Psychology and Aging © 2014 American Psychological Association

2014, Vol. 29, No. 3, 706 –716 0882-7974/14/$12.00 http://dx.doi.org/10.1037/a0037507

Video Game Training Enhances Cognition of Older Adults:

A Meta-Analytic Study
Pilar Toril, José M. Reales, and Soledad Ballesteros
The National University of Distance Education

It has been suggested that video game training enhances cognitive functions in young and older adults.
However, effects across studies are mixed. We conducted a meta-analysis to examine the hypothesis that
training healthy older adults with video games enhances their cognitive functioning. The studies included
in the meta-analysis were video game training interventions with pre- and posttraining measures. Twenty
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

experimental studies published between 1986 and 2013, involving 474 trained and 439 healthy older
This document is copyrighted by the American Psychological Association or one of its allied publishers.

controls, met the inclusion criteria. The results indicate that video game training produces positive effects
on several cognitive functions, including reaction time (RT), attention, memory, and global cognition.
The heterogeneity test did not show a significant heterogeneity (I2 ⫽ 20.69%) but this did not preclude
a further examination of moderator variables. The magnitude of this effect was moderated by method-
ological and personal factors, including the age of the trainees and the duration of the intervention. The
findings suggest that cognitive and neural plasticity is maintained to a certain extent in old age. Training
older adults with video games enhances several aspects of cognition and might be a valuable intervention
for cognitive enhancement.

Keywords: aging, cognitive functions, meta-analysis, moderating factors, video game training

The proportion of people aged over 65 is increasing worldwide Aging is associated with declines in many cognitive processes.
(United Nations, 2010). Given this increasing longevity and the However, current findings from longitudinal studies have found
cognitive and physical declines that occur with aging, researchers older age a time of decline, stability, and even growth (Baltes &
are investigating ways to promote independent living, delaying Lindenberger, 1997; Rönnlund, Lövden, & Nilsson, 2008). Spe-
cognitive decline as much as possible (for reviews see Hertzog, cifically, declines occur with age in several processes including
Kramer, Wilson, & Lindenbergen, 2008; Park & Reuter-Lorenz, processing speed, attention, executive control, working memory,
2009). To this end, efforts are being made to investigate the and episodic memory (e.g., Hoyer & Verhaeghen, 2006; Nilsson,
potential of new information and communication technologies 2003; Rönnlund, Nyberg, Bäckman, & Nilsson, 2005; Salthouse,
(ICT) to improve cognitive functioning (Bond, Wolf-Wilets, 1996). In contrast, other crystalized cognitive functions as general
Fiedler, & Burr, 2001) and quality of life in older adults (Balles- knowledge, verbal abilities (e.g., Bialystok & Craik, 2006; Hedden
teros, Toril, Mayas, Reales, & Waterworth, 2014; Leung & Lee, & Gabrieli, 2004) and implicit memory (e.g., Ballesteros, Bischof,
2005; Peter et al., 2013). The reduction in the number of social Goh, & Park, 2013; Ballesteros & Reales, 2004; Mitchell & Bruss,
relations, the deterioration of physical abilities and the decline of 2003; Wiggs, Weisberg, & Martin, 2006) are mostly preserved or
cognitive functioning are important burdens in old age (Meijer, even improve.
Van Boxtel, Van Gerven, Van Hooren, & Jolles, 2009). A number of recent studies have shown that positive changes in
older adults= cognition can occur after training with video games
(Anguera et al., 2013; Buschkuehl et al., 2008; Nouchi et al.,
2012). Although findings on this topic are sparse, evidence sug-
Pilar Toril, Department of Basic Psychology, The National University of
Distance Education; José M. Reales, Department of Methodology of the Social
gests that the older brain retains considerable plasticity. In other
Sciences, Studies on Aging and Neurodegenerative Diseases Research Group, words, it has the ability to increase its capacity in response to
The National University of Distance Education; Soledad Ballesteros, Depart- experience. The observed increase in neural volume in response to
ment of Basic Psychology, Studies on Aging and Neurodegenerative Diseases cognitive training is an important indicator of brain change (see
Research Group, The National University of Distance Education. Boyke, Driemeyer, Gaser, Büchel, & May, 2008; Park & Bischof,
This work was supported by grants from the Spanish Government 2013). Encouraged by previous findings showing that cognitive
(PSI2010-21609-C02-01) and a Madrid Community Biomedical Research training interventions can improve cognition in healthy older
grant to S. Ballesteros (S2010/BMD-2349), and a predoctoral research grant to adults (Ball et al., 2002; Basak, Boot, Voss, & Kramer, 2008;
Pilar Toril awarded to the Studies on Aging and Neurodegenerative Diseases Boot, Kramer, Simons, Fabiani, & Gratton, 2008; Willis et al.,
Consolidated Research Group (The National University of Distance Educa-
2006), there is a growing interest in video game training as an
tion). We are very grateful to three anonymous reviewers for their helpful and
very constructive comments that helped us to improve the article.
effective way of enhancing important aspects of cognition and
Correspondence concerning this article should be addressed to Soledad neural plasticity in older adults (e.g., Anguera et al., 2013; Boot,
Ballesteros, Department of Basic Psychology, The National University of Blakely, & Simmons, 2011; Nouchi et al., 2012; Prakash et al.,
Distance Education, Juan del Rosal, 10, 28040 Madrid, Spain. E-mail: 2012). Cognitive training can be defined as an intervention that
mballesteros@psi.uned.es provides structured practice on tasks relevant to different aspects

706
 VIDEO GAME TRAINING IN AGING: A META-ANALYSIS 707

of cognitive functioning, such as memory, attention, language, or over SR is that it involves statistical tests on individual effect sizes
executive functions. Theoretically, these studies suggest that the producing, among other things, significance statistics, confidence
older human brain maintains some neural plasticity (e.g., Pascual- intervals, and heterogeneity indexes that enhance the information
Leone, Amedi, Fregni, & Merabet, 2005; Raz et al., 2005) al- provided by the analysis. For example, Kueider et al. (2012)
though not to the same degree as young adults (e.g., Bialystok & outlined the difficulties posed by the variability and length of
Craik, 2006; Lee et al., 2008; Li, Brehmer, Shing, Werkle-Bergner, video game interventions in the results obtained. Meta-analysis is
& Lindenberger, 2006). thus an appropriate tool to elucidate the conflicting results pub-
In our technological society, computer-based training programs lished so far in this field. The existence of several outcomes and
and video games have interested researchers as a tool for improv- moderator variables in the original studies does not preclude the
ing and/or maintaining perceptual and cognitive functions in older possibility of conducting a meta-analysis. Borenstein, Hedges,
adults. However, so far, scientific evidence of the potential of these Higgins, and Rothstein (2009) suggested that this kind of meta-
interventions is mixed, as reviewed below. analysis is feasible although it involves great complexity. The
present meta-analytic study is thus intended as a refinement and an
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Effects of Training With Video Games on Cognitive extension of the SR conducted by Kueider et al. (2012).
This document is copyrighted by the American Psychological Association or one of its allied publishers.

One of the main difficulties in obtaining a clear picture of the

Functions in Older Adults
effects of video game training is the great variability of several key
A video game is an electronic game that involves human inter- features of the intervention studies, including the type of video
action with a computer by means of a user interface that generates game used, the type of cognitive process assessed, the way in
visual and auditory feedback. Video games can be classified in which these cognitive processes were evaluated, and personal
different nonoverlapping categories. The first category is “serious characteristics of the trainees. Presumably, the mixed results re-
games” (Gopher, Weil, & Bareket, 1994), primarily designed to ported in the literature can be attributed to this variability. For
convey information or a learning experience of some sort to the example, some studies reported positive results (e.g., Anguera et
game player. A second category comprises “educational games,” al., 2013; Goldstein et al., 1997; Torres, 2008), whereas others did
such as Brain Age and Brain Training (McDougall & House, not find any cognitive improvement (e.g., Ackerman et al., 2010;
2012). Video games can also be classified as simple or non-action Owen et al., 2010). Meta-analysis is a good tool to determine the
games and complex games. variables responsible for the discrepancies reported in the inter-
The use of video games as a way to enhance cognitive func- vention studies (Hertzog et al., 2008).
tioning in healthy older adults has an advantage over traditional
cognitive training programs in that they are relatively inexpensive,
Scope, Aims, and Hypotheses of the Meta-Analysis
enjoyable, and fun (Zelinski & Reyes, 2009). Video games include
images, movement, sound, and feedback. All these characteristics The main aim of this meta-analytic study was to investigate the
are more attractive and rewarding than printing materials. More- extent to which cognitive training with video games enhances
over, a recent study (Allaire, McLaughlin, Trujillo, Whitlock, & cognitive functions in healthy older adults. It synthesizes the effect
Laporte, 2013) showed that regular and occasional video game sizes obtained in video game training studies conducted to inves-
players reported significantly higher levels of wellbeing and less tigate transfer effects. Furthermore, we were interested to find out
depression than nonvideo game players. For these reasons, video which training variables might explain the wide variability of
games have been used as a tool for training young and older adults, results reported in the literature. We also tried to identify the
and are considered to provide a good context for cognitive enrich- specific variables involved in the modulation of the effect sizes
ment (Achtman, Green, & Bavelier, 2008; Green & Bavelier, obtained in the intervention studies. The first step was to review
2008). Early studies with older adults revealed improvements in the published literature in order to identify the main variables that
several cognitive functions after video game training, including modulate the effect. We selected six main variables thought to
processing speed (Clark, Lamphear, & Riddick, 1987; Dustman, contribute to the results: (a) the type of video game used, (b) the
Emmerson, Steinhaus, Shearer, & Dustman, 1992; Goldstein et al., duration of the training program, (c) the number of games used in
1997), intelligence (Drew & Waters, 1986), visuomotor coordina- the training program, (d) the type of program, (e) the type of
tion (Drew & Waters, 1986), attention (Belchior, 2008), and global control group, and (f) the age of the participants. These variables
cognitive function (Torres, 2008). However, other studies did not were coded for each study and introduced as covariates in the
find a significant transfer of training to measures of cognitive and analysis. The dependent variables used to compute the effect sizes
perceptual functioning (e.g., Ackerman, Kanfer, & Calderwood, are known as outcomes. In the literature on video game interven-
2010; Boot et al., 2013; Owen et al., 2010). The transfer of video tion studies, there are a variety of outcomes related to cognitive
game training to untrained cognitive functions is critical and has processes. We combined the outcomes to compute a single effect
important practical significance. size for each study. To examine the effect of the moderator
Recently, Kueider, Parisi, Gross, and Rebok (2012) conducted a variables, the outcomes were classified into five broad cognitive
systematic review (SR) to examine the effectiveness of computer- categories: memory, attention, RT, executive functions, and global
based cognitive interventions in cognitively healthy older adults. cognitive function.
The SR method attempts to answer a theoretical question by To code the type of video game used in each study, we at-
analyzing empirical published studies in the field of interest to tempted to classify them. This was not an easy task as there has
obtain a summary of the results in terms of effect size. Meta- been a spectacular development in this field during the last decade.
analysis, however, provides a better way of analyzing the results of Briefly, the first video games were simple or non-action games
several individual studies. The main advantage of meta-analysis (e.g., Pac-Man or Donkey Kong). Their use as cognitive training
 708 TORIL, REALES, AND BALLESTEROS

tools improved performance on speed of processing tasks but not ducted between 1986 and 2013 using several combinations of the
on executive function or working memory tasks (Clark et al., 1987; following keywords: “aging,” “older adults,” “video game train-
Dustman et al., 1992; Goldstein et al., 1997). Later, more complex ing,” “memory,” and “cognitive function.” We also performed a
action games were developed in which planning and strategic manual search and cross-referencing of original articles. We re-
factors played a predominant role. Video games such as Medal of stricted the search to articles written in English. As the video game
Honor and Space Fortress are classified as complex games that business developed with the computer industry in the late 1980s,
activate many perceptual and cognitive skills. Some studies have the first article found was published in 1986. After reading the
used this type of video game in their interventions with older articles obtained from the electronic search, we found additional
adults (e.g., Basak et al., 2008; Stern et al., 2011). The type of related papers not obtained in the previous step. The titles and
video game used to train older adults is thus an important variable abstracts of the original articles were screened for potential inclu-
in video game training studies. Zelinski and Reyes (2009) classi- sion in the study.
fied video games as simple versus complex. The former does not Selection criteria. Studies were included if they met the fol-
involve complex cognitive demands, yet the latter requires the lowing criteria: (a) they involved only healthy older adults; (b)
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

concurrent recruitment of many perceptual and cognitive skills. they reported pre- and postevaluation results of the same cognitive
This document is copyrighted by the American Psychological Association or one of its allied publishers.

We followed this approach when classifying the types of video outcomes; (c) the trained group received only video game training;
game used in the intervention studies. Two complementary mod- and (d) the studies reported all the descriptive statistics necessary
erators were also considered in our analysis: duration of training, to compute the d effect size index and its confidence interval. The
and number of games used in the intervention. The amount of electronic and manual search yielded a total of 60 articles, but only
training, as well as the diversity of skills involved, might also 20 fulfilled the inclusion criteria and were included in the meta-
explain some of the variance of the results even when the type of analysis. Figure 1 shows the flow diagram with the search char-
video game is the same. We also included in the meta-analysis acteristics and the inclusion criteria.
other two moderator variables, type of program (commercial video Of the 20 articles included in the present study, two (Ackerman
games vs. “brain training” programs designed specifically to im- et al., 2010 and Cassavaugh & Kramer, 2009) did not include a
prove cognitive functions) and type of control group (active con- control group. However, it was possible to compute the effect size
trol group vs. passive control group). Finally, the age of the indexes corresponding to these two studies using the formula for a
participants as a moderator variable is also important as older prepost design without a control group (see Equation 1).
adults vary considerably in their cognitive abilities and health
status. People aged between 60 and 70 might have preserved their
functionalities but people older than that might experience the
burdens of aging more profoundly.
d ⫽ c(n ⫺ 1)
Y Pre ⫺ ៮
៮ Y Post
SPre 冉
⫽ 1⫺
3
4n ⫺ 5 冊 ៮
Y Pre ⫺ ៮
SPre
Y Post
(1)

We hypothesized that there would be an overall cognitive im- Equation 1 has the c term included multiplicatively to correct
provement after training, with some variance in the amount of bias in Cohen’s d (Carlson & Schmidt, 1999; Hedges & Olkin,
improvement depending on the cognitive processes assessed (out- 1985), n is the number of participants, ៮
Y is the mean of the
comes). Moreover, with regard to the combined effect size of the dependent variable with the subscript signaling the phase (Pre:
individual studies, we hypothesized that complex games would pretest, Post: posttest) and S is the standard deviation of the same
yield greater improvement than simple games because the former dependent variable.
involve a broader range of skills. Our second hypothesis was that The other 18 studies were coded using Equation 2 (prepost
there would be a positive relationship between the overall cogni- experimental design with two groups, experimental and control,
tive improvement and the number of training sessions. The third and a continuous dependent variable).
hypothesis was that there would also be a positive relationship
between the number of video games the participants were trained
冉 3
冊共 ៮ Pre ⫺ Y Post兲 ⫺ 共Y Pre ⫺ Y Post 兲
Y Exp ៮Exp ៮Cont ៮Cont

冑
to play and improvements in cognition. The fourth hypothesis was d⫽ 1⫺
4(nExp ⫹ nCont) nExp共SExp
Pre 兲 ⫹ nCont共SPre 兲
2 Cont 2
that the interaction with the experimenter would be beneficial. We
tested this hypothesis by assessing the effect of the type of control nExp ⫹ nCont ⫺ 2
group included in the study. The fifth hypothesis was that brain
(2)
training specifically designed to improve cognition would be more
effective than commercial video games. Finally, we hypothesized Equation 2 has the same meaning as Equation 1 in the sub-
that cognitive improvements, while remaining significant, would scripts, superscripts, and the symbols used. So, the first multipli-
decrease with age. Consequently, the age covariant would have a cative term between brackets is a bias correction of Cohen’s d
detrimental effect on cognitive improvement. (Hedges & Olkin, 1985), n is the number of participants on each
group (signaled by their subscripts) and superscripts Exp and Cont
Method meaning “Experimental group” and “Control Group,” respectively.
Standardized Cohen’s d statistics were computed from the pre- and
Literature Search postintervention means and standard deviations of each group for
each outcome variable using the formulas outlined above (Hedges
A systematic search strategy was used to identify relevant & Olkin, 1985). The characteristics of these studies are presented
studies to be included in this meta-analysis. The MEDLINE, in Table 1.
Psyc-Info, and Google Scholar databases were searched to identify Characteristics of training interventions. As expected, the
relevant studies. Periodic searches of these databases were con- video games used as training platforms in the 20 studies included
 VIDEO GAME TRAINING IN AGING: A META-ANALYSIS 709

With regard to the type of video game, these were classified as

simple versus complex. A game was coded as complex if it
involved multiple cognitive processes and complex cognitive de-
mands. The number of video games was also classified with two
categories: few (one to six games) and many (seven to 12 games).
Another factor considered in the present study was the duration of
the intervention; two levels were selected: short (between 1 and 6
weeks) and long (between 7 and 12 weeks). Type of program was
coded with two levels (“brain training” program and video games).
The variable type of control group was also coded with two levels
(active and passive). The last variable was the age of the partici-
pants. Unfortunately, the published studies did not provide con-
sistent information about the age of the participants, some report-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

ing the age range and others the mean age of participants.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Consequently, we decided to code age as a categorical variable

with two levels: (a) from 60 to 70 years of age, and (b) from 71 to
80 years of age (when the mean age or the half-split range point
reported was within these ranges). Finally, we coded the outcome
measures of the studies as cognitive processes (RT, executive
functions, memory, attention, and global cognitive functioning)
based on the category of tests used at prepost assessment.
Table 2 displays the moderator variables and their coded levels
in the present study.

Outcomes
The 20 studies included in this meta-analysis reported multiple
outcomes on several cognitive functions that we classified as
memory, attention, RT, executive functions, and overall cognitive
function, although not all of them reported the same cognitive
functions. Moreover, even when studies assessed the same cogni-
tive functions, they used different tests to evaluate them. The
difficulty of comparing different measures of the same cognitive
function using different tests was overcome by converting the raw
statistics to Cohen’s d using a common metric for all the outcomes.
To evaluate interrater agreement, the first author (P.T.) and the
Figure 1. Flow diagram of studies considered and ultimately included in second author (J.M.R) coded the data from the 20 original studies
the meta-analysis. using the same codebook. There was no disagreement regarding
the moderator variables (Pearson=s correlation r ⫽ 1). The level of
agreement between the two coders for outcomes (memory, atten-
in the present investigation differed in several characteristics. It is tion, RT, global cognitive function, and executive functions) was
important to note that none of the video games used in these computed as the correlation between the effect sizes (d) for the
intervention studies, except those used in Anguera et al.’s (2013) outcomes between the two assessments. The interrater correlation
study, were originally designed to improve cognition in older for the main outcomes was r ⫽ .97 (p ⬍ .001), indicating that the
adults. The studies used a wide range of video game genres (see agreement between the coders was very high, both for moderator
Table 1); some used commercial video games such as Medal of and outcome variables.
Honor, Pac Man, Donkey Kong, Tetris, Crystal Castle, and so
forth, while others used a combination of classic cognitive tasks
Meta-Analytic Procedure
taken from commercial packages (e.g., Nintendo Brain Training,
Brain Age, Big Brain Academy, etc.). This implies that the indi- The analyses were conducted using the Comprehensive Meta-
vidual studies trained different cognitive processes as an unavoid- Analysis software (Borenstein, Hedges, Higgins, & Rothstein,
able consequence. Two other relevant differences were the number 2005). Effect sizes were computed for each individual study and test
of training sessions and the number of video games used in the using Cohen’s d for experimental designs with pre- and posttests and
same intervention study. These two characteristics are likely to experimental and control groups. As mentioned above, the only
affect the intensity of the training program. We coded these char- exceptions were the studies by Cassavaugh and Kramer (2009) and
acteristics (type of video game, duration or number of training by Ackerman, Kanfer, and Calderwood (2010) in which we used
sessions, type of program, type of control group and number of only the treatment group. As all the studies reported data on more
video games included in the training program) as moderator vari- than one individual test in the same cognitive area, we computed
ables. the mean. The effect size and its standard deviation (SD) for each
 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.

710
Table 1
Characteristics of the Studies Included in the Meta-Analysis

STUDY AGE Control N Video games Duration (training) Significant findings

Ackerman et al., 2010 50–71 -------------- 78 Wii Big Brain Academy 4 weeks: 5x/weeks No significant transfer of training from Wii
practice or reading tasks to measures of
cognitive and perceptual speed abilities.
Anguera et al., 2013 M ⫽ 67 Active/nonactive 46 Neuroracer 4 weeks Training enhanced cognitive control in older
adults. These benefits were extended to
untrained abilities.
Basak et al., 2008 63–75 No contact 39 Rise of Nations 4–5 weeks: 3x/weeks EG improved memory, executive function, and
visuospatial abilities.
Belchior, 2008 67–84 Tetris or no contact 58 UFOV or Medal of Honor 2 weeks: 2–3/week EG improved processing speed more than control
group.
Boot, 2013 M ⫽ 74 No contact 40 Brain Age 12 weeks Cognitive abilities did not improve.
Boot, 2013 M ⫽ 74 No contact 34 Mario Kart 12 weeks Cognitive abilities did not improve.
Bozoki et al., 2013 60–80 Online activities 60 Online video games 6 weeks Overall analysis did not show transfer effects. The
effect sizes were relatively small.
Cassavaugh & Kramer, 2009 M ⫽ 71.7 ------------- 21 Computer training program 2–3 weeks: 8 sessions EG improved reaction time.
Clark et al., 1987 57–83 No contact 14 Pac Man or Donkey Kong 7 weeks: 120 min/ EG improved reaction time.
week
Drew & Waters, 1986 61–78 Contact 13 Atari Crystal Castle 8 weeks: 12x/week EG improved psychomotor speed and global
cognition.
Dustman et al., 1992 62–71 Movie or no contact 60 Breakout, Galaxian Frogger 11 weeks: 3x/week EG improved reaction time.
Kaboom, PacMan, . . .
Goldstein et al., 1997 72–85 No contact 22 SuperTetris 5 weeks: 300 min/ EG improved reaction time. EG and CG also
week improved executive functions but there were no
differences between groups.
Maillot et al., 2012 65–75 No contact 32 Nintendo Wii 12 weeks EG group improved more than CG on measures
of physical function, executive control, and
processing speed functions, but not on
TORIL, REALES, AND BALLESTEROS

visuospatial measures.
McDougall & House, 2012 M ⫽ 74 No contact 41 Nintendo Brain Training 6 weeks EG improved in Digit Span Test and other tests.
Nouchi et al., 2012 M ⫽ 69 Tetris 28 Brain Age 4 weeks EG improved executive functions and processing
speed.
Peretz et al., 2011 60–77 Computer games 121 C. Personal Coach 12 weeks: 3x/week EG and CG improved focused and saturated
attention, memory recognition, and mental
flexibility.
Sosa, 2011 M ⫽ 74 Nonactive 31 Brain Age 5 weeks: 1/week EG improved syllable (time), arithmetic (time),
and Stroop tests.
Stern et al., 2011 M ⫽ 66 Active or nonactive control 60 Space Fortress 12 weeks: 36 hr One measure of executive control showed
improvements in EG.
Torres, 2008 60–86 Muscle relaxation/no contact 43 Super Granny, Zoo Keeper, 8 weeks: 1/week EG showed less cognitive decline than CG
Penguin Push, Bricks,
memory games
Van Muijden et al., 2012 60–77 Documentary group 72 Anagram, Falling bricks 7 weeks/24.5 hr Modest support for the potential of video game
training to improve cognitive functions in older
people.
Note. EG ⫽ experimental group; CG ⫽ control group; UFOV ⫽ useful field of view; Control refers to the control group activity.
 VIDEO GAME TRAINING IN AGING: A META-ANALYSIS 711

Table 2 this result could be due to a low power (Borenstein et al., 2009).
Variables Analyzed For this reason, we calculated the effect of the moderator variables.
Regarding the type of video game (simple vs. complex), the
Variables Levels heterogeneity between groups was not significant, Q (1) ⫽ 0.55,
Age of participants 60–70 years 71–80 years p ⬎ .05. The mean effect size for simple video games was 0.42,
Training duration Short (1–6 weeks) Long (7–12 weeks) 95% CI [0.25, 0.58], Z ⫽ 5.00, p ⬍ .01, and for complex games
Type of video game Simple Complex it was 0.33, 95% CI [0.18, 0.48], Z ⫽ 4.38, p ⬍ .01.
Number of games Few (1–6) Many (7–12) The second variable was the duration of training with two levels
Type of control group Active Passive
Type of program Brain training Video games (short vs. long). The heterogeneity between groups was significant,
Q (1) ⫽ 3.73, p ⫽ .05, I2 ⫽ 73.19%. The mean effect size for short
training was 0.49, 95% CI [0.32, 0.67], Z ⫽ 5.59, p ⬍ .01; and for
long training it was 0.26, 95% CI [0.09, 0.43], Z ⫽ 3.03, p ⬍ .01.
study and each cognitive function were calculated. Finally, we The third moderator variable analyzed in the present meta-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

combined the effect sizes of the same cognitive function in each analysis was the number of video games (few vs. many) used in the
This document is copyrighted by the American Psychological Association or one of its allied publishers.

study to overcome the statistical problem of assigning more weight interventions. The heterogeneity between groups was not signifi-
to studies with more individual outcomes. We used these mean cant, Q (1) ⫽ 0.37, p ⬎ .05. The mean effect size for many games
scores for each cognitive function as outcomes instead of raw was 0.30, 95% CI [0.07, 0.54], Z ⫽ 2.56, p ⬍ .05, and for few
effect sizes computed for each individual test in each study. games it was 0.39, 95% CI [0.26, 0.51], Z ⫽ 6.12, p ⬍ .01.
Moreover, five of the 20 final individual studies included in the The fourth variable analyzed was type of program (brain train-
meta-analysis had subgroups within the study. In all cases, the ing vs. video game). The heterogeneity between groups was not
subgroups compared the same experimental (trained) group with significant, Q (1) ⫽ 0.27, p ⬎ .05. The mean effect size for brain
different control groups. As not all the studies reported data on the training was 0.34, 95% CI [0.17, 0.50], Z ⫽ 4.04, p ⬍ .01, and for
same subgroups, we treated each subgroup as a separate study. video games was 0.40, 95% CI [0.25, 0.55], Z ⫽ 5.25, p ⬍ .01.
Mean effect sizes and confidence intervals were estimated using The next moderator variable analyzed was type of control group
the fixed effect model. By convention, an absolute effect size of (active vs. passive). We analyzed this variable in two ways: (a)
0.2 or less is considered small, an absolute value between 0.2 and including all the studies that had a control group, excluding only
0.6 is considered moderate, and an estimated value equal to or those without control group (two studies); and (2) analyzing the
greater than 0.6 is considered large. A 95% confidence interval effect sizes of those studies that used both an active and a passive
(CI) was calculated for each effect size to establish whether it was control group, excluding those studies that had just a passive
statistically different from zero. We examined the variation in control or no control group. The results of the first analysis showed
effect sizes between studies using a standardized scale, based on that the heterogeneity between groups was not significant, Q (1) ⫽
the Q index of homogeneity (Hedges & Olkin, 1985). To estimate 0.54, p ⬎ .05. The mean effect size for active control was 0.27,
the proportion of the observed variance that reflects real differ- 95% CI [0.09, 0.45], Z ⫽ 3.03, p ⬍ .01, and for passive control
ences among studies, we calculated I2. was 0.37, 95% CI [0.19, 0.55], Z ⫽ 4.0, p ⬍ .01 The results
Thus, a total of 20 studies contributed data to this meta-analysis, corresponding to the second analysis showed that the heterogene-
comprising 474 trained older adults and 439 control participants. ity between groups was not significant, Q(1) ⫽ 0.004, p ⬎ .05. The
Full details of all the studies included in this meta-analysis are mean effect size for active control was 0.36, 95% CI [0.06, 0.66],
provided in the References section, marked with an asterisk. A Z ⫽ 2.40, p ⬍ .05, and for passive control 0.37, 95% CI [0,07,
major concern in meta-analytic studies is the existence of publi- 0.68], p ⬍ .05. The difference between active and passive control
cation bias. We used funnel plots to assess the relationship be-
tween sample size and effect size (Egger, Smith, Schneider, &
Minder, 1997). As shown below, publication bias does not seem to
affect the validity of the overall effect size obtained in the present
study.

Results

Effect Size Estimates: Mean Effects and

Test of Heterogeneity
An overall effect size was calculated incorporating all effect
sizes. The mean effect size was calculated incorporating all effects
sizes. The mean effect size was 0.37 (ET ⫽ 0.05) with a 95%
confidence interval of between 0.26 and 0.48. Effect sizes showed
no significant heterogeneity, Q (19) ⫽ 23.95, p ⬎ .05, I2 ⫽ Figure 2. Mean global effect size (d) and 95% CI corresponding to the 20
20.69%. The forest plot is shown in Figure 2. individual studies combining effect sizes within each study. On the left are
Although we obtained a nonsignificant p value for heterogeneity shown the articles identified by the first author’s name and publication
and did not therefore reject the null hypothesis, we are aware that year.
 712 TORIL, REALES, AND BALLESTEROS

was not statistically significant but they were significantly differ- applied the Egger test (Egger et al., 1997). This test is an un-
ent from zero. weighted regression that takes the precision of each study as the
Finally, the last moderator variable was age. The heterogeneity independent variable and effect size as the dependent variable. The
of effect sizes for both age groups was highly significant, Q (1) ⫽ Egger test was not significant for the intercept of the regression
4.50, p ⬍ .01, I2 ⫽ 77.77%. The mean effect size for the 60- to model, intercept ⫽ 0.74; t(8) ⫽ 1.03, p ⫽ 0.31. The t test for the
70-year-old group was 0.30, 95% CI [0.16, 0.44], Z ⫽ 4.27, p ⬍ null hypothesis of an intercept equal to zero showed a statistically
.01, and for the 71- to 80-year-old group 0.57, 95% CI [0.34, 0.79], nonsignificant result. We can therefore assume that publication
Z ⫽ 4.98, p ⬍ .01. Table 3 summarizes these results. bias is not a threat to the validity of the overall effect size. Second,
To assess the interaction between age and duration of training, we calculated the fail-safe Nfs index (Becker, 2005). The result of
we computed the correlation between these two variables. The this analysis suggests that in order to cancel the global mean effect
result showed that the association between these variables was not obtained in our meta-analysis (d ⫽ 0.37), it would be necessary to
statistically significant (r ⫽ ⫺0.12, p ⫽ .60). This suggests that have 207 nonpublished studies with null effects not included in the
grouping certain levels of training duration with certain levels of meta-analysis. These results also suggest that the effect obtained in
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

participants= age did not bias the results obtained in this meta- the present study is not affected by publication bias. Thus, com-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

analysis. bined results from these methods suggest that the likelihood of
Next, we present the results of the analysis of the combined publication bias is minimal.
outcomes for each cognitive function. The heterogeneity between
cognitive functions was significant, Q (5) ⫽ 13.20, p ⬍ .05, and Discussion
we therefore report the analysis using the random effect model for
outcomes. We give below the statistics for those cognitive func- As far as we know, this is the first meta-analysis of the effects
tions that showed a significant treatment effect. The effect size for of video game training on improving cognitive functioning in
RT, d ⫽ 0.63, 95% CI [0.42, 0.84], Z ⫽ 5.93, p ⬍ .01; memory, healthy older adults. The present meta-analytic study investigated
d ⫽ 0.39, 95% CI [0.01, 0.64], Z ⫽ 3.08, p ⬍ .05; attention, d ⫽ whether video game training enhances cognitive functions in older
0.37, 95% CI [0.17, 0.57], Z ⫽ 3.67, p ⬍ .01; and overall cognitive adults by reanalyzing individual studies published on this topic
function, d ⫽ 0.38, 95% CI [0.13, 0.62], Z ⫽ 3.07, p ⬍ .05, were since 1986. The overall meta-analysis unambiguously revealed
all significant. Only executive functions did not reach statistical that training older adults with video games improves cognition.
significance (p ⬎ .05). These results are displayed in Table 4. The main findings can be summarized as follows: (a) video game
The present results suggest that the effects of the interventions training in older adults produces positive effects on several cog-
depend on variables such as the age of the participants and the nitive functions that decline with aging; (b) several methodological
duration (number of sessions) of the intervention. Specifically, this and personal factors have moderator effects; (c) among the ana-
analysis suggests that short interventions (from 1 to 6 weeks) are lyzed variables, the age of the participants and the number of
more effective than long interventions (from 7 to 12 weeks). sessions in the training program were significant in modifying the
Furthermore, age has a significant effect, suggesting that the oldest effect size of the interventions. These moderator variables may
adults (71– 80 years) benefit more from video game training than explain, in part, the variability of the results reported so far in the
younger participants (60 –70 years). literature on this topic.
Overall, the results of this meta-analysis confirm our main
hypothesis that video game training improves cognitive function-
Evaluation of Publication Bias
ing in older adults. However, the present results do not confirm the
To evaluate possible publication bias, we performed a funnel specific hypotheses regarding age, duration of training, type of
plot (see Figure 3). The symmetry of the graph suggests the program, number, and type of video games. In fact, we predicted
absence of publication bias. The interpretation of a funnel plot is greater improvement with longer training interventions, but the
subjective, so we performed other quantitative tests of publication results showed that short training is a better option for this type of
bias (Begg & Mazumbar, 1994; Egger et al., 1997). First, we intervention with older adults. Moreover, cognitive improvements

Table 3
Results of the Significant Moderator Variables

Variable Level d (SE) Q I2 Z 95% CI

Age 60–70 0.30 (0.07) Q(1) ⫽ 4,50, p ⬍ .01 77.77 % 4.27, p ⬍ .01 [0.16, 0.44]
71–80 0.57 (0.11) 4.98, p ⬍ .01 [0.34, 0.79]
Training duration Short 0.49 (0.08) Q(1) ⫽ 3.73, p ⫽ .05 73.19 % 5.59, p ⬍ .01 [0.32, 0.67]
Long 0.26 (0.08) 3.03, p ⬍ .01 [0.09, 0.43]
Type of game Simple 0.42 (0.08) Q(1) ⫽ 0.55, p ⬎ .05 5.00, p ⬍ .01 [0.25, 0.58]
Complex 0.33 (0.07) 4.38, p ⬍ .01 [0.18, 0.48]
Number of games Few 0.39 (0.06) Q(1) ⫽ 0.37, p ⬎ .05 6.12, p ⬍ .01 [0.26, 0.51]
Many 0.30 (0.12) 2.56, p ⬍ .01 [0.07, 0.54]
Type of program Video games 0.40 (0.07) Q(1) ⫽ 0.27, p ⬎ .05 5.25, p ⬍ .01 [0.25, 0.55]
Brain training 0.34 (0.08) 4.04, p ⬍ .01 [0.17, 0.50]
Type of control group Active 0.27 (0.09) Q(1) ⫽ 0.54, p ⬎ .05 3.03, p ⬍ .01 [0.09, 0.45]
Passive 0.37 (0.09) 4.00, p ⬍ .01 [0.19, 0.55]
 VIDEO GAME TRAINING IN AGING: A META-ANALYSIS 713

Table 4 a greater effect size in the oldest adults (71– 80 years). Our results
Effect Sizes (D), Z, and CI Corresponding to Each Cognitive do not rule out the possibility that the performance of younger
Process (Outcomes) participants (60 –70 years) may be relatively high at the start of the
training program. Notwithstanding, it is worth stressing that al-
Outcomes d (SE) Z 95% CI though older people may benefit extensively from video game
Memory 0.39 (0.12) 3.08, p ⬍ .01 [0.01, 0.64] training, they use new technologies and video games in particular
Attention 0.37 (0.10) 3.67, p ⬍ .01 [0.17, 0.57] less than other members of the population, even when they can
Reaction time 0.63 (0.10) 5.93, p ⬍ .01 [0.42, 0.84] obtain a greater benefit.
Cognitive function 0.38 (0.12) 3.07, p ⬍ .01 [0.13, 0.62] The results of the current study show that there are no signifi-
Executive functions 0.16 (0.13) 1.20, p ⬎ .05 [⫺0.10, 0.42]
cant differences between few and many video games used in the
training phase, although there is a nonsignificant trend indicating
that few games are better than many. In relation to the social
due to video game training increased rather than decreased with interaction between experimenter and trainee that occurs in the
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

age; the oldest adults (71– 80 years) improved more after training experimental group and in the active control group but not in the
This document is copyrighted by the American Psychological Association or one of its allied publishers.

than younger participants (60 –70 years). We also predicted greater passive control group, the results suggest a minor, although non-
cognitive improvements with complex games than with simple significant beneficial effect on cognition. This effect might be due
games. The results, however, showed no significant differences to motivational factors. This result, however, must be taken with
between simple and complex games. Similar results emerged for caution due to the small number of studies (five studies) that
the number of video games used in the intervention. included passive and active control groups in the same study.
The current results have both theoretical and practical implica- Future investigation in this field should include at least these two
tions. First, duration of training was a significant variable. Our types of control groups (active and passive) as a better control. It
results indicate that the training effects are greater when training is should be noted that the means showed in Table 3 for the variable
of short duration (1– 6 weeks) than when it is long (7–12 weeks). type of control group correspond to standardized mean differences
This finding has practical implications as many intervention pro- of the experimental versus active control group and the experi-
grams spend a great amount of time training older participants on mental versus passive control group. This implies that if the active
the assumption that longer training will produce better results. The control group had an effect, the computed d for the first contrast
results of this meta-analysis do not support this assumption. Our will be lower than that of the second contrast. The reason is that the
own experimental work with senior citizens (Ballesteros et al., in performance of the active control group will be closer to the
press; Mayas, Andrés, Parmentier, & Ballesteros, 2014) suggests experimental group than to the passive control group. The present
that long training schedules lead to loss of motivation. Training results show that there are no significant differences between
sessions may be quite exciting at first but older adults get tired and “brain training” programs and video games. The same happened
bored in the last sessions. It seems that what motivates older with type of game. There are no significant differences between
participants to practice the games in the later sessions is not the simple and complex games, although there is a nonsignificant
training per se but the affective link or personal relationship trend indicating that simple games produce more benefits than
established with the experimenter. Moreover, despite evidence complex games.
suggesting a significant effect size of training on cognitive func- So far, we have considered the moderator variables included in
tioning, older people do not seem to perceive its functionality in the meta-analysis. An important issue, however, is whether the
their daily life. Apart from the motivational factors outlined above, effects of training older people with video games are transferable
the present results might be explained in terms of the Temporal to untrained tasks. This is the so-called “transfer effect.” The
Discounting Hypothesis. Temporal discounting (Green, Fry, & question is whether the effects of training with video games
Myerson, 1994) refers to the phenomenon that future rewards are transfer to cognitive processes such as memory, attention, execu-
less valuable than immediate rewards. This temporal discounting is
greater in elderly participants whose expected life-time is shorter.
This means that the effort put into learning a new task must be
balanced with the expected reward of acquiring the new skill.
When the time needed to learn and improve in the video game task
begins to be relatively long with respect to the expected reward,
the motivation to continue training decreases. The reason may be
that the anticipated rewards are small compared with the immedi-
ate cost of attending the training sessions.
The second significant moderator variable that appeared in our
analysis was the age of the trainees. The results suggest that the
benefits of training increases as participants get older. This result
is relevant for applied purposes and could be explained by the
larger training gains in people with lower baseline scores. In other
words, the oldest adults (71– 80 years) start the training program
with lower cognitive functioning scores, but they show greater Figure 3. Funnel plot of standard errors and effect sizes (g) of the 20
improvement after training than the younger participants (60 –70 studies of healthy older adults trained with video games included in the
years of age). The combination of these two effects may produce meta-analysis.
 714 TORIL, REALES, AND BALLESTEROS

tive functions, RT, and global cognitive function. Excluding ex- Future Research
ecutive functions that did not show a significant effect different
More research is needed to study effective ways of maintaining
from zero, effect sizes revealed an interesting pattern in the other
cognition and of improving the quality of life of older adults. The
cognitive functions. The results of this meta-analytic study suggest
findings regarding the moderating effects of the age of the trainees
that attention is perhaps the cognitive function that improved most
and the duration of the training program suggest that these factors
following training. This finding is in agreement with the results of require further investigation. However, the main issue for future
a recent study (Mayas et al., 2014). In this study, the trained group, research is how to improve the transfer effects of video games on
but not the control group, received 20 1-hr training sessions with cognitive functioning in older adults, especially executive func-
non-action video games and were evaluated before and after the tions. A fruitful new approach is the incorporation of neuroimag-
intervention using a cross-modal visual-auditory oddball task mea- ing data to identify the mental processes that operate in multiple
suring alertness and distraction. The results showed that training task domains (Anguera et al., 2013; Basak, Voss, Erickson, Boot,
reduced distractibility by improving attention filtering (a function & Kramer, 2011; Lustig, Shah, & Reuter-Lorenz, 2009; Prakash et
that declines with age and is largely dependent on frontal regions) al., 2012). These mental processes might be targeted with specific
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

and also improved alertness supported by a larger brain network. cognitive tasks directed at these processes and evaluating their
This document is copyrighted by the American Psychological Association or one of its allied publishers.

The present study also showed that speed of processing, as- performance on another task (Dahlin et al., 2008; Persson &
sessed by RT tasks, improved significantly after training. This Reuter-Lorenz, 2008).
finding is in line with a recent review (Bavelier, Green, Pouget, &
Schrater, 2012). Most video games demand rapid responses and Limitations
the maintenance in long-term memory of information relevant to
the task at hand ready to be used when required. This is reflected It is important to stress that publication bias is a potentially
in the significant effect of video game training on memory. The serious threat to the validity of the meta-analysis. Publication bias
concerns the issue of missing data. However, the statistical anal-
significant effect of training on global cognitive function also
ysis suggests that the results of this meta-analysis are reasonably
shows, although to a lesser extent, that some components of
good and that the likelihood of publication bias is minimal.
training transfer to general cognitive processes as assessed by
Finally, it should be noted other possible limitations, including
standardized psychometric intelligence tests.
the lack of documented effects on latent factors, which makes it
Interestingly, executive functions did not show a significant difficult to rule out effects due to specific strategies, the lack of
effect of training (although there was a trend in this direction). This reported effects on abilities related to daily life, and the difficulty
result is in agreement with the study of Dahlin, Nyberg, Bäckman, to rule out possible effects due to motivational factors rather than
and Neely (2008), who found that the transfer of computer training neurocognitive plasticity.
to the updating of information in working memory was limited to
young participants. They concluded that older participants have
Conclusion
more limited neural plasticity, reducing their ability to generalize
an executive skill (information updating). The results of the pres- In summary, although video game training has positive effects
ent meta-analysis are also congruent with those of a study from our for older adults, the benefits do not transfer to all cognitive
laboratory (Ballesteros et al., in press). A recent meta-analysis functions. Moreover, the positive effects are moderated by per-
conducted with children, young adults, adults, and older adults also sonal and methodological factors. For instance, the age of the
yielded negligible effects for executive functions (Powers, Brooks, trainee and the duration of the training program are important
Aldrich, Palladino, & Alfieri, 2013). factors that have to be considered in this type of intervention and
Overall, our results are in line with those of Kueider et al. (2012) should be the focus of future research. The results of this meta-
who also found that video game training improved RT and global analytic study suggest neurocognitive plasticity in the aging hu-
cognition but that it was less efficient for improving executive man brain, as training with video games is found to enhance
cognitive performance on several untrained functions.
functions. This meta-analysis extends the systematic review car-
References marked with an asterisk indicate studies included in
ried out by Kueider et al. (2012) by including a greater number of
the meta-analysis.
studies. In order to ascertain the consistency of our results, we
compared them with those obtained by Kueider and colleagues in
the meta-analytic review. The mean effect sizes of both studies are References
similar. For instance, our mean effect size for RT was d ⫽ 0.63 References marked with an asterisk indicates studies included in the
versus d ⫽ 0.77 reported by Kueider and colleagues. These results meta-analysis.
suggest that video game training improves memory and other
Achtman, R. L., Green, C. S., & Bavelier, D. (2008). Video games as a tool to
cognitive functions in healthy older adults and have important train visual skills. Restorative Neurology and Neuroscience, 26, 435– 446.
practical implications. Helping older adults to maintain cognitive ⴱ
Ackerman, P. L., Kanfer, R., & Calderwood, C. (2010). Use or lose it? Wii
health may reduce the risk of dependency. However, the video brain exercise practice and reading for domain knowledge. Psychology
game industry should create attractive and useful games specifi- and Aging, 25, 753–766. doi:10.1037/a0019277
cally designed for the elderly, in other words, video games that Allaire, J. C., McLaughlin, A. C., Trujillo, A., Whitlock, L., & Laporte, L.
meet their perceived needs and which may not be the same as those (2013). Successful aging through digital games: Socioemotional differ-
ences between older adult gamers and non gamers. Computers in Human
that young people find attractive and engaging.
Behaviour, 29, 1302–1306. doi:10.1016/j.chb.2013.01.014
 VIDEO GAME TRAINING IN AGING: A META-ANALYSIS 715
ⴱ
Anguera, J. A., Boccanfuso, J., Rintoul, J. L., Al-Hashimi, O., Faraji, F., control. Acta Psychologica, 129, 387–398. doi:10.1016/j.actpsy.2008.09
Janowich, J., . . . Gazzaley, A. (2013). Video game training enhances .005
cognitive control in older adults. Nature, 501, 97–101. doi:10.1038/ Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. (2005).
nature12486 Comprehensive meta-analysis (Version 2.0). [Computer software].
Ball, K., Berch, D. B., Helmer, K. F., Jobe, J. B., Leveck, M. D., Marsiske, Englewood, NJ: Biostat.
M., . . . Willis, S. L. (2002). Effects of cognitive training interventions Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. (2009).
with older adults: A randomized controlled trial. Journal of the Ameri- Introduction to meta-analysis. Chichester, UK: Wiley. doi:10.1002/
can Medical Association, 288, 2271–2281. doi:10.1001/jama.288.18 9780470743386
.2271 Boyke, J., Driemeyer, J., Gaser, C., Büchel, C., & May, A. (2008). Training
Ballesteros, S., Bischof, G. N., Goh, J. O., & Park, D. C. (2013). Neural induced brain structure changes in the elderly. The Journal of Neuro-
correlates of conceptual object priming in young and older adults: An science, 28, 7031–7035. doi:10.1523/JNEUROSCI.0742-08.2008
ⴱ
event related functional magnetic resonance imaging study. Neurobiol- Bozoki, A., Radovanovic, M., Winn, B., Heeter, C., & Anthony, J. (2013).
ogy of Aging, 34, 1254 –1264. doi:10.1016/j.neurobiolaging.2012.09.019 Effects of a computer based cognitive exercise program on age related
Ballesteros, S., Prieto, A., Mayas, J., Toril, P., Pita, C., Ponce de León, L., cognitive decline. Archives of Gerontology and Geriatrics, 57, 1–7.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Reales, J. M., & Waterworth, J. A. (in press). Training older adults with doi:10.1016/j.archger.2013.02.009
This document is copyrighted by the American Psychological Association or one of its allied publishers.

non-action video games enhances cognitive functions that decline with Buschkuehl, M., Jaeggi, S. M., Hutchison, S., Perrig-Chiello, P., Däpp, C.,
aging: A randomized controlled trial. Frontiers in Aging Neuroscience. Müller, M., . . . Perrig, W. J. (2008). Impact of working memory training
Ballesteros, S., & Reales, J. M. (2004). Intact haptic priming in normal on memory performance in old-old adults. Psychology and Aging, 23,
aging and Alzheimer’s disease. Evidence for dissociable memory sys- 743–753. doi:10.1037/a0014342
tems. Neuropsychologia, 42, 1063–1070. doi:10.1016/j.neuropsycholo- Carlson, K. D., & Schmidt, F. L. (1999). Impact of experimental design on
gia.2003.12.008 effect size: Findings from the research literature on training. Journal of
Ballesteros, S., Toril, P., Mayas, J., Reales, J. M., & Waterworth, J. A. Applied Psychology, 84, 851– 862. doi:10.1037/0021-9010.84.6.851
ⴱ
(2014). The role of a new ICT in supporting successful aging. Geron- Cassavaugh, N. D., & Kramer, A. F. (2009). Transfer of computer based
technology, 13, 39 – 48. training to simulated driving in older adults. Applied Ergonomics, 40,
Baltes, P. B., & Lindenberger, U. (1997). Emergence of a powerful 943–952. doi:10.1016/j.apergo.2009.02.001
ⴱ
connection between sensory and cognitive functions across the adult life Clark, J. E., Lamphear, A. K., & Riddick, C. C. (1987). The effects of
span: A new window to the study of cognitive aging? Psychology and videogame playing on the response selection processing of elderly
Aging, 12, 12–21. doi:10.1037/0882-7974.12.1.12 adults. Journal of Gerontology, 42, 82– 85. doi:10.1093/geronj/42.1.82
ⴱ
Basak, C., Boot, W. R., Voss, M. W., & Kramer, A. F. (2008). Can training Dahlin, E., Nyberg, L., Bäckman, L., & Neely, A. (2008). Plasticity of
in a real time strategy video game attenuate cognitive decline in older executive functions in young and older adults: Immediate training gains,
adults? Psychology and Aging, 23, 765–777. doi:10.1037/a0013494 transfer and long term maintenance. Psychology and Aging, 23, 720 –
Basak, C., Voss, M. W., Erickson, K., Boot, W. R., & Kramer, A. F. 730. doi:10.1037/a0014296
ⴱ
(2011). Regional differences in brain volume predict the acquisition of Drew, B., & Waters, J. (1986). Video games: Utilization of a novel
skill in a complex real-time strategy video game. Brain and Cognition, strategy to improve perceptual motors skills and cognitive functioning in
76, 407– 414. doi:10.1016/j.bandc.2011.03.017 the no institutionalized elderly. Cognitive Rehabilitation, 4, 26 –31.
ⴱ
Bavelier, D., Green, C. S., Pouget, A., & Schrater, P. (2012). Brain plasticity Dustman, R. E., Emmerson, R. Y., Steinhaus, L. A., Shearer, D. E., &
through the life span: Learning to learn and action video games. Annual Dustman, T. J. (1992). The effects of videogame playing on neuropsy-
Review of Neuroscience, 35, 391– 416. doi:10.1146/annurev-neuro-060909- chological performance of elderly individuals. Journal of Gerontology,
152832 47, 168 –171. doi:10.1093/geronj/47.3.P168
Becker, B. J. (2005). Failsafe N or file drawer number. In H. R. Rothstein, Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias in
A. J. Sutton, & M. Borenstein (Eds.), Publication bias in meta-analysis: meta-analysis detected by simple, graphical test. British Medical Jour-
Prevention, assessment and adjustments (pp. 111–125). Chichester, UK: nal, 315, 629 – 634. doi:10.1136/bmj.315.7109.629
ⴱ
Wiley. Goldstein, J., Cajko, L., Oosterbroek, M., Michielsen, M., Van Houten,
Begg, C. B., & Mazumbar, M. (1994). Operating characteristics of a rank O., & Saavedra, F. (1997). Videogame and elderly. Social Behavior and
correlation test for publication bias. Biometrics, 50, 1088 –1101. doi: Personality, 25, 345–352. doi:10.2224/sbp.1997.25.4.345
10.2307/2533446 Gopher, D., Weil, M., & Bareket, T. (1994). Transfer of skill from a
ⴱ
Belchior, P. D. C. (2008). Cognitive training with video games to computer game trainer to flight. Human Factors, 36, 387– 405.
improve driving skills and driving safety among older adults [disser- Green, C. S., & Bavelier, D. (2008). Exercising your brain: A review of
tation]. Dissertation Abstract International: The Sciences and Engineer- human brain plasticity and training induced learning. Psychology and
ing, 68(9 –B). Aging, 23, 692–701. doi:10.1037/a0014345
Bialystok, E., & Craik, F. I. M. (2006). Lifespan cognition: Mechanisms of Green, L., Fry, A. F., & Myerson, J. (1994). Temporal discounting and
change. Oxford, UK: Oxford University Press. doi:10.1093/acprof:oso/ preference reversals in choice between delayed outcome. Psychonomic
9780195169539.001.0001 Bulletin and Review, 1, 383–389. doi:10.3758/BF03213979
Bond, G. E., Wolf-Wilets, V., Fiedler, F. E., & Burr, R. L. (2001). Hedden, T., & Gabrieli, J. D. E. (2004). Insights into the ageing mind: A
Computer aided cognitive training of the aged: A pilot study. Clinical view from cognitive neuroscience. Nature Reviews Neuroscience, 5,
Gerontologist, 22, 19 – 42. doi:10.1300/J018v22n02_03 87–96. doi:10.1038/nrn1323
Boot, W. R., Blakely, D. P., & Simmons, D. (2011). Do action video games Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis.
improve perception and cognition? Frontiers in Psychology, 2, 1– 6. Orlando, FL: Academic Press.
ⴱ
Boot, W. R., Champion, M., Blakely, D. P., Wright, T., Souders, D. J., & Hertzog, C., Kramer, A. F., Wilson, R., & Linderbergen, U. (2008).
Charness, N. (2013). Video games as a means to reduce age-related Enrichment effects on adult cognitive development. Can the functional
cognitive decline: Attitudes, compliance, and effectiveness. Frontiers in capacity of older adults be preserved and enhanced? Psychological
Psychology, 4, 31. doi:10.3389/fpsyg.2013.00031 Science in the Public Interest, 9, 1– 65.
Boot, W. R., Kramer, A., Simons, D. J., Fabiani, M., & Gratton, G. (2008). Hoyer, W. J., & Verhaeghen, P. (2006). Memory aging. In J. Birren & W.
The effects of video game playing on attention, memory and executive Schaie (Eds.), Handbook of the psychology of aging (pp. 209 –232).
 716 TORIL, REALES, AND BALLESTEROS

Burlington, MA: Elsevier Academic Press. doi:10.1016/B978- A randomized double-blind prospective trial cognitive stimulation. Neu-
012101264-9/50013-6 roepidemiology, 36, 91–99. doi:10.1159/000323950
Kueider, A. M., Parisi, J., Gross, A. L., & Rebok, G. W. (2012). Comput- Persson, J., & Reuter-Lorenz, P. A. (2008). Gaining control training
erized cognitive training with older adults: A systematic review. PloS executive function and far transfer of the ability to resolve interference.
ONE, 7, e40588. Psychological Science, 19, 881– 888. doi:10.1111/j.1467-9280.2008
Leung, L., & Lee, P. S. N. (2005). Multiple determinants of life quality: .02172.x
The role of internet activities, use of new media, social support and Peter, C., Kreisner, A., Schröter, M., Hyosun, K., Bieber, G., Öhberg, F.,
leisure activities. Telematics and Informatics, 22, 161–180. doi:10.1016/ Hoshi, K., . . . Ballesteros, S. (2013). AGNES: Connecting people in a
j.tele.2004.04.003 multimodal way. Journal on Multimodal User Interactions, 7, 229 –245.
Li, S.-C., Brehmer, Y., Shing, Y. L., Werkle-Bergner, M., & Lindenberger, U. doi:10.1007/s12193-013– 0118-z
(2006). Neuromodulation of associative and organizational plasticity across Powers, K. L., Brooks, P. J., Aldrich, N. J., Palladino, M., & Alfieri, L.
the life span: Empirical evidence and neurocomputational modeling. Neu- (2013). Effects of video-game play on information processing: Meta-
roscience & Biobehavioral Reviews, 30, 775–790. doi:10.1016/j.neubiorev analytic investigation. Psychonomic Bulletin & Review, 20, 1055–1079.
.2006.06.004 Prakash, R. S., De León, A., Mourany, L., Lee, H., Voss, M. W., Boot,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Li, S.-C., Schmiedek, F., Huxhold, O., Röcke, C., Smith, J., & Lindenberger, W. R., . . . Kramer, A. F. (2012). Examining neural correlates of skill
acquisition in a complex videogame training program. Frontiers in
This document is copyrighted by the American Psychological Association or one of its allied publishers.

U. (2008). Working memory plasticity in old age: Practice gain, transfer,

and maintenance. Psychology and Aging, 23, 731–742. doi:10.1037/ Human Neuroscience, 6, 115, 1–11.
a0014343 Raz, N., Lindenberger, U., Rodrigue, K. M., Kennedy, K. M., Head, D.,
Lustig, C., Shah, P., & Reuter-Lorenz, P. (2009). Aging, training and the Williamson, A., . . . Acker, J. D. (2005). Regional brain changes in aging
brain: A review and future directions. Neuropsychology Review, 19, healthy adults: General trends, individual differences and modifiers.
504 –522. doi:10.1007/s11065-009-9119-9 Cerebral Cortex, 15, 1676 –1689. doi:10.1093/cercor/bhi044
ⴱ
Maillot, P., Perrot, A., & Hartley, A. (2012). Effects of interactive physical Rönnlund, M., & Nilsson, L.-G. (2008). Cross-sectional versus longitudi-
nal age gradients of Tower of Hanoi Performance: The role of practice
activity videogame training on physical and cognitive function in older
effects and cohort differences in education. Aging, Neuropsychology and
adults. Psychology and Aging, 27, 589 – 600. doi:10.1037/a0026268
Cognition, 15, 40 – 67.
Mayas, J., Andrés, P., Parmentier, F., & Ballesteros, S. (2014). Plasticity of
Rönnlund, M., Nyberg, L., Bäckman, L., & Nilsson, L.-G. (2005). Stabil-
attentional functions in older adults after non-action video game train-
ity, growth and decline in adult life span development of declarative
ing: A randomized controlled trial. PloS ONE, 9, e92269. doi:10.1371/
memory: Cross-sectional and longitudinal data from a population-based
journal.pone.0092269
ⴱ study. Psychology and Aging, 20, 3–18. doi:10.1037/0882-7974.20.1.3
McDougall, S., & House, B. (2012). Brain training in older adults: Evidence
Salthouse, T. A. (1996). The processing speed theory of adult age differ-
of transfer to memory span performance and pseudo-Matthew effects.
ences in cognition. Psychological Review, 103, 403– 428. doi:10.1037/
Aging, Neuropsychology, and Cognition, 19, 195–221. doi:10.1080/
0033-295X.103.3.403
13825585.2011.640656 ⴱ
Sosa, G. (2011). The impact of video game intervention in the cognitive
Meijer, W. A., Van Boxtel, M. P., Van Gerven, P. W., Van Hooren, S. A.,
functioning, self- efficacy, self-esteem and video games attitude of older
& Jolles, J. (2009). Interaction effects of education and health status on
adults. A final project. CUG Theses and Dissertations.
cognitive change: A 6 years follow up of the Maastricht Aging Study. ⴱ
Stern, Y., Blumen, H., Rich, L., Richards, A., Herzberg, G., & Gopher, D.
Aging & Mental Health, 13, 521–529. doi:10.1080/13607860902860821 (2011). Space Fortress game training and executive control in older
Mitchell, D. B., & Bruss, P. J. (2003). Age differences in implicit memory: adults: A pilot intervention. Aging, Neuropsychology and Cognition, 18,
Conceptual, perceptual o methodological. Psychology and Aging, 18, 653– 677. doi:10.1080/13825585.2011.613450
807– 822. doi:10.1037/0882-7974.18.4.807 ⴱ
Torres, A. (2008). Cognitive effects of video games on older people.
Nilsson, L.-G. (2003). Memory function in normal aging. Acta Neuro- ICDVRAT, 19, 191–198.
logica Scandinavica, 179, 7–13. doi:10.1034/j.1600-0404.107.s179.5.x United Nations, Department of Economic and Social Affairs. (2010).
ⴱ
Nouchi, R., Taki, Y., Takeuchi, H., Hashizume, H., Akitsuki, Y., Shige- World Population prospects: The 2010 revision, volume II: Demo-
mune, Y., . . . Kawashima, R. (2012). Brain training game improves graphic profiles. ST/ESA/SER. A/317.
executive functions and processing speed in the elderly: A randomized ⴱ
Van Muijden, J., Band, G., & Hommel, B. (2012). Online games training
controlled trial. PloS ONE, 7, e29676. doi:10.1371/journal.pone aging brains: Limited transfer to cognitive control functions. Frontiers in
.0029676 Human Neuroscience, 6, 1–13.
Owen, A. M., Hampshire, A., Grahn, J. A., Stenton, R., Dajani, S., Burns, Wiggs, C. L., Weisberg, J., & Martin, A. (2006). Repetition priming across
A. S., . . . Ballard, C. G. (2010). Putting brain training to test. Nature, the adult lifespan—The long and short of it. Aging, Neuropsychology,
465, 775–778. doi:10.1038/nature09042 and Cognition, 13, 308 –325. doi:10.1080/138255890968718
Park, D. C., & Bischof, G. (2013). The aging mind: Neuroplasticity in Willis, S. L., Tennstedt, S. L., Marsiske, M., Ball, K., Elias, J., Koepke,
response to cognitive training. Dialogues in Clinical Neuroscience, 15, K. M., . . . Wright, E. (2006). Long-term effects of cognitive training on
109 –119. everyday functional outcomes in older adults. Journal of the American
Park, D. C., & Reuter-Lorenz, P. (2009). The adaptive brain: Aging and Medical Association, 296, 2805–2814. doi:10.1001/jama.296.23.2805
neurocognitive scaffolding. Annual Review of Psychology, 60, 173–196. Zelinski, E. M., & Reyes, R. (2009). Cognitive benefits of computer games
doi:10.1146/annurev.psych.59.103006.093656 for older adults. Gerontechnology, 8, 220 –235. doi:10.4017/gt.2009.08
Pascual-Leone, A., Amedi, A., Fregni, F., & Merabet, L. B. (2005). The .04.004.00
plastic human brain cortex. Annual Review of Neuroscience, 28, 377–
401. doi:10.1146/annurev.neuro.27.070203.144216 Received October 31, 2013
ⴱ Revision received June 20, 2014
Peretz, C., Korczyn, A., Shatil, E., & Aharonson, V. (2011). Computer
based personalized cognitive training versus classical computer games: Accepted June 23, 2014 䡲