The effects of video game playing on attention, memory, and executive control
Walter R. Boot*, Arthur F. Kramer, Daniel J. Simons, Monica Fabiani, Gabriele Gratton
Beckman Institute, Department of Psychology, University of Illinois, Urbana-Champaign, Urbana, IL 61801, USA
a r t i c l e i n f o
Received 6 March 2008
Received in revised form 7 September 2008
Accepted 8 September 2008
Available online 16 October 2008
a b s t r a c t
Expert video game players often outperform non-players on measures of basic attention and perfor-
mance. Such differences might result from exposure to video games or they might reflect other group dif-
ferences between those people who do or do not play video games. Recent research has suggested a
causal relationship between playing action video games and improvements in a variety of visual and
attentional skills (e.g., [Green, C. S., & Bavelier, D. (2003). Action video game modifies visual selective
attention. Nature, 423, 534–537]). The current research sought to replicate and extend these results by
examining both expert/non-gamer differences and the effects of video game playing on tasks tapping a
wider range of cognitive abilities, including attention, memory, and executive control. Non-gamers
played 20+ h of an action video game, a puzzle game, or a real-time strategy game. Expert gamers and
non-gamers differed on a number of basic cognitive skills: experts could track objects moving at greater
speeds, better detected changes to objects stored in visual short-term memory, switched more quickly
from one task to another, and mentally rotated objects more efficiently. Strikingly, extensive video game
practice did not substantially enhance performance for non-gamers on most cognitive tasks, although
they did improve somewhat in mental rotation performance. Our results suggest that at least some dif-
ferences between video game experts and non-gamers in basic cognitive performance result either from
far more extensive video game experience or from pre-existing group differences in abilities that result in
a self-selection effect.
? 2008 Elsevier B.V. All rights reserved.
Recent research suggests that playing video games, even for a
relatively short period of time, improves performance on a number
of tasks that measure visual and attentional abilities. In fact, a
number of studies have found that having participants play action
video games for as few as 10 h can improve performance on labo-
ratory tasks that, on the surface, are dissimilar to the games they
were asked to play (e.g., Feng, Spence, & Pratt, 2007; Green & Bave-
lier, 2003, 2006a, 2006b, 2007). Thus, video game experience ap-
pears to improve basic skills that can be applied to novel tasks
The recent surge of interest in video games as a means to im-
prove basic perceptual and cognitive abilities builds on earlier vi-
deo game findings. For example, playing video games such as
Donkey Kong and Pac Man was found to significantly improve
the reaction times of older adults as compared to controls who
did not play (Clark, Lanphear, & Riddick, 1987). In 1989, a special
issue of Acta Psychologica was devoted to Space Fortress, a video
game specifically designed by cognitive psychologists as a training
and research tool (Donchin et al., 1989). The skills acquired while
playing Space Fortress appeared to transfer to other tasks as well.
For example, young adults who played Space Fortress performed
better than controls on a test of physics knowledge (Frederiksen
& White, 1989), and Israeli Air Force flight school cadets who
played Space Fortress significantly outperformed a no-game con-
trol group on actual flight performance, suggesting that skills
learned from the game transferred to flight control (Gopher, Weil,
& Bareket, 1994). In Space Fortress, players must focus attention to
multiple demanding and overlapping component tasks, so im-
proved flight performance might result from improved attentional
control. Space Fortress training was considered so successful that it
was subsequently added to the training program of the Israeli Air
Force. Similarly, helicopter pilots trained on Space Fortress outper-
formed pilots trained on a helicopter flight simulation game (Hart
& Battiste, 1992).
More recently, Green and Bavelier (2003), Green and Bavelier
(2006a, Green and Bavelier (2006b, Green and Bavelier (2007) pro-
vided evidence that video game playing can improve performance
on a number of attentional and perceptual tasks. Both action video
game players and non-video game players who were given 10 h of
action video game experience (a first-person shooter called Medal
of Honor, Electronic Arts) demonstrated superior performance in
0001-6918/$ - see front matter ? 2008 Elsevier B.V. All rights reserved.
* Corresponding author. Current address: Florida State University, Department of
Psychology, 1107 W. Call Street, Tallahassee, FL 32306-4301, USA. Tel.: +1 850 645
8734; fax: +1 850 644 7739.
E-mail address: email@example.com (W.R. Boot).
Acta Psychologica 129 (2008) 387–398
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/actpsy
the attentional blink task, a measure of attentional flexibility over
time (Green & Bavelier, 2003). Similarly, performance in the Func-
tional Field of View (FFOV) task, a measure of the breadth of visual
attention, was also improved by the video game experience. Habit-
ual action video game players also demonstrated an increased flan-
ker effect, indicating greater visual resources, and participants who
played an action game for only 10 h demonstrated a greater subi-
tizing capacity as well. Importantly, participants who played Tetris,
a puzzle game with similar motor components compared to Medal
of Honor but fewer attentional demands, did not improve on these
tasks. Action video game experience also resulted in the ability to
monitor a greater number of moving objects simultaneously
(Green & Bavelier, 2006a). Most surprisingly, video game players
and participants who practiced video games demonstrated a
change in visual acuity, or the ability to make fine discrimination
judgments of visually crowded stimuli (Green & Bavelier, 2007).
Additional evidence for superior performance of video game ex-
perts comes from the work of Castel, Pratt, and Drummond
(2005) that shows that in a number of tasks, video game players
demonstrate faster response times than non-gamers.
When considered together, these studies suggest that playing
video games promise improved performance in a wide variety of
situations; the transfer of video game experience appears to be
broad. The transfer tasks described above (helicopter and jet pilot-
ing, the FFOV, the attentional blink task, etc.) were different from
the games participants played both in the displays and the re-
quired responses, but transfer still occurred. Moreover, improve-
ments in skills occurred quickly, in some instances with as little
as 10 h of game experience.
The purpose of this study was to determine whether video
game benefits are restricted to visual and attentional tasks, or
whether improvements might be broader. In addition to measures
of visual attention, we assessed the effect of video game playing on
a number of memory, reasoning, and executive control tasks. We
also examined the effect of game type. Participants either played
a fast-paced action game, a slower-paced strategy game, or a puz-
zle game. It is easy to imagine how a highly complex strategy game
might better improve executive control, planning, and memory as
compared to a first-person shooter or puzzle game. Successful
game play would seem to require such skills (e.g., memory for
where enemies and resources are located, switching between sev-
eral tasks as the complex demands of the game changes, remem-
bering the complex sequence of events required to attain
multiple simultaneous goals). Thus, we expected game-specific ef-
fects based on the nature of the video game participants were
asked to play. Broadly, we expected the action game to improve vi-
sual/attentional skills, the strategy game to improve executive con-
trol skills, and the puzzle game to improve some spatial skills. Both
longitudinal and cross-sectional approaches were taken. The longi-
tudinal approach had participants complete more than 20 h of vi-
deo game practice, and perceptual and cognitive abilities were
assessed before practice, after ten h of practice, and then after
21 h of practice. Additionally, the cross-sectional approach com-
pared the abilities of expert gamers and non-video game players.
To preview the results, although a number of non-gamer/expert
differences were found, these differences, for the most part, did
not emerge even after 20+ h of video game experience.
2.1.1. Cross-sectional groups (expert vs. non-gamers)
Eleven expert video game players and ten non-video game play-
ers were recruited from the Urbana-Champaign community. Par-
ticipants were considered experts if they played seven or more
hours of video games per week for the past two years. Experts were
selected such that they had high levels of expertise with action vi-
deo games such as Halo, Grand Theft Auto and Unreal Tournament.
However, expert video game players also had experience with a
wide variety of game genres including role-playing, strategy, and
sports games. Non-gamers were selected such that they played vi-
deo games one hour a week or less. During initial participant
recruitment, it was found that males were much more likely to re-
port having video game expertise, and thus cross-sectional groups
were restricted to male participants.
2.1.2. Longitudinal groups (game practice and passive control)
Eighty-two college students and members of the Urbana-Cham-
paign community participated in the longitudinal portion of the
study. To maximize the likelihood of observing improvements, all
participants in the longitudinal groups were non-gamers and re-
ported playing less than one hour of video games a week over
the past 2 years. We included as many participants as possible
who reported playing 0 h/week. Females were much more likely
to report being non-gamers, and thus longitudinal groups were pri-
marily female (potential gender implications are discussed later).
All participants were right-handed and demonstrated normal
visual acuity and normal color vision. All reported no major med-
ical or psychological conditions. Demographics for each group are
listed in Table 1.
Potential participants were contacted either through flyers
posted in campus buildings and businesses or through advertise-
ments posted to online bulletin boards. People responding to these
flyers and advertisements completed a survey of their video game
habits. Sixty-three participants in the longitudinal portion of the
study were randomly assigned to one of three video game practice
conditions: (a) Medal of Honor, an action game (n = 20), (b) Rise of
Nations, a strategy game (n = 23), or (c) Tetris, a puzzle game
(n = 20). An additional nineteen participants were assigned to a
no-practice control group.
Four Pentium 4-based PCs were used for the majority of cogni-
tive testing and game playing. These computers were connected to
21-inch monitors. Additionally, one eMAC with a 17-inch monitor
was used for two of the cognitive tasks. For all testing and game
playing sessions, seating was adjusted so that participants were
approximately 57 cm from the monitor. All PC-based tasks were
programmed with the E-prime software (Psychology Software
Tools, www.pstnet.com). All Mac tasks were programmed with
the Vision Shell software (http://www.visionshell.com/).
Demographic information for participants in each group, standard deviations are in
Mean age Years of EdMale/female
Differential shading identifies groups with the same comparison. MOH, Medal of
Honor; TET, Tetris; RON, Rise of Nations.
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
2.3. The games
2.3.1. Medal of honor
Allied Assault (Electronic Arts, 2002) is a World War II-based
first-person shooter game. Players are presented with a first-per-
son, egocentric view of a complex virtual environment. Like most
first-person shooters, Medal of Honor (MOH) primarily focuses
on combat. Players complete a number of missions in which they
must kill enemies and avoid being killed. MOH is a fast-paced
game that is perceptually demanding. The game requires players
to successfully localize enemies (enemies can occur almost any-
where on screen, including in the far distance and in the periphery)
and to deal with multiple enemies simultaneously or within close
temporal proximity. Previous research (Green & Bavelier, 2003)
found MOH to improve a number of visual and attentional abilities,
so we expected MOH experience to improve performance in tasks
that might involve these skills. MOH experience might also im-
prove visuo-spatial working memory (remembering location and
identity of objects in the environment) and performance in tests
of executive function (switching between various goals such as
killing enemies, navigating, and locating supplies).
Tetris (ValuSoft, 2004) is a popular puzzle game in which play-
ers rotate and move blocks descending from the top of the screen
so that these blocks form lines at the bottom of the screen. After a
complete line with no gaps is formed, the line disappears. If no
lines are formed, the blocks pile higher and higher until the block
pile reaches the top of the screen, at which point the game ends
and the player loses. The goal is to keep the game going as long
as possible by forming complete lines. As the game progresses,
the blocks descend faster, giving players less time to choose where
to place each block. Like MOH, Tetris is a fast-paced game. How-
ever, Tetris does not involve the same attentional demands as
MOH. In Tetris, players are concerned with only one moving object
at a time, whereas in MOH players attend to several enemies
simultaneously. We used a version of Tetris that minimized plan-
ning by eliminating any preview of upcoming blocks (see Green
& Bavelier, 2003). Previous research found Tetris improved some
spatial skills (De Lisi & Wolford, 2002; Sims & Mayer, 2002), but Te-
tris experience did not appear to transfer beyond those specific
tasks as much as MOH did.
Rise of Nations (Microsoft, 2003) is a highly complex real-time
strategy game in which players develop and defend a civilization.
In Rise of Nations (RON), players initially manage a single city of
a primitive society. They collect resources to advance their civiliza-
tion and build new structures, eventually building more cities and
exploring the map. The resources (e.g., wood, stone, oil, food, and
money) can be invested in a number of ways such as building a lar-
ger and stronger army or investing in technology. While players
are building their civilization one or multiple computer players
are building their own civilizations and these other civilizations
occasionally attack and try to take over the player’s cities. The
game is won by either taking control of 70% of the map or com-
pletely conquering all other civilizations. This game was chosen
over several other similar strategy games since RON offered the
most complexity and the most opportunity for participants to en-
gage in strategic behavior. Multi-tasking is an important compo-
nent of game play since players must manage several cities,
collect needed resources, defend their cities, and also attack enemy
cities, all at the same time. Players must also switch among these
tasks in response to their current situation (i.e., participants might
need to switch from building a city to defending a city if a city
comes under attack). Unlike MOH, game play tends to be slower
and focuses more on planning and resource management rather
than quick actions. Consequently, we predicted that RON would
improve performance on tasks requiring executive control. Work-
ing memory also plays an important role in the game as partici-
pants need to keep complex goals and information in mind as
they build their civilization. Spatial working memory may be espe-
cially important since participants must remember spatial loca-
tions on a large map of cities, buildings, and resources.
2.4. Practice and testing schedule
Participants in the longitudinal practice groups completed fif-
teen game sessions in the laboratory over a period of four to five
weeks. The duration of thirteen of these game sessions were
1.5 h, but the duration of the first and last session was only 1 h
(the remaining .5 h of those sessions were devoted to completing
a portion of the cognitive battery described later. Participants as-
signed to the MOH and RON groups started game practice by com-
pleting a game tutorial. Given the relative simplicity of Tetris,
participants were given a brief explanation of the game but did
not complete a tutorial. This schedule resulted in a total practice
time of 21.5 h for each participant in each of the longitudinal game
groups. Given the number of practice hours required and the num-
ber of participants, all participants did not start in the study simul-
taneously, but instead participants were run in waves of 15 to 20
participants. Each wave did not overlap with the previous wave.
Thus, all data were collected in the span of approximately six
Participants in the MOH group played straight through the
game. Although difficulty was not explicitly adjusted, game play
becomes more difficult as the game progresses. At the end of each
session, game progress was saved and participants began the next
session at this point. A variety of game scenarios were created for
the RON group to play. These scenarios increased in difficulty, with
participants first competing against one computer enemy, and la-
ter switching to two enemies after the first few games. These sce-
narios varied the nationality of the player and computer players,
and also the geography of the map (each nationality within the
game has particular strengths and weaknesses). If a participant
did not finish a game within the session, the game was saved
and participants started from this point at the start of the next ses-
sion. Participants in the Tetris group played each game until com-
pletion, after which a new game was started. Participants started a
new game at the beginning of each session.
Game performance was recorded for each participant. At the
end of each mission, MOH lists the number of enemies killed, the
number of hits taken, firing accuracy, and a number of other statis-
tics. At the end of each RON scenario, scores relating to military
and civic accomplishments are listed. Tetris reports two scores (to-
tal score and number of lines) at the end of each game. If a Tetris
game was not completed due the session ending, the score until
that point was recorded and it was noted that this game was
incomplete. On the final practice session, participants in the
MOH and RON groups repeated the first scenario/mission com-
pleted at the beginning of the study to measure the degree of
improvement in game performance.
Participants in the control group played no video games, but
were tested on all cognitive tests three times. The time between
each testing session matched that of participants who were in
the MOH, Tetris, or RON game groups.
2.5. Expert and non-gamer schedule
Expert and non-gamer participants received no game practice
and were not asked to play video games in the laboratory. These
participants received only the cognitive tasks described below
once (whereas the longitudinal groups received the battery three
times: once before, once half-way through game practice, and fi-
nally once after all game practice was completed). Expert and
non-gamer participants were not tested on the Tower of London
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
or the Ravens tests described below; otherwise they completed the
same tasks, in the same order, as the longitudinal groups.
2.6. The cognitive battery
Longitudinal participants (including the control group) com-
pleted a battery of cognitive tests three times. The majority of each
cognitive battery could be administered during two 1.5 to 2 h
assessment sessions. Additionally, the first half hour of the first
and last game practice sessions was also used for cognitive testing
for participants in the game groups. The battery included a number
of tasks that were completed in a fixed order, with each task taking
8–30 min to complete. Tasks fell into three general categories: (1)
visual and attentional tasks, (2) spatial processing and spatial
memory tasks, and (3) executive control tasks. Three of the vi-
sual/attentional tasks included in the battery (functional field of
view, attentional blink, and enumeration) were those in which
Green and Bavelier (2003) found to improve after ten h of MOH
experience. All tasks are described below.
2.6.1. Visual and attentional tasks
22.214.171.124. Functional field of view. Participants searched for a white
triangle within a circle (4.3? diameter) among square distractors
(4.3? ? 4.3?) in a briefly presented (12 ms) display (see Green &
Bavelier, 2003). Search items were arrayed in eight radial arms,
and targets occurred with equal probability on each arm at eccen-
tricities of 10, 20, or 30? from fixation. The search display was fol-
lowed by a bright, colorful mask (100 ms). After this mask, a
response screen containing lines representing the radial arms of
the search display appeared and participants click the arm that
had the target. After one block of practice trials (24 trials), partic-
ipants completed 120 test trials.
126.96.36.199. Attentional blink (Raymond, Shapiro, & Arnell, 1992). Partic-
ipants viewed a rapid sequence of letters (approximately 1? high)
on a gray background at the center of the screen and reported
two things about each letter sequence: (1) the identity of the one
white letter in the sequence of black letters and (2) whether an
X was present sometime after the white letter (50% of trials). Each
letter appeared for 12 ms, followed by an 84 ms blank interval be-
fore the next letter. The sequence varied in length from 16 to 22
letters, with the white letter appearing unpredictably after either
the 7th, 10th, or 13th letter. In this task, participants often fail to
report the X when it appears approximately three items after the
first target. Green and Bavelier (2003) found improved detection
of the X following videogame experience, with the largest advan-
tage occurring when the X occurred 3, 4, 5, or 6 letters after the
first target, so we used these lags between the white letter and
the X (with an equal number of trials for each). Participants com-
pleted 15 practice trials and 144 test trials.
188.8.131.52. Enumeration. Participants viewed briefly presented arrays
of 1–8 dots (diameter .25?) and indicated how many dots appeared.
Dots were randomly positioned in a 7 ? 7 matrix (7? ? 7?) with the
exception that no dot could appear at the center location. Each trial
started with a small fixation point at the center of the screen
(900 ms), followed by a blank screen (600 ms), and then by the test
array (50 ms). Participants entered the number of dots that ap-
peared using the number keys at the top of the keyboard, after
which the next trial began. Participants completed 32 practice tri-
als followed by 160 test trials.
184.108.40.206. Multiple object tracking (Pylyshyn & Storm, 1988). Partici-
cles (targets). After a button press, the red items turned green and
were identical to the distractors. Participants then pressed the right
arrow key and the circles started moving. Additional presses of the
right arrow increased the speed of the objects, and pressing the left
arrow slowed the objects. Participants tried to find the speed at
which the circles moved as fast as possible while they could still
& Franconeri, 2005). If participants lost track of one or more target
circles they could show the targets (i.e., make the targets red again),
slow the items down, and then hide them again. When participants
found the correct speed they pressed the space bar. Participants re-
als were averaged. Participants were then tested on their ability to
keep track of three out of ten circles moving at this average speed.
During test, participants saw three red circles and seven green cir-
cles. Then the red circles turned green and all circles moved at the
speed participants set. After 8 s, one circle in the display turned
red, and participants were asked whether the red circle was a target
(one of the initially red circles) or a distractor circle.
220.127.116.11. Visual short-term memory (Luck & Vogel, 1997). Participants
viewed displays containing colored lines (red, green, blue, pink,
and black) at different orientations (vertical, horizontal, tilted to
the left, or tilted to the right). Each line measured .2? ? 1.6? and
had a center-to-center distance of at least 3.5?. Participants first
viewed a display containing 2, 4, or 6 lines for 100 ms. This mem-
ory display was followed by a blank screen for 900 ms, and then a
test display. On half of all trials, one item in the test display either
changed color or orientation compared to the memory display, and
participants indicated whether anything changed. Accuracy was
emphasized over speed. Participants completed 24 practice trials
and 144 test trials.
2.6.2. Spatial processing and spatial memory
18.104.22.168. Spatial 2-back. Participants viewed displays in which letters
appeared one at a time at different spatial locations and they
pressed one key if the letter was in the same location as the letter
presented two items previously and a different key if it was in a
different location (e.g., Braver et al., 1997). Letters measured
roughly 1.75? and they appeared at one of ten equally spaced loca-
tions around an imaginary circle with a diameter of 15.5? (the cen-
ter to center distance of adjacent locations was 5.0?). Each letter
appeared for 500 ms with an inter-stimulus interval of 2000 ms.
On 75% of trials the letter location was different from the location
of the item presented 2-items back, and on 25% of trials the loca-
tion was the same. Both speed and accuracy were stressed, and
participants completed 100 trials.
22.214.171.124. Corsi block-tapping task (Corsi, 1972). Participants viewed
displays of nine gray squares (2.3? ? 2.3?) in an irregular pattern
on the computer screen. One at a time these items could change
from gray to white, then back to gray. At the end of each trial they
tried to click the boxes in the same order that they had changed.
Participants completed four trials of sequence length 3, then four
of length 4, etc. until they had completed four trials with length
9. We emphasized accuracy over speed.
126.96.36.199. Mental rotation (Cooper & Shepard, 1973). Participants tried
to determine whether two simultaneously presented shapes were
the same or different. They responded as quickly and as accurately
as possible by pressing one of two keys. The shape on the right was
either the same shape or a mirror image of the shape on the left,
and the two shapes differed in orientation by 0, 45, 90, 135, 180,
225, 270, or 315?. These shapes were based on those appearing
in Tetris. Each measured approximately 2.4 ? 2.4? and was pre-
sented 3 d? from the center of the screen. Two shapes (the Z and
the backwards Z shape) only appeared at orientations of 0, 45,
90, and 135 since any further rotation would cause the shape to ro-
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
tate into itself. All other shapes could appear equally often at each
rotation from 0 to 315?. Participants completed 30 practice trials
followed by 128 test trials.
2.6.3. Executive control and reasoning
188.8.131.52. Task switching (Pashler, 2000). Participants completed a
task that required them to switch between judging whether a
number (1, 2, 3, 4, 6, 7, 8, or 9) was odd or even and judging
whether it was low or high (i.e., smaller or larger than 5). Numbers
were presented individually for 1500 ms against a pink or blue
background at the center of the screen, with the constraint that
the same number did not appear twice in succession. If the back-
ground was blue, participants used one hand to report as quickly
as possible whether the letter was high (‘‘X” key) or low (‘‘Z”
key). If the background was pink, participants used their other
hand to report as quickly as possible whether the number was
odd (‘‘N” key) or even (‘‘M” key). Participants completed four single
task blocks (2 blocks of odd/even and 2 blocks of high/low) of 30
trials each. They then completed a practice dual task block in
which they switched from one task to the other every five trials
for 30 trials. Finally, they completed a dual task block of 160 trials
during which the task for each trial was chosen randomly.
184.108.40.206. Tower of London (Tunstall, 1999). Participants viewed an
apparatus with three pegs and four discs and tried to rearrange
the discs on the pegs to match a target arrangement shown in a
picture. Only one disk could be moved at a time – a move consisted
of shifting a disk from one peg to another. Participants tried to
complete each problem within a certain number of moves, and
had three chances to solve each problem. Participants attempted
nine problems of increasing difficulty, and they were told that
accuracy was important rather than speed. Solving a problem cor-
rectly on the first attempt merited three points. With each addi-
tional attempt, the number of points awarded was decreased by
one. This test was administered once at the beginning of the study
and once at the end of the study.
220.127.116.11. Working memory operation span (Turner & Engle, 1989). Par-
ticipants solved math problems (e.g., IS (9/1) + 2 = 9?) while simul-
taneously trying to remember sets of 3–6 words. After each set of
3–6 words, participants were asked to recall the words in the set.
Since this test was administered three times, three versions of
the test were used.
18.104.22.168. Ravens matrices (Raven, Court, & Raven, 1990). Participants
completed a version of the Ravens Advanced Matrices. This test in-
volved presenting participants with a complex visual pattern with
a piece cut out of it. The task of the participant was to find the
missing piece that completed the pattern. The full version of the
Ravens was divided into three sub-tests of approximately equal
difficulty, with each test containing 12 items. During the first test-
ing battery, participants were given 5 min to complete a practice
version of the test before the first actual test. Participants were gi-
ven 20 min to complete each 12 item test, once at the beginning of
the study, once in the middle, and once at the end of the study.
All group means for each cognitive task, when not reported in
graphical form, are available in the form of an online appendix
(see Appendix A). Given the large number of tasks and analyses,
we discuss only critical effects and interactions. All ANOVA terms
not reported in the results section are also available online. The
central questions are whether expert and non-gamers differed in
their perceptual and cognitive abilities, and whether practice on
a particular game had a differential effect on task performance
on the tasks in the assessment battery over time — that is, whether
video game experience led to greater task improvement, and
whether the tasks affected by game experience differed depending
on the type of video game. These questions were tested by exam-
ining whether expert and non-gamer performance differed on the
tasks in the assessment battery, and second by testing within the
game practice groups whether performance on the assessment bat-
tery tasks interacted with group (CONTROL, MOH, TETRIS, RON)
and assessment session (Session1, Session 2, Session 3). To reduce
the influence of within-participant outliers, we analyzed median
rather than mean response times unless otherwise noted. Further-
more, if it was apparent from the data that a participant confused
the response mappings or simply did not understand the task, data
from that participant were not included in the analysis of that task.
Practice blocks were not included in any of the analyses.
3.1. Game performance
Before analyzing the effects of transfer from the video games to
the tasks in the assessment battery, we first examined whether
practice led to improvement on the practiced game. Performance
on the first level was compared before and after practice.1For
MOH and RON players, the first level followed the game tutorials.
Because Tetris had no tutorial, performance on the first day was trea-
ted as a ‘‘tutorial,” and the second session was treated as the first
All game groups improved significantly. For MOH, firing accu-
racy improved significantly (45% and 52% pre- and post-practice,
respectively, t(18) = 3.36, p < .01) as did the ratio of hits taken to
enemies killed (.72 and 1.01 pre-and post-practice, respectively,
t(18) = 2.43, p < .05). Following practice, MOH participants hit tar-
gets more accurately and were more efficient. RON players also im-
proved significantly. Initially, only 57% of participants achieved
victory on the first scenario they were asked to complete. At the
end of the practice period, 97% were victorious with the same sce-
nario. Scores generally decreased overall. However, this was simply
due to participants achieving victory too quickly to gain many
points. Whereas initially it took participants on average 146 min
to complete the scenario, post-practice it only took participants
56 min(t(19) = 4.27, p < .001). Althoughscores generally decreased,
two scores increased significantly. Participants’ ‘‘territory” score, an
index of how much land participants controlled at the end of the
game, improved (473 and 583 pre-and post-practice, respectively,
t(19) = 2.81, p < .05). Additionally, participants’ ‘‘wonders” scores
improved as well (329 and 749 pre-and post-practice, respectively,
t(19) = 3.10, p < .01). Withan advancedenoughsociety, participants
may choose to build world wonders and are awarded points for
doing so. Thus, with practice, participants were able to build larger,
more advanced societies in less time, and also achieve victory more
often. Finally, following practice, Tetris participants showed a large
and significant improvement in total score (167,328 and 302,635
pre-and post-practice, respectively, t(19) = 3.31, p < .01).2
3.2. Visual and attentional tasks
3.2.1. Functional field of view task
expert group to outperform the non-gamer group, and for MOH
practice to differentially improve participants’ ability to detect
1Due to a computer error, game scores were lost for 1 MOH participant. Data from
three RON participants were recorded incompletely and were not included in the
analysis of game scores.
2Total score is reported, but all three measures of Tetris performance (score, level,
number of lines) were highly correlated.
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
andlocalize the target.Althoughexpert videogameplayers enjoyed
an advantage in this task compared to non-gamers, this effect did
not reach significance (F(1, 19) = .95, p = .34). Furthermore, target
eccentricity did not interact with expertise (F(2, 38) = 1.30, p = .29).
The performance of all longitudinal groups improved (F(2,
152) = 120.79, p < .001). However, participants who received
MOH experience did not improve more than participants who re-
ceived experience on Tetris or no video game experience at all:
groupdidnot interact significantly
(F(6,152) = 1.57, p = .16) nor with testing session and eccentricity
(F(12, 304) = .49, p = .92).
3.2.2. Attentional blink
Based on earlier results (Green & Bavelier, 2003) we expected
participants who played MOH to show the largest reduction in
the attention blink effect and for experts to outperform non-ga-
mers. That is, we expected them to show improved accuracy in
detecting T2 (the second target) in the RSVP stream given that T1
(the first target) was correctly detected. Data from one expert par-
ticipant were lost due to computer error.
Similar to FFOV data, although data were in the predicted direc-
tion, the difference between expert and novice groups was not sig-
nificant when T2 data were entered into an ANOVA with lag (3, 4,
5, 6) as a within-participant factor and group (expert vs. non-ga-
mer) as a between-participant factor (F(1, 19) = 1.75, p = .20). Addi-
tionally, there was no reliable lag x group interaction (F(3,
57) = 1.07 , p = .37), nor did experts outperform non-gamers in T1
detection (99% vs. 95% respectively, F(1, 19) = 1.79, p = .20).
Longitudinal data were entered into an ANOVA with T2 lag and
testing session as within participant factors and group as a be-
tween participant factor. Participants demonstrated the classic
attention blink effect. Accuracy improved across testing sessions
(F(2, 156) = 69.63, p < .001), especially for early lags as indicated
by a significant testing session x lag interaction (F(6, 468) = 5.71,
p < .001). However, group did not interact with testing session
(F(6, 156) = .90, p = .50) nor with the testing session and lag
(F(18, 468) = .76, p = .75). Thus, our data provide no evidence that
participants in the MOH group gain any significant advantage from
MOH experience over Tetris experience, RON experience, or no-
game experience at all.
To rule out the possibility that differences in T1 performance
masked T2 improvements, we compared T1 accuracy across testing
sessions and groups. This analysis revealed no reliable effects of
group (F(3, 78) = .62, p = .60), testing session (F(2, 156) = 1.96,
p = .15), or group x testing session interaction (F(6, 156) = 1.31,
p = .26). Thus, the longitudinal groups did not differ reliably in their
ability to detect T1.
Based on earlier results (Green & Bavelier, 2006b), we expected
experts to outperform non-gamers and for MOH experience to en-
hance the ability to report the number of dots in a briefly presented
display. Again, although numerically, experts outperformed non-
gamers, this effect was not significant (F(1, 19) = 2.10, p = .17),
nor did groups interact with the number of items displayed (F(7,
133) = 1.10, p = .37). Next, we examine the effect of game practice.
Benefits of practice, if present, should occur when the number of
objects in the display exceeds the subitizing range, as indicated
by a session x group x number of objects interaction. An ANOVA re-
vealed decreasing accuracy as the number of objects increased
(F(7, 539) = 132.18, p < .001), no effect of testing session (F(2,
154) = .50, p = .61), and no effect of group (F(3, 77) = .15, p = .93).
Critically, these factors did not interact significantly (F(42,
1078) = 1.15, p = .24) and no other interaction with group ap-
proached statistical significance (all p’s > .80).
3.2.4. Multiple object tracking
The primary measure of interest in this task is the speed at
which participants could track three items while maintaining near
perfect accuracy. Three participants were not included in the cur-
rent analysis due to experimenter error (two CONTROL, one RON).
Experts far outperformed non-gamers in their ability to track at
higher speeds (Fig. 1, F(1, 19) = 15.82, p < .001). Accuracy did not
differ significantly for expert and non-gamers (F(1, 19) = .77,
p = .39, 94% and 98%, respectively). To further ensure that this
speed effect was not influenced by a speed-accuracy trade-off (in
the sense that experts were setting the speed higher at the cost
of less accurate performance), the same analysis was performed
for participants who only performed at 100% accuracy during test-
ing. This analysis also resulted in a significant expertise effect (F(1,
14) = 12.7, p < .01).
Although experts outperformed non-gamers, a video game
advantage was not evident in the longitudinal groups. Participants
generally set the speed faster with repeated testing (F(3,75) = 4.96,
p < .05), but no significant group x session interaction was pres-
ent(F(6, 150) = .73 p = .63).3
3.2.5. Visual short-term memory
The primary measure in this task is the accuracy of change
detection accuracy. Experts far out performed non-gamers in this
task, especially in the large set size condition as indicated by a sig-
nificant group effect (F(1, 18) = 19.72, p < .001) and a significant
group by set size interaction (F(2, 36) = 3.63, p < .05, Fig. 2).4How-
ever, there was no effect of group (F(3, 78) = .64, p = .59), nor did
Speed Setting (degrees/sec)
Fig. 1. Expert vs. non-gamer performance in the multiple object tracking task.
Speed setting represents the maximum speed participants could track three items
perfectly. Error bars represent plus and minus 1 SEM.
3Participants were tested on their ability to track multiple objects at the speed
setting they selected. Accuracy data suggest that some of the observed speed
increases in the longitudinal groups might be due to an increasingly liberal bias in
setting the speed. The proportion correct during this testing phase did not vary by
group (F(3, 75) = 1.16, p = .33), but participants tended to be less accurate across
sessions (F(2, 150) = 2.99, p = .05), with accuracy dropping from 97% at session1 to
95% at session3. This pattern was consistent across groups, with no significant testing
session x group interaction (F(6, 150) = .50, p = .81); the speed/accuracy trade-off did
not differentially affect one group more than another.
4One expert was excluded from this analysis. This participant appeared to have
rested his hands too heavily on the response keys, resulting in the response keys
being always depressed and chance performance.
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
156) = 1.38, p = .23) nor with session and set size (F(12, 312) = .91,
p = .53).
membership interactwith assessmentsession(F(6,
3.3. Spatial processing and spatial memory tasks
3.3.1. Spatial 2-back task
The primary measures of interest are the speed and accuracy in
determining whether the spatial location of a letter was the same
as the location of the letter presented two items earlier. Given the
spatial aspects of Tetris, and to some degree RON, we predicted dif-
ferential improvement on this spatial task. We also predicted that
experts would outperform non-gamers in this task.
Experts showed a trend to be faster than non-gamers in this
task (514 vs. 687 ms, respectively, F(1, 19) = 4.18, p = .06), but were
no more accurate (90% vs. 90%, F(1, 19) < 1, p = .99). For the longi-
tudinal groups, response times decreased with testing session (F(2,
156) = 44.22, p < .001) but the groups did not differ (F(3, 78) = .23,
p = .89) and groups did not interact with testing session (F(6,
156) = .65, p = .69). In short, participants in all groups were faster
with practice on the task, and game experience did not lead to dif-
ferential improvements in response speed.
In terms of accuracy, there was no significant difference be-
tween groups (F(3, 78) = 1.9, p = .14). Unlike response times, accu-
racy remained relatively constant over repeated testing (F(2,
156) = 1.52, p = .22). Again, groups did not interact with testing
session (F(6, 156) = .38, p = .89), meaning that game experience
had no differential effect on accuracy. Overall, these results do
not support the prediction that Tetris can differentially improve
spatial memory ability.
3.3.2. Corsi block-tapping task
The primary measure in this task is proportion correctly re-
called for each sequence at each set size.
No significant difference was observed between expert and
non-gamers (F(1, 18) = .71, p = .41) , and in general, performance
was better for smaller set sizes (F(5, 90) = 57.04, p < .001). Overall,
the performance of longitudinal participants improved with re-
peated testing (F(2, 154) = 7.38, p < .01) and was better for smaller
set sizes (F(6, 462) = 766.46, p < .001). These factors did not inter-
act significantly (F(12, 924) = 1.21, p = .27). Additionally, there
was no effect of group (F(3, 77) = .88, p = .48), and groups did not
interact with set size (F(18, 462) = .75, p = .76), with testing session
(F(6, 154) = 1.43, p = .21), or with set size and testing session (F(36,
924) = .67, p = .93). Video game experience did not enhance perfor-
mance beyond what occurred solely due to repeated testing on this
task. More specifically, Tetris, which involves a spatial memory
component, did not enhance performance beyond the no-game
3.3.3. Mental rotation
The primary measures in this task are the speed and accuracy of
performance as a function of the extent of rotation required by the
display. Given that Tetris requires mental rotation and that the
shapes used in this task were modeled on those from Tetris, Tetris
experience should lead to differential improvements in perfor-
mance over time.
For expert and non-gamers, response times showed a classic
mental rotation pattern, with slower responses up to 180? of rota-
tion in either direction (Fig. 3, F(7, 133) = 45.61, p < .001). Although
showing a definite trend, expert video game players were not sig-
nificantly faster than non-gamers (F(1, 19) = 2.51, p = .13). Degree
of rotation and groups did not interact (F(1, 19) = .63, p = .73).
Accuracy data were similar, indicating a trend for video game play-
ers to be more accurate (F(7, 133) = 2.95, p = .10). Rotation and
Fig. 2. Expert vs. non-gamer performance in the visual short-term memory task.
Error bars represent plus and minus 1 SEM.
0 4590 135 180 225 270 315
Response Times (ms)
0 4590 135 180 225 270 315
Fig. 3. (A) Response times for expert vs. non-gamers in the mental rotation task. (B)
Accuracy for expert vs. non-gamers in the mental rotation task. Error bars represent
plus and minus 1 SEM.
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
groups did not interact (F(7, 133) = .44, p = .88). Note that a one-
tailed test would be justified giving the directional nature of our
hypothesis, and such a test would lead to a near-significant and
significant expertise effect for response time and accuracy, respec-
tively. A composite measure of performance (Townsend & Ashby,
1983) that takes both speed and accuracy into account revealed a
19) = 6.05, p = .03). Given that this benefit was general (i.e., it did
not interact with rotation, F(1, 19) = 1.03, p = .42), it does not ap-
pear to be related to mental rotation per se, but just to an overall
advantage in processing and response speed.
As with the experts and non-gamers, the response times of lon-
gitudinal groups showed a classic mental rotation effect (F(7,
546) = 184.01, p < .001). Overall, participants were faster with re-
peated testing (F(2, 156) = 107.57, p < .001), and the testing session
interacted with the extent of rotation such that response latencies
improved more for greater extents of rotation (F(14, 1092) = 11.24,
p < .001). Interestingly, though, this pattern interacted with the
group (F(42, 1092) = 1.64, p = .01) — Tetris players improved most,
especially when the shape required the most rotation (Fig. 4).5
Accuracy data mirrored RT data except for the absence of either
a session by group interaction (F(6, 156) = 1.15, p = .34) or a session
by group by rotation interaction (F(42, 1092) = 1.05 p = .39). Thus,
the group effect was evident in the RT but not in the accuracy data.
These results suggest that Tetris playing did differentially enhance
performance on mental rotation (see also De Lisi & Wolford, 2002;
Sims & Mayer, 2002). Note, however, that transfer in this case is
limited, given that the mental rotation task was both visually
and conceptually similar to the video game.
3.4. Executive function tasks
For this analysis three participants were excluded due to failure
to comply with instructions (one TETRIS, two RON), two were ex-
cluded due to poor math performance (one TETRIS, one MOH),
and two were excluded due to experimenter error (two CONTROL).
The primary measure in this task was the number of correctly re-
called words on each test. Expert gamers did not differ from novice
gamers (M = 33.5 and 35.8 respectively; F(1, 19) = .59, p = .45).
Overall, performance improved across testing sessions for the lon-
gitudinal groups (F(2, 140) = 15.42, p < .001). Critically, training
groups did not interact with testing session (F(6, 140) = 1.02,
p = .41), suggesting no effect of video game experience.
3.4.2. Tower of London
The primary measure in this task was based on the number of
times participants solved the problem correctly and how many at-
tempts they needed to do so. The task was completed only at the
beginning and end of the game practice period. This task was not
included in the expert/novice task battery. Overall, participants
performed better the second time they performed the task (F(1,
78) = 63.89, p < .001). However, there was no differential improve-
ment as a function of group (F(3, 78) = .722, p = .54). Thus, game
experience did not modulate this improvement.
3.4.3. Task switching
The primary measure in this task is switch cost during the dual
task blocks: the difference in performance for trials when the pre-
ceding trial involved the same task and those when the preceding
trial was of the other task. Switch costs were calculated by sub-
tracting the response time for non-switch trials from the response
time for switch trials. It was clear from error rates that three par-
ticipants in the expert group, and one in the novice group confused
the response mappings during at least on block of the task. Addi-
tionally, in the longitudinal groups four participants confused re-
sponse mappings (one TETRIS, one CONTROL, two RON). Data
from these participants were excluded. Experimenter error re-
sulted in the loss of data from two participants (one MOH, one
RON). It was predicted that switch costs should decrease more
for participants given practice on videogames that require task
switching compared to practice on games that do not or no game
experience, and for expert video game players to outperform
Experts showed a smaller switch cost compared to non-gamers
(F(1, 15) = 5.87, p < .05, Fig. 5). Group (expert vs. non-gamers)
interacted with task (high/low vs. odd/even), suggesting that ex-
perts demonstrated a smaller cost than non-gamers, but primarily
for the easier high/low task (F(1, 15) = 4.40, p = .05). A reduction in
switch cost was not the result of lower accuracy, given there was
no overall effect of group (F(1, 15) = .08, p = .79) or of an interaction
between group and task F(1, 15) = .82, p = .39).
Turning to the longitudinal groups, switch costs diminished
with repeated testing (F(2, 144) = 11.84, p < .001). However, the
groups did not differ (F(3, 72) = 2.07, p = .12) and groups did not
interact with testing session F(6, 144) = .73, p = .63) or with testing
session and task (F(6, 144) = .82, p = .55). Thus, there is no evidence
that video game experience had any effect on the ability to quickly
switch between two tasks.
In general, accuracy data mirrored the response time data, with
switch costs generally declining across sessions (F(2, 144) = 91,
p < .01). Again, there was no effect of group (F(3, 72) = 1.19,
0 4590 135
RT Improvement (ms)
Fig. 4. Response time improvement in the mental rotation task for each rotation
and group comparing session 1 and session 3. Error bars represent plus and minus 1
5The observant reader might notice in the online appendix that although Tetris
players improved most, they were initially slower compared to the other three
groups. This is almost exclusively due to three participants in the Tetris group with
abnormally long response times and steep mental rotation slopes in this task.
Excluding these three participants from analysis equates response time baselines
almost perfectly, and the time x group x rotation interaction remains significant (F(42,
1050) = 1.56, p < .05). From Session 1 to Session 3, this interaction is evident by Tetris
players improving significantly more compared to RON players (at the .05 level) at
rotations of 45, 90, 180, 270, and 315?, MOH players at rotations of 90, 180, and 315?,
and CONTROL participants at the 45? rotation. Thus it is unlikely that differences in
initial baseline performance can explain differential improvement in the mental
rotation task. No significant group effects or interactions with group were observed in
the accuracy data.
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
p = .32) and group did not interact with assessment session (F(6,
144) = 1.59, p = .15).
3.4.5. Ravens matrices
For each 12-problem test, training participants were given one
point for each problem solved correctly. Ravens scores were en-
tered into an assessment session (session 1, session 2, session 3)
by group ANOVA. This ANOVA revealed no effect of session (F(2,
156) = .25, p = .78), no effect of group (F(3, 78) = .60, p = .62), and
no interaction between session and group (F(6, 156) = 1.17,
p = .33). Thus, game experience did not appear to modulate
4. Discussion and conclusions
Two basic questions were investigated in our study. First,
whether video-game experience results in broad and differential
improvements in perceptual and cognitive abilities, as could be
identified by an examination of the interactions between longitu-
dinal groups, assessment session and task. In addition, we were
also interested in establishing whether habitual gamers (experts)
and non-gamers differ in their performance on the same battery
of transfer tasks used with the longitudinal groups. Because ex-
perts and non-gamers differed greatly in the amount of time they
had spent playing video games, a comparison of these two groups
can in principle address the question of whether some differences
may only emerge with much longer game practice periods that can
be afforded by an experimental study.
In a number of tasks, video game experts outperformed non-ga-
mers. Experts were able to track objects moving at greater speeds,
perform more accurately in a visual short-term memory test,
switch between tasks more quickly, and make decisions about ro-
tated objects more quickly and accurately. However, with the
exception of Tetris, practicing video games for twenty-one h was
not enough to engender benefits in non-video game players. Action
or strategy game practice regimens did not significantly improve
performance on any of the transfer tasks over and above improve-
ment related to performing the transfer task multiple times (i.e.
the control group).
Interestingly, even tasks in which video game experience has
been found to be beneficial in the past did not reveal significant vi-
deo game effects, including the FFOV task, the attention blink task,
and the enumeration blink task (e.g., Feng et al., 2007; Green &
Bavelier, 2003; Green & Bavelier, 2006a; Green & Bavelier,
2006b). The fact that experts did not perform significantly better
on these tasks as compared to non-gamers (although the expert
data do indicate clear trends, see online data appendix) suggests
that differences between our tasks and tasks used previously by
other researchers may be playing an important role. For example,
it is known that factors such as the intensity of the mask relative
to the intensity of the target and the time between the offset of
the target and offset of the mask can have large effects on the effec-
tiveness of a visual mask (Fehrer & Smith, 1962; Macknik & Living-
stone, 1998). This may account for the discrepancy between our
FFOV results and the results of others. The masking stimulus in
our FFOV task was of a much shorter duration (100 ms) and was
bright and colorful (compared to the black and gray mask used
by Green & Bavelier, 2003; Green & Bavelier, 2006a). Masking in
our task may have been more similar to masking by light as com-
pared to metacontrast masking or pattern masking (see Breitmeyer
& Ög ˘men, 2006 for a review of visual masking). Additionally, the
nature of the mask may have led to stronger backwards masking,
which is supported by the much lower initial performance com-
pared to previous reports (e.g., Green & Bavelier, 2006a). Thus
the masking of the target may have been quantitatively or qualita-
tively different from the masking which is amenable to video game
expertise effects. Future work needs to determine the exact mech-
anisms that allow video game experts and those trained to play vi-
deo games to better localize briefly presented targets in the
periphery, perhaps by varying stimulus properties of the task.
Interestingly, although the mask was very effective, performance
improved substantially with repeated testing. Participants were
learning something about how to better perform this task, which
may have obscured our ability to detect video game experience
Other task differences may also have played an important role.
In the attention blink task, we restricted the lag between the first
target and the second target to be between lags 3 and 6, where
the greatest video game effects have been observed previously
(Green & Bavelier, 2003). However, we may have unintentionally
reduced our ability to observe video game training effects by
reducing the temporal uncertainty of the second target. It is more
difficult to explain the lack of significant training effects in the enu-
meration task given the simplicity of the task and the close replica-
tion of experimental paradigm. Our version of the task only
included set sizes from 1 to 8, whereas previous video game exper-
iments have used set sizes up to 12 (but only analyzed and re-
ported data up to 8). Additionally, in our task, accuracy was
emphasized over speed for two reasons: 1) video game players
have faster key press responses overall (Orosy-Fildes & Allan,
1989), and in general may be more familiar with the layout of
the numeric keys of the keyboard and 2) Green and Bavelier
(2003), Green and Bavelier (2006b) have reported effects of game
experience on enumeration for accuracy only. Our emphasis on
accuracy is in contrast to previous work emphasizing speed or re-
sponse, which may give video game players an additional advan-
tage as compared to non-gamers (Green & Bavelier, 2003; Green
& Bavelier, 2006b). As for the multiple-object tracking task, we
did observe expert/non-gamer differences but did not observe
game practice effects. However, it is important to note the large
differences between the MOT paradigm we used, which measured
tracking speed, and the paradigm used by Green and Bavelier
(2006b), which measured tracking capacity.
Could differences in the schedule of game practice sessions have
an effect on the degree to which practice transferred to perceptual
and cognitive tasks? It is possible, but in our opinion, unlikely. We
had participants visit the laboratory four times a week (on separate
days), for 1.5 h each session, while Green and Bavelier (e.g., Green
& Bavelier, 2006b) had participants complete ten 1-hour sessions
within the span of 15 days. Thus, sessions were more spaced-out
Switch Cost (ms)
Fig. 5. Expert vs. non-gamer performance in the task switching task (response time
cost). Error bars represent plus and minus 1 SEM.
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
in time for our participants (although participants played more
during each individual session, and completed a greater number
of sessions). However, many results suggest that in a number of
different contexts, a more spaced or distributed practice schedule
is actually more effective (e.g., Baddeley & Longman, 1978; She-
bilske, Corrington, & Jordan, 1994). Currently, little is known
regarding how the schedule of video game practice affects transfer
to other perceptual and cognitive abilities, but this is an interesting
and meaningful question for future research.
Could gender be an important factor in our observation of no dif-
ferential task improvement as a function of video game practice?
Whereas previous studies have used a mixed group of males and fe-
males (e.g., Green & Bavelier, 2003; Green & Bavelier, 2006a; Green
& Bavelier, 2006b), our longitudinal game groups were comprised
mainly of female participants. Gender differences in spatial cogni-
tion are well known (e.g., Casey, Nuttall, Pezaris, & Benbow, 1995;
Geary, Saults, Liu, & Hoard, 2000; Terlecki & Newcombe, 2005).
However, it is very unlikely that the lack of transfer observed is
the result of a primarily female sample. Feng et al. (2007) observed
a larger game practice benefit for female participants in the FFOV
task and a mental rotation task as compared to male participants.
In fact, video game training was able to completely eliminate gen-
der differences in the FFOV task (while in the mental rotation task,
differenceswere reducedsubstantially). It is interestingto notethat
our own FFOV accuracy data appear to replicate this pattern of gen-
der differences. When non-gamer males and females are compared
directly, males significantly outperform females (F(1,88) = 11.07,
p < .01, Ms = .35 vs. .25 for males and females, respectively). Yet,
action game practice did not differentially improve performance
in our study. Ruling out gender differences as a factor suggests
again that stimulus and task factors are likely the explanation for
why we observed little transfer of practice while previous studies
have found substantial and significant video game effects.
Could methodological shortcomings of our study have an im-
pact on our ability to observe transfer effects? This is a possibility,
but again, we feel that this is unlikely. Given the multitude of tasks
we asked participants to perform, it is possible that similar tasks in
the cognitive battery trained each other and this may have been
responsible for the general response time and accuracy improve-
ments across sessions we observed. These general improvements
may have masked any effect of video game practice. Furthermore,
multiple assessment sessions (three rather than the usual two typ-
ically used by Green and Bavelier) may have had an influence as
well. Our purpose in choosing three assessment points was to gain
information about the rate of improvement over time and the
dose-response function of video game experience. However, great-
er experience with the transfer tasks themselves might have
masked game effects (although if this were the case, we would
have observed differential improvements from session 1 to session
2, which were not present). The quite substantial improvements in
response time and accuracy that were often observed (e.g., in the
FFOV task, the attention blink task, and the mental rotation task)
from simply repeated testing may have practical implications.
When it comes to improving performance on a particular task,
practice or repeated experience with that task may prove the most
efficient route compared to a general video game practice
The inability to observe transfer effects and strong expertise ef-
fects in tasks similar or nearly identical to tasks that have exhibited
training effects in the past may be important in understanding the
nature of video game effects and their ultimate practical implica-
tion. Whereas it is difficult to pinpoint the exact reason for our reg-
imens inability to produce significant improvements on these
tasks, our results suggest that there exist important boundary con-
ditions on the effectiveness of the use of video games to improve
performance on other tasks. What may seem like inconsequential
procedural and stimulus changes can significantly alter the degree
to which video game experience transfers to other tasks. This raises
concerns regarding whether video game practice may transfer, not
only to laboratory tasks, but also to the complex and dynamic tasks
we perform every day outside the laboratory. It is also unclear ex-
actly how efficient video game interventions, with the purpose of
improving perceptual and cognitive performance, may be in cer-
tain situations. It would be ideal to have data on how much video
game experience is required to improve performance on a transfer
task to a certain level as compared to how much practice on the ac-
tual transfer task is required to reach the same level of perfor-
mance. To our knowledge, the only evidence published in a
journal of video game experience transferring to complex, real-
world tasks has been the case of Space Fortress training improving
the flight performance of Israeli air force pilots (Gopher et al.,
For the tasks in which video game experts outperformed non-
gamers, twenty-one h of video game training did not produce these
effects except for the case of mental rotation. It is entirely possible
that, in order to see these benefits, many more hours of experience
are necessary (but see Green & Bavelier, 2003 in which only 10 h of
training were needed to achieve significant transfer). Most experts
in our expert group reported playing video games starting from
very early childhood. This means that it is likely that these partic-
ipants had tens of thousands of hours of video game experience
when we tested them. Of course it is important to note that other
explanations may account for the higher transfer task performance
by video game experts. For example, video game experts may dem-
onstrate superior perceptual, attentional, and cognitive skills due
to self-selection. These skills may encourage video game expertise
given that they are required for successful gaming. Other non-cau-
sal mechanisms might be speculated as well, including differences
in households in which video game systems are present. Video
game expertise may in fact be a complicated confluence of causal
and non-causal variables.
Although it may be the case that video games can produce
broad transfer to a number of tasks, it is important to understand
why transfer might be broad (or appears to be broad). The answer
to this question has important implications for theories of learning
and expertise. One potential explanation is that video game train-
ing encourages flexible strategies and results in general improve-
ments in attentional control, which can in turn be applied to a
number of different tasks, as is the case with variable priority
training (e.g., Fabiani et al., 1989; Gopher, Weil, & Siegel, 1989;
Kramer, Larish, & Strayer, 1995). The complexity of modern video
games, which typically have many different goals and sub-goals,
would appear to encourage strategies centered on dynamically
shifting attention to different elements of the game. Another pos-
sibility is that transfer occurs due to the same skills being critical
to both the game and transfer task. However, the key may be that
broad transfer is observed because complex video games require
the same skills to be executed in a variety of different contexts
(Schmidt & Bjork, 1992). As an example, action games require par-
ticipants to track multiple moving enemies, much in the same way
that participants must track circles in a multiple object-tracking
task. However, in the game different enemies might have many dif-
ferent speeds or ways of moving, encouraging participants to gen-
eralize the skills they learn in the gaming context to novel stimuli.
Answering such questions would require either the deconstruction
of complex video games into simpler sub-games or strategy
manipulations to encourage participants to emphasize certain as-
pects of the game over others, much in the same way Space For-
tress training and transfer has been studied (e.g., Fabiani et al.,
The exact mechanisms of improved performance on video
games and transfer tasks is still somewhat uncertain (i.e., what
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
is it that players actually learn from playing video games). In an
interesting series of papers Maglio and colleagues (Kirsh & Ma-
glio, 1994; Maglio & Kirsh, 1996; Maglio, Wenger, & Copeland,
2008) examined expertise in Tetris. Tetris expertise was related
to strategy shifts that utilized epistemic actions, or actions that
decreased mental computations (e.g., rotating Tetris blocks on
the screen rather than in the mind). Given that our mental rota-
tion task did not allow for the use of such epistemic actions, it is
unlikely that this strategy shift is the explanation for improved
performance in the transfer mental rotation task. However, it
is possible that participants used some combination of epistemic
actions and more efficient mental rotation to improve their Te-
tris performance, and our transfer task tapped only mental rota-
tion ability. Unfortunately, our battery of cognitive tasks did not
have a clear task in which measures of epistemic actions are
available, but it would be interesting to examine whether Tetris
training might make participants to utilize such strategies in
other tasks as well. Strategies learned during video game play
(in addition to improved visual and attention processing) may
be an important and neglected factor in explaining video game
To date, video game training appears to be one of the more
interesting and promising means to improve perceptual, atten-
tional, and cognitive abilities. One of its promises is that, compared
to traditional training, it can be engaging and entertaining. This has
led some companies to begin to market video games for the spe-
cific purpose of improving cognition. For example, Nintendo adver-
tises Big Brain AcademyTMas a game that ‘‘trains your brain with a
course load of mind-bending activities across five categories: think,
memorize, analyze, compute, and identify” (Nintendo, 2008a).
Players are assigned a ‘‘brain weight” score based on their perfor-
mance, with a higher brain weight corresponding to better perfor-
mance. Nintendo states that these puzzles are ‘‘designed to help
you increase the weight of your mighty brain”. Brain AgeTM, an-
other game developed by Nintendo that appears specifically tar-
geted to older adults, makes the claim that players can ‘‘train
their brain in minutes a day” (Nintendo, 2008b). Like Big Brain
AcademyTM, players initially receive a performance score known
as their ‘‘brain age”. This score is based on response speed and
accuracy on a number of simple reaction time and perceptual tests.
As players train on these tasks their ‘‘brain age” score decreases,
implying that the training transfers broadly to other aspects of
However, it is clear that much more research needs to be con-
ducted before researchers might recommend a certain game to an
individual to improve performance on a task of interest. For
example, it should be noted that the reaction time, logic, and per-
ceptual training offered by these ‘‘brain training” games bear a
striking resemblance to the training tasks of the ACTIVE trial (Ball
et al., 2002), which produced little meaningful immediate transfer
to real-world tasks (in contrast to the visual search training par-
adigm developed by Ball and colleagues which transfers to driv-
ing performance, Roenker, Cissell, Ball, Wadley, & Edwards,
2003). Thus, buying one of these games for the purpose of
improving ones cognitive abilities may be premature. While our
laboratory has demonstrated that older adults who practice the
game Rise of Nations show significant improvements in tasks
measuring memory, task-switching ability, reasoning ability, and
spatial skills, (Basak, Boot, Voss, & Kramer, in press), future re-
search should investigate whether these gains transfer to com-
plex, real-world tasks. It should also investigate the dose-
response curve for video game experience and potential benefits,
the relationship between the nature of the game and the nature
of transfer, and the exact mechanism or mechanisms that im-
prove task performance, including potential interactions with
individual differences and age.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.actpsy.2008.09.005.
Alvarez, G. A., & Franconeri, S. L. (2005). How many objects can you track? Evidence
for a flexible tracking resource [Abstract]. Journal of Vision, 5(8), 641a, http://
Baddeley, A. D., & Longman, D. J. A. (1978). The influence of length and frequency of
training session on the rate of learning to type. Ergonomics, 21, 627–635.
Ball, K., Berch, D. B., Helmers, K. F., Jobe, J. B., Leveck, M. D., Marsiske, M., et al.
(2002). Effects of cognitive training interventions with older adults: A
randomized controlled trial. Journal of the American Medical Association, 288,
Basak, C., Boot, W. R., Voss, M., & Kramer, A. F. (in press). Can training in a real-time
strategy videogame attenuate cognitive decline in older adults? Psychology and
Braver, T. S., Cohen, J. D., Nystrom, L. E., Jonides, J., Smith, E. E., & Noll, D. C. (1997). A
parametric study of prefrontal cortex involvement in human working memory.
NeuroImage, 5, 49–62.
Breitmeyer, B. G., & Ög ˘men, H. (2006). Visual masking: Time slices through conscious
and unconscious vision. Oxford: Oxford University Press.
Casey, M. B., Nuttall, R., Pezaris, E., & Benbow, C. P. (1995). The influence of spatial
ability on gender differences in mathematics college entrance test scores across
diverse samples. Developmental Psychology, 31, 697–705.
Castel, A. D., Pratt, J., & Drummond, E. (2005). The effects of action video game
experience on the time course of inhibition of return and the efficiency of visual
search. Acta Psychologica, 119, 217–230.
Clark, J. E., Lanphear, A. K., & Riddick, C. C. (1987). The effects of videogame playing
on the response selection processing of elderly adults. Journals of Gerontology,
Cooper, L., & Shepard, R. (1973). Chronometric studies of the rotation of mental
images. In W. Chase (Ed.), Visual Information Processing (pp. 135–142). New
York, NY: Academic Press.
Corsi, P. M. (1972). Human memory and the medial temporal region of the brain.
Dissertation Abstracts International, 34(02), 819B.
Donchin, E., Fabiani, M., & Sanders, A. (Eds.). (1989). The learning strategies
program: An examination of the strategies in skill acquisition [Special issue].
Acta Psychologica, 71, 1–311.
De Lisi, R., & Wolford, J. L. (2002). Improving children’s mental rotation accuracy
with computer game playing. The Journal of Genetic Psychology, 163, 272–282.
Fabiani, M., Buckley, J., Gratton, Gl, Coles, M. H., Donchin, E., & Logie, R. (1989). The
training of complex task performance. Acta Psychologica, 71, 259–299.
Fehrer, E., & Smith, E. (1962). Effects of luminance ration on masking. Perceptual and
Motor Skills, 14, 243–253.
Feng, J., Spence, I., & Pratt, J. (2007). Playing an action video game reduces gender
differences in spatial cognition. Psychological Science, 18, 850–855.
Frederiksen, J. R., & White, B. Y. (1989). An approach to training based upon
principled task decomposition. Acta Psychologica, 71, 89–146.
Geary, D. C., Saults, S. J., Liu, F., & Hoard, M. K. (2000). Sex differences in spatial
cognition, computational fluency, and arithmetical reasoning. Journal of
Experimental Child Psychology, 77, 337–353.
Gopher, D., Weil, M., & Bareket, T. (1994). Transfer of skill from a computer game
trainer to flight. Human Factors, 36, 387–405.
Gopher, D., Weil, M., & Siegel, D. (1989). Practice under changing priorities: An
approach to the training of complex skills. Acta Psychologica, 71, 147–177.
Green, C. S., & Bavelier, D. (2003). Action video game modifies visual selective
attention. Nature, 423, 534–537.
Green, C. S., & Bavelier, D. (2006a). Effect of action video games on the spatial
distribution of visuospatial attention. Journal of Experimental Psychology: Human
Perception and Performance, 1465–1468.
Green, C. S., & Bavelier, D. (2006b). Enumeration versus multiple object tracking:
The case of action video game players. Cognition, 101, 217–245.
Green, C. S., & Bavelier, D. (2007). Action video game experience alters the spatial
resolution of attention. Psychological Science, 18, 88–94.
Hart, S. G., & Battiste, V. (1992). Flight test of a video game trainer. In Proceedings of
the Human Factors Society 26th Meeting (pp. 1291–1295). New York: MacMillan.
Kirsh, D., & Maglio, P. P. (1994). On distinguishing epistemic from pragmatic action.
Cognitive Science, 18, 513–549.
Kramer, A. F., Larish, J. F., & Strayer, D. L. (1995). Training for attentional control in
dual task settings: A Comparison of young and old adults. Journal of
Experimental Psychology: Applied, 1, 50–76.
Luck, S. J., & Vogel, E. K. (1997). The capacity of visual working memory for features
and conjunctions. Nature, 390, 279–281.
Macknik, S. L., & Livingstone, M. S. (1998). Neuronal correlates of visibility and
invisibility in the primate visual system. Nature Neuroscience, 1, 44–149.
Maglio, P. P., & Kirsh, D. (1996). Epistemic action increases with skill. In Proceedings
of the eighteenth annual conference of the cognitive science society (pp. 391–396).
Mahwah, NJ: Erbaum.
Maglio, P. P., Wenger, M. J., & Copeland, A. M. (2008). Evidence for the role of self-
priming in epistemic action: Expertise and the effective use of memory. Acta
Psychologica, 127, 72–88.
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398
Nintendo., (2008a). What is big brain academy? http://www.bigbrainacademy.com/ Download full-text
Nintendo., (2008b). What is brain age? http://www.brainage.com/launch/what.jsp.
Orosy-Fildes, C., & Allan, R. W. (1989). Psychology of computer use: XII. Videogame
play: Human reaction time to visual stimuli. Perceptual and Motor Skills, 69,
Pashler, H. (2000). Task switching and multi-task performance. In S. Monsell & J.
Driver (Eds.), Attention and performance XVIII: Control of cognitive processes
(pp. 277–309). Cambridge: MIT Press.
Pylyshyn, Z. W., & Storm, R. W. (1988). Tracking multiple independent targets:
Evidence for a parallel tracking mechanism. Spatial Vision, 3, 179–197.
Raven, J., Court, J. H., & Raven, J. (1990). Manual for Raven’s progressive matrices and
vocabulary scales. Research supplement no. 3 (2nd ed.). American and International
Norms. Oxford: Oxford Psychologists Press.
Raymond, J. E., Shapiro, K. L., & Arnell, K. M. (1992). Temporary suppression of visual
processing in an RSVP task: An attentional blink? Journal of Experimental
Psychology: Human Perception and Performance, 18, 849–860.
Roenker, D., Cissell, G., Ball, K., Wadley, V., & Edwards, J. (2003). Speed of processing
and driving simulator training result in improved driving performance. Human
Factors, 45, 218–233.
Schmidt, R. A., & Bjork, R. A. (1992). New conceptualizations of practice. Common
principles in three paradigms suggest new concepts for training. Psychological
Science, 3(4), 207–217.
Shebilske, W. L., Corrington, K., & Jordan, J. A. (1994). Massed versus distributed
practice in complex skill acquisition. In Proceedings of the human factors and
ergonomics society 38th annual meeting (pp. 1215–1219). Cambridge: Cambridge
Sims, V. K., & Mayer, R. E. (2002). Domain specificity of spatial expertise: The case of
video game players. Applied Cognitive Psychology, 16(1), 97–115.
Terlecki, M. S., & Newcombe, N. S. (2005). How important is the digital divide? The
relation of computer and videogame usage to gender differences in mental
rotation ability. Sex Roles, 53, 433–441.
Townsend, J. T., & Ashby, F. G. (1983). Stochastic modeling of elementary psychological
processes. Cambridge: Cambridge University Press.
Tunstall, J. R. (1999). Improving the utility of the tower of London, a neuropsychological
test of planning. Doctoral dissertation. Australia: Griffith University.
Turner, M., & Engle, R. (1989). Is memory capacity task dependent? Journal of
Memory and Language, 28(2), 127–154.
W.R. Boot et al./Acta Psychologica 129 (2008) 387–398