strategy performed as well as fourteen participants who trained for 40 sessions without strategy instructions.”
Quote in the paperSubstantial claim - Findingown data
Performance slopes abruptly rise following the discovery of the efficient strategy.
“Their slopes abruptly rise following this discovery.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Findingown data
Six of the 14 participants in the no-instruction group explicitly indicated that they used the efficient 1-update strategy starting from a particular training week.
“However, six (out of 14) participants explicitly indicated that they used the efficient 1-update strategy starting from a particular training week (see Fig. 6).”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Findingown data
Participants who did not report discovering a strategy during their 8 training weeks trained 8 weeks without discovering the efficient strategy.
“Plots of participants who did not report discovering a strategy during their 8 training weeks begin at “−8”, indicating that they trained 8 weeks without discovering the efficient strategy.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants who did not report using the efficient strategy did not use it.
- Findingown data
Explicit reports of the efficient strategy are associated with a sharp rise in the performance curve.
“Explicit reports of the strategy are associated with a sharp rise in the performance curve.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Findingown data
In this study, performance during the first session was not a good predictor of learning rate.
“We do not find evidence for that in our group of participants; performance during the first session is not a good predictor of learning rate, though perhaps mean performance during the first session already includes some learning.”
Quote in the paperNot a substantial claim - Findingown data
Large individual differences in training-induced improvement within the no-instruction group delineate participants who discovered an efficient strategy spontaneously from those who did not.
“Additionally, we showed that the large individual differences in training-induced improvement within a no-strategy group delineate the participants who discovered an efficient strategy spontaneously versus those who did not.”
Quote in the paperSubstantial claim - Interpretationreasoning
Raising the target level of significance would have rendered the reported transfer nonsignificant.
“raising the target level of significance would have rendered the reported transfer nonsignificant (e.g., Anguera et al., 2013).”
Quote in the paperNot a substantial claim - Interpretationreasoning
Differences in transfer may stem from the mere existence of a training protocol rather than from the specific training protocol of the experimental group.
“Therefore, differences in transfer may stem from the mere existence of a training protocol rather than from the specific training protocol of the experimental group”
Quote in the paperSubstantial claim - Premisecitation
The typical size of trained groups is small (~15 per group).
“Since the typical size of trained groups is also small (~15 per group; Chooi & Thompson, 2012; De Simoni & von Bastian, 2018; Gibson et al., 2012; Redick et al., 2013; Thompson et al., 2013)”
Quote in the paperSubstantial claim0 of 3 sources checked - Interpretationreasoning
Switching the strategy from updating all slots to updating only the attended slot, with the index looping over the number of items-first-second-third-first-reduces the WM resources required for attaining the same level of success.
“As with letters, switching the strategy from updating all slots to updating only the attended slot, with the index looping over the number of items—first-second-third-first—reduces the WM resources required for attaining the same level of success.”
Quote in the paperSubstantial claim - Premisecitation
Studies that used active control groups typically found either small near-only transfer or no transfer at all.
“Studies that used active control groups (trained with a similarly demanding task and a similar reward protocol) typically found either small near-only transfer (Linares, Borella, Teresa, Id, & Carretti, 2019) or no transfer at all (Jakoby, Raviv, Jaffe-dax, Lieder, & Ahissar, 2019).”
Quote in the paperSubstantial claim0 of 2 sources checked- cites Linares et al., 2019
Not checked in the free run.
- cites Jakoby et al., 2019 source
Not checked in the free run.
- Study design and conductcitation
Statistical power analysis shows that to get the power of 0.9 with α = 0.01, based on the results of the first and the last session of the no-instruction group (Jakoby et al., 2019), at least 10 participants are needed per group.
“Statistical power analysis shows that to get the power of 0.9 with α = 0.01, based on the results of the first and the last session of the no-instruction group (Jakoby et al., 2019), at least 10 participants are needed per group (Table 1).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Jakoby et al., 2019 source
Not checked in the free run.
- Premisecitation
The shift in the attended slot in WM does not put load on WM.
“The shift in the attended slot in WM does not put load on WM (Myers, Chekroud, Stokes, & Nobre, 2018).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Myers, Chekroud, Stokes, & Nobre, 2018
Not checked in the free run.
- Interpretationown data
This training study provides further support for the strategy mediation theory of working memory improvement compared with the capacity theory.
“This training study provides further support for the strategy mediation theory of WM improvement (Dunning & Holmes, 2014; Peng & Fuchs, 2017), compared with the capacity theory (Engle & Kane, 2003).”
Quote in the paperSubstantial claim0 of 3 sources checked - Interpretationown data
n-back training allows the discovery of strategies that enable better performance with the same WM resources.
“We conclude that n-back training allows the discovery of strategies that enable better performance with the same WM resources.”
Quote in the paperSubstantial claim- Rests on an assumption: Training does not increase WM capacity.
- Premisecitation
The efficient strategy identified in this study is the same or very similar to the strategy described previously for n-back with letters and with digits.
“In fact, and as mentioned before, our strategy is the same or very similar to the strategy described previously for n-back with letters (Laine et al., 2018) and with digits (Fellman et al., 2020) in their studies characterizing the continuous effect of training with and without an explicit strategy.”
Quote in the paperNot a substantial claim0 of 2 sources checked - Interpretationown data
Training-induced improvement in the n-back task can be fully explained by the discovery of a task-specific strategy.
“Our results indicate that training-induced improvement in the n-back task can be fully explained by the discovery of a task-specific strategy.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Rests on an assumption: Participants who did not report using the efficient strategy did not use it.
- Interpretationreasoning
Efficient strategies need not reduce the total number of operations per trial; they need only reduce the operations that put load on working memory.
“Importantly, there is no need to reduce the total number of operations per trial—only the operations that put load on WM.”
Quote in the paperSubstantial claim - Premisecitation
Unlike n-back tasks that use items that can be easily named, such as letters or digits, spatial tasks cannot benefit from subvocal rehearsal of the items.
“Unlike n-back tasks that use items that can be easily named, like letters or digits, spatial tasks cannot benefit from subvocal rehearsal of the items (see Chooi & Logie, 2020).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Chooi & Logie, 2020
Not checked in the free run.
- Interpretationown data
The high cross-participant variability in the no-instruction group is largely explained by the difference in improvement between those who discovered the efficient strategy and substantially improved and those who did not, whose improvement ranged from small to none.
“Thus, the high cross-participant variability in the no-instruction group (see Fig. 4) is largely explained by the difference in improvement between those who discovered the efficient strategy, and substantially improve, and those who did not, whose improvement ranges between small to none at all.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Rests on an assumption: Participants who did not report using the efficient strategy did not use it.
- Interpretationreasoning
Performance in tasks with the same structure as the trained task may be improved.
“Thus, performance in tasks with the same structure may be improved (Fellman et al., 2020; Redick, 2019), as previously reported (Linares et al., 2019).”
Quote in the paperSubstantial claim0 of 3 sources checked - Interpretationown data
The self-reported strategy described by participants maps directly to the implementation of the efficient 1-update strategy with n = 4.
“Figure 5 illustrates how this account directly maps to the implementation of the efficient 1-update strategy with n = 4.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Study design and conductown data
The verbal reports of eight of the 14 participants in the no-instruction group were too vague to determine or rule out any specific strategy.
“Analyzing the self-reports of participants in the no-instruction group was complex, because the verbal reports of eight (out of 14) participants were too vague to determine or rule out any specific strategy.”
Quote in the paperNot a substantial claim - Interpretationreasoning
This study adds to recent studies by specifying an efficient strategy for the spatial n-back task.
“Our study adds to recent studies (Fellman et al., 2020; Laine et al., 2018) in specifying an efficient strategy for the case of the spatial n-back task.”
Quote in the paperSubstantial claim0 of 2 sources checked - Interpretationreasoning
The finding that improvement is explained by task-specific strategy discovery explains previous findings of no transfer, or only very-near transfer, to other types of the n-back task.
“This finding explains previous findings of no transfer, or only very-near transfer to other types of the n-back task (e.g., Jakoby et al., 2019; Linares et al., 2019).”
Quote in the paperSubstantial claim0 of 2 sources checked- cites Jakoby et al., 2019 source
Not checked in the free run.
- cites Linares et al., 2019
Not checked in the free run.
- Interpretationreasoning
Improvement in the n-back task does not indicate a general enhancement of working memory capacity.
“Hence, improvement does not indicate a general enhancement of WM capacity.”
Quote in the paperSubstantial claim - Interpretationown data
The increase in cross-participant variability during training results from substantially different rates of improvement across participants, particularly when no explicit instructions are given.
“This pattern results from the substantially different rates of improvement across participants, particularly when no explicit instructions are given.”
Quote in the paperSubstantial claim - Premisecitation
Scanning through items without changing their slot in working memory does not add to working memory load.
“For example, scanning through items without changing their slot in WM does not add to WM load (Myers et al., 2018).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Myers et al., 2018
Not checked in the free run.
- Interpretationown data
The cross-participant variability in the no-instruction group results from different success rates in discovering the efficient task-specific 1-update strategy.
“However, Figs. 4 and 5 show that this cross-participant variability results from different success rates in discovering the efficient task-specific 1-update strategy.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Study design and conductown data
Participants in the strategy-instruction group all indicated in questionnaires that they understood and used the instructed strategy during the two postinstruction sessions.
“Participants in the strategy-instruction group indicated in the questionnaires that they all understood and used the instructed strategy during the two postinstruction sessions.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Rests on an assumption: The strategy-instruction group's better performance is due to their use of the instructed strategy rather than other differences between groups.
- Interpretationreasoning
Training-based improvement probably results from using a set of more efficient task-specific operations.
“Hence, training-based improvement probably results from using a set of more efficient task-specific operations.”
Quote in the paperSubstantial claim - Interpretationreasoning
Training working memory does not open a common bottleneck.
“Training WM does not open a common bottleneck.”
Quote in the paperSubstantial claim - Interpretationreasoning
When the same operations are administered to the same sequences of stimuli repeatedly, as in word reading, working memory operations are replaced with chunking and schemas.
“When the same operations are administered to the same sequences of stimuli repeatedly, as in word reading, we replace the WM operations with chunking and schemas.”
Quote in the paperSubstantial claim - Interpretationreasoning
When the crux of the task is using operations on untrained stimuli sequences, as in reading nonwords, chunking cannot replace online computations.
“But when the crux of the task is using the operations on untrained stimuli sequences (as in reading nonwords), chunking cannot replace online computations.”
Quote in the paperSubstantial claim - Interpretationreasoning
Successfully trained individuals do not perform the same operations faster or better.
“Successfully trained individuals do not perform the same operations faster or better.”
Quote in the paperSubstantial claim - Interpretationreasoning
Successfully trained individuals change the set of operations used to solve the task.
“They change the set of operations used to solve the task.”
Quote in the paperSubstantial claim
- May combine two claims. 28
- The quote only partly states this claim. 1
- Classified as probably not a substantial claim. 26
May combine two claims.
Findingown dataThe replies of the no-instruction group indicate that successful learners discovered the same strategy and their improvement was associated with this discovery.
“Their replies indicate that successful learners discovered the same strategy and their improvement was associated with this discovery.”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataWithin three sessions, performance of the strategy-instruction group reached the level attained by the no-instruction group only after 25-40 sessions, and did not significantly differ from the no-instruction group's final performance during the fortieth session.
“In fact, within three sessions performance of the strategy-instruction group reached the level attained by the no-instruction group only after 25–40 sessions, and did not significantly differ from the no-instruction group’s final performance during the fortieth session (strategy-instruction group third session: mean n = 4.3, SD = .66, 95% CI [3.95, 4.65]; no-instruction group 40th session: mean n”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataAll participants in the strategy-instruction group improved, though not to the same extent.
“All participants improved in the strategy-instruction group, though not to the same extent.”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown dataIndividual changes in n during training in the no-instruction group ranged between −0.376 and 2.72, while in the strategy-instruction group all changes were positive.
“Changes in n of individuals in the no-instruction group range between −0.376 and 2.72, while in the strategy instruction group, all changes are positive.”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataMean performance in the second session significantly differed between the two groups, with the n of the strategy-instruction group being significantly higher than that of the no-instruction group.
“Mean performance in the second session significantly differed between the two groups, with the n of the strategy-instruction group (mean n = 3.25, SD = .43, 95% CI [3.02, 3.48]) being significantly higher than that of the no-instruction group (mean n = 2.56, SD = .58, 95% CI [2.36, 2.82]; p = .002, Cohen’s d = 1.29)”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataSix participants in the no-instruction group discovered the efficient strategy: four within their first training week (first five sessions), one in the third week, and one in the fourth week.
“Four of them discovered this strategy within their first training week—that is, within their first five sessions (and hence their plots begin at “0”)—and the other two discovered it later, one in the third week and the other in the fourth week.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
May combine two claims.
Findingown dataThere was a significant main effect of group on n-back performance (F(1) = 17.47, p < .0001, ηp2 = .4), with the strategy-instruction group showing significantly better performance overall than the no-instruction group.
“a significant main effect of group, F(1) = 17.47, p < .0001, ηp2 = .4, indicating different performance levels in the two groups, with the strategy-instruction group showing significantly better performance overall;”
Quote in the paperSubstantial claimThe quote only partly states this claim.
Findingown dataThere was a significant session × group interaction (F(2) = 26.7, p < .0001, ηp2 = .68), indicating faster improvement in the strategy-instruction group than in the no-instruction group across the first three sessions.
“and crucially—a significant interaction between session and group, F(2) = 26.7, p < .0001, ηp2 = .68, indicating faster improvement in the strategy-instruction group.”
- Also stated in: § Improvement was substantially faster in the strategy-instruction group · plain
“The initial and end points are similar, but the strategy-instruction group improved much faster.”
Quote in the paperSubstantial claim- Also stated in: § Improvement was substantially faster in the strategy-instruction group · plain
May combine two claims.
Findingown dataIn the no-instruction group, improvement differed significantly between those who explicitly deciphered the efficient strategy and those who did not (p = .00067 in a Mann-Whitney U test), with no overlap between the degree of improvement of individuals in the two subgroups.
“In the no-instruction group, improvement differed significantly between those who explicitly deciphered the efficient strategy and those who did not (p = .00067 in a Mann–Whitney U test), with no overlap between the degree of improvement of individuals in the two subgroups.”
Quote in the paperSubstantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Rests on an assumption: Participants who did not report using the efficient strategy did not use it.
May combine two claims.
Findingown dataCross-participant variability in n-back performance was similar between groups in the first session (strategy-instruction SD = .37, no-instruction SD = 0.44) and increased for both groups during training, with a greater increase in the no-instruction group (strategy-instruction SD = .66, no-instruction SD = 1.13).
“Cross-participant variability is similar for both groups in the first session (strategy-instruction SD = .37, no-instruction SD = 0.44 in the no-instruction group), and it increases for both groups during training, with a greater increase observed in the no-instruction group (strategy-instruction SD = .66, no-instruction SD = 1.13).”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenThe same efficient strategy was repeated across participants.
“Though to the best of our understanding, the same efficient strategy was repeated across participants, we do not claim that this is the only possible strategy.”
Quote in the paperNot a substantial claim- Rests on an assumption: Participants' self-reports accurately reflect their strategy use.
- Rests on an assumption: The efficient strategy reported by different participants is the same strategy.
May combine two claims.
Premisenone givenThe no-contact group is not given monetary (or equivalent) rewards or stimulating personal attention, both of which positively impact performance.
“The no-contact group is not given monetary (or equivalent) rewards or stimulating personal attention, both of which positively impact performance.”
Quote in the paperSubstantial claim0 of 3 sources checkedMay combine two claims.
PremisecitationDifferentiating lures from targets is difficult-participants tend to press the button upon detecting a repetition, even with different intervals.
“Differentiating lures from targets is difficult—participants tend to press the button upon detecting a repetition, even with different intervals (Duncan, 2003).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Duncan, 2003 source
Not checked in the free run.
May combine two claims.
InterpretationcitationThe efficient 1-update strategy requires, at most, updating the content of one WM slot (compared with n slots in the naïve strategy), keeping track of which position is now relevant, and an attention shift, which does not occupy additional WM resources.
“Thus, the strategy requires, at most, updating the content of one WM slot (compared with n slots in the naïve strategy), keeping track of which position is now relevant, and an attention shift, which does not occupy additional WM resources (Myers et al., 2018).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Myers et al., 2018
Not checked in the free run.
Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).
PremisecitationThe importance of a strategy that reduces WM requirements has been gradually acknowledged.
“Indeed, the importance of a strategy that reduces WM requirements has been gradually acknowledged (Redick, 2019).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Redick, 2019 source
Not checked in the free run.
May combine two claims.
PremisecitationMost of the participants in the prior no-instruction training improved significantly in the spatial n-back task, but showed no transfer to other WM tasks.
“Most of those participants improved significantly in this task, but showed no transfer to other WM tasks.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Jakoby et al., 2019 source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductown dataThe data of one participant from the no-instruction group were excluded from the analysis because her performance on the task before training was an extreme outlier (z score of over 2.5 in each session).
“The data of one participant from the no-instruction group were excluded from the analysis (data of 14 participants are reported) because her performance on the task before training was an extreme outlier (z score of over 2.5 in each session).”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductreasoningThe choice of 14 participants in the strategy-instruction group was aimed to match the number of participants who had previously formed the no-instruction group.
“The choice of 14 participants in the strategy-instruction group was aimed to match the number of participants who had previously formed the no-instruction group.”
Quote in the paperNot a substantial claimMay combine two claims.
PremisereasoningThe magnitude of transfer, when such was reported, is usually small, and is difficult to dissociate from a null result, since it is not resilient to correction for multiple comparisons.
“The magnitude of transfer, when such was reported, is usually small, and is difficult to dissociate from a null result, since it is not resilient to correction for multiple comparisons.”
Quote in the paperSubstantial claim0 of 3 sources checkedClassified as probably not a substantial claim.
PremisecitationThe use of a task-specific strategy may facilitate training-induced improvement.
“The few recent studies that addressed this question suggest that the use of a task-specific strategy may facilitate training-induced improvement (Fellman et al., 2020; Laine, Fellman, Waris, & Nyman, 2018; Linares et al., 2019; Redick et al., 2013).”
Quote in the paperSubstantial claim0 of 3 sources checkedMay combine two claims.
PremisecitationThe strategy described by Laine et al. (2018) for letters includes no shifts in WM representation; rather than shifting the items in WM slots, it shifts the slot to which attention is allocated in the given WM representation.
“By contrast, the strategy described by Laine et al. (2018) for letters (see Fig. 1, right column) and its parallel for the spatial task that was derived by us (see Fig. 2, right column) includes no shifts in WM representation. Rather than shifting the items in WM slots, it shifts the slot to which attention is allocated in the given WM representation.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Laine et al. (2018) source
Not checked in the free run.
Classified as probably not a substantial claim.
PremisereasoningThe naive strategy requires an update in the content of all WM slots-a shift in the slots of all n items in memory upon each stimulus presentation.
“Thus, the naive strategy requires an update in the content of all WM slots—a shift in the slots of all n items in memory upon each stimulus presentation.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim.
PremisecitationReliable transfer of WM training occurs only when the new task is very similar in structure ("near") to the trained task and requires similar cognitive routines.
“Gathercole, Dunning, Holmes, and Norris (2019) concluded that reliable transfer of WM training occurs only when the new task is very similar in structure (“near”) to the trained task and requires similar cognitive routines.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Gathercole, Dunning, Holmes, and Norris (2019) source
Not checked in the free run.
Probably not a substantial claim: may not pass S4 (based).
PremisereasoningNaïve participants can typically perform well with n = 1 and n = 2, but find n ≥ 3 extremely challenging.
“Naïve participants can typically perform well with n = 1 and n = 2, but find n ≥ 3 extremely challenging.”
Quote in the paperNot a substantial claimMay combine two claims.
PremisecitationMost training studies found substantial improvement in the trained task, with little to no transfer to untrained tasks.
“Most training studies found substantial improvement in the trained task, with little to no transfer to untrained tasks.”
- Also stated in: § Abstract · hedged
“Training WM unequivocally yields improvement in the trained task, but the generalization of this benefit has been heatedly debated (e.g., Redick, 2019).”
Quote in the paperSubstantial claim0 of 3 sources checked- Also stated in: § Abstract · hedged
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
Interpretationown dataThe excluded participant's first self-report indicates the discovery of the efficient strategy, and we believe she had discovered the strategy early in the first session.
“Her first self-report indicates the discovery of the efficient strategy. We believe that she had discovered the strategy early in the first session.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim. May combine two claims.
Study design and conductcitationSince improvement was larger than three standard deviations (Jakoby et al., 2019), 14 participants per group were sufficient in both the previously studied group (no-instruction) and the newly added one (strategy-instruction group).
“Since improvement was larger than three standard deviations (Jakoby et al., 2019), 14 participants per group were sufficient in both the previously studied group (no-instruction) and the newly added one (strategy-instruction group).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Jakoby et al., 2019 source
Not checked in the free run.
May combine two claims.
PremisecitationThe effect size of transfer to untrained tasks, if any, is small: ~0.3 standard deviations in methodologically weaker studies, ~0.01 in methodologically sound studies.
“However, the effect size of transfer to untrained tasks, if any, is small: ~0.3 standard deviations in methodologically weaker studies, ~0.01 in methodologically sound studies (Melby-Lervåg et al., 2016; Redick, 2019).”
Quote in the paperSubstantial claim0 of 2 sources checkedClassified as probably not a substantial claim.
InterpretationreasoningThe task becomes difficult with n ≥ 3 because participants need to update the content of n positions (slots) in WM following each item presentation.
“The reason the task becomes difficult with n ≥ 3 is that participants need to update the content of n positions (slots) in WM following each item presentation.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim.
Premisecitation"Far" transfer is more characteristic of studies without an active-control group.
“Others (e.g., Jacoby & Ahissar, 2013, 2015; Melby-Lervåg et al., 2016; Redick, 2019; Sala & Gobet, 2017; Simons et al., 2016) noted that “far” transfer is more characteristic of studies without an active-control group.”
Quote in the paperSubstantial claim0 of 3 sources checkedClassified as probably not a substantial claim.
InterpretationreasoningThe same n-updates versus 1-update strategy applies to the spatial task.
“When the task is spatial (see Fig. 2), spatial locations of stimuli need to be retained in WM. Thus, the same n-updates versus 1-update strategy applies to the spatial task.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductnone givenAll the reported results remain statistically significant when including this participant.
“Importantly, all the reported results remain statistically significant when including this participant.”
Quote in the paperNot a substantial claimMay combine two claims.
PremisecitationLures increase WM load and the requirements of cognitive control.
“We included lures because it has been previously shown that lures increase WM load and the requirements of cognitive control (e.g., Redick & Lindsey, 2013; Szmalec, Verbruggen, & Kemps, 2011).”
Quote in the paperSubstantial claim0 of 2 sources checked- cites Redick & Lindsey, 2013 source
Not checked in the free run.
- cites Szmalec, Verbruggen, & Kemps, 2011
Not checked in the free run.
Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).
Interpretationnone givenAll efficient strategies across working memory tasks are expected to have something in common.
“Yet, theoretically, we expect all efficient strategies across WM tasks to have something in common.”
Quote in the paperNot a substantial claimMay combine two claims.
InterpretationreasoningPerformance in untrained tasks will improve only when the discovered efficient strategy applies and its relevance is transparent to the participants.
“Based on our results, we predict that performance in untrained tasks will improve only when the discovered efficient strategy applies, and its relevance is transparent to the participants.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim.
Study design and conductown dataFour independent reviewers, who were familiar with the task yet blind to participants' performance, evaluated based on each week's reply of each participant whether she or he had used the efficient 1-update strategy.
“To decide which strategy had been used and whether it was modified with training, we asked four independent reviewers, who were familiar with the task yet blind to participants’ performance, to evaluate based on each week’s reply of each participant whether she or he had used the efficient 1-update strategy, and if so.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S4 (based).
Premisenone givenMost of the literature on strategy mediation theory does not specify the strategies that would be efficient for working memory tasks.
“Yet most of the literature of strategy mediation theory does not specify the strategies that would be efficient for WM tasks.”
Quote in the paperNot a substantial claimMay combine two claims.
InterpretationreasoningPerformance in most other working memory tasks, and even in n-back tasks for which the efficient strategy is difficult to implement, is not expected to benefit from the discovery of this strategy.
“However, performance in most other WM tasks, and even in n-back tasks for which the strategy is difficult to implement (Jakoby et al., 2019), is not expected to benefit from the discovery of this strategy.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Jakoby et al., 2019 source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationLarge cross-participant variability has also been observed in previous no-instruction training studies.
“Large cross-participant variability was also observed in previous (no-instruction) training studies (e.g., Jaeggi, Buschkuehl, Jonides, & Shah, 2011).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Jaeggi, Buschkuehl, Jonides, & Shah, 2011 source
Not checked in the free run.
Probably not a substantial claim: may not pass S4 (based).
Interpretationnone givenAll efficient strategies reduce the number of manipulations in working memory per trial compared with the naïve strategy.
“We assert that all efficient strategies reduce the number of manipulations in WM per trial, compared with the naïve strategy.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
InterpretationcitationThis account is in line with the proposal of Gathercole et al. (2019) that training-induced transfer occurs only when participants have acquired a new complex cognitive skill during training and when that skill can be applied to a novel task.
“This account is in line with the proposal of Gathercole et al. (2019), who claim that training-induced transfer occurs only when participants have acquired a new complex cognitive skill during training, and when that skill can be applied to a novel task.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Gathercole et al. (2019) source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductnone givenThe answers to all questionnaires were read and analyzed only after the experiment had ended, so that participants would not be affected by the experimenters' expectations.
“The answers to all questionnaires were read and analyzed only after the experiment had ended, so that participants would not be affected by the experimenters’ expectations.”
Quote in the paperNot a substantial claimMay combine two claims.
InterpretationreasoningThe description of rehearsal or chunking does not capture the unique structure of working memory tasks, which are designed to require online manipulations that cannot be organized into fixed chunks, or at least not in any straightforward manner.
“The description of rehearsal or chunking (McNamara & Scott, 2001; Peng & Fuchs, 2017; Turley-Ames & Whitfield, 2003) does not capture the unique structure of WM tasks, which are designed to require online manipulations that cannot be organized into fixed chunks, or at least not in any straightforward manner.”
Quote in the paperSubstantial claim0 of 3 sources checkedProbably not a substantial claim: may not pass S2 (checkable), S4 (based).
InterpretationreasoningThis type of change is likely to underlie the acquisition of all expertise.
“This type of change is likely to underlie the acquisition of all expertise.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S4 (based).
Interpretationnone givenSubvocal verbalization can help memorize which of the n items in working memory should be updated in each step.
“Still, subvocal verbalization can help memorizing which of the n items in WM should be updated in each step.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThere is some evidence that individuals with larger working memory pretraining are the ones who benefit more from training.
“There is some evidence that individuals with larger WM pretraining are the ones who benefit more from training (Foster et al., 2017; Redick, 2019; Wiemers, Redick, & Morrison, 2019).”
Quote in the paperNot a substantial claim0 of 3 sources checkedMay combine two claims.
PremisecitationFellman et al. (2020) taught participants a specific strategy for the n-back task with digits via the Internet and found that using this strategy facilitated initial learning.
“For example, Fellman et al. (2020) taught participants a specific strategy for the n-back task with digits via the Internet, and found that using this strategy facilitated initial learning.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Fellman et al. (2020) source
Not checked in the free run.
May combine two claims.
InterpretationcitationInstruction was beneficial in Fellman et al. (2020), but its impact was smaller than in the present study, perhaps due to the nature of Internet training or the difference in task stimuli.
“Thus, instruction was beneficial, but its impact was smaller than in our case, perhaps due to the nature of Internet training, or due to the difference in task stimuli.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Fellman et al. (2020) source
Not checked in the free run.
May combine two claims.
PremisecitationIn Fellman et al. (2020), the advantage of explicit instruction was partial: the strategy group continued to improve throughout the 12 training sessions, and performance of the tutored and nontutored groups was similar following the first few training sessions.
“However, the advantage of explicit instruction, observed during the first few sessions, was partial, as the strategy group continued to improve throughout the 12 training sessions, and performance of the tutored and nontutored groups was similar following the first few training sessions.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Fellman et al. (2020) source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
InterpretationreasoningThe enhanced efficiency of instruction in this study may be partly due to the effort put on strategy clarity (presenting instructions with a video clip) and verification of participants' understanding.
“Another potential contribution to the enhanced efficiency of instruction in our study, which yielded the equivalent of more than 25 uninstructed training sessions following only three sessions, was the effort we put on strategy clarity (presenting instructions with a video clip [https://youtu.be/-21tuZQNMMQ]), and verification of participants' understanding.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S2 (checkable). May combine two claims.
InterpretationreasoningBetter understanding of these task-specific strategies may both teach us about the structure of working memory and facilitate performance in tasks that heavily load on limited working memory resources.
“Better understanding of these task-specific strategies may both teach us about the structure of WM and facilitate performance in tasks that heavily load on our limited WM resources.”
Quote in the paperNot a substantial claim
The paper relies on this without support.
Training does not increase WM capacity.
“We hypothesized that training does not increase WM capacity, but instead provides opportunities to develop an efficient task-specific strategy.”
The paper relies on this without support.
Participants' self-reports accurately reflect their strategy use.
“Participants in the strategy-instruction group indicated in the questionnaires that they all understood and used the instructed strategy during the two postinstruction sessions.”
“However, six (out of 14) participants explicitly indicated that they used the efficient 1-update strategy starting from a particular training week (see Fig. 6).”
The paper relies on this without support.
Participants who did not report using the efficient strategy did not use it.
“Plots of participants who did not report discovering a strategy during their 8 training weeks begin at “−8”, indicating that they trained 8 weeks without discovering the efficient strategy.”
The paper relies on this without support.
The strategy-instruction group's better performance is due to their use of the instructed strategy rather than other differences between groups.
“Participants in the strategy-instruction group indicated in the questionnaires that they all understood and used the instructed strategy during the two postinstruction sessions.”
“a significant main effect of group, F(1) = 17.47, p < .0001, ηp2 = .4, indicating different performance levels in the two groups, with the strategy-instruction group showing significantly better performance overall;”
The paper relies on this without support.
The efficient strategy reported by different participants is the same strategy.
“Though to the best of our understanding, the same efficient strategy was repeated across participants, we do not claim that this is the only possible strategy.”
“In fact, and as mentioned before, our strategy is the same or very similar to the strategy described previously for n-back with letters (Laine et al., 2018) and with digits (Fellman et al., 2020) in their studies characterizing the continuous effect of training with and without an explicit strategy.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The pattern of decreased activation is compatible with a training-related rebalancing of recruitment within a fronto-parietal-cerebellar circuit.
“Together, these patterns are compatible with a training-related rebalancing of recruitment within a fronto–parietal–cerebellar circuit, although attenuation could reflect strategy change, habituation, reduced post-training control demands, or increased neural efficiency79.”
From Meta-analysis of computerised working memory training: behavioural gains, training parameters, transfer mechanisms, and neural correlates (2026) · cites this paper as [79]
The Discussion and conclusions reports that improvement on the n-back task can be explained by a switch to an efficient strategy, supporting strategy change as a possible explanation. It does not address decreased brain activation or rebalancing of recruitment in a fronto–parietal–cerebellar circuit; the citation therefore does not support that neural interpretation. (LLM, conf=0.84)
Experimental — accuracy not yet measuredAttenuation of activation could reflect strategy change, habituation, reduced post-training control demands, or increased neural efficiency.
“Together, these patterns are compatible with a training-related rebalancing of recruitment within a fronto–parietal–cerebellar circuit, although attenuation could reflect strategy change, habituation, reduced post-training control demands, or increased neural efficiency79.”
From Meta-analysis of computerised working memory training: behavioural gains, training parameters, transfer mechanisms, and neural correlates (2026) · cites this paper as [79]
The Discussion and conclusions (page unavailable) links training-related improvement to switching to strategies that reduce working-memory manipulations, supporting strategy change as a possible explanation. Our copy does not connect those strategies to attenuated neural activation or discuss habituation, reduced post-training control demands, or increased neural efficiency as explanations for attenuation. (LLM, conf=0.87)
Experimental — accuracy not yet measuredBehavioural improvements may partly reflect practice-related gains and strategy optimisation rather than durable increases in cognitive capacity.
“Finally, behavioural outcomes were largely task-based, and observed improvements may partly reflect practice-related gains and strategy optimisation rather than durable increases in cognitive capacity79.”
From Meta-analysis of computerised working memory training: behavioural gains, training parameters, transfer mechanisms, and neural correlates (2026) · cites this paper as [79]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Two different references have identical journal pages and identifiers
MinorReferences 10 and 24 give different titles and authors for an article at the same journal, volume, issue, and pages. They also list the same PubMed and PMC identifiers.
“Bejjanki VR Zhang R Li R Pouget A Green CS Lu Z Action video game play facilitates the development of better perceptual templates Proceedings of the National Academy of Sciences of the United States of America 2014 111 47 16961 16966 10.1073/pnas.1417056111 25385590 PMC4250112”
“Green CS Bavelier D Learning, attentional control, and action video games Proceedings of the National Academy of Sciences of the United States of America 2014 111 47 16961 16966 10.1016/j.cub.2012.02.012 25385590 PMC4250112”
Cited prior study does not match the described n-back study
MinorThe paper describes Jakoby et al. (2019) as a study of spatial n-back training and transfer to other working-memory tasks, but its reference entry describes auditory-frequency training and transfer to pitch-discrimination tasks.
“The data of this group have been previously published, in a study aimed at assessing transfer to other WM tasks, which found no transfer (Jakoby et al., 2019).”
“Jakoby H Raviv O Jaffe-dax S Lieder I Ahissar M Auditory frequency discrimination is correlated with linguistic skills, but its training does not improve them or other pitch discrimination tasks”
Cites
The magical number 4 in short-term memory: A reconsideration of mental storage capacity
Cowan, N. (2001) · Behavioral and Brain Sciences · doi:10.1017/s0140525x01003922 · Cited by 6,779 · cited 1× · see citing works
“The number of items which can be explicitly accessed and manipulated, i.e. WM capacity, is extremely limited and poses a strict bottleneck to human cognition (Cowan, 2001).”
BackgroundBaddeley, A. (1992) · Science · doi:10.1126/science.1736359 · Cited by 5,258 · see citing works
Working memory: Looking back and looking forward
Baddeley, A. (2003) · Nature Reviews Neuroscience · doi:10.1038/nrn1201 · Cited by 5,200 · cited 1× · see citing works
“Working memory (WM) is defined as the ability to simultaneously retain and manipulate information within short time periods (Baddeley, 1992b, 2003).”
BackgroundWorking memory, short-term memory, and general fluid intelligence: A latent-variable approach.
Engle, R. W., Tuholski, S. W., Laughlin, J. E., & Conway, A. R. A. (1999) · Journal of Experimental Psychology: General · doi:10.1037/0096-3445.128.3.309 · Cited by 3,078 · see citing works
Improving fluid intelligence with training on working memory
Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0801268105 · Cited by 2,187 · see citing works
When and where do we apply what we learn?: A taxonomy for far transfer
Barnett, S. M., & Ceci, S. J. (2002) · Psychological Bulletin · doi:10.1037/0033-2909.128.4.612 · Cited by 2,076 · see citing works
Is working memory training effective? A meta-analytic review
Melby-Lervåg, M., & Hulme, C. (2013) · Developmental Psychology · doi:10.1037/a0028228 · Cited by 1,782 · see citing works
Video game training enhances cognitive control in older adults
Anguera, J. A., Boccanfuso, J., Rintoul, J. L., Al-Hashimi, O., Faraji, F., Janowich, J., Kong, E., Larraburo, Y., Rolle, C., Johnston, E., et al. (2013) · Nature · doi:10.1038/nature12486 · Cited by 1,666 · see citing works
Do “brain-training” programs work?
Simons, D. J., Boot, W. R., Charness, N., Gathercole, S. E., Chabris, C. F., Hambrick, D. Z., & Stine-Morrow, E. A. L. (2016) · Psychological Science in the Public Interest · doi:10.1177/1529100616661983 · Cited by 1,183 · see citing works
Executive Attention, Working Memory Capacity, and a Two-Factor Theory of Cognitive Control
Engle, R. W., & Kane, M. J. (2003) · Psychology of Learning and Motivation · doi:10.1016/s0079-7421(03)44005-x · Cited by 1,104 · see citing works
Is working memory training effective?
Shipstead, Z., Redick, T. S., & Engle, R. W. (2012) · Psychological Bulletin · doi:10.1037/a0027473 · Cited by 1,101 · see citing works
The concurrent validity of the <i>N</i> -back task as a working memory measure
Jaeggi, S. M., Buschkuehl, M., Perrig, W. J., & Meier, B. (2010) · Memory · doi:10.1080/09658211003702171 · Cited by 981 · see citing works
Melby-Lervåg, M., Redick, T. S., & Hulme, C. (2016) · Perspectives on Psychological Science · doi:10.1177/1745691616635612 · Cited by 910 · see citing works
Meta-analysis of action video game impact on perceptual, attentional, and cognitive skills.
Bediou, B., Adams, D. M., Mayer, R. E., Tipton, E., Green, C. S., & Bavelier, D. (2018) · Psychological Bulletin · doi:10.1037/bul0000130 · Cited by 807 · see citing works
Short- and long-term benefits of cognitive training
Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Shah, P. (2011) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1103228108 · Cited by 769 · see citing works
Learning, Attentional Control, and Action Video Games
Green, C. S., & Bavelier, D. (2012) · Current Biology · doi:10.1016/j.cub.2012.02.012 · Cited by 713 · see citing works
Improving fluid intelligence with training on working memory: a meta-analysis
Au, J., Sheehan, E., Tsai, N., Duncan, G. J., Buschkuehl, M., & Jaeggi, S. M. (2015) · Psychonomic Bulletin & Review · doi:10.3758/s13423-014-0699-x · Cited by 671 · see citing works
Redick, T. S., Shipstead, Z., Harrison, T. L., Hicks, K. L., Fried, D. E., Hambrick, D. Z., Kane, M. J., & Engle, R. W. (2013) · Journal of Experimental Psychology: General · doi:10.1037/a0029082 · Cited by 617 · see citing works
Working-memory capacity explains reasoning ability—and a little bit more
Süß, H., Oberauer, K., Wittmann, W. W., Wilhelm, O., & Schulze, R. (2002) · Intelligence · doi:10.1016/s0160-2896(01)00100-3 · Cited by 591 · see citing works
Karbach, J., & Verhaeghen, P. (2014) · Psychological Science · doi:10.1177/0956797614548725 · Cited by 589 · see citing works
Working memory training revisited: A multi-level meta-analysis of n-back training studies
Soveri, A., Antfolk, J., Karlsson, L., Salo, B., & Laine, M. (2017) · Psychonomic Bulletin & Review · doi:10.3758/s13423-016-1217-0 · Cited by 350 · see citing works
Continental-scale animal tracking reveals functional movement classes across marine taxa
Brodie, S. J., Lédée, E. J. I., Heupel, M. R., Babcock, R. C., Campbell, H. A., Gledhill, D. C., Hoenner, X., Huveneers, C. P. M., Jaine, F. R. A., Simpfendorfer, C. A., et al. (2018) · Scientific Reports · doi:10.1038/s41598-018-21988-5 · Cited by 348 · see citing works
Does Working Memory Training Transfer? A Meta-Analysis Including Training Conditions as Moderators
Schwaighofer, M., Fischer, F., & Bühner, M. (2015) · Educational Psychologist · doi:10.1080/00461520.2015.1036274 · Cited by 332 · see citing works
Complex span and n-back measures of working memory: A meta-analysis
Redick, T. S., & Lindsey, D. R. B. (2013) · Psychonomic Bulletin & Review · doi:10.3758/s13423-013-0453-9 · Cited by 305 · see citing works
Working memory capacity and strategy use
McNamara, D. S., & Scott, J. L. (2001) · Memory & Cognition · doi:10.3758/bf03195736 · Cited by 296 · see citing works
Strategy training and working memory task performance
Turley-Ames, K. (2003) · Journal of Memory and Language · doi:10.1016/s0749-596x(03)00095-0 · Cited by 289 · see citing works
Hitch, G. J., Towse, J. N., & Hutton, U. M. Z. (2001) · Journal of Experimental Psychology General · doi:10.1037/0096-3445.130.2.184 · Cited by 271 · see citing works
Placebo effects in cognitive training
Foroughi, C. K., Monfort, S. S., Paczynski, M., McKnight, P. E., & Greenwood, P. M. (2016) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1601243113 · Cited by 270 · see citing works
Video game training does not enhance cognitive ability: A comprehensive meta-analytic investigation
Sala, G., Tatlidil, K. S., & Gobet, F. (2018) · Psychological Bulletin · doi:10.1037/bul0000139 · Cited by 257 · see citing works
Working memory training involves learning new skills
Gathercole, S. E., Dunning, D. L., Holmes, J., & Norris, D. (2019) · Journal of Memory and Language · doi:10.1016/j.jml.2018.10.003 · Cited by 225 · see citing works
Working memory training in typically developing children: A meta-analysis of the available evidence.
Sala, G., & Gobet, F. (2017) · Developmental Psychology · doi:10.1037/dev0000265 · Cited by 216 · cited 1× · see citing works
“It seems that “far” transfer is found only in methodologically flawed studies (Melby-Lervåg et al., 2016; Redick, 2019; Sala & Gobet, 2017).”
BackgroundWorking memory training does not improve intelligence in healthy young adults
Chooi, W.-T., & Thompson, L. A. (2012) · Intelligence · doi:10.1016/j.intell.2012.07.004 · Cited by 209 · see citing works
Failure of Working Memory Training to Enhance Cognition or Intelligence
Thompson, T. W., Waskom, M. L., Garel, K.-L. A., Cardenas-Iniguez, C., Reynolds, G. O., Winter, R., Chang, P., Pollard, K., Lala, N., Alvarez, G. A., et al. (2013) · PLoS ONE · doi:10.1371/journal.pone.0063614 · Cited by 208 · cited 2× · see citing works
“…of trained groups is also small (∼ 15 per group, Chooi & Thompson, 2012; De Simoni & von Bastian, 2018; Gibson et al., 2012; Redick et al., 2013; Thompson et al., 2013), raising the target level of significance would have rendered the reported transfer non-significant (e.g. Anguera et al.,…”
BackgroundAction video game play facilitates the development of better perceptual templates
Bejjanki, V. R., Zhang, R., Li, R., Pouget, A., Green, C. S., Lu, Z., & Bavelier, D. (2014) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1417056111 · Cited by 190 · see citing works
Swanson, H. L. (2004) · Memory & Cognition · doi:10.3758/bf03195856 · Cited by 139 · cited 1× · see citing works
“Indeed, WM capacity is strongly correlated with fluid intelligence (Engle, Laughlin, Tuholski, & Conway, 1999; Süß, Oberauer, Wittmann, Wilhelm, & Schulze, 2002) and with academic achievements (Baddeley, 1992a; Bayliss, Jarrold, Baddeley, & Gunn, 2005; Hitch, Towse, & Hutton, 2001; Swanson, 2004).”
Control of interference during working memory updating.
Szmalec, A., Verbruggen, F., Vandierendonck, A., & Kemps, E. (2011) · Journal of Experimental Psychology: Human Perception and Performance · doi:10.1037/a0020365 · Cited by 133 · see citing works
Does working memory training promote the use of strategies on untrained working memory tasks?
Dunning, D. L., & Holmes, J. (2014) · Memory & Cognition · doi:10.3758/s13421-014-0410-5 · Cited by 129 · see citing works
Why does working memory span predict complex cognition? Testing the strategy affordance hypothesis
Bailey, H., Dunlosky, J., & Kane, M. J. (2008) · Memory & Cognition · doi:10.3758/mc.36.8.1383 · Cited by 121 · see citing works
Au, J., Gibson, B. C., Bunarjo, K., Buschkuehl, M., & Jaeggi, S. M. (2020) · Journal of Cognitive Enhancement · doi:10.1007/s41465-020-00164-6 · Cited by 91 · see citing works
Working memory updating and binding training: Bayesian evidence supporting the absence of transfer.
De Simoni, C., & von Bastian, C. C. (2018) · Journal of Experimental Psychology: General · doi:10.1037/xge0000453 · Cited by 84 · see citing works
The relationship between short‐term memory and working memory: Complex span made simple?
Bayliss, D. M., Jarrold, C., Baddeley, A., & Riby, D. M. (2004) · Memory · doi:10.1080/09658210344000332 · Cited by 78 · see citing works
Do the effects of working memory training depend on baseline ability level?
Foster, J. L., Harrison, T. L., Hicks, K. L., Draheim, C., Redick, T. S., & Engle, R. W. (2017) · Journal of Experimental Psychology: Learning, Memory, and Cognition · doi:10.1037/xlm0000426 · Cited by 76 · see citing works
Peng, P., & Fuchs, D. (2017) · Journal of Learning Disabilities · doi:10.1177/0022219415594609 · Cited by 73 · see citing works
The early effects of external and internal strategies on working memory updating training
Laine, M., Fellman, D., Waris, O., & Nyman, T. J. (2018) · Scientific Reports · doi:10.1038/s41598-018-22396-5 · Cited by 71 · see citing works
The Hype Cycle of Working Memory Training
Redick, T. S. (2019) · Current Directions in Psychological Science · doi:10.1177/0963721419848668 · Cited by 69 · cited 8× · see citing works
“Indeed, the need for a strategy that reduces WM requirements has been gradually acknowledged (Redick, 2019).”
BackgroundResultInfluentialBenefits of flexible prioritization in working memory can arise without costs.
Myers, N. E., Chekroud, S. R., Stokes, M. G., & Nobre, A. C. (2017) · Journal of Experimental Psychology Human Perception & Performance · doi:10.1037/xhp0000449 · Cited by 68 · cited 1× · see citing works
“For example, scanning through items without modifying their slot in WM does not add to load (Myers et al., 2018).”
BackgroundThe role of strategy use in working memory training outcomes
Fellman, D., Jylkkä, J., Waris, O., Soveri, A., Ritakallio, L., Haga, S., Salmi, J., Nyman, T. J., & Laine, M. (2020) · Journal of Memory and Language · doi:10.1016/j.jml.2019.104064 · Cited by 64 · see citing works
The Influence of Individual Differences in Cognitive Ability on Working Memory Training Gains
Wiemers, E. A., Redick, T. S., & Morrison, A. B. (2019) · Journal of Cognitive Enhancement · doi:10.1007/s41465-018-0111-2 · Cited by 48 · see citing works
Gibson, B. S., Kronenberger, W. G., Gondoli, D. M., Johnson, A. C., Morrissey, R. A., & Steeger, C. M. (2012) · Journal of Applied Research in Memory and Cognition · doi:10.1016/j.jarmac.2012.06.005 · Cited by 44 · see citing works
Linares, R., Borella, E., Lechuga, M. T., Carretti, B., & Pelegrina, S. (2019) · PLOS ONE · doi:10.1371/journal.pone.0211321 · Cited by 38 · cited 4× · see citing works
“We asked – what is it that participants learn which enables their substantial improvement in a challenging updating task designed to require limited-capacity online manipulations? The few recent studies that addressed this question suggest that the use of a task-specific strategy may facilitate training-induced improv…”
BackgroundResultInfluentialTang, M. F., Dickinson, J. E., Visser, T. A. W., & Badcock, D. R. (2015) · Journal of Vision · doi:10.1167/15.13.4 · Cited by 38 · see citing works
Intelligence tests predict brain response to demanding task events
Duncan, J. (2003) · Nature Neuroscience · doi:10.1038/nn0303-207 · Cited by 37 · cited 1× · see citing works
“Differentiating lures from targets is difficult – participants tend to press the button upon detecting a repetition, even with different intervals (Duncan, 2003).”
BackgroundWhat does It take to Show that a Cognitive Training Procedure is Useful?
Jacoby, N., & Ahissar, M. (2013) · Progress in Brain Research · doi:10.1016/b978-0-444-63327-9.00004-7 · Cited by 30 · see citing works
Self-citationJacoby, N., & Ahissar, M. (2015) · Journal of Vision · doi:10.1167/15.10.6 · Cited by 24 · see citing works
Self-citationAction Video Games DO NOT Promote Visual Attention
Roque, N. A., & Boot, W. R. (2018) · Springer eBooks · doi:10.1007/978-3-319-95495-0_9 · Cited by 16 · cited 1× · see citing works
“This claim had been challenged, and the findings regarding the advantages of strategic video games have been questioned (e.g. Roque & Boot, 2018; Sala, Tatlidil, & Gobet, 2018).”
ResultChanges in Error Patterns during N-back Training Indicate Reliance on Subvocal Rehearsal
Chooi, W.-T., & Logie, R. H. (2020) · Memory & Cognition · doi:10.3758/s13421-020-01066-w · Cited by 10 · see citing works
Jakoby, H., Raviv, O., Jaffe-Dax, S., Lieder, I., & Ahissar, M. (2019) · Journal of Experimental Psychology: General · doi:10.1037/xge0000573 · Cited by 6 · see citing works
Self-citation(2017) · Developmental Psychology · doi:10.1037/dev0000265.supp · Cited by 1 · see citing works
Flagged first, then by citations
2 of the 60 references couldn’t be matched to a known paper.
Cited by
Sorted by citations
The domain-specific approach of working memory training
Peng, P., & Swanson, H. L. (2022) · Developmental Review · doi:10.1016/j.dr.2022.101035 · Cited by 41 · see citing works
The State of Cognitive Control in Language Processing
Ness, T., Langlois, V. J., Kim, A. E., & Novick, J. M. (2023) · Perspectives on Psychological Science · doi:10.1177/17456916231197122 · Cited by 37 · see citing works
“…suggesting that working memory training does not produce broad increases in cognitive ability or general fluid intelligence (e.g., Harrison et al., 2013; Malinovitch et al., 2021; Melby-Lervåg et al., 2016; Simons et al., 2016; Soveri et al., 2017; Sprenger et al., 2013; Thompson et al., 2013).”
The Role of Executive Function in Reading Development and Reading Intervention
Peng, P. (2023) · Mind Brain and Education · doi:10.1111/mbe.12375 · Cited by 18 · see citing works
Strategy use and its evolvement in word list learning: a replication study
Laine, M., Fellman, D., Eräste, T., Ritakallio, L., & Salmi, J. (2024) · Royal Society Open Science · doi:10.1098/rsos.230651 · Cited by 9 · see citing works
“However, performance improvements in previous studies where participants received memory strategy instructions speak for a causal relationship between strategy employment and performance (e.g. [ 3,4,7 – 10,42”
Spontaneous memory strategies in a videogame simulating everyday memory tasks
Laine, M., Jylkkä, J., Ritakallio, L., Eräste, T., Kangas, S., Hering, A., Zuber, S., Kliegel, M., Fellman, D., & Salmi, J. (2023) · Quarterly Journal of Experimental Psychology · doi:10.1177/17470218231183958 · Cited by 10 · see citing works
“…in memory performance through manipulation of participants’ strategic choice (e.g., Bailey et al., 2014; Borella et al., 2017; Carretti et al., 2007; Fellman et al., 2020; Forsberg et al., 2020; Laine et al., 2018; Malinovitch et al., 2021; McNamara & Scott, 2001; Turley-Ames & Whitfield, 2003).”
BackgroundAsymmetric negative transfer effects of working memory training
Ni, N., Gathercole, S. E., Norris, D. G., & Saito, S. (2023) · Memory & Cognition · doi:10.3758/s13421-023-01412-8 · Cited by 9 · see citing works
“This strategy mediation account captures training outcomes as the improved efficacy of unchanged WM capacity (Dunning & Holmes, 2014; Fellman et al., 2020; Forsberg et al., 2020; Laine et al., 2018; Malinovitch et al., 2021).”
BackgroundIzhak, S. B., & Michal Lavidor, B. (2023) · Journal of Cognition and Development · doi:10.1080/15248372.2023.2172413 · Cited by 9 · see citing works
Ritakallio, L., Fellman, D., Jylkkä, J., Waris, O., Lönnroth, N., Nervander, R., Salmi, J., & Laine, M. (2021) · Journal of Cognitive Enhancement · doi:10.1007/s41465-021-00235-2 · Cited by 7 · see citing works
Shao, W., Zhou, H., Qi, Y., Zhu, Z., Zhang, T., Chen, Y., Chen, Y., & Yu, X. (2024) · Journal of Experimental Child Psychology · doi:10.1016/j.jecp.2024.105950 · Cited by 6 · see citing works
Malinovitch, T., Albouy, P., Zatorre, R. J., & Ahissar, M. (2022) · Cerebral Cortex · doi:10.1093/cercor/bhac175 · Cited by 5 · see citing works
Self-citationN-Back training in middle adulthood: evidence for transfer only to structurally similar task
Juras, L., Sedlar, M. K., Vrhovski, G., & Vranić, A. (2024) · Current Psychology · doi:10.1007/s12144-024-05758-3 · Cited by 5 · see citing works
“Recent research highlighted that training gains in structurally similar tasks can be attributed to the acquisition of task-specific strategies (Forsberg et al., 2020; Malinovitch et al., 2021 ).”
Self-reported strategy use in working memory tasks
Ritakallio, L., Fellman, D., Salmi, J., Jylkkä, J., & Laine, M. (2024) · Scientific Reports · doi:10.1038/s41598-024-54160-3 · Cited by 5 · see citing works
“…have received explicit WM strategy instructions indicate subsequent performance improvement when compared with non-instructed controls 6,11,12,15–17,26 – 28 In conclusion, the present results show highly varied self-generated strategy use in WM tasks that is strongly related to task outcomes,…”
ResultLi, G., Liu, Y., & Chen, A. (2026) · npj Digital Medicine · doi:10.1038/s41746-026-02478-9 · Cited by 2 · see citing works
“Finally, beha-viouraloutcomeswerelargelytask-based,andobservedimprovementsmay partlyre fl ect practice-related gains and strategy optimisation rather than durable increases in cognitive capacity 79 .”
InfluentialBurke, I., Chesser, C., Brown, C. P. K., Watkins, R., Butterworth, P., Olofsson, J., Laver, K., Hampstead, B. M., & Bahar‐Fuchs, A. (2025) · Alzheimer s & Dementia Translational Research & Clinical Interventions · doi:10.1002/trc2.70120 · Cited by 2 · see citing works
“Thus, the opportunity to develop efficient task-specific strategies through daily training tasks 29 was enhanced by the multimodal OMT condition.”
Zetsche, U., Neumann, P., Bürkner, P., Renneberg, B., Koster, E. H. W., & Hoorelbeke, K. (2023) · doi:10.31234/osf.io/ydvqa · Cited by 2 · see citing works
Lin, Y.-T., & Leber, A. B. (2024) · doi:10.31234/osf.io/shfqc · Cited by 2 · see citing works
Short-Term Practice Modulates ERP Components Without Behavioral Change in a Short-ISI Go/NoGo Task
Sugawara, Y., Matsuda, Y., Kurokawa, R., Kosuge, R., Kudoh, S., Akaiwa, M., Saito, H., Sasaki, T., & Sugawara, K. (2025) · Brain Sciences · doi:10.3390/brainsci15111208 · Cited by 1 · see citing works
“The effects of task practice have been shown to depend on factors such as task difficulty and the amount of practice, including its duration [26,48,49].”
Understanding the relations between cognition and academic skills in the context of instruction
Peng, P. (2025) · Advances in child development and behavior · doi:10.1016/bs.acdb.2024.10.003 · Cited by 2 · see citing works
Jugar videojuegos de estrategia en tiempo real tiene efectos positivos en la memoria de trabajo
López-Riquelme, G. O., Soto-Alba, E. E., Mata-Bautista, C. R., & Marzoug, R. (2022) · Revista Iberoamericana ConCiencia. · doi:10.32654/concienciaepg/eds.especial-7
Cognitive Strategy Instruction in Vocabulary Learning Task: the Role of IntelligencePreprint
Izmalkova, A. I. (2024) · Preprints.org · doi:10.20944/preprints202405.0852.v1 · Cited by 1 · see citing works
Pelatihan Pengembangan Kelembagaan Kelompok Usaha Bersama (KUBE) di Kecamatan Moramo Utara
Arsyad, M. R., Hos, J., Bahtiar, B., Upe, A., Supiyah, R., Roslan, S., & Anggraini, D. (2022) · Indonesian Journal of Community Services · doi:10.47540/ijcs.v1i2.631
Semantic Bimodal Presentation Differentially Slows Working Memory Retrieval
Cheng, J., Li, J., Wang, A., & Zhang, M. (2023) · Brain Sciences · doi:10.3390/brainsci13050811
“The presence of lures, along with the fact that all stimuli in the experiment were a perfect bimodal repe‑ tition of samples, allowed us to exclude the possibility that participants simply relied on a sense of familiarity to achieve a correct response [31].”
BackgroundInfluentialMinimal training duration inducing near transfer in two-phase working memory training paradigm
Ni, N., Tsubomi, H., & Saito, S. (2025) · Quarterly Journal of Experimental Psychology · doi:10.1177/17470218251369061
“This view is supported by the findings that spontaneous strategy use was associated with WM task performance and training outcomes (e.g., Laine et al., 2018; Malinovitch et al., 2021; Waris et al., 2021).”
Jiang, H., Johnstone, S. J., Li, Y. Y., et al. (2025) · Research Square · doi:10.21203/rs.3.rs-7508412/v1
Jiang, H., Johnstone, S. J., Li, Y. Y., Lu, J., & Mao, B.-B. (2026) · npj Science of Learning · doi:10.1038/s41539-026-00415-9
Introduction to Cognitive Psychology
Cominelli, L., & Lucchiari, C. (2026) · Springer eBooks · doi:10.1007/978-3-032-21273-3_1
Cognitive Strategy-based neuromodulation optimizes neural communication to improve working memory
Tian, Y., Tan, C., Jiang, Y., Li, P., & Ma, W. (2026) · NeuroImage · doi:10.1016/j.neuroimage.2026.122104
- Authors
- Malinovitch, T., Jakoby, H., & Ahissar, M.
- Affiliations
- Hebrew University of Jerusalem (IL)
- Jerusalem Multidisciplinary College (IL)
- Corresponding author
- Malinovitch, Tamar
- Venue
- Psychonomic Bulletin & Review
- Published
- 2020-10-15
- Volume
- 28
- Issue
- 2
- Pages
- 526-536
- ISSN
- 1069-9384 (ISSN-L)
- 1531-5320
- PMC
- PMC8062341
- OpenAlex
- W3092950725
- Semantic Scholar
- 5428e8101d63a7f85010ce636635173848a7bc19
- Cited by
- 29
- References
- 60
- Citations by year
- 2026: 6 · 2025: 5 · 2024: 6 · 2023: 7 · 2022: 4 · 2021: 1
- FWCI
- 2.90
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Locations
- Psychonomic Bulletin & Review (publishedVersion, cc-by)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Funders
- European Commission (HORIZON2020)
- Israel Science Foundation (1650/17, 2425/15)
- Azrieli Foundation (2425/15)
- Canadian Institutes of Health Research
- H2020 European Research Council (833694)
- Topics
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Cognitive Abilities and Testing
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Cognitive Functions and Memory
- Life Sciences › Neuroscience › Cognitive Neuroscience › Neural and Behavioral Psychology Studies
- Keywords
- working memory, n-back task, cognitive training, strategy learning, strategy transfer, strategy discovery, reasoning skills
- MeSH terms
- Adult, Humans, Memory, Short-Term, Practice, Psychological, Psychomotor Performance, Thinking
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-08.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-08.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
