Content uploaded by Ulman Lindenberger
Author content
All content in this area was uploaded by Ulman Lindenberger on Nov 30, 2015
Content may be subject to copyright.
BRIEF REPORT
Younger Adults Show Long-Term Effects of Cognitive Training on Broad
Cognitive Abilities Over 2 Years
Florian Schmiedek
Max Planck Institute for Human Development, Berlin, Germany,
and German Institute for International Educational Research
(DIPF), Frankfurt am Main, Germany
Martin Lövdén
Max Planck Institute for Human Development, Berlin, Germany;
Karolinska Institutet; and Stockholm University
Ulman Lindenberger
Max Planck Institute for Human Development, Berlin, Germany
In the COGITO study (Schmiedek, Lövdén, & Lindenberger, 2010), 101 younger adults practiced 12
tests of perceptual speed, working memory, and episodic memory for over 100 daily 1-hr sessions. The
intervention resulted in positive transfer to broad cognitive abilities, including reasoning and episodic
memory. Here, we examine whether these ability-based transfer effects are maintained over time. Two
years after the end of the training, 80 participants returned for follow-up assessments of the compre-
hensive battery of transfer tasks. We found reliable positive long-term transfer effects for reasoning and
episodic memory, controlling for retest effects by including participants from the original control group.
This shows, for the first time, that intensive cognitive training interventions can have long-term broad
transfer at the level of cognitive abilities.
Keywords: cognitive training, cognitive abilities, transfer effects, latent change score models, long-term
effects
Attempts to improve cognitive functioning with training inter-
ventions have a long history in psychology. For many years,
interventions used strategy instruction and practice on tasks from
psychometric test batteries of cognitive abilities, and at most these
interventions produced transfer effects (i.e., improvements on un-
trained tasks) that must be considered narrow (Noack, Lövdén,
Schmiedek, & Lindenberger, 2009). More recently, however, cog-
nitive training research has produced a number of findings that
paint a more positive picture of the effectiveness of practice-
induced changes of cognitive functioning. The most promising
findings come from trainings that (a) build on self-guided practice,
rather than instruction of strategies (cf. Hofland, Willis, & Baltes,
1981); (b) focus on the core capacities of working memory (WM;
e.g., Dahlin, Stigsdotter-Neely, Larsson, Bäckman, & Nyberg,
2008;Jaeggi, Buschkuehl, Jonides, & Perrig, 2008;Klingberg et
al., 2005; see Morrison & Chein, 2011, for review) or executive
functions like task switching (Karbach & Kray, 2009); and (c) use
computerized setups that adapt task difficulties to a continuously
challenging level. Holding individualized task-difficulty up high
creates a continuous mismatch of cognitive demands and individ-
ual functional supplies. Such mismatches, if present for a pro-
longed period, could have the potential to improve cognitive
processing efficiency rather than merely exploiting the available
behavioral flexibility with effective, but typically task-specific,
strategies (Lövdén, Bäckman, Lindenberger, Schaefer, & Schmie-
dek, 2010). As of recently, failed replications of WM training
studies have also been reported (Chooi & Thompson, 2012; Redick
et al., 2012), and critical reviews on WM training have appeared
(Melby-Lervåg & Hulme, 2013;Shipstead, Hicks, & Engle, 2012;
Shipstead, Redick, & Engle, 2012). Thus, the jury on the effec-
tiveness and efficiency of cognitive training is still out and await-
ing further empirical evidence that allows evaluating its useful-
ness.
This article was published Online First July 14, 2014.
Florian Schmiedek, Center for Lifespan Psychology, Max Planck Insti-
tute for Human Development, Berlin, Germany, and German Institute for
International Educational Research (DIPF), Frankfurt am Main; Martin
Lövdén, Center for Lifespan Psychology, Max Planck Institute for Human
Development, Berlin, Germany, and Aging Research Center, which be-
longs to the Karolinska Institutet and Stockholm University; Ulman Lin-
denberger, Center for Lifespan Psychology, Max Planck Institute for
Human Development, Berlin, Germany.
The COGITO Study was supported by the Max Planck Society, includ-
ing a grant from the innovation fund of the Max Planck Society (M.FE.
A.BILD0005); the Sofja Kovalevskaja Award (to Martin Lövdén) of the
Alexander von Humboldt Foundation donated by the German Federal
Ministry for Education and Research (BMBF); the German Research
Foundation (DFG; KFG 163); and the German Federal Ministry for Edu-
cation and Research (BMBF; CAI). Ulman Lindenberger was supported by
the Gottfried Wilhelm Leibniz Award of the DFG.
Correspondence concerning this article should be addressed to Florian
Schmiedek, German Institute for International Educational Research
(DIPF), Schloßstr. 29, 60486 Frankfurt am Main, Germany. E-mail:
schmiedek@dipf.de
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Developmental Psychology © 2014 American Psychological Association
2014, Vol. 50, No. 9, 2304–2310 0012-1649/14/$12.00 http://dx.doi.org/10.1037/a0037388
2304
To be of practical relevance for everyday competencies,
training-induced changes need to meet two criteria. First, changes
need to be located at the level of broad cognitive abilities, that is,
they have to reach beyond the acquisition of task-specific skills.
Second, changes need to be enduring, that is, maintained for some
time after the training intervention has ended (cf. Sternberg, 2008).
Ideally, training interventions enhance the long-term trajectory of
cognitive development, foster success in educational and profes-
sional settings, and extend the period in old age during which
individuals are able to live independently (Hertzog, Kramer, Wil-
son, & Lindenberger, 2008).
Empirically, the first criterion can be evaluated by investigating
the range of transfer effects. Effects observed on individual trans-
fer tasks, however, provide only weak evidence for improvements
in general cognitive abilities. If an ability (e.g., reasoning) had
indeed improved, one would expect that performance on indicator
tasks (e.g., Raven’s Advanced Progressive Matrices; Raven &
Horn, 2009) of this ability should improve. However, because
performance on observed tasks can be influenced by factors be-
yond the underlying ability, like measurement error or task-
specific skills, the practice of relying on individual indicators of a
given ability can easily lead to false positive findings (e.g., im-
provements due to the acquisition of task-specific skills) as well as
negative findings (e.g., due to lack of power because of improve-
ments in ability being blurred by task-specific variance and mea-
surement error) regarding the question of whether the underlying
ability has improved.
Therefore, studies on transfer of training need to investigate
whether transfer can be discerned at the level of cognitive abilities
(Lövdén et al., 2010;Noack et al., 2009;Schmiedek, Lövdén, &
Lindenberger, 2010;Shipstead et al., 2012). This requires assess-
ing transfer with broad selections of heterogeneous tasks that cover
the range of the target ability in a comprehensive manner and test
changes at the level of common factors of these tasks. Such
common factors represent sources of variance that are shared
across tasks and are therefore free from measurement error and
task-specific influences. Demonstrating transfer at this level pro-
vides a more solid basis for concluding that ability has improved
than focusing on the task level.
Using data from the COGITO study, in which 101 younger and
103 older adults practiced a battery of 12 cognitive tasks over 100
daily sessions, Schmiedek et al. (2010) could show that a cognitive
intervention can result in transfer at the ability level for reasoning
(i.e., fluid intelligence) and episodic memory in healthy younger
adults. In addition, transfer was observed on a factor of WM tasks
in both age groups. The tasks comprising this factor were struc-
turally similar to the trained ones but differed in task content.
Transfer of training was not reliable for reasoning and episodic
memory in the older adults, and for perceptual speed as well as for
a factor of complex span tasks of WM in both age groups.
Regarding the criterion of temporal preservation, there is evi-
dence that improvements can be maintained up to several years,
particularly for improvements on the trained tasks (e.g., Ball et al.,
2002) and for specific strategies and skills (e.g., Brehmer et al.,
2008;Klauer & Phye, 2008;Stigsdotter-Neely & Bäckman, 1993).
For long-term transfer effects, empirical evidence is scarcer. There
is some indication that transfer effects can be maintained up to 18
months (e.g., Borella, Carretti, Riboldi, & De Beni, 2010;Dahlin,
Nyberg, Bäckman, & Stigsdotter-Neely, 2008;Holmes, Gather-
cole, & Dunning, 2009;Li et al., 2008). Regarding the question of
transfer breadth, earlier studies are of limited value because they
were either confined to near transfer or to single indicator tasks per
target ability.
It is completely unknown whether transfer at the level of latent
ability factors induced by cognitive interventions can be main-
tained over longer periods of time (e.g., years). The COGITO
study provides an opportunity to address this question because
participants of the training and control groups came back for
follow-up assessments of the transfer tasks about 2 years after
posttest. Sample sizes at follow-up were sufficiently large to
investigate long-term transfer effects at the ability level using
latent change score models (McArdle, 2009;McArdle & Prindle,
2008). These models have the advantage of allowing to directly
test transfer effects at the latent factor level, which no longer
contains task-specific sources of variance or measurement error
(see Figure 1). We predicted that the pattern of positive transfer at
the factor level at follow-up (i.e., changes from pretest to
follow-up for the training group minus corresponding changes for
the control group) that we observed at posttest would be main-
tained at follow-up. As no reliable transfer effects for the abilities
of episodic memory and reasoning could be demonstrated for the
older adults at posttest, we restricted our analyses to the younger
adults.
Method
Participants and Procedure
During the training phase, 101 younger adults (51.5% women,
M
age
⫽25.6 years, SD
age
⫽2.7, range: 20–31 years) completed an
average of 101 practice sessions (SD ⫽2.6, range: 87–109).
Participants in the no-contact control group were 44 younger
adults (47.7% women, M
age
⫽25.2 years, SD
age
⫽2.5, range:
21–29 years). Before and after the training, participants completed
pre- and posttests during 10 sessions that consisted of 2–2.5 hr of
comprehensive cognitive test batteries and self-report question-
naires. On average, time elapsing between pre- and posttest was
197 versus 193 days for the training and control groups, respec-
tively. Additional information on sample characteristics and study
dropout can be found in Schmiedek, Lövdén, and Lindenberger
(2010) and Schmiedek, Bauer, Lövden, Brose, and Lindenberger
(2010).
The cognitive assessment of the posttest sessions was repeated
at the 2-year follow-up (time from posttest to follow-up: M
time
⫽
755 days, Mdn ⫽749 days, range: 679–927 days, for the training
group; M
time
⫽745 days, Mdn ⫽742 days, range: 693–798 days,
for the control group). Participation rates at follow-up were satis-
factory (80 younger adults in the training and 32 in the control
group, corresponding to 79% and 73% of the original sample sizes,
respectively). Comparisons of pretest performance on the transfer
tasks and on the Digit-Symbol Substitution Test (Wechsler, 1981)
showed that the follow-up sample did not differ significantly from
the dropouts between posttest and follow-up (ps⬎.05), with the
exception of numerical reasoning, for which the follow-up sample
had significantly higher performance at pretest than did the dropouts,
t(99) ⫽2.22, p⫽.028. The present analyses were confined to the
follow-up sample. Within this sample, pretest differences on the trans-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2305
LONG-TERM EFFECTS OF COGNITIVE TRAINING
fer tasks and the Digit-Symbol Substitution Test between the
trained and control groups were not significant (ps⬎.05).
Tasks
In each session, participants practiced 12 different computerized
tasks with two to eight blocks each. For perceptual speed, those
were three two-choice reaction tasks (odd vs. even numbers;
consonants vs. vowels; symmetric vs. asymmetric figures) and
three comparison tasks (two strings of digits/consonants, or two
three-dimensional figures). For episodic memory, tasks required
participants to memorize word lists, number–word pairs, or object
positions in a grid. WM tasks were adapted versions of the alpha
span, numerical memory updating, and spatial n-back tasks (for
details of all tasks, see Schmiedek, Lövdén, & Lindenberger,
2010). Difficulty levels for the choice-reaction, episodic memory,
and WM tasks were individualized using different presentation
times based on pretest performance.
Transfer tasks included computerized tasks as well as 27 tasks
from the paper-and-pencil Berlin Intelligence Structure (BIS) test
(Jäger, Süß, & Beauducel, 1997). The three near transfer WM tasks
were based on the same three paradigms as the practiced WM tasks,
but used different content material. The far transfer WM tasks were
established complex span tasks (reading span, counting span, and
rotation span). For episodic memory, one computerized word
paired-associates task and nine tasks from the BIS (three for each
content domain) were used. Transfer in reasoning was measured
with 15 items from the Raven’s Advanced Progressive Matrices
(Raven & Horn, 2009) as well as with nine tasks from the BIS,
three for each content domain.
Data Analysis
Effect sizes (d) for single tasks were calculated as mean pre-post
(pre-follow-up) differences in accuracy divided by the SD of the
experimental group at pretest. Net effects provided in Table 1 were
obtained by subtracting the effect sizes for the control from those
of the training group. Whether these net effects were statistically
significant was investigated by testing the interaction of occasion
and group with linear mixed effect models (using PROC MIXED
in SAS 9.3; Kenward-Roger degrees of freedom; see Littell, Mil-
liken, Stroup, Wolfinger, & Schabenberger, 2006) that allowed for
different variances at pre- and posttest (Ftests for the interaction
are provided in Table 1). Effects at the latent level were analyzed
with latent change score models (McArdle, 2009;McArdle &
Prindle, 2008). In these models, latent factors were defined by a set
of transfer tasks. Improvements at the latent factor level were
captured by the means of latent change score factors (see Figure 1).
In order for these means to be readily interpretable, it is necessary
that factor loadings and intercepts are constrained to be equal
across occasions and experimental groups (strong measurement
invariance). Here, we even aimed for strict measurement invari-
ance (i.e., residual variances also fixed across occasions and ex-
perimental groups). Tests of whether mean changes at the latent
Figure 1. Latent change score model for modeling training-induced changes at the latent factor level. Squares
represent observed variables, circles represent latent factors, and the triangle serves to represent information
regarding means and intercepts. A1–A3, B1–B3, C1–C3 ⫽observed indicator variables A, B, and C (i.e., tasks
of one ability) measured at the three time points; F1–F3: latent factor of ability at the three time points; LC1:
latent change factor from pretest to posttest; LC2: latent change factor from pretest to follow-up; ␣: latent mean
of ability factor at pretest; : mean change of latent ability factors from pre- to posttest; ␥: mean change of latent
ability factors from pretest to follow-up; ␦: variance (individual differences) in latent ability at pretest; variances
of the latent change factors was fixed to zero, because they were not significant. Loadings of observed variables
on latent factors, intercepts of observed variables, and residual variances were fixed to be the same across the
three time points and across training and control groups (i.e., strict measurement invariance). Residuals for the
same observed variable were allowed to correlate across time points.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2306 SCHMIEDEK, LÖVDÉN, AND LINDENBERGER
factor level were significant were conducted by comparing
the ⫺2LL of models in which means of the latent change factor
were estimated separately for the training and control groups with
models in which both means were constrained to be equal, result-
ing in a
2
test with one df. Testing whether effects at follow-up
differed from those at posttest were conducted by comparing the
unconstrained model to one in which the differences training
minus control were constrained to be equal for both latent change
factors, resulting in a
2
test with one df.
Model fits were acceptable for reasoning,
2
(75) ⫽83.91,
root-mean-square error of approximation (RMSEA) ⫽.05, and
episodic memory,
2
(75) ⫽93.61, RMSEA ⫽.07, but not for the
model of WM near transfer tasks, even if only strong measurement
invariance was modeled,
2
(60) ⫽106.73, RMSEA ⫽.12. We
therefore refrain from interpreting results for WM at the latent
factor level.
Latent effect sizes were calculated by dividing the latent mean
differences by the latent SDs at pretest. For analyses of the BIS
test, tasks were parceled for each ability construct by calculating
composites of standardized scores for the three tasks of each
content domain. As these scores were thus already standardized
based on pretest SDs, mean differences are in effect-size metric
and do not need to be divided by SDs.
Results
In the following, we focus on long-term transfer effects at the
latent factor level and restrict our analyses to those transfer effects
for which we found significant results at posttest for the younger
adults (Schmiedek, Lövdén, & Lindenberger, 2010); that is, for
latent factors of reasoning and episodic memory. Results on trans-
fer effects at the observed task level are reported in Table 1.
For the latent factor of reasoning, there was a significant inter-
action of experimental group and occasion,
2
(2) ⫽15.54, p⬍
.001. The latent net effect sizes were .17,
2
(1) ⫽7.41, p⫽.006,
at posttest and .23,
2
(1) ⫽14.57, p⬍.001, at follow-up. The
difference of these effects was not reliable,
2
(1) ⫽1.12, ns.As
shown in Figure 2, this was due to relative stability of latent means
for both the trained and the control group. For the latent factor of
episodic memory, there was a significant interaction of experimen-
tal group and occasion,
2
(2) ⫽31.45, p⬍.001. The latent net
effect sizes were .47,
2
(1) ⫽30.48, p⬍.001, at posttest and .18,
2
(1) ⫽3.88, p⫽.041, at follow-up. The difference of these
effects was reliable,
2
(1) ⫽11.54, p⬍.001. The reduction of the
effect was mainly due to a reduction of the effect in the trained
group (see Figure 2).
In sum, the results at the latent factor level show that the
improvements at the ability level for reasoning and episodic mem-
ory were (a) significant at posttest for the reduced follow-up
sample, (b) significant at the 2-year follow-up, and (c) signifi-
cantly reduced at follow-up, in comparison to transfer at posttest,
for episodic memory, but not for reasoning. Group differences in
motivation are unlikely to be the cause of these effects, as self-
reported motivation to work on the tasks did not differ signifi-
cantly between the training and control groups (see Figure 3).
Discussion
The present results show that far transfer to broad cognitive
abilities can be maintained over several years. The sizes of the
observed reliable effects were not large. However, their breadth
renders them beneficial for a number of real-life outcomes. As
reasoning and episodic memory are abilities of high predictive
validity for everyday competency (Tucker-Drob, 2011), even
small effects can have a substantial impact on performance in
educational, professional, and leisure activity settings. Training
interventions that lead to small effects of wide scope and high
temporal stability may pay off more than interventions that lead to
strong but specific effects that do not last for long.
Regarding reasoning, transfer effects at follow-up were signif-
icant at the observed task as well as at the latent ability level and
of comparable size as at posttest. While for episodic memory,
Table 1
Transfer Effects for Follow-Up Sample and Individual Tasks at Posttest and Follow-Up
Task Pre-post net
effect size Pre-Post ⫻
Experimental Group Pre-follow-up net
effect size Pre-Follow-up ⫻
Experimental Group
Working memory—Near
Animal span .02 F(1, 110) ⫽0.01, ns ⫺.06 F(1, 110) ⫽0.11, ns
N-back numerical .41 F(1, 110) ⫽6.21, p⫽.014 .46 F(1, 110) ⫽9.07, p⫽.003
Memory updating spatial .07 F(1, 124) ⫽0.18, ns ⫺.05 F(1, 124) ⫽0.06, ns
Working memory—Far
Reading span .00 F(1, 124) ⫽0.00, ns .31 F(1, 124) ⫽1.72, ns
Counting span .03 F(1, 124) ⫽0.03, ns .24 F(1, 124) ⫽1.24, ns
Rotation span .08 F(1, 124) ⫽0.28, ns .04 F(1, 124) ⫽0.08, ns
Reasoning
Verbal .12 F(1, 110) ⫽1.38, ns .22 F(1, 110) ⫽4.14, p⫽.044
Numerical .25 F(1, 110) ⫽5.40, p⫽.022 .32 F(1, 110) ⫽7.11, p⫽.009
Figural/spatial .23 F(1, 110) ⫽3.68, ns .27 F(1, 110) ⫽7.30, p⫽.008
Raven .21 F(1, 109) ⫽1.58, ns .40 F(1, 107) ⫽3.90, ns
Memory
Verbal .49 F(1, 110) ⫽17.09, p⬍.0001 .15 F(1, 110) ⫽1.68, ns
Numerical .53 F(1, 110) ⫽11.15, p⫽.001 .16 F(1, 110) ⫽1.20, ns
Figural/spatial .20 F(1, 110) ⫽3.42, ns .21 F(1, 110) ⫽3.43, ns
Word pairs .22 F(1, 110) ⫽2.20, ns .16 F(1, 110) ⫽0.92, ns
Note. Pre ⫽pretreatment; post ⫽posttreatment.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2307
LONG-TERM EFFECTS OF COGNITIVE TRAINING
transfer effects were not significant anymore at the observed task
level for verbal, numerical, and figural-spatial memory at
follow-up (see Table 1), the effect at the level of their common
factor was reduced in comparison to the posttest effects, but still
maintained reliable. This further demonstrates the usefulness of
investigating transfer at the latent factor level. At the observed task
level, performance is measured with imperfect reliability due to
measurement error and might be influenced by task-specific strat-
egies that have been acquired during the training, but could not be
reactivated in an effective manner after 2 years. As the latent level
only captures sources of variance that have a general influence on
all indicator tasks of the factor, general effects, if present, are more
easily detectable there.
How did transfer to broad cognitive abilities come about, and
how was it maintained over the considerable period of 2 years? We
hold that plasticity at the neural level requires a sustained chal-
lenge of the cognitive system produced by a mismatch between
cognitive demands and functional supplies (Lövdén et al., 2010).
The breadth (12 heterogeneous tasks that differed in content and
paradigms), intensity (high difficulty due to adjustment to individ-
ual performance levels), and dosage (100 sessions of about 1 hr
duration) of the training fulfills this requirement and could thereby
lead to plastic brain changes, for example, in gray matter (Dra-
ganski et al., 2006), white matter (Scholz, Klein, Behrens, &
Johansen-Berg, 2009), and neurotransmitter systems (Bäckman et
al., 2011;McNab et al., 2009). For a subsample of COGITO
participants, Lövdén, Bodammer, et al. (2010) have found indica-
tions of improved white-matter microstructure as well as increased
volumes of the anterior corpus callosum at posttest. Little is known
about the temporal stability of plastic neural changes, and we do
Figure 2. Latent means and associated standard errors for the training and control groups at pretest, posttest,
and follow-up. Training group shown with solid lines, control with dashed lines. A: latent factor of reasoning;
B: latent factor of episodic memory. As the indicator tasks of the latent factors were standardized by SDsat
pretest, latent means are in effect size metric.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2308 SCHMIEDEK, LÖVDÉN, AND LINDENBERGER
not know whether and how they help to preserve positive transfer
in broad cognitive abilities.
In addition to plastic changes at the neural level, we also need
to consider rather complex reciprocal effects among the develop-
mental trajectories of cognitive and other psychological variables.
Improved cognitive abilities may open opportunities in the educa-
tional and professional paths of younger adults that in turn lead to
continuously raised levels of cognitive demand, which may help to
perpetuate the beneficial effects of the training. Similarly, in-
creased cognitive capacities might lead to an increased need for
cognition (Cacioppo, Petty, Feinstein, & Jarvis, 1996) or openness
to experience (Jackson, Hill, Payne, Roberts, & Stine-Morrow,
2012) that makes participants seek and face cognitive challenges in
their lives. Findings of long-term benefits of early education pro-
grams that sometimes last decades after the intervention programs
have ended (Barnett, 2011) underscore the importance of taking a
developmental perspective on cascading outcomes of training in-
terventions.
The finding that latent transfer effects were reduced at follow-up
for episodic memory, but not reasoning, speaks to the possibility
that the acquisition of general strategies might also have contrib-
uted to the findings for episodic memory at posttest. Besides the
influence of task-specific strategies, which should not influence
findings at the latent factor level, our participants might also have
acquired and practiced more general strategies, like mental imag-
ery, that are supportive for a broad selection of episodic memory
tasks. Difficulties with an ad-hoc reactivation of these strategies at
follow-up might explain the reduction of transfer effects. As no
reasoning tasks were included in the training and as potential
strategies used with the practiced WM tasks are much less likely to
be of help for performance on the transfer reasoning tasks, a
strategy-based explanation of the transfer to reasoning is difficult
to entertain.
In sum, the present findings provide room for cautious optimism
(cf. Hertzog et al., 2008). Cognitive trainings can produce transfer
effects that are sufficiently large in scope and stable over time to
justify the considerable effort that is needed to produce them.
Future studies should hold up the proposed standard of investigat-
ing transfer at the level of latent ability factors and improve on the
investigation of the mechanisms that produce transfer and main-
tenance. Future research will need to take close and continuous
looks at postintervention developmental trajectories on behavioral,
social, and neural dimensions to better understand the conditions
under which cognitive training interventions can trigger a cascade
of changes that result in improved or maintained cognitive com-
petence.
References
Bäckman, L., Nyberg, L., Soveri, A., Johansson, J., Andersson, M., Dahlin,
E.,...Rinne, J. O. (2011, August 5). Effects of working-memory
training on striatal dopamine release. Science, 333, 718. doi:10.1126/
science.1204978
Ball, K., Berch, D. B., Helmers, K. F., Jobe, J. B., Leveck, M. D.,
Marsiske, M.,...Willis, S. L. (2002). Effects of cognitive training
interventions with older adults: A randomized controlled trial. Journal of
the American Medical Association, 288, 2271–2281. doi:10.1001/jama
.288.18.2271
Barnett, W. S. (2011, August 19). Effectiveness of early educational
intervention. Science, 333, 975–978. doi:10.1126/science.1204534
Borella, E., Carretti, B., Riboldi, F., & De Beni, R. (2010). Working
memory training in older adults: Evidence of transfer and maintenance
effects. Psychology and Aging, 25, 767–778. doi:10.1037/a0020683
Brehmer, Y., Li, S.-C., Straube, B., Stoll, G., von Oertzen, T., Müller, V.,
& Lindenberger, U. (2008). Comparing memory skill maintenance
across the life span: Preservation in adults, increase in children. Psy-
chology and Aging, 23, 227–238. doi:10.1037/0882-7974.23.2.227
Cacioppo, J. T., Petty, R. E., Feinstein, J. A., & Jarvis, W. B. G. (1996).
Dispositional differences in cognitive motivation: The life and times of
individuals varying in need for cognition. Psychological Bulletin, 119,
197–253. doi:10.1037/0033-2909.119.2.197
Chooi, W.-T., & Thompson, L. A. (2012). Working memory training does
not improve intelligence in healthy young adults. Intelligence, 40, 531–
542. doi:10.1016/j.intell.2012.07.004
Dahlin, E., Nyberg, L., Bäckman, L., & Stigsdotter-Neely, A. (2008).
Plasticity of executive functioning in young and older adults: Immediate
training gains, transfer, and long-term maintenance. Psychology and
Aging, 23, 720–730. doi:10.1037/a0014296
Dahlin, E., Stigsdotter-Neely, A., Larsson, A., Bäckman, L., & Nyberg, L.
(2008, June 13). Transfer of learning after updating training mediated by
the striatum. Science, 320, 1510–1512. doi:10.1126/science.1155466
Draganski, B., Gaser, C., Kempermann, G., Kuhn, H. G., Winkler, J.,
Büchel, C., & May, A. (2006). Temporal and spatial dynamics of brain
structure changes during extensive learning. Journal of Neuroscience,
26, 6314–6317. doi:10.1523/JNEUROSCI.4628-05.2006
Hertzog, C., Kramer, A. F., Wilson, R. S., & Lindenberger, U. (2008).
Enrichment effects on adult cognitive development: Can the functional
capacity of older adults be preserved and enhanced? Psychological
Science in the Public Interest, 9, 1–65. doi:10.1111/j.1539-6053.2009
.01034.x
Hofland, B. F., Willis, S. L., & Baltes, P. B. (1981). Fluid intelligence
performance in the elderly: Intraindividual variability and conditions of
Figure 3. Self-reported motivation to work on the tasks at pretest, post-
test, and follow-up for the training and control groups. Participants an-
swered the question “I tried to do well on the tasks” on an 8-point scale
(0 ⫽does not apply at all, 7⫽does apply very well) at the end of the
session in which they had worked on the Berlin Intelligence Structure test.
This information was available on all three occasions for 71 participants
from the training and 31 of the control group participants. Solid and broken
lines show means for the trained and control group, respectively. Error bars
denote standard errors. While the main effect of occasion was significant,
F(2, 202) ⫽4.69, p⫽.010, neither the main effect of group, F(1, 201) ⫽
2.88, ns, nor the interaction of group and occasion, F(2, 202) ⫽0.03, ns,
was reliable.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2309
LONG-TERM EFFECTS OF COGNITIVE TRAINING
assessment. Journal of Educational Psychology, 73, 573–586. doi:
10.1037/0022-0663.73.4.573
Holmes, J., Gathercole, S. E., & Dunning, D. L. (2009). Adaptive training
leads to sustained enhancement of poor working memory in children.
Developmental Science, 12, F9–F15. doi:10.1111/j.1467-7687.2009
.00848.x
Jackson, J. J., Hill, P. L., Payne, B. R., Roberts, B. W., & Stine-Morrow,
E. A. L. (2012). Can an old dog learn (and want to experience) new
tricks? Cognitive training increases openness to experience in older
adults. Psychology and Aging, 27, 286–292. doi:10.1037/a0025918
Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008). Im-
proving fluid intelligence with training on working memory. Proceed-
ings of the National Academy of Sciences, USA, 105, 6829–6833.
doi:10.1073/pnas.0801268105
Jäger, A. O., Süß, H.-M., & Beauducel, A. (1997). Der Berliner
Intelligenzstruktur-Test (BIS-Test; Form 4) [Berlin Intelligence Struc-
ture Test (BIS-Test; Form 4)]. Göttingen, Germany: Hogrefe.
Karbach, J., & Kray, J. (2009). How useful is executive control training?
Age differences in near and far transfer of task-switching training.
Developmental Science, 12, 978–990. doi:10.1111/j.1467-7687.2009
.00846.x
Klauer, K. J., & Phye, G. D. (2008). Inductive reasoning: A training
approach. Review of Educational Research, 78, 85–123. doi:10.3102/
0034654307313402
Klingberg, T., Fernell, E., Olesen, P. J., Johnson, M., Gustafsson, P.,
Dahlström, K.,...Westerberg, H. (2005). Computerized training of
working memory in children with ADHD—A randomized, controlled
trial. Journal of the American Academy of Child & Adolescent Psychi-
atry, 44, 177–186. doi:10.1097/00004583-200502000-00010
Li, S.-C., Schmiedek, F., Huxhold, O., Röcke, C., Smith, J., & Linden-
berger, U. (2008). Working memory plasticity in old age: Practice gain,
transfer, and maintenance. Psychology and Aging, 23, 731–742. doi:
10.1037/a0014343
Littell, R. C., Milliken, G. A., Stroup, W. W., Wolfinger, R. D., &
Schabenberger, O. (2006). SAS for mixed models (2nd ed.). Cary, NC:
SAS Institute.
Lövdén, M., Bäckman, L., Lindenberger, U., Schaefer, S., & Schmiedek, F.
(2010). A theoretical framework for the study of adult cognitive plas-
ticity. Psychological Bulletin, 136, 659–676. doi:10.1037/a0020080
Lövdén, M., Bodammer, N. C., Kühn, S., Kaufmann, J., Schütze, H.,
Tempelmann, C.,...Lindenberger, U. (2010). Experience-dependent
plasticity of white-matter microstructure extends into old age. Neuro-
psychologia, 48, 3878–3883. doi:10.1016/j.neuropsychologia.2010.08
.026
McArdle, J. J. (2009). Latent variable modeling of differences and changes
with longitudinal data. Annual Review of Psychology, 60, 577–605.
doi:10.1146/annurev.psych.60.110707.163612
McArdle, J. J., & Prindle, J. J. (2008). A latent change score analysis of a
randomized clinical trial in reasoning training. Psychology and Aging,
23, 702–719. doi:10.1037/a0014349
McNab, F., Varrone, A., Farde, L., Jucaite, A., Bystritsky, P., Forssberg,
H., & Klingberg, T. (2009, February 6). Changes in cortical dopamine
D1 receptor binding associated with cognitive training. Science, 323,
800–802. doi:10.1126/science.1166102
Melby-Lervåg, M., & Hulme, C. (2013). Is working memory training
effective? A meta-analytic review. Developmental Psychology, 49, 270–
291. doi:10.1037/a0028228
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, 46–60. doi:
10.3758/s13423-010-0034-0
Noack, H., Lövdén, M., Schmiedek, F., & Lindenberger, U. (2009). Cog-
nitive plasticity in adulthood and old age: Gauging the generality of
cognitive intervention effects. Restorative Neurology and Neuroscience,
27, 435–453. doi:10.3233/RNN-2009-0496
Raven, J. C., & Horn, R. (2009). Raven’s Progressive Matrices and
Vocabulary Scales: Standard progressive matrices: Test manual. Göt-
tingen, Germany: Hogrefe.
Redick, T. S., Shipstead, Z., Harrison, T. L., Hicks, K. L., Fried, D. E.,
Hambrick, D. Z.,...Engle, R. W. (2013). No evidence of intelligence
improvement after working memory training: A randomized, placebo-
controlled study. Journal of Experimental Psychology: General, 142,
359–379. doi:10.1037/a0029082
Schmiedek, F., Bauer, C., Lövdén, M., Brose, A., & Lindenberger, U.
(2010). Cognitive enrichment in old age: Web-based training programs.
GeroPsych, 23, 59–67. doi:10.1024/1662-9647/a000013
Schmiedek, F., Lövdén, M., & Lindenberger, U. (2010). Hundred days of
cognitive training enhance broad cognitive abilities in adulthood: Find-
ings from the COGITO study. Frontiers in Aging Neuroscience, 2, 1–10.
doi:10.3389/fnagi.2010.00027
Scholz, J., Klein, M. C., Behrens, T. E. J., & Johansen-Berg, H. (2009).
Training induces changes in white-matter architecture. Nature Neuro-
science, 12, 1370–1371. doi:10.1038/nn.2412
Shipstead, Z., Hicks, K. L., & Engle, R. W. (2012). Cogmed working
memory training: Does the evidence support the claims? Journal of
Applied Research in Memory & Cognition, 1, 185–193. doi:10.1016/j
.jarmac.2012.06.003
Shipstead, Z., Redick, T. S., & Engle, R. W. (2012). Is working memory
training effective? Psychological Bulletin, 138, 628–654. doi:10.1037/
a0027473
Sternberg, R. J. (2008). Increasing fluid intelligence is possible after all.
Proceedings of the National Academy of Sciences, USA, 105, 6791–
6792. doi:10.1073/pnas.0803396105
Stigsdotter-Neely, A., & Bäckman, L. (1993). Long-term maintenance of
gains from memory training in older adults: Two 3 1/2-year follow-up
studies. Journals of Gerontology: Series B, 48, 233–237.
Tucker-Drob, E. M. (2011). Neurocognitive functions and everyday func-
tions change together in old age. Neuropsychology, 25, 368–377. doi:
10.1037/a0022348
Wechsler, D. (1981). Wechsler Adult Intelligence Scale: Revised manual
(WAIS–R). New York, NY: Psychological Corporation.
Received February 4, 2013
Revision received January 23, 2014
Accepted May 20, 2014 䡲
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2310 SCHMIEDEK, LÖVDÉN, AND LINDENBERGER
