Sei sulla pagina 1di 7

Computers in Human Behavior 29 (2013) 632638

Contents lists available at SciVerse ScienceDirect

Computers in Human Behavior


journal homepage: www.elsevier.com/locate/comphumbeh

Computerized working memory training: Can it lead to gains in


cognitive skills in students?
Tracy Packiam Alloway a,, Vanessa Bibile b, Gemma Lau b
a
b

Department of Psychology, University of North Florida, 1 UNF Drive, Jacksonville, FL 32240, United States
Department of Psychology, University of Stirling, United Kingdom

a r t i c l e

i n f o

Article history:
Available online 16 January 2013
Keywords:
Computerized training
Working memory
IQ
Academic attainment

a b s t r a c t
Given that working memory is an important cognitive skill that is linked to academic success, there is
increasing attention given to exploring ways to support working memory problems in struggling students. One promising approach is computerized training, and the aim of the present study focused on
whether computerized working memory training could result in both near and far transfer training
effects; and whether such effects would be maintained over time. Students were allocated into one of
three groups: Nonactive Control, Active Control, where they trained once a week (WMT-Low frequency);
Training group, where they trained four times a week (WMT-High frequency). All three groups were
tested on measures of working memory, verbal and nonverbal ability, and academic attainment before
training; and re-tested on the same measures after training, as well as 8 months later. The data indicate
gains in both verbal and visuo-spatial working memory tasks for the high-frequency Training group.
Improvements were also evidenced in tests of verbal and nonverbal ability tests, as well as spelling, in
the high-frequency Training group. There were some maintenance effects when students were tested
8 months later. Possible reasons for why the computerized working memory training led to some far
transfer effects in the high-frequency Training group are included in the discussion.
2012 Elsevier Ltd. All rights reserved.

1. Introduction
Working memory is the term used to refer to a higher-order skill
to allocate attentional resources despite distraction or interference
(Baddeley, 1996; Cowan, 2006; Engle, Tuholski, Laughlin, & Conway, 1999). Working memory increases considerably throughout
childhood (Alloway & Alloway, 2012) and there is a considerable
degree of individual variation at each age point (Alloway &
Gathercole, 2006). For example, at 6.5 years, the 10th centile is
close to the mean for the 4.5 year old sample, and the 90th centile
approximates to the mean performance level for 9.5 year old children. Thus, in an average class, we would expect to see working
memory capacity differences corresponding to 5 years of normal
development between the highest and lowest scoring individuals.
Individual differences in the capacity of working memory have
important consequences for childrens ability to acquire knowledge and new skills, and decits are linked with learning difculties (see Cowan & Alloway, 2008; Swanson & Alloway, 2011, for
reviews). For example, in a large-scale screening study, one in
ten students were found to have working memory decits that
led to below average performance in language and math (Alloway,

Corresponding author. Tel.: +1 904 620 1699.


E-mail address: t.alloway@unf.edu (T.P. Alloway).
0747-5632/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.chb.2012.10.023

Gathercole, Kirkwood, & Elliott, 2009). One explanation for how


working memory decits impact learning is that it creates a bottleneck that prevents a child from acquiring the necessary knowledge
for subsequent learning (Alloway & Gathercole, 2006). It appears
that children with working memory decits are not able to catch
up with their peers and continue to struggle throughout their academic career, even after IQ has been statistically accounted
(Alloway, 2009; also Alloway & Alloway, 2010). There is increasing
evidence that working memory decits are evident in a range of
learning difculties, from reading disorders (Gathercole, Alloway,
Willis, & Adams, 2006), math difculties (Geary, Hoard, & Hamson,
1999), ADHD (Martinussen, Hayden, Hogg-Johnson, & Tannock,
2005), and motor impairments (Alloway, 2007). It is therefore a
priority to nd avenues to overcome working memory difculties
in order that a students learning can be maximized.
In response to clearly established links between low working
memory and a multitude of cognitive, social and educational
decits, increasing efforts have been directed towards the development of effective working memory interventions. Traditionally,
working memory was viewed as genetically xed (Kremen,
Jacobsen, Xian, et al., 2007), and thus unable to be modied by
an individuals environmental experiences or opportunities. However, an increasing body of recent studies has challenged this notion, demonstrating that considerable cerebral plasticity exists
within the developing brain and that working memory capacity

T.P. Alloway et al. / Computers in Human Behavior 29 (2013) 632638

may potentially be improved by environmental intervention and


support (see Buschkuehl, Jaeggi, & Jonides, 2012; Jaeggi,
Buschkuehl, Jonides, & Shah, 2011; Morrison & Chein, 2011, for
reviews). Such ndings highlight important opportunities for
improving the learning, behavioral and social opportunities of
the large number of children with working memory problems.
One approach that is gaining research momentum is computerized working memory training. To date, there are a range of working memory training programs. Some programs mimic working
memory tests and require the individual to process and store information for brief periods. These included remembering numbers in
backwards order or shape locations on a grid (Klingberg, van de
Molen). Others require the individual to update information and
keep it in mind for brief periods, like the n-back task (Jaeggi,
Buschkuehl, Jonides, & Perrig, 2008). A key question that training
studies typically address is the issue of near and far transfer. Near
transfer effects refer to improvements in an area related to the domain that was training. In the present context, the near transfer effects refer to an improvement in working memory scores. Far
transfer effects refer to the extent that training can impact other
related abilities. Tasks used to measure far transfer effects for
working memory training are primarily nonverbal ability tests
(Alloway, 2012a; Jaeggi et al., 2008; Schmiedek, Lovden, &
Lindenberger, 2010). Theoretically, the issue of far transfer effects
is critical in demonstrating the potential efcacy of a training program. Near transfer effects could be written off as a form of practice effects, particularly if the near transfer test closely mirrors
the training task (e.g., Klingberg et al., 2005). However, evidence
of far transfer effects suggests that working memory is a
domain-general resource that impacts a range of other cognitive
skills, including as nonverbal ability (Jaeggi et al., 2008) and reading comprehension (Chein & Morrison, 2010).
However, far transfer effects can be elusive, and not all studies
on working memory have found signicant effects. The majority of
studies have reported near transfer effects of training to working
memory tasks (Klingberg et al., 2005; Thorell, Lindquist, Bergman
Nutley, Bohlin, & Klingberg, 2009; Van der Molen, Van Luit, Van
der Molen, Klugkist, & Jongmans, 2010). The far transfer effects
that have been reported include IQ scores and behavior ratings in
ADHD samples (parents only, not teachers; Klingberg et al.,
2005), nonverbal ability in college students (Jaeggi et al., 2008),
verbal ability scores in students with learning difculties (Alloway,
2012), and reading comprehension (Chein & Morrison, 2010). However, there are cases where there were no far transfer effects, for
example to academic attainment immediately following training
(Holmes, Gathercole, & Dunning, 2009) or to real-world settings
in older adults (Owen et al., 2010; see Morrison & Chein, 2011,
for an excellent review). Furthermore, the extent to which working
memory training can transfer to educational outcomes is as yet
unclear (Healy, Wohldmann, Sutton, & Bourne, 2006).
Thus, the aim of the present study was to investigate both near
and far transfer effects of computerized working memory training
in a developmental population. Specically, we were interested in
whether working memory training would result in transfer effects
within an educational setting, measured by standardized tests of
verbal ability and academic attainment. The rationale behind this
approach was that these tasks rely in part on working memory
(see Swanson & Alloway, 2011 for a review), and it is possible that
by training working memory, some improvements may indeed
transfer to more knowledge-driven domains, like verbal ability
and attainment.
Another area of interest was whether any potential gains would
be maintained over time, without additional booster sessions.
Though the majority of studies have not reported long-term maintenance effects of training (see Morrison & Chein, 2011), there are

633

some positive improvements that remain. For example, in children


near transfer effects 3 months later (Klingberg, Forssberg, & Westerberg, 2002) and in older adults near transfer effects 3 months
later (Li et al., 2008). There were near and some far transfer effects
reported 6 months later in children (Holmes et al., 2009); and in
young adults near transfer effects 18 months later (Dahlin et al.,
2008); It is important to address the issue of maintenance effects
as it can help clarify whether any initial gains were indeed stable.
Thus, in the present study, we included an additional testing phase,
8 months post-training.
As is typical in this type of research, we employed the following
testing paradigm: a pre-test of cognitive skills (tests of working
memory, verbal ability, and academic attainment), followed by
an 8-week computerized working memory training program, and
then a post-test of the same cognitive skills. In addition, all students were also followed up 8 months later and re-tested on the
same cognitive tests.
In addition to a nonactive control group, we also included an active control group that only engaged with the training program on
a low-frequency basis. The inclusion of both nonactive and active
control groups is in response to a general criticism of Hawthorne
effect in the working memory training research (see Chein &
Morrison, 2010; Dahlin et al., 2008; Li et al., 2008; Schmiedek
et al., 2010). The Hawthorne effect refers to assumption that people
with adjust their behavior in accordance to their perceived expectations (see Green & Bavelier, 2008). In a training study, a nonactive control group may be aware that they are not receiving any
training and thus inadvertently perform poorly on the cognitive
tests. Likewise, the training cohort may invest in trying hard and
thus, exhibit good performance. It is hoped that the inclusion of
an active control group in the present study would circumvent
some of these issues. In addition, researchers were also blind to
which group the participants were assigned to when they scored
the cognitive tests at all time points.

2. Method
2.1. Participants
There were 94 students classied as having learning difculties
who participated in this study. The following criteria for learning
difculties was met: (a) they had a signicantly greater difculty
in learning than the majority of children of the same age; (b) they
were under compulsory school age and would fall within the definition specied in (a) if special educational provision was not
made for them; and (c) they were not regarded as having a learning
difculty solely because the language or form of language of their
home was different from the language in which they will be
taught. All participating students were receiving special educational support as a result of their learning difculties, which included reading difculties and language impairments. Special
educational provision refers to educational provision which is
additional to, or otherwise different from, the educational provision made generally for children of their age in schools maintained
by the Local Authority, other than special schools, in the area (SEN
Code of Practice 2001, 1:3 and the Education Act 1996, Section 312). All students were attending mainstream schools and
were allocated to one of three groups: nonactive control (n = 39;
12 males; M age = 10.11 years), active control, who engaged in
low-frequency training of once a week (WMT-Low; n = 32; 14
males; M age = 10.06 years); and Training group, who engaged in
high-frequency training of four times a week (WMT-High; n = 23;
14 males; M age = 11.2 years).

634

T.P. Alloway et al. / Computers in Human Behavior 29 (2013) 632638

There were three testing phases in the present study. In Phase 1,


there was an initial assessment of performance in tests of working
memory, ability, and attainment (pre-test). Training commenced
within 1 week after these assessments and lasted for 8 weeks. In
Phase 2, all participating students completed the same set of
assessments within 2 weeks of completing the training (post-test).
Finally, in Phase 3, participants were followed up 8 months later
and completed the same set of assessments (follow-up).

Objective Reading Dimensions (WORD; Wechsler, 1993). The math


test was taken from The Wechsler Objective Numerical Dimensions (WOND; Wechsler, 1996) and consisted of 28 questions,
relating to general arithmetic as well as fractions. These two subtests were selected for use in the present as they form the current
WIAT-Revised Test (WIAT-II-A) and represent highly valid and reliable measures of academic attainment. Testretest reliabilities for
spelling and maths are .91 and .85, respectively. Raw scores were
converted to standard scores (M = 100; SD = 15) based on the normative sample.

2.3. Cognitive assessments

2.4. Computerized training program

2.3.1. Working memory


Working memory was measured using a beta version of a standardized memory assessment, the Automated Working Memory
Assessment-II (AWMA; Alloway, 2012b). All test trials began with
two items, and increased by one item in each block, until the participant was unable to recall three correct trials at a particular
block. There were four trials in each block and the number of correct trials was scored for each participant. The move forward and
discontinue rules, as well as the scoring, were automated by the
program. The experimenter explained the task to the child and
conrmed that they understood the task. The children also
watched a computer-led demonstration, followed by a set of practice trials before beginning the test trials.
There were two verbal working memory tests. In the Backwards
Digit Recall test, the individual recalled a sequence of spoken digits
in the reverse order. In Processing Letter Recall, the participant
views a letter in red that stays on the computer screen for 1 s.
Another letter in black immediately follows this on the screen. Participants verify whether the black letter was the same as the red
letter by clicking on a box marked either Yes or No on the screen.
They then click on the red letters they saw in the correct sequence.
Visual working memory was tested using a shape recall test.
The participant views a colored shape in a 4  4 grid. The shape
and grid disappear and another shape appears in the center of
the computer screen. They verify whether those two shapes were
the same color and shape by clicking on a box marked either
Yes or No on the screen. Then they have to remember the location of the rst shape on the 4  4 grid in the correct sequence.
The stimuli in all working memory tests were randomized so no
stimulus sequence was repeated to avoid potential practice effects.
Scoring was based on the number of correct responses, and raw
scores were converted into standard scores (based on a normative
sample; Alloway, 2007), where 100 is the mean and 85 is the standard deviation. Testretest reliability of the AWMA was established in a random selection of the normative sample tested on
two separate occasions, 4 weeks apart. The reliability coefcient
for the verbal working memory tests was .86 and for the visuospatial working memory test, it was .84 (Alloway, 2007).

The training compromised of Jungle Memory (2008), a webbased working memory training program aimed at 716 year-old
children. The program uses three interactive computer games with
up to 30 levels of difculty in each game to train working memory.
Each game trains different aspects of working memory and provides the student with regular feedback of progress. Game 1 involves memory for and later use of word endings, Game 2
involves mental rotation of letters and Game 3 involves sequential
memory of mathematical solutions. Motivational features in the
program included positive verbal feedback, a display of the users
best scores, percentile rankings, and the number of super Monkeys
collected as a result of successfully completing the training levels.
Students in both the low- and high-training frequency groups
received individual usernames and passwords to log into the program. There was a maximum of 10 trials per session per game
and the students performed 30 trials on each day of training. The
main difference between the Active Control (WMT-Low) and Training (WMT-High) groups was the frequency in which they used the
training program. All features of the training were the same for
both groups. The students in the WMT-Low completed the training
program once a week over an 8-week period. They completed 24
sessions on average for all three memory games (eight sessions
per game). In contrast, the WMT-High used the program four times
a week and completed 84 sessions on average for all three memory
games (28 sessions per game) over an 8-week period.
Treatment delity was monitored by teacher and parent feedback and was also veried by log data generated from the training
program. Each time a student logged in and completed all 10 trials
for the three games, the program recorded their participation in a
log, which was stored on a database on the server. The experimenter (VB) regularly reviewed the log le to verify compliance.
Reminder emails and phone calls were also included regularly
throughout the 8-week period. Aside from the regularity of training, the training program, including access to the range of difculty
in the levels and the motivational features, were identical for both
these groups. The students in the control group carried on with
their regular classroom activities and did not participate in any
training program.

2.2. Procedure

2.3.2. General ability


Two subtests from the Wechsler Abbreviated Scale of Intelligence (WASI, Wechsler, 1999) were administered. In the Vocabulary test (verbal ability), the student had to provide a short
denition of a given word. In the Matrix test (nonverbal ability),
the student was shown incomplete patterns and had to select the
shape to complete the sequence. The WASI has UK norms and
has been shown to correlate well with full-scale IQ scores. However, due to time restrictions, only the Training group (WMT-High)
completed the Matrix test at testing Phases 1 and 2.

3. Results
The pre- and post-test assessments were scored by researchers
who were blinded to the student group allocation. Means and SDs
on the cognitive measures for the Training and nonactive and active control groups from all three testing phases are provided in
Table 1. All cognitive assessments were scored as standard scores
(M = 100, SD = 15).
3.1. Phase 1: Pre-training

2.3.3. Academic attainment


Standardized measures of spelling and arithmetic were also
administered. The spelling test was taken from The Wechsler

In order to conrm that both the Training and nonactive and active Control groups did not differ signicantly with respect to the

635

cognitive measures pre-training, we carried out a series of one-way


ANOVAs. There were no signicant differences between students
in the control and training groups between the cognitive measures,
except for Spelling (the alpha level was set to .008 for multiple
comparisons): Verbal working memory [F(2, 91) = 2.75, p = .07];
Visuo-spatial working memory [F(2, 91) = .145, p = .24]; Vocabulary [F(2, 91) = 3.80, p = .03]; Math [F(2, 91) = 2.60, p = .08]; Spelling [F(2, 91) = 22.95, p < .001].
Note: Alpha level for comparisons between pre- and post-training assessments was set to .01 for multiple comparisons. Sample sizes are indicated for Phases 1 and 2/Phase 3 (8-month follow-up).

M (SD)
101.28 (13.49)
107.89 (13.78)
95 (16.48)
113.27 (14.73)
83.44 (17.08)
93.36 (17.19)
p
.000
.005
.004
.001
.01
.40
M (SD)
102.30 (14.11)
101.70 (15.47)
91.13 (16.12)
97.70 (19)
82.17 (11.63)
93.26 (17.45)
M (SD)
91.31 (12.61)
85 (14.14)
84.38 (10.89)
106.83 (13.6)
99.5 (9.65)
94.04 (12.43)
p
.50
.39
.82

.89
.82
M (SD)
95.44 (9.60)
95.35 (15.96)
94 (11.85)

99.3 (11.62)
96.59 (11.22)
Verbal WM
Visuo-spatial WM
IQ: vocab (Verbal)
IQ: matrix (nonverbal)
Spelling
Math

M (SD)
94.02 (12.75)
97.12 (15.81)
94.38 (14.15)

99.9 (10.86)
96.77 (10.88)

M (SD)
87.74 (13.91)
95.95 (14.14)
90.50 (13.21)

89.44 (14.03)
90.19 (10.99)

M (SD)
91.47 (11.38)
94.81 (16.43)
86.56 (15.02)

89.19 (19.09)
88.81 (12.34)

p
.15
.58
.02

.35
.50

M (SD)
82.34 (13.7)
88.42 (13.62)
78 (16.3)
92.42 (18.83)
88.37 (17.1)
87.37 (14.47)

M (SD)
87.26 (12.99)
90.04 (19.66)
84.13 (15.31)
88.13 (23.45)
78.48 (11.37)
90.83 (18.78)

8-month
follow up
(n = 23/n = 11)

Pre- to
post-training
Pre-training
Pre- to
post-training
Post-training
Pre-training
Pre- to
post-training
Post-training
Pre-training

Active control (WMT-Low)

8-month
follow up
(n = 39/n = 24)
Nonactive control
Measures

Table 1
Group proles and means (standard scores) for cognitive tests at two testing times as a function of group.

(n = 32/n = 19)

8-month
follow up

Training group WMT-High

Post-training

T.P. Alloway et al. / Computers in Human Behavior 29 (2013) 632638

3.2. Phase 2: Post-training


A series of mixed ANOVAs were conducted on the cognitive
scores pre- and post-training for the Training and nonactive and
active Control groups. The following patterns were established.
For verbal working memory, there was a signicant difference in
performance between the two times [F(1, 88) = 20.86, p < .001,
g2p = .19] and approaching signicance as a function of group
[F(2, 88) = 2.65, p = .06, g2p = .06]; the interaction was signicant
[F(2, 88) = 6.63, p = .002, g2p = .13]. For visuo-spatial working
memory, there was a signicant difference in performance between the two times [F(1, 88) = 3.95, p = .05, g2p = .04] but not as
a function of group [F(2, 88) 6 1]; and the interaction was signicant [F(2, 88) = 7.23, p = .001, g2p = .14]. For both working memory
tests, the training gains were signicantly greater for the WMTHigh group compared to nonactive control and WMT-Low groups
(p < .01).
For vocabulary, there was not a signicant difference in performance as a function of group [F(2, 91) = 2.43, p = .09]; or between
the two times [F(1, 91) 6 1, p = .40]; but the interaction was significant [F(2, 91) = 8. 02, p = .001, g2p = .15]. The WMT-High group
performed signicantly better than the WMT-Low group posttraining (p < .05). For spelling, there was a signicant difference
in performance as a function of group [F(2, 91) = 17.44, p < .001,
g2p = .28]; but not between the two times [F(1, 91) = 1.03,
p = .31] ; the interaction was not signicant [F(2, 91) = 1. 85,
p = .16]. For math, there was a signicant difference in performance as a function of group [F(2, 91) = 3.07, p < .05, g2p = .06];
but not between the two times [F(1, 91) 6 1] ; the interaction
was not signicant [F(2, 91) = 1. 04, p = .34].
To examine the gains as a function of working memory training,
we subtracted the pre-test scores from the post-test scores and
compared the difference in scores (Time 2 Time 1) as a function
of group. In Table 2, scores below 0 indicate that the group performed worse at the post-test. Scores above 0 indicate improvements that the group made after training. The Nonactive Control
group showed no signicant improvements between the two testing times across all cognitive tests. The WMT-Low group (Active
Control) also showed no signicant improvements in scores from
pre- to post-training across all cognitive measures. In contrast,
the WMT-High (Training) group demonstrated signicant
improvements in all areas (working memory, verbal and nonverbal
ability, and spelling; p < .01), except for math. This selective pattern of enhancement suggests that the frequency of working memory training can make an impact when considering gains in
cognitive skills.
3.3. Phase 3: 8-Month follow-up
A series of mixed ANOVAs were conducted on the cognitive
scores pre-training and 8 months later for the control and training
groups. The following patterns were established. For verbal working memory, there was no signicant difference in performance
between the two times [F(1, 46) 6 1] but a signicance group
difference [F(2, 46) = 3.31, p = .04, g2p = .13]; the interaction was
signicant [F(2, 46) = 4.58, p = .015, g2p = .16]. For visuo-spatial
working memory, there was no signicant difference in

636

T.P. Alloway et al. / Computers in Human Behavior 29 (2013) 632638

Table 2
Difference between pre-test scores and post-test scores (Time 2Time 1) as a function
of group.
Measures

Nonactive
control

Active control
(WMT-Low)

Training group
(WMT-High)

Verbal WM
Visuo-spatial WM
IQ: vocab (Verbal)
IQ: matrix (Nonverbal)
Spelling
Math

15*
12*
7*
10*
4*
2

2
0

1
0

1
4

0
1

Note: Data for the Matrix test are only available for the WMT-High group.
*
Indicates a signicant difference between pre- and post-training scores (p < .01).

performance between the two times [F(1, 46) = 1] but there was a
signicant group difference [F(2, 46) = 3.11, p = .05, g2p = .12]; and
the interaction was signicant [F(2, 46) = 7.31, p = .002, g2p = .24].
Post hoc comparisons (p < .05, adjusted for multiple comparisons)
indicated that in the verbal working memory test, the WMT-High
group performed signicantly better than the WMT-Low group
8 months later. In the visuo-spatial working memory test, the
training gains were greater in Phase 3 for the WMT-High group
compared to both the Control and WMT-Low groups.
For vocabulary, there was not a signicant difference in performance as a function of group [F(2, 49) 6 1]; but there was a significant difference between the two times [F(1, 49 = 8.47, p < .005,
g2p = .15]; and the interaction was signicant [F(2, 49) = 10.98,
p = .001, g2p = .31]. Post hoc comparisons (p < .05, adjusted for
multiple comparisons) indicated that in the WMT-High group performed signicantly better than the WMT-Low group 8 months
later.
For spelling, there was a signicant difference in performance as
a function of group [F(2, 49) = 7.46, p < .001, g2p = .23]; but not between the two times [F(1, 419 = 1.43, p = .24] ; the interaction was
not signicant [F(2, 49) = 3. 09, p = .06]. WMT-High group performed signicantly better than the control group 8 months later.
For math, there was no signicant difference in performance as
a function of group [F(2, 51) = 1.78, p = .18]; but between the two
times [F(1, 51) = 5.05, p = .03, g2p = .09] ; the interaction was signicant [F(2, 91) = 4.25, p = .02, g2p = .14]. Post hoc comparisons
indicated that both the Control and WMT-Low groups performed
signicantly worse on the math test 8 months later, compared to
their pre-training score. The WMT-High training group did show
improvements at the 8-month follow up, though this was not
signicant.
4. Discussion
The present study focused on whether computerized working
memory training could result in both near and far transfer training
effects; and whether such effects would be maintained over time.
With respect to near transfer effects, the ndings suggest that
there were gains in untrained tests of both verbal and visuo-spatial
working memory for the high-frequency training group compared
to the nonactive and active (WMT-Low) Control groups. It is worth
noting that the surface features of the stimuli in the training task
were different from that of the working memory test, thus the
gains made in verbal and visuo-spatial working memory are unlikely due to a practice effect or improvements in test-taking skills
(Bors & Vigneau, 2003). It is possible that these gains could be explained by the nature of the training program (Ackerman, 1987), as
students had to engage in multiple executive processes required
for working memory tasks, such as monitoring current information, managing two tasks simultaneously, inhibiting irrelevant
stimuli, and updating items for recall. The nding that frequent

training improved both verbal and visuo-spatial scores suggests


that working memory skills are a domain-general resource, and
not just specialized for handling only one type of information
(Alloway, Gathercole, & Pickering, 2006; Bayliss, Jarrold, Gunn, &
Baddeley, 2003; Kane et al., 2004).
Looking next at far transfer effects, the data indicate that the
high training group performed signicantly better in verbal ability
compared than the low training group, in Phase 2. We also looked
at the potential increase in performance (as measured by standard
scores) from Phase 1 to Phase 2, as a function of group. The data
suggest that while performance did not improve signicantly for
the nonactive controls and Low-Training frequency group, the
high-training group performed signicantly better in tests of verbal and nonverbal ability (IQ), as well as in Spelling.
Finally, the data from the present study indicate that there were
some maintenance effects. In particular, performance in both verbal and visuo-spatial working memory, as well as in verbal ability
and spelling, was maintained in the high training group. This pattern is noteworthy, given that performance between the three
groups was similar at the baseline (Phase 1).
There are a number of possibilities for why the computerized
working memory training led to some far transfer effects in the
high training group. One explanation for the improvements evident immediately after training is that there greater efciency of
neuronal responses (Westerberg & Klingberg, 2007). The idea of
neural efciency is that when individuals do well in an activity,
fewer neurons are recruited compared to when they are learning
a task. A recent review of neural activity as a result of cognitive
training suggested that working memory training can result in decreased activation in the prefrontal cortex, an area of the brain
associated with working memory skills. This pattern is thought
to reect more efcient use of working memory, as a function of
practice and expertise (Buschkuehl et al., 2012).
Improved neural activity, as a result of working memory training, could have a twofold benet. First, given the importance of
working memory in a range of cognitive skills (see Cowan & Alloway, 2008), this greater efciency can result in improved performance in a range of untrained tests, including general ability
measures. This cascade effect ts in with the idea of working memory skills like a bottleneckthose with poor working memory
struggle to acquire the necessary knowledge to succeed in the
classroom. However, the data from the present study suggest that
regular and targeted training can widen the bottleneck, thus allowing the student a greater platform for learning and forming new
knowledge episodes.
Another potential benet of improved neural activity is an increase in targeted goal-directed behavior. Working memory has
also been described as a controller, a cognitive resource that can
keep a goal in mind, bring in cognitive resources from different
parts of the brain, and also manage incoming information (see
Alloway & Alloway, 2012, for an overview). This process has been
associated with prefrontal cortex functioning (Fuster, 1997), and
goal-maintenance and attentional focus pervades many higherlevel processes that are relevant to everyday activities, such as
driving (Basak & Zelinski, 2012). In the present study, feedback
from students indicated that regular training made them more
aware of how working memory impacts learning and the majority
reported that they now applied strategies in the classroom when
activities became too difcult. This shift from disengaging from
education when they were overloaded, to generating alternate
solutions to meet their learning goals suggests that they were able
to use their working memory skills with greater efciency to direct
their behavior.
An alternate explanation for the far transfer effects evidenced in
the present study could be due to an expectancy effect, where the
training group was aware that they were investing in a training

T.P. Alloway et al. / Computers in Human Behavior 29 (2013) 632638

activity. Thus, they may have exerted extra effort during the posttraining assessments, as a result of the expectancy and investment
during the training phase (see Morrison & Chein, 2011). It is hoped
that the inclusion of the low training group may have mitigated
this effect as they also engaged in training for 8 weeks, but nonetheless, the expectancy effect remains a possibility. However, it
may be that, given the role of working memory in goal-directed
behavior, training can also lead to increased motivation.
An alternative approach to support working memory difculties
is to introduce changes to the students environment to reduce the
detrimental effects of memory overload. Suggested ways include
simplifying the processing demands of classroom activities and
providing effective learning strategies (Gathercole & Alloway,
2008). Studies on strategy use in the classroom offers some
support in using such complimentary methods to help students
with working memory decits. For instance, rehearsal training
has been shown capable of improving working memory in
individuals with down syndrome (Conners, Rosenquist, Arnett,
Moore, & Hume, 2008; Conners, Rosenquist, & Taylor, 2001) and
fetal alcohol spectrum disorder (Loomes, Rasmussen, Pei, Manji,
& Andrew, 2008). In typically developing pre-schoolers, using mental abacus training and music training had positive effects on
visuo-spatial memory and verbal memory, respectively (Lee, Lu,
& Ko, 2007).
In summary, the present study offers supporting evidence that
computerized working memory training can lead to transfer gains
in untrained cognitive tests of ability and attainment, when used
regularly. Perhaps the way forward in supporting the working
memory needs of students is to offer a combination of working
memory training and targeted classroom strategies.

References
Ackerman, P. L. (1987). Individual differences in skill learning: An integration of
psychometric and information processing perspectives. Psychological Bulletin,
102, 327.
Alloway, T. P. (2007). Working memory, reading and mathematical skills in children
with developmental coordination disorder. Journal of Experimental Child
Psychology, 96, 2036.
Alloway, T. P. (2009). Working memory, but not IQ, predicts subsequent learning in
children with learning difculties. European Journal of Psychological Assessment,
25, 9298.
Alloway, T. P. (2012a). Can interactive working memory training improving
learning? Journal of Interactive Learning Research, 23, 111.
Alloway, T. P. (2012b). Automated Working Memory Assessment-II (AWMA-II).
London: Pearson Assessment.
Alloway, T. P., & Alloway, R. G. (2010). Investigating the predictive roles of working
memory and IQ in academic attainment. Journal of Experimental Child
Psychology, 106, 2029.
Alloway, T. P., & Alloway, R. G. (2012). Working memory: The connected intelligence.
New York: Psychology Press.
Alloway, T. P., & Gathercole, S. E. (2006). How does working memory work in the
classroom? Educational Research and Reviews, 1, 134139.
Alloway, T. P., Gathercole, S. E., Kirkwood, H. J., & Elliott, J. E. (2009). The cognitive
and behavioural characteristics of children with low working memory. Child
Development, 80, 606621.
Alloway, T. P., Gathercole, S. E., & Pickering, S. J. (2006). Verbal and visuo-spatial
short-term and working memory in children: Are they separable? Child
Development, 77, 16981716.
Baddeley, A. D. (1996). Exploring the central executive. Quarterly Journal of
Experimental Psychology, 49A, 528.
Basak, C., & Zelinski, E. (2012). A hierarchical model of working memory and its
change in healthy older adults. In T. P. Alloway & R. G. Alloway (Eds.), Working
memory: The connected intelligence. New York: Psychology Press.
Bayliss, D. M., Jarrold, C., Gunn, M. D., & Baddeley, A. D. (2003). The complexities of
complex span: Explaining individual differences in working memory in children
and adults. Journal of Experimental Psychology: General, 132, 7192.
Bors, D. A., & Vigneau, F. (2003). The effect of practice on Ravens advanced
progressive matrices. Learning and Individual Differences, 13, 291312.
Buschkuehl, M., Jaeggi, S., & Jonides, J. (2012). Neuronal effects following working
memory training. Developmental Cognitive Neuroscience, 25, S167S179.
Chein, J., & Morrison, A. (2010). Expanding the minds workspace. Training and
transfer effects with a complex working memory span task. Psychonomic
Bulletin and Review, 17, 193199.

637

Conners, F. A., Rosenquist, C. J., Arnett, L., Moore, M. S., & Hume, L. E. (2008).
Improving memory span in children with down syndrome. Journal of Intellectual
Disability Research, 52, 244255.
Conners, F. A., Rosenquist, C. J., & Taylor, L. A. (2001). Memory training for children
with down syndrome. Downs Syndrome, Research and Practice, 7, 2533.
Cowan, N. (2006). Within uid cognition: Fluid processing and uid storage?
Behavioral and Brain Sciences, 29, 129130.
Cowan, N., & Alloway, T. P. (2008). The development of working memory in
childhood. In M. Courage & N. Cowan (Eds.), Development of memory in infancy
and childhood (2nd ed., pp. 303342). Hove, England: Psychology Press.
Dahlin, E., Neely, A. S., Larsson, A., Backman, L., & Nyberg, L. (2008a). Transfer of
learning after updating training mediated by the striatum. Science, 320,
15101512.
Dahlin, E., Nyberg, L., Backman, L., & Neely, A. S. (2008b). Plasticity of executive
functioning in young and older adults: Immediate training gains, transfer, and
long-term maintenance. Psychology and Aging, 23, 720730.
Engle, R. W., Tuholski, S. W., Laughlin, J. E., & Conway, A. R. A. (1999). Working
memory, short-term memory, and general uid intelligence: A latent-variable
approach. Journal of Experimental Psychology: General, 128, 309331.
Fuster, J. M. (1997). The Prefrontal Cortex (3rd ed.). New York: Raven Press.
Gathercole, S. E., & Alloway, T. (2008). Working memory and learning: A practical
guide. London: Sage Publications.
Gathercole, S. E., Alloway, T. P., Willis, C., & Adams, A. M. (2006). Working memory
in children with reading disabilities. Journal of Experimental Child Psychology, 93,
265281.
Geary, D. C., Hoard, M. K., & Hamson, C. O. (1999). Numerical and arithmetical
cognition: Patterns of functions and decits in children at risk for a
mathematical disability. Journal of Experimental Child Psychology, 74, 213239.
Green, C. S., & Bavelier, D. (2008). Exercising your brain: A review of human brain
plasticity and training-induced learning. Psychology and Aging, 23, 692701.
Healy, A. F., Wohldmann, E. L., Sutton, E. M., & Bourne, L. E. (2006). Specicity effects
in training and transfer of speeded responses. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 32, 534546.
Holmes, J., Gathercole, S. E., & Dunning, D. L. (2009). Adaptive training leads to
sustained enhancement of poor working memory in children. Developmental
Science, 12, 915.
Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008). Improving uid
intelligence with training on working memory. Proceeding of the National
Academy of Science, USA, 105, 68296833.
Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Shah, P. (2011). Short- and long-term
benets of cognitive training. Proceedings of the National Academy of Sciences of
the USA, 108, 1008110086.
Jungle Memory (2008). Memosyne Ltd.
Kane, M. J., Hambrick, D. Z., Tuholski, S. W., Wilhelm, O., Payne, T. W., & Engle, R. W.
(2004). The generality of working memory capacity: A latent variable approach
to verbal and visuo-spatial memory span and reasoning. Journal of Experimental
Psychology: General, 133, 189217.
Klingberg, T., Fernell, E., Olesen, P. J., Johnson, M., Gustafsson, P., Dahlstrom, K., et al.
(2005). Computerised training of working memory in children with ADHD A
randomized, controlled trial. Journal of the American Academy of Child and
Adolescent Psychiatry, 44, 177186.
Klingberg, T., Forssberg, H., & Westerberg, H. (2002). Training of working memory in
children with ADHD. Journal of Clinical and Experimental Neuropsychology, 24,
781791.
Kremen, W. S., Jacobsen, K. C., Xian, H., et al. (2007). Genetics of verbal working
memory process: A twin study of middle-aged men. Neuropsychology, 21,
569580.
Lee, Y., Lu, M., & Ko, H. (2007). Effects of skill training on working memory capacity.
Learning and Instruction, 17, 336344.
Li, S. C., Schmiedek, F., Huxhold, O., Rocke, C., Smith, J., & Lindenberger, U. (2008).
Working memory plasticity in old age: Practice gain, transfer, and maintenance.
Psychology and Aging, 23, 731742.
Loomes, C., Rasmussen, C., Pei, J., Manji, S., & Andrew, G. (2008). The effect of
rehearsal training on working memory span of children with fetal alcohol
spectrum disorder. Research in Developmental Disabilities, 29, 113124.
Martinussen, R., Hayden, J., Hogg-Johnson, S., & Tannock, R. (2005). A meta-analysis
of working memory impairments in children attention-decit/hyperactivity
disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 44,
377384.
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 and Review, 18, 4660.
Owen, A. M., Hampshire, A., Grahn, J. A., Stenton, R., Dajani, S., Burns, A. S., et al.
(2010). Putting brain training to the test. Nature, 465, 775778.
Schmiedek, F., Lovden, M., & Lindenberger, U. (2010). Hundred days of cognitive
training enhance broad abilities in adulthood: Findings from the COGITO study.
Frontiers in Aging Neuroscience, 2, 110.
Swanson, H. L., & Alloway, T. P. (2011). Working memory, learning, and academic
achievement. In K. Harris, T. Urdan, & S. Graham (Eds.), APA Educational
Psychology Handbook (Vol. 1).
Thorell, L. B., Lindquist, S., Bergman Nutley, S., Bohlin, G., & Klingberg, T. (2009).
Training and transfer effects of executive functions in preschool children.
Developmental Science, 12, 106113.
Van der Molen, M. J., Van Luit, J. E. H., Van der Molen, M. W., Klugkist, I., & Jongmans,
M. J. (2010). Effectiveness of a computerised working memory training in

638

T.P. Alloway et al. / Computers in Human Behavior 29 (2013) 632638

adolescents with mild to borderline intellectual disabilities. Journal of


Intellectual Disability Research, 54, 433447.
Wechsler, D. (1993). Wechsler objective reading dimensions. London: Pearson
Assessment.
Wechsler, D. (1996). Wechsler objective numerical dimensions. London: Pearson
Assessment.

Weschler, D. (1999). Weschler abbreviated scale of intelligence. London: Pearson


Assessment.
Westerberg, H., & Klingberg, T. (2007). Changes in cortical activity after training of
working memory: A single-subject analysis. Physiology and Behavior, 92,
186192.

Potrebbero piacerti anche