Vision Research 61 (2012) 132–143

Contents lists available at SciVerse ScienceDirect

Vision Research

journal homepage: www.elsevier .com/locate /v isres

Neural bases of selective attention in action video game players

D. Bavelier a,b,c,⇑, R.L. Achtman a,b,1, M. Mani a, J. Föcker a,b

a Rochester Center for Brain Imaging, Rochester, NY 14627-8917, USAb Department of Brain & Cognitive Sciences, University of Rochester, Rochester, NY 14627, USAc FPSE, University of Geneva, Geneva, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history:Received 1 June 2011Received in revised form 7 August 2011Available online 16 August 2011

Keywords:Action video gamesBrain plasticityVisual attentionfMRIPerceptual loadFronto-parietal network

0042-6989/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.visres.2011.08.007

⇑ Corresponding author at: Department of Brain & Cof Rochester, RC 270268, Meliora Hall, Rochester, N5854429216.

E-mail address: daphne@cvs.rochester.edu (D. Bav1 Present address: Nazareth College, Rochester, NY 1

Over the past few years, the very act of playing action video games has been shown to enhance severaldifferent aspects of visual selective attention, yet little is known about the neural mechanisms that medi-ate such attentional benefits. A review of the aspects of attention enhanced in action game players sug-gests there are changes in the mechanisms that control attention allocation and its efficiency (Hubert-Wallander, Green, & Bavelier, 2010). The present study used brain imaging to test this hypothesis by com-paring attentional network recruitment and distractor processing in action gamers versus non-gamers asattentional demands increased. Moving distractors were found to elicit lesser activation of the visualmotion-sensitive area (MT/MST) in gamers as compared to non-gamers, suggestive of a better early fil-tering of irrelevant information in gamers. As expected, a fronto-parietal network of areas showed greaterrecruitment as attentional demands increased in non-gamers. In contrast, gamers barely engaged thisnetwork as attentional demands increased. This reduced activity in the fronto-parietal network that ishypothesized to control the flexible allocation of top-down attention is compatible with the proposal thataction game players may allocate attentional resources more automatically, possibly allowing more effi-cient early filtering of irrelevant information.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Selective attention is fundamental to allowing task-relevantinformation to guide behavior, while reducing the impact of irrel-evant or distracting information. Many paradigms have beendeveloped with the goal of quantitatively measuring visual selec-tive attention (Carrasco & Yeshurun, 1998; Eckstein, Pham, &Shimozaki, 2004; Eriksen & Eriksen, 1974; Lavie, 1997; Treisman& Gelade, 1980). These paradigms range from visual search to flan-ker compatibility, measuring the efficiency with which targets areselected and irrelevant, potentially distracting, stimuli are ignored.Recently, playing fast-paced action video games has been shown toenhance several different aspects of selective visual attention ascompared to control games (Green & Bavelier, 2003; Hubert-Wal-lander, Green, & Bavelier, 2010 for a review). The present studyasks how such changes in behavior may be instantiated at the neu-ral level by comparing action video game players (VGPs) to individ-uals who do not play such games (NVGPs). We first review theaspects of attention that have been shown to be modified in VGPsas the design of the present study was based on this body of work.

ll rights reserved.

ognitive Sciences, UniversityY 14627-0268, USA. Fax: +1

elier).4618, USA.

It was first demonstrated that VGPs outperform NVGPs in selec-tive attention by using the Useful Field of View (UFOV) paradigm ini-tially developed by Ball and collaborators. This task requires subjectsto distribute their attention widely over the screen and locate aperipheral target while ignoring irrelevant distractors (Feng, Spence,& Pratt, 2007; Green & Bavelier, 2003; Sekuler & Ball, 1986; Spenceet al., 2009). Enhanced spatial selective attention in gamers has beenshown more recently using different types of search tasks, such asthe Swimmer task (West et al., 2008) or difficult visual search tasks(Hubert-Wallander, Green, & Bavelier, 2010; but see Castel, Pratt, &Drummond, 2005 for a different result). Interestingly, some of thesetasks include a condition where participants perform a peripherallocalization task while simultaneously discriminating betweentwo possible shapes located at fixation. This version of the task re-quires spatial selective attention as well as divided attention. Undersuch conditions, VGPs outperformed NVGPs on both the peripheraltask and the central task (Green & Bavelier, 2006a). Thus, both selec-tive attention over space as well as divided attention is enhanced inVGPs.

VGPs not only exhibit better selective attention over space, theyalso exhibit enhanced selective attention to objects. For example,VGPs can track a greater number of dynamic, moving objects ascompared to NVGPs (Dye & Bavelier, 2010; Green & Bavelier,2003, 2006b; Trick, Jaspers-Fayer, & Sethi, 2005). This skill requiresthe ability to allocate attention to several objects and to do so effi-ciently for several seconds. Another aspect of selective attention

D. Bavelier et al. / Vision Research 61 (2012) 132–143 133

also found to change in VGPs is the deployment of attention intime, or the ability to select a target from distractors presentedin a temporal sequence. Using an Attentional Blink paradigm(Shapiro, 1994), limits on the dynamic allocation of visual atten-tion were compared in VGPs and NVGPs. VGPs exhibited much lessof a blink than NVGPs, with a number of VGPs exhibiting no blinkwhatsoever, indicating that their attention recovers more quicklyover time (Green & Bavelier, 2003).

Importantly, the causal effect of action game play on several ofthese aspects of visual selective attention has been establishedthrough training studies in which naïve subjects are required toplay either action-packed, fast-paced video games or controlgames. Those asked to play action games showed greater atten-tional gains from pre-to post-test than those asked to play controlgames. This was shown when testing spatial selective attention(Feng et al., 2007; Green & Bavelier, 2003, 2006a; Spence et al.,2009), selective attention to objects (Cohen, Green, & Bavelier,2007; Green & Bavelier, 2003, 2006b) as well as selective attentionover time (Cohen, Green, & Bavelier, 2007; Green & Bavelier, 2003).

The attentional skills mentioned above primarily involve goal-directed, top-down attention. This begs the question of whetherother aspects of attention may be equally modified by action gameplay. Although stimulus-driven, exogenous attention is certainlyengaged while playing action games, it seems that the capacityand dynamics of exogenous attention are less susceptible to the ef-fects of playing action video games. Exogenous cues were found toinduce equivalent performance enhancement in VGPs and NVGPsleading to similar cue-validity effects and comparable inhibitionof return2 (Castel, Pratt, & Drummond, 2005; Hubert-Wallander,Green, Sugarman, & Bavelier, 2011). Thus, not all aspects of attentionare equally modified in VGPs. Reports that VGPs show reducedattentional capture as compared to NVGPs could suggest less exoge-nous pull in VGPs; however, the available data are also consistentwith the proposal that VGPs have better top-down attentional con-trol allowing them to either limit or recover faster from the distract-ing effect of abrupt onsets (Chisholm et al., 2010; but see West et al.,2008 for a different view). In line with the proposal of greater top-down selective attention in VGPs, a recent electrophysiological studyby Mishra et al. (2011) reported greater suppression of distracting,unattended information in VGPs. Participants were presented withfour rapid serial visual presentation streams in a steady-state visu-ally evoked potential design allowing one to recover the cortical re-sponses to the task-relevant attended stream as well as to thedistracting, unattended streams. Under these high load conditions,VGPs and NVGPs similarly processed the attended streams, but VGPsmore efficiently suppressed the unattended streams. Notably, thisgreater suppression was associated with faster reaction times. Great-er distractor suppression may be a possible mechanism for more effi-cient executive and attentional control (Clapp et al., 2011 in olderadults; Serences et al., 2004; Toepper et al., 2010). The present workbuilds on the findings of these earlier studies to further our under-standing of the mechanisms that may be at play in the attentionalenhancements noted in VGPs.

The present study directly compares VGPs and NVGPs by using avisual search paradigm contrasting an easy versus a more difficultsearch, while concurrently measuring the impact of search difficultyon the processing of irrelevant motion information (Lavie, 2005). Asmost behavioral changes documented so far point to improvementsin top-down attention after action gaming, we expected to observechanges in the dorsal fronto-parietal network, whose role in the con-trol and regulation of attention is well-established (Corbetta & Shul-man, 2002; Hopfinger, Buonocore, & Mangun, 2000). To recruit this

2 Speed and accuracy with which an object is detected are first briefly enhancedafter the object is attended, and then hindered. This hindrance has been termed‘inhibition of return’.

network, the present design varies the difficulty of target selectionusing small search arrays under two different perceptual load condi-tions. In addition, the present study takes advantage of the well-doc-umented attentional modulation of neural activity in visuo-perceptual areas such as MT/MST to compare distractor processingin action gamers and non-gamers (Rees, Frith, & Lavie, 1997).

Subjects were presented with a ring of shapes and asked to de-cide whether there was a square or a diamond among the shapespresented. On each trial, there could only be one target (either asquare or a diamond). By manipulating the homogeneity of theother shapes in the ring, two levels of difficulty were used (seeFig. 1). In the low load condition, all non-target shapes were circlesallowing the target to pop-out and thus be easily discriminated; inthe higher load condition, three different filler shapes were usedleading to a more heterogeneous display making the target dis-crimination more difficult. Under this high load condition, we ex-pected increased recruitment of fronto-parietal networks ascompared to the low load condition. Of interest was the differencebetween VGPs and NVGPs in recruiting this network as search dif-ficulty increased. Importantly, we selected rather easy search tasks(the low load effectively corresponds to a pop-out situation and thehigh load is just slightly more difficult) as we were aiming for rel-atively comparable increase in reaction times across groups fromlow to high attentional load. Indeed, while it is the case that VGPshave faster search rates than NVGPs (Hubert-Wallander et al.,2011), relatively matched increase in RTs between two levels ofdifficulty can still be found when using very easy searches. Byusing the low load condition as the baseline, any group differencesin BOLD signal between VGPs and NVGPs could then be attributedto their group status, rather than a significantly greater increase indifficulty from low to high load in one group and not the other.

Concurrent to this main search task, irrelevant patches of ran-dom dots (either moving or static) were presented to examine dis-tractor suppression. Previous work from Lavie and collaboratorshas shown that as the perceptual load of the main search task in-creases, distractors receive fewer processing resources, therebyresulting in smaller activation of MT/MST by irrelevant movingpatterns (Lavie, 2005; Rees, Frith, & Lavie, 1997). While this patternof results was predicted for both VGPs and NVGPs, the amount ofactivation in MT/MST triggered by irrelevant moving stimuli wasexpected to differ across populations. Greater attentional controlshould allow more efficient suppression of task-irrelevant motion(see for example, Mishra et al., 2011). By contrasting the neuralcorrelates of motion processing in MT/MST in VGPs and NVGPs,the present study allowed us to directly compare how much pro-cessing irrelevant distractors may undergo in each population.

2. Material and methods

2.1. Participants

Participants were 26 naïve males (18–26 years, mean age20.5 years) who were trained on the task prior to the scanning ses-sion. Participants were placed in one of two groups, video gameplayers (VGPs,n = 12) or non-video game players (NVGPs,n = 14),according to their responses to a questionnaire designed to estab-lish the frequency of action video game usage in the 12 monthsprior to testing. For each video game which participants reportedplaying, they were asked how often they had played that gamein the previous 12 months, and for how long they had played itduring a typical session. The criterion to be considered a VGPwas a minimum of 5 h per week (on average) of action video gameplay over the previous year. It is important to note that only expe-rience with action video games counted towards this requirement.Action video games are played from the first-person perspective

Fig. 1. Example of stimuli. Subjects had to indicate with a button press which of two targets (square or diamond) appeared in the ring of shapes while maintaining fixation onthe centre cross. In the low load condition the remaining seven shapes in the annulus were circles and in the high load condition the remaining shapes were a mixture ofcircles and other shapes (e.g., triangles, trapezoids, houses). The distractors, patches of moving or static dots, were placed on both the left and right, either inside (centraldistractors) or outside (peripheral distractors) the annulus of shapes.

134 D. Bavelier et al. / Vision Research 61 (2012) 132–143

and feature fast motion while requiring vigilant monitoring of theperiphery and simultaneous tracking of multiple objects, puttingdivided attention at a premium. An abridged list of the games re-ported as played by the VGP group includes Halo, Counterstrike,Gears of War, and Call of Duty. The criterion to be considered aNVGP was one or less hours per week of action video game playover the previous year. (Note that some NVGPs did play other kindsof games, such as board games, puzzle games, card games, strategygames or social games). All studies were performed with the in-formed consent of the participants and were approved by the Uni-versity of Rochester’s Research Subject Review Board.

2.2. Behavioral training prior to brain imaging

All participants were trained on the task in a one-hour session inthe week prior to their scanning session. The stimulus conditionsduring the training session were similar to those used in the scanner(described below). This training session was used to familiarize thesubjects with the experiment and ensure that they were performingabove 90% correct on the task before being scanned. Participantswere instructed to be as fast and as accurate as possible. Two ofthe 14 NVGPs did not meet our performance criteria, leaving12 NVGPs and 12 VGPs who were scanned. We note that this proce-dure of training all subjects on the task untill they achieved high per-formance could only weaken possible differences between groups.

2.3. MR image acquisition

Magnetic resonance images were acquired with a Siemens Trio3T MRI and a Siemens CP head coil. To minimize head motion andhelp reduce cumulative head drift during the scanning session,foam padding was used to support the head and neck.

Thirty-one T2�-weighted gradient echo (GE) echo-planar imag-ing (EPI) axial slices covering the entire brain were acquired every3 s (TE = 51 ms, flip angle = 90�, voxel dimension = 4 mm3, inter-leaved slices). One hundred measurements (time frames) were ac-quired for each run. Nine fMRI scans were performed for eachparticipant.

Three-dimensional, T1-weighted anatomical MR images (aMRI)were acquired in the same session. This aMRI was an MPRAGE se-quence (TR = 2020 ms, TE = 3.93 ms, flip angle = 12�, 256 � 256 ma-trix, 1 mm3 resolution).

2.4. Visual stimuli and procedure

The visual stimuli were generated in MATLAB on a MacintoshG4 running OS9, displayed using a JVC DLA-SXS21E projector andpresented on a rear-projection screen placed at the back of themagnet bore. Viewing distance was 0.8 m and the screen wasviewed using a mirror mounted above the eyes at an angle of �45�.

The stimulus was composed of eight shapes (each subtending�1�) presented along an annulus (5� radius – see Fig. 1). Partici-pants were asked to fixate on a centre cross and identify whethera square or diamond target was present in the annulus of shapes.Participants responded by pressing one of two buttons on anMR-compatible response box. Task difficulty was manipulated byincreasing the number of different shapes while keeping the num-ber of overall shapes constant (eight), so as to control for the num-ber of abrupt onsets in the visual display. We measured accuracyand reaction time during two levels of task difficulty (low/highload). Low load trials consisted of the target (square or diamond)with all other filler shapes being circles. High load trials consistedof the target, three different shapes selected randomly from a set of12 possibilities (e.g., triangles, trapezoids, houses – in various ori-entations), and four circles. The shapes were presented every 1 sfor 120 ms (inter-trial interval = 880 ms). They were light gray ona dark gray background (contrast 60%).

In the peripheral condition, the distractors (patches of movingor static dots) were positioned on both sides of the central fixationspot, outside the annulus of shapes. The dots in the distractorpatches were continuously presented and were alternately movingor static in 18 s intervals.

A central condition, identical to the peripheral one except thatthe distractors (patches of moving or static dots) were placed with-in the annulus of shapes, was also used (see Fig. 1). Of interest wasthe comparison between the processing or suppression of the cen-tral and peripheral distractors. Would these distractors be pro-cessed to the same extent and in the same way?

The middle of the central and peripheral distractor patcheswere equally spaced from the target annulus and the size of thepatches were scaled according to the nasal cortical magnificationfactor as calculated by Romano and Virsu (1979). Central distrac-tors were positioned 1.6� from fixation with a diameter of 1.8�,peripheral distractors were 8.4� from fixation with a diameter of4.6�.

D. Bavelier et al. / Vision Research 61 (2012) 132–143 135

Each functional scanning run consisted of both low and highload trials presented in a block design (block length = 36 s forone load level) with the distractors (patches of dots) alternatingfrom moving to static (or vice versa) every 18 s. The block orderas well as the initial moving or static state of the distractors foreach block was randomized. Each subject performed 8 functionalruns – 4 runs of peripheral distractors were intermixed with 4 runsof central distractors. See Fig. 2 for the time course of sample scan-ning runs.

We did not record eye movements, but instead trained partici-pants to maintain fixation on a central point. In the behaviouraltraining session performed outside the magnet prior to the scan-ning session, participants were instructed to fixate. Although theeccentric location of the shapes within the annulus could have trig-gered eye movements, the use of a search annulus with a centralfixation cross presented for only 120 ms ensured that subjectscould only perform the task well while fixating the central fixationcross. Along with the behavioral performance, we will see that thebrain imaging data show that the subjects were centrally fixated.Indeed, activity along the calcarine sulcus related to the distractors(central/peripheral) was as one might predict and only possible ifthe subjects were centrally fixating.

In the same scanning session, along with these eight functionalscanning runs, each subject also did a separate motion localizer runused to define his area MT/MST, which is sensitive to visual mo-tion. The stimulus for the motion localizer consisted of a full field(12� radius) of white dots on a black background (100% contrast).The dots were alternately moving (radially) or static in 18 s inter-vals. Subjects were asked to fixate the centre of the screen (a bluefixation dot) and were not required to make any response.

2.5. Behavioral data analysis

Reaction time (RT) and accuracy measures were collected dur-ing the scanning session. RTs from erroneous responses were notincluded in the behavioral analysis (erroneous responses includedincorrect responses as well as trials where RTs < 250 ms – of whichthere were fewer than 1%). There were no anomalously long RTs asthe maximum RT of 1000 ms was determined by the 1 s inter-trialinterval. A repeated measures 2 � 2 � 2 analysis of variance

Fig. 2. Sample scanning run. Each scanning run lasted 5 min and included eight, 36-s blothere were four blocks of the two load levels presented in a randomized block order. Dur18-s intervals. The moving or static state of the dots was also randomized within a blocdistractors. Run order was randomized across subjects.

(ANOVA) was performed between the groups (VGP/NVGP), withload level (low/high) and distractor eccentricity (central/periphe-ral) as within group variables.

2.6. MR image analysis

Image analysis was performed using tools from the FMRIB Soft-ware Library (FSL, version 4.0, FMRIB, Oxford, UK, www.fsl.ox.a-c.uk/fsl, see also Smith et al., 2004; Woolrich et al., 2009). Thefirst 4 time frames of each functional run were discarded to removeany start-up magnetization transients in the data. The followingpreprocessing techniques were applied: motion correction usingMCFLIRT (no participant moved more than 2 mm in any directionand rotations were less than 1.3�) (Jenkinson et al., 2002); field-map-based EPI unwarping using PRELUDE + FUGUE (Jenkinson,2003; Jenkinson, Wilson, & Jezzard, 2004); slice-timing correctionusing Fourier-space time-series phase-shifting; non-brain removalusing BET (Smith, 2002); spatial smoothing using an isotropic 3DGaussian kernel (full-width-half-maximum = 5 mm) to attenuatehigh frequency noise; grand mean-based intensity normalizationof all volumes by the same factor; and nonlinear high-pass tempo-ral filtering with a 50 s cut-off. Statistical analyses were then car-ried out using FEAT 5.63, the FMRI Expert Analysis Tool.

2.6.1. Single subject analysesFor each run within each subject, we first used FILM (FMRIB’s

Improvised Linear Model) based on a general linear model (GLM)with prewhitening for correlated errors (Woolrich et al., 2001).Regressors or explanatory variables (EVs) were used to model thefollowing three conditions in the first level model: (i) low load withmoving distractors (MotionLow), (ii) high load with moving dis-tractors (MotionHigh), and (iii) high load with static distractors(StaticHigh). The condition of low load with static distractors (Sta-tic-Low) was used as the baseline. To identify the brain areas acti-vated under different conditions, a set of contrasts were derivedfrom the EVs. Co-efficients were computed for each contrast forthe four peripheral and four central runs for each subject. Registra-tion to high-resolution and/or standard images (MNI-152 tem-plate) was carried out using FLIRT (Jenkinson & Smith, 2001).

cks as well as an initial dummy block of 12 s (not depicted in the figure). In each runing each block the distractors (patches of dots) were alternately moving or static ink. There were eight scanning runs: 4 with central distractors and 4 with peripheral

Fig. 3. Behavioral results. RT data plotted by load level with % correct data noted on each bar for the peripheral distractor and the central distractor conditions respectively.There was no significant difference across load level or distractor eccentricity in terms of accuracy (percentages at the base of the histograms), but there was a main effect ofload on reaction time. In both experiments, the high load condition produced longer RTs. Importantly, VGPs and NVGPs were comparably slowed down from low to high load,indicating similar increase in difficulty across groups.

136 D. Bavelier et al. / Vision Research 61 (2012) 132–143

Then, the four peripheral runs within a session for a single sub-ject were combined using general linear model with fixed effects(likewise for the four central runs) for the same set of contrasts.The co-efficient of contrasts was computed using the fixed effectsmodel by forcing the random effects variance to zero with FLAME(FMRIB’s Local Analysis of Mixed Effects) (Beckmann, Jenkinson, &Smith, 2003; Woolrich et al., 2004). Z (Gaussianized T/F) statisticimages were thresholded using clusters determined by Z > 2.3and a corrected cluster significance threshold of p = 0.05 (Worsleyet al., 1992).

2.6.2. Whole-brain analysesWe first performed within-group, whole brain analyses by pool-

ing the subjects into two groups, NVGPs and VGPs. For each group,we computed the difference in brain activations between high loadand low load by defining a contrast of [MotionHigh � Static-Low] � [MotionLow� StaticLow]. The co-efficients of the contrastcomputed for the central and peripheral runs in the single subjectanalyses were used as inputs to study the load difference effect with-in each group. Data were modeled with mixed effects using FLAMEstage 1 (Beckmann, Jenkinson, & Smith, 2003; Woolrich et al.,2004). Z (Gaussianised T/F) statistic images were thresholded usingclusters determined by Z > 3.0 and a (corrected) cluster significancethreshold of p = 0.05 (Worsley et al., 1992) for each of the group.

We then performed between-group, whole brain analyses toidentify those brain areas where high versus low load differencesdiffer across groups. The inputs were the co-efficients of contrastcomputed using the central and peripheral runs from each groupas described in the paragraph above. Again the analysis was carriedout using a mixed effects model with FLAME stage 1. Z (Gaussian-ised T/F). Statistic images were thresholded using clusters deter-mined by Z > 2.0 and a (corrected) cluster significance thresholdof p = 0.05 (Worsley et al., 1992). Here contrast masking was usedto limit group differences to those areas showing significant differ-ences for HighLoad versus LowLoad in each of the groups.

The activations were quite extensive for the high versus lowload conditions both within and between groups, with clusters of-ten spanning multiple anatomically and functionally distinct re-gions. To decompose these large clusters of activation, we firstidentified all anatomical regions activated in these comparisons,then extracted the results from anatomically-defined regions ofinterest (ROIs) based on the work of Tzourio-Mazoyer et al.(2002). The boundaries of these areas were slightly distorted toavoid attributing activation at the fringe of a well-delineatedcluster to structures other than the main center of mass of the

cluster. ROIs from each hemisphere included frontal areas – thesuperior frontal sulcus encompassing the frontal eye field, the mid-dle frontal gyrus, the inferior frontal gyrus, the supplementary mo-tor area, and the dorsal anteriors cingulate cortex; parietal areas –superior parietal cortex including the dorsal part of the intra-pari-etal sulcus, the intra-parietal sulcus proper, and the cuneus andprecuneus; occipital areas – superior, middle and inferior occipitalcortices, as well as the cerebellum and basal ganglia structures (seeTables 1 and 2).

Finally, the contrast between central and peripheral distractorconditions were computed separately for VGPs and NVGPs. Con-junction analyses were then performed to determine areas com-mon to these contrasts in VGPs and NVGPs by multiplyingbinarized versions of the thresholded statistical maps. This analysiswas used to verify our analyses as central and peripheral distrac-tors are known to evoke activation at different locations in the vi-sual cortices. Z (Gaussianized T/F) statistic images werethresholded using Z > 3 and a corrected cluster significance thresh-old of p = 0.05 (Worsley et al., 1992).

2.6.3. Processing of motion distractors – regions of interest analysisGiven our design, brain activation differences in area MT/MST

were of particular interest. To allow for individual variation inthe location and magnitude of response, each subject’s MT/MSTwas functionally defined using his motion localizer scan and anMT/MST region-of-interest (ROI) was drawn based on the func-tional data and known localization of MT/MST. These ROIs werethen applied to each subject’s signal change images. Percent BOLDsignal change was computed using StaticLow as a baseline for allcontrasts of interest, resulting in two main contrasts for low loadand high load, respectively: (1) MotionLow � StaticLow and (2)[MotionHigh � StaticLow] � [StaticHigh � StaticLow]. PercentBOLD signal change for the different conditions were extractedon an individual subject basis and used as the dependent measurein the statistical analyses reported below.

3. Results

3.1. Behavioral results

A 2 � 2 � 2 ANOVA with distractor eccentricity (central/periph-eral) and load (low/high) as within-subject variables and group(VGP/NVGP) as the between subject variable was carried out onthe percent correct data. The only significant effect was an

Table 1Regions showing greater activation for high load than low load in NVGPs and VGPs. SFS: Superior Frontal Sulcus, SMA: Supplementary Motor Area, IPS: Intraparietal Sulcus. ‘L’ and‘R’ stand for Left and Right respectively. For this and all other tables, coordinates are given in the Montreal Neurological Institute stereotaxic space.

ROI NVGP VGP

X Y Z Max Z Vol. (mm3) X Y Z Max Z Vol. (mm3)

FrontalSFS

L �27 �7 55 4.60 2840R

Middle frontalLR 48 8 29 4.64 5872

SMAL �3 7 51 4.94 2288R 6 6 54 4.79 2440

Dorsal anterior cingulateL �6 13 38 3.71 352R 8 14 38 4.32 952

ParietalSuperior parietal/dorsal IPS

L �23 �65 51 5.27 7248 �22 �61 51 4.41 4256R 24 �64 55 4.87 5128 24 �61 54 4.37 2024

IPSL �35 �51 45 5.24 6032 �30 �54 48 4.16 1744R 36 �50 47 4.96 3848 28 �57 48 4.27 776

Cuneus/precuneusL �12 �68 56 4.07 1152 �14 �61 50 3.11 48R 16 �70 49 4.01 1024 18 �65 45 3.39 72

OccipitalSuperior occipital

L �22 �76 34 5.04 1888 �24 �73 32 4.32 448R 26 �71 37 4.98 2976 26 �66 42 4.01 648

Middle occipitalL �28 �76 23 5.55 4432 �31 �78 19 5.18 5288R 32 �77 24 5.14 3464 34 �77 21 4.38 1784

Inferior occipital/middle temporalL �45 �65 �10 4.71 3512 �42 �72 10 4.90 3568R

CerebellumCerebellum

L 5 �76 25 4.09 776R 83 �76 25 4.48 1592

D. Bavelier et al. / Vision Research 61 (2012) 132–143 137

interaction between eccentricity and group [F(1,22) = 11.8,p < 0.002, partial eta squared = 0.36] as NVGPs were slightly lessaccurate under the peripheral condition whereas VGPs wereequally accurate under both conditions (NVGP_Peripheral = 95.2%,NVGP_Central = 96.15%; VGP_Peripheral = 96.5%; VGP_Cen-tral = 96.05% – Fig. 3). All other p’s > 0.18.

A similar 2 � 2 � 2 ANOVA with reaction times as the dependentvariable was carried out. Main effects of load [High Load = 574 ms,Low Load = 506 ms, F(1,22) = 276.4, p < .0001, partial etasquared = .93] and of group [VGPs = 514 ms, NVGPs = 566 ms,F(1,22) = 10.8, p < 0.003, partial eta squared = .33] confirmed fasterRTs in the low load than in the high load condition as well as fasterRTs in VGPs as compared to NVGPs (Fig. 3). In addition a triple inter-action between group, load and eccentricity [F(1,22) = 8.9, p < 0.007,partial eta squared = 0.29] was significant, indicating differential ef-fects of load and group as a function of eccentricity (Fig. 3). The costof going from low to high load displays was slightly greater underperipheral motion than central motion for NVGPs, whereas VGPsexhibited the opposite pattern. The source of this triple interactionremains unclear; as it was not predicted, it would have to be furtherconfirmed before being interpreted. When considered along withthe accuracy data, the overall pattern of results suggests that irrele-vant motion in the visual periphery is more disrupting in NVGPs

than VGPs. None of the other effects were significant; in particularthere was no interaction between load and group (p > .8), eccentric-ity and group (p > .9) or load by eccentricity (p > .7).

In sum, we found faster RTs in VGPs than NVGPs in the face ofcomparable accuracy replicating past reports on how action gameplay affects speed and accuracy (Dye, Green, & Bavelier, 2009a;Green, Pouget, & Bavelier, 2010). In addition, increasing the searchdifficulty from the low to high load displays increased reactiontimes by about 70 ms in both VGPs and NVGPs indicating equal in-crease in difficulty from low to high load in the two populationscompared. Equivalent increases in reaction times from low to highload is an important characteristic of this study that asks how VGPsand NVGPs may differ as attentional difficulty increases. Indeed, inthe face of a similar change in RTs from low to high load in eachgroup, differences between groups are more likely to arise fromgroup status rather than just a lesser increase in task difficultyfor the VGP group.

3.2. Impact of load increase on attentional networks – fMRI wholebrain analysis

The whole brain analyses looked for differences in brain activa-tion as load was increased from low to high between VGPs and

Table 2Between-group analyses revealed overall greater activation in NVGPs than in VGPs for the high versus low load contrast. No areas showed greater activation for VGPs than NVGPs.

ROI NVGP > VGP VGP > NVGP

X Y Z Max Z Vol. (mm3) X Y Z Max Z Vol. (mm3)

FrontalSFS

L �27 �8 55 4.26 2776R 33 �10 57 3.08 1904

Middle frontalLR 48 7 30 3.74 4904

IFG/insulaLR 41 20 8 3.44 1024

Frontal poleLR 42 37 14 2.74 736

SMAL �4 4 51 4.09 1800R 6 3 55 3.54 2360

Dorsal anterior cingulateL �6 13 38 3.53 728R 5 8 43 3.21 224

ParietalSuperior parietal/dorsal IPS

LR 10 �72 52 2.86 864

Cuneus/precuneusL �9 �57 50 2.03 16R 14 �69 50 3.08 712

OccipitalSuperior occipital

LR 25 �73 38 2.91 1528

Middle occipitalLR 29 �76 28 2.93 344

SubcorticalPutamen/insula

LR 30 13 4 2.93 192

Fig. 4. Pattern of activation as attentional load is increased for NVGPs (in green, a)and for VGPs (in blue, b). VGPs show much reduced recruitment of the fronto-parietal network as compared to NVGPs (see p. 5 for statistical parameters).

138 D. Bavelier et al. / Vision Research 61 (2012) 132–143

NVGPs. This effect was studied by asking which brain structureschange their activation level under the high load condition usingthe low load condition as a baseline. These analyses were first per-formed separately for VGPs and NVGPs; then between-group anal-yses were performed to characterize those brain regions that differbetween groups. Although separate analyses were performed forperipheral and central distractor runs, the patterns of brain activitywere very similar in the two groups across distractor eccentricities.

Therefore, the effect of load will be presented averaged across cen-tral and peripheral distractors runs.

As expected, NVGPs showed activation in a network of fronto-parietal areas as load increased (Table 1). This network includedactivation bilaterally in the superior frontal and inferior frontalareas, the pre-central and post-central gyri as well as the SMA(supplementary motor area). Importantly, a large activation wasnoted in the dorsal anterior cingulate. Thus, both midline and lat-eral frontal areas showed greater recruitment as task difficulty wasincreased. Parietal activations were seen bilaterally in the inferiorparietal cortex, the superior parietal cortex extending medially tothe cuneus/precuneus. Finally, marked activation was noted in vi-sual areas including superior and middle occipital areas bilaterally,as well as along the left inferior and middle temporal gyri.Although a similar network of areas was recruited in VGPs as taskdifficulty increased (Table 1), the recruitment of the fronto-parietalnetwork was much less marked. Of note, there was no significantactivation in frontal areas (medial or lateral, see Fig. 4). Bilateralparietal activation was restricted to a smaller region in the inferiorand superior parietal lobules. The bulk of the activation in the VGPswas limited to visual areas including superior and middle occipitalgyri bilaterally, and the left inferior temporal gyrus.

Importantly, between-group analyses confirmed significantlygreater recruitment of the fronto-parietal network in NVGPs than

Fig. 5. Greater activation for NVGPs than VGPs was noted as load increased in areasof the fronto-parietal network (a). Difference in percent bold changes between highand low load is plotted for five regions of interest in the attentional network systemin NVGPs and in VGPs (b). SFS = Superior Frontal Sulcus, MFG = Middle FrontalGyrus, IFG = Inferior Frontal Gyrus, Cing = Cingulum, IPS = Intraparietal Sulcus.

Fig. 6. Pattern of activation for central versus peripheral distractors computedthrough a conjunction analysis of VGPs and NVGPs activation maps (see p. 5, topparagraph for statistical parameters). Across groups, greater activation for centraldistractors was noted more posteriorly along the calcarine sulcus, whereas greateractivation for peripheral distractors was observed more anteriorly, as predicted bythe known retinotopic organization of early visual areas.

D. Bavelier et al. / Vision Research 61 (2012) 132–143 139

in VGPs as task difficulty increased (Table 2; Fig. 5). This differencewas especially marked in frontal areas including the superior fron-tal cortex, inferior and middle frontal gyri, as well as the SMA, andthe dorsal anterior cingulate cortex. Greater activation in NVGPswas also noted in parietal areas and especially the right superiorparietal lobe and its extension to the right cuneus and precuneus.Finally, visual areas (occipital lobe) themselves were also more ac-tive in NVGPs than VGPs as illustrated by the significantly greateractivation in right superior and middle occipital gyri. Greater acti-vation in NVGPs was also noted in the right insula and the rightputamen.

3.3. Impact of eccentricity on recruitment of visual areas – fMRI wholebrain analyses

We performed analyses to confirm the predicted pattern of re-sults as a function of distractor eccentricity. A conjunction analysisof VGP and NVGP data for the contrast between the central versusthe peripheral moving distractor conditions confirmed a very sim-ilar pattern of results in both groups. Notably, activation along thecalcarine fissure was observed and showed a more posterior focusfor central than peripheral distractor conditions as expected(Fig. 6). Activation was also noted bilaterally in the lingual gyrus.In the case of the central distractors, the activation also extendedlaterally covering part of the middle and inferior occipital sulcusin both the left and right hemisphere. These analyses confirmedall expected patterns of activation given the known retinotopicorganization of visual cortices.

3.4. Processing of motion distractors – region of interest results

Given our research question, we also characterized how theprocessing of irrelevant motion was altered from low to high loadin VGPs and NVGPs. To do so, we turned to a region-of-interestanalysis, known to have greater sensitivity than whole brain anal-

yses. Indeed, the latter requires normalizing all brains into a com-mon template, potentially muddying boundaries of areasfunctionally defined such as MT/MST.

This analysis focused on MT/MST to assess how the irrelevantmotion was processed as a function of load and group. A2 � 2 � 2 ANOVA using percent change in MT/MST as the depen-dent variable showed a main effect of distractor eccentricity[F(1,22) = 31.58, p < 0.0001, partial eta squared = 0.59] due to low-er activation from peripheral than central motion, and a near sig-nificant effect of load [F(1,22) = 4.13, p = 0.054, partial etasquared = 0.16] reflecting greater activation under low than highload. In addition, eccentricity interacted with load [F(1,22) = 6.06,p = 0.022, partial eta squared = 0.22] as well as with group, albeitweakly so [F(1,22) = 3.53, p < 0.075, partial eta squared = 0.14].There was a larger effect of load when motion distractors were pre-sented peripherally compared to centrally, as well as a larger groupdifference under central than peripheral motion distractors (Fig. 7).None of the other effects were significant, all p’s > .1. The differenteffects of load as a function of eccentricity led us to carry out sep-arate 2 � 2 analyses for central and peripheral moving distractors.

For peripheral motion, a main effect of load was observed[F(1,22) = 7.98, p = 0.01, partial eta squared = 0.27] replicating thework of Rees, Frith, and Lavie (1997). No other effect was signifi-cant (all p’s > 0.5). For central motion, a different pattern of resultswas observed. A main effect of group was observed [F(1,22) = 4.55,p = 0.044, partial eta squared = 0.17] reflecting lower activation inVGPs than NVGPs, but no other effects were present (allp’s > 0.8), and in particular there was no main effect of load.

Importantly, analysis of motion localizer runs indicated no dif-ference across groups (mean percent BOLD signal change and stan-dard deviation, VGPs = 0.98%, ±0.38; NVGPs = 1.05%, ±0.30,t(21) = �0.48, p = 0.64). Thus, it is not the case that VGPs alwaysshow reduced activation in MT/MST as compared to NVGPs whenviewing a moving stimulus (motion localizer stimulus).

4. General discussion

4.1. Summary

The present study was designed to compare the neural net-works underlying attentional processing in VGPs and NVGPs. Inparticular we aimed at identifying the neural networks recruitedas attentional load is increased and at characterizing the fate of dis-tractors under different attentional loads in these two groups. Forthis purpose, a visual search paradigm was used alternating be-tween an easy and a more attentionally demanding task, while dis-tractors, either static or moving random-dot displays, appeared inthe visual field.

Fig. 7. Percent BOLD signal change in area MT/MST as a function of load and the eccentricity (a-peripheral distractors, b-central distractors) of the irrelevant motion patch.

140 D. Bavelier et al. / Vision Research 61 (2012) 132–143

VGPs were faster at performing the search tasks than NVGPs rep-licating previous reports in the literature. To more cleanly isolategroup effects as attentional demands increase, the two levels ofsearch tasks used in this study were selected to be relatively easyto lead to comparable increase in RTs from low to high load in eachpopulation. Indeed, although VGPs show faster search rates thanNVGPs (Hubert-Wallander et al., 2011), this difference is difficultto capture when only considering easy searches. Therefore, by usingthe low load condition as our baseline, we could compare therecruitment of the fronto-parietal attentional network in VGPs andNVGPs as attentional demands increased reaction times by about70 ms in each group. Despite this matched increase in attentionaldifficulty, we observed a significantly lesser recruitment of the fron-to-parietal attentional network in VGPs as compared to NVGPs.

In a separate analysis focused on the motion specific area MT/MST, we evaluated the fate of alternatively moving and static dis-tractors as a function of attentional load in each population. Over-all, we found that irrelevant motion leads to lesser activation inVGPs than NVGPs. We discuss these results in more details below.

4.2. Processing of irrelevant motion distractors in gamers and non-gamers

4.2.1. Group effectThis study takes advantage of the perceptual load paradigm to

measure the fate of irrelevant and unattended distractors acrosseccentricity in different populations. Alternations of static and mov-ing stimuli, although irrelevant to the task, led to activation in MT/MST. Overall, VGPs showed less recruitment of MT/MST than NVGPsduring task performance. It is important to note that VGPs showedlesser recruitment of MT/MST than NVGPs during task performance,whereas no difference in activation was noted during the MT/MSTlocalizer. This confirms that the decreased activation in VGPs doesnot reflect a generalized baseline difference between the two popu-lations. Lesser activation in VGPs suggests that they may suppressirrelevant motion distractors more efficiently than NVGPs. Bettersuppression of distracting information in VGPs has been reported re-cently using steady-state evoked potential and a very different de-sign (Mishra et al., 2011). Thus, efficient distractor suppressionmay be a common mechanism that contributes to the superiorattentional capabilities of VGPs.

One may ask how these results relate to the proposal in the lit-erature that VGPs may benefit from greater attentional resources(Green & Bavelier, 2003). As proposed by Lavie and collaborators(Lavie, 2005), irrelevant, peripheral moving distractors tend to pro-duce greater MT/MST activation when they receive more atten-tional resources. This could have led one to expect greaterrecruitment of MT/MST in VGPs, rather than lesser recruitment

as observed here. Indeed, a number of behavioral studies have re-ported that distractors typically receive more processing resourcesin VGPs as exemplified by a greater impact of distractor identity onthe main task reaction times in VGPs (Dye, Green, & Bavelier,2009b; Green & Bavelier, 2006a). The present study cannot directlyaddress this issue, however, since the motion distractors were nottask-relevant, and thus could not compete for responses. This pre-vented the quantification of distractor identity on reaction times.Further research will be needed to clarify this point.

4.2.2. Eccentricity effectsAs initially shown by Lavie and collaborators (Lavie, 2005), irrel-

evant, peripheral moving distractors produced greater MT/MSTactivation when the perceptual load of the target task was low ascompared to high. A common interpretation of this effect is thata low perceptual load task does not exhaust all attentional re-sources, allowing some to spread to irrelevant distractors in theperiphery. The present study replicates this peripheral distractoreffect, which is observed in both VGPs and NVGPs. In addition,Fig. 7a illustrates that the high perceptual load task leads to equalactivation of MT/MST in all participants, whereas at low load, a(non-significant) trend can be seen for greater activation in NVGPsthan in VGPs. This pattern also mirrors a recent report by Forsterand Lavie (2007) in which these authors document that a sureway to equate performance across groups that differ in their re-sources is to use a high perceptual load for all.

Notably, the use of central distractors (as compared to peripheraldistractors) led to a different pattern of results. First, central motionled to an overall greater level of activation in MT/MST than periph-eral motion. Although the stimuli were corrected for cortical magni-fication, this was to be expected. Central motion typically elicitsgreater response in MT/MST than peripheral motion (Bavelieret al., 2000; Beauchamp, Cox, & DeYoe, 1997), and there is ample evi-dence that foveal and para-foveal vision is functionally unique, bothbecause of its anatomical organization and of its connectivity (Mas-sey, 2006; Ruff et al., 2006). Accordingly, when using distractors thatare task-relevant and thus either compatible or incompatible withthe target response, greater compatibility effects are typically ob-served from central as compared to peripheral distractors. This ac-cords with our result of greater activation in MT/MST for centralrather than peripheral distractors. Second, there was no effect ofload on MT/MST activation when central distractors were used.Thus, unlike the condition with peripheral distractors, activationtriggered by central distractors was not modulated by the resourcesconsumed by the primary task. This finding was unexpected. Severalbehavioral studies have established that varying the perceptual loadmodulates the size of the compatibility effect for both peripheral aswell as central distractors (Beck & Lavie, 2005; Green & Bavelier,

D. Bavelier et al. / Vision Research 61 (2012) 132–143 141

2006a; Proksch & Bavelier, 2002). In the present study, the periphe-ral condition produces the expected load effect on MT/MST activa-tion in both populations, and the only difference between thecentral and peripheral conditions is the eccentricity of the distrac-tors. Thus, the lack of modulation by perceptual load for the centralcondition is unlikely to reflect a problem with the load manipulationor an oddity of the participants studied. It may be that peripheral re-sources deplete before central ones, and that given an even morechallenging target task, a modulation of central distractors by per-ceptual load would also be observed.

4.3. Attentional networks in action gamers and non-gamers

As the perceptual load of the search task was increased, thefronto-parietal network of areas known to be involved in the allo-cation and control of attention was robustly activated. Perceptualload in this study was manipulated by changing the salience ofthe target stimulus, but not the number of stimuli presented. In-deed, the number of search elements was kept constant ensuringan equal amount of sensory stimulation across loads, but the het-erogeneity of the search array was increased so as to render thesearch for the target more difficult in the high load condition (Dun-can & Humphreys, 1989). This manipulation successfully changedthe difficulty of the task as exemplified by longer RTs in the highload condition. Importantly, this manipulation lengthened RTs inNVGPs and VGPs by a similar amount (about 70 ms) suggestingan equivalent increment in difficulty between low and high loadsacross groups (VGPs – Low Load = 480 ms, High Load = 548 ms;NVGPs – Low Load = 531 ms, High Load = 601 ms). Despite thisbehavioral similarity, marked differences in brain activation werenoted across groups as load increased.

In accordance with the known brain networks of attention,strong activation was noted in both goal-directed and stimulus-driven attentional systems. In particular, NVGPs exhibited strongbilateral recruitment of the superior frontal sulcus as well as pari-etal areas along the intra-parietal sulcus and the dorsal anteriorcingulate gyrus. Recent studies suggest the anterior cingulategyrus to be involved in stimulus driven shifts of attention andselective target processing (Hopfinger, Buonocore, & Mangun,2000; Shulman et al., 2009, 2010). Moreover, its fundamental rolein cognitive control has been observed in a variety of studies(Braver & Barch, 2006; Dosenbach et al., 2006; Schulz et al.,2011). While this network is typical of goal-directed attentionalcontrol, marked activation was also noted in right frontal areasincluding the middle and inferior frontal gyri which have beenassociated with sensory salience and its filtering (Corbetta, Patel,& Shulman, 2008). Crucially, this fronto-parietal network of areaswas much less recruited in VGPs who exhibited reduced activationthroughout all frontal and parietal areas (Fig. 5). This was reflectedin the between-group statistics showing that no region was moreactivated in VGPs than NVGPs as load increased. In contrast, signif-icantly greater activation was noted in NVGPs throughout the net-work of areas considered as load was increased.

Lesser activation in VGPs is consistent with the proposal thatVGPs develop more efficient attentional processes as a result of theirgaming activity, allowing them to allocate attention in a less effort-ful manner (Hubert-Wallander et al., 2010). We acknowledge thatthe present work only contrasts VGPs and NVGPs, and thus doesnot directly establish a causal effect of action video game play onthe reduced attentional network recruitment noted in VGPs. It isworth noting, however, that the point of departure for this study in-cluded several different training studies that established a causal ef-fect of action game play on visuo-spatial selective attention, asengaged in the present fMRI experimental design (Green & Bavelier,2003, 2006a, 2007). The present aim was to investigate the neuralbases of this attentional enhancement. Future training studies will

certainly be valuable in consolidating that link. Yet, by characteriz-ing the neural mechanisms by which selective attention enhance-ment is attained in VGPs (who are likely to have experienced morethan the limited training regimen employed by prior training stud-ies), this work documents how a typical action gamer’s attentionalsystem ends up benefiting from his/her action game play.

A working hypothesis for future work is that lesser recruitmentof attention-related areas is a signature of greater attentional con-trol. A similar proposal was advanced by Brefczynski-Lewis et al.(2007) in a study of the neural bases of meditation, a state knownto enhance attention regulation (Jha, Krompinger, & Baime, 2007;Lutz et al., 2008; Tang & Posner, 2009). In the case of the presentstudy, one could object and argue for an alternative accountwhereby the lesser recruitment of fronto-parietal areas in VGPsas load increased may have been due to the fact that the high loadcondition was easier for VGPs than NVGPs. It is correct that abso-lute level of difficulty was not matched across subjects – VGPswere faster than NVGPs, a now well-established signature of per-formance in the VGPs group (Dye, Green, & Bavelier, 2009a; Green,Pouget, & Bavelier, 2010). However, the use of the low load condi-tion as a baseline in the analyses and the focus of our analyses onthe contrast between high and low load should protect against thisalternative account. Indeed, the present design ensured that the in-crease in difficulty between low and high load was matched acrossgroups. Thus although VGPs were faster overall than NVGPs, thetwo groups were similarly slowed down by the change in percep-tual load from low to high. This comparable change in behavioracross loads should have led to a comparable change in recruit-ment of attentional network; yet it did not, supporting the pro-posal of a change in attentional efficiency.

Preliminary functional connectivity analyses of the fronto-parie-tal network provide further support for this view (See Supplemen-tary Data). Seeding from parietal areas revealed no majordifferences in functional connectivity between VGPs and NVGPs.However, seeding from frontal areas (e.g., dorsal anterior cingulateand right middle frontal gyrus) revealed enhanced functional con-nectivity in VGPs to a distinctive network of areas. A largely overlap-ping network of areas was observed to be functionally connected tothe dorsal anterior cingulate and middle frontral gyrus in VGPs andNVGPs. These included, in addition to the two seed areas, the supe-rior parietal cortex, the supra-marginal gyrus, the SMA, the pre-cen-tral gyrus, the insular cortex, and interestingly the anteriorprefrontal cortex. Along with the superior parietal cortex, the insulaand the pre-central gyrus, this anterior prefrontal area showed sig-nificantly greater connectivity with the anterior cingulate and themiddle frontal gyrus in VGPs than in NVGPs. Higher regulatory cog-nitive functions have been assigned to the anterior prefrontal cortex(Gilbert et al., 2006; Vincent et al., 2008 for an overview). Koechlinet al. (1999), for example, observed activation in this area when par-ticipants were instructed to achieve a main goal that required per-forming several subgoals along the way. The view that the anteriorprefrontal cortex is recruited when participants have to monitormultiple task sets and switch among them is further supported bystudies of problem solving (Ramnani & Owen, 2004), decision mak-ing (Vincent et al., 2008) or task-set switching (Braver, Reynolds, &Donaldson, 2003). Whether greater efficiency of this network mayaccount for some of the selective attention enhancement noted ingamers should be a fruitful avenue of research to understand themechanisms by which attention and executive functions may be en-hanced in future studies.

Overall, the results are consistent with the proposal that en-hanced attentional skills in VGPs may proceed through an autom-atization of the resource allocation process, resulting in lesserrecruitment of the fronto-parietal network that mediates suchattention allocation. The view that automatization of processingresults in diminished cortical recruitment is echoed in the

142 D. Bavelier et al. / Vision Research 61 (2012) 132–143

literature across domains including those of motor, verbal and per-ceptual learning as well as learning at a more executive level(Beauchamp et al., 2003; Erickson et al., 2007; Poldrack et al.,2005; Puttemans, Wenderoth, & Swinnen, 2005; Raichle et al.,1994; Shadmehr & Holcomb, 1997; see Clare Kelly and Garavan(2005) for a review). This is not to say that VGPs would not engagefronto-parietal networks of attention under any circumstance.Rather the working hypothesis is that it would take a much greaterburden of task difficulty before they do so. This view is consistentwith the behavioral literature on action gamers that documentsenhanced performance in tasks that require primarily efficientand flexible allocation of attentional resources (Hubert-Wallanderet al., 2010) and indicate that such behavioral enhancement maybe mediated through greater automatization of resource allocationand in turn more efficient suppression of irrelevant or distractinginformation in VGPs.

Acknowledgments

We thank S. Hillyard and J. Mishra-Ramanathan for helpfulcomments on the manuscript. We are indebted to Ted Jacquesfor help with manuscript and figure preparation, and Daniel Colefor help with Table preparation. This work was supported by Na-tional Institutes of Health Grant EY016880 and the Office of NavalResearch Grant N00014-07-1-0937.3 to D.B., as well as award P30EY001319 from the National Eye Institute. The content of the man-uscript is solely the responsibility of the authors and does not nec-essarily represent the official views of the National Eye Institute orthe National Institutes of Health.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.visres.2011.08.007.

References

Bavelier, D., Tomann, A., Hutton, C., Mitchell, T., Liu, G., Corina, D., et al. (2000).Visual attention to periphery is enhanced in congenitally deaf individuals.Journal of Neuroscience, 20, 1–6.

Beauchamp, M., Cox, R., & DeYoe, E. (1997). Graded effects of spatial and featuralattention on human area MT and associated motion processing areas. Journal ofNeurophysiology, 78(1), 516–520.

Beauchamp, M. H., Dagher, A., Aston, J. A. D., & Doyon, J. (2003). Dynamic functionalchanges associated with cognitive skill learning of an adapted version of theTower of London task. NeuroImage, 20(3), 1649–1660.

Beck, D. M., & Lavie, N. (2005). Look here but ignore what you see: Effects ofdistractors at fixation. Journal of Experimental Psychology: Human Perception andPerformance, 31(3), 592–607.

Beckmann, C., Jenkinson, M., & Smith, S. M. (2003). General multi-level linearmodelling for group analysis in FMRI. NeuroImage, 20, 1052–1063.

Braver, T. S., & Barch, D. M. (2006). Extracting core components of cognitive control.Trends in Cognitive Sciences, 10(12), 529–532.

Braver, T. S., Reynolds, J. R., & Donaldson, D. I. (2003). Neural mechanisms oftransient and sustained cognitive control during task switching. Neuron, 39,713–726.

Brefczynski-Lewis, J. A., Lutz, A., Schaefer, H. S., Levinson, D. B., & Davidson, R. J.(2007). Neural correlates of attentional expertise in long-term meditationpractitioners. Proceedings of the National Academy of Sciences of the United Statesof America, 104(27), 11483–11488.

Carrasco, M., & Yeshurun, Y. (1998). The contribution of covert attention to the set-size and eccentricity effects in visual search. Journal of Experimental Psychology:Human Perception and Performance, 24(2), 673–692.

Castel, A. D., Pratt, J., & Drummond, E. (2005). The effects of action video gameexperience on the time course of inhibition of return and the efficiency of visualsearch. Acta Psychologica, 119, 217–230.

Chisholm, J. D., Hickey, C., Theeuwes, J., & Kingstone, A. (2010). Reduced attentionalcapture in action video game players. Attention, Perception, & Psychophysics,72(3), 667–671.

Clapp, W. C., Rubens, M. T., Sabharwal, J., & Gazzaley, A. (2011). Deficit in switchingbetween functional brain networks underlies the impact of multitasking onworking memory in older adults. Proceedings of the National Academy of Sciences,108(17), 7212–7217.

Clare Kelly, A. M., & Garavan, H. (2005). Human functional neuroimaging of brainchanges associated with practice. Cerebral Cortex, 15, 1089–1102.

Cohen, J. E., Green, C. S., & Bavelier, D. (2007). Training visual attention with videogames: Not all games are created equal. In H. O’Neil & R. Perez (Eds.), Computergames and adult learning (pp. 205–227). Elsevier.

Corbetta, M., Patel, G., & Shulman, G. L. (2008). The reorienting system of the humanbrain: From environment to theory of mind. Neuron, 58(3), 306–324.

Corbetta, M., & Shulman, G. (2002). Control of goal-directed and stimulus-drivenattention in the brain. Nature Reviews in Neuroscience, 3(3), 201–215.

Dosenbach, N. U. F., Visscher, K. M., Palmer, E. D., Miezin, F. M., Wenger, K. K., Kang,H. C., et al. (2006). A core system for the implementation of set tasks. Neuron, 50,799–812.

Duncan, J., & Humphreys, G. W. (1989). Visual search and stimulus similarity.Psychological Review, 96, 433–458.

Dye, M. W., Green, C. S., & Bavelier, D. (2009a). Increasing speed of processing withaction video games. Current Directions in Psychological Science, 18(6), 321–326.

Dye, M. W. G., & Bavelier, D. (2010). Differential development of visual attentionskills in school-age children. Vision Research, 50(4), 452–459.

Dye, M. W. G., Green, C. S., & Bavelier, D. (2009b). The development of attentionskills in action video game players. Neuropsychologia, 47, 1780–1789.

Eckstein, M., Pham, B. T., & Shimozaki, S. S. (2004). The footprints of visual attentionduring search with 100% valid and 100% invalid cues. Vision Research, 44,1193–1207.

Erickson, K. I., Colcombe, S. J., Wadhwa, R., Bherer, L., Peterson, M. S., Scalf, P. E., et al.(2007). Training-induced functional activation changes in dual-task processing:An fMRI study. Cerebral Cortex, 17, 192–204.

Eriksen, B. A., & Eriksen, C. W. (1974). Effects of noise letters upon the identificationof a target letter in nonsearch task. Perception & Psychophysics, 16(1), 143–149.

Feng, J., Spence, I., & Pratt, J. (2007). Playing an action videogame reduces genderdifferences in spatial cognition. Psychological Science, 18(10), 850–855.

Forster, S., & Lavie, N. (2007). High perceptual load makes everybody equal:Eliminating individual differences in distractibility with load. PsychologicalScience, 18(5), 377–381.

Gilbert, S. J., Spengler, S., Simons, J. S., Steele, J. D., Lawrie, S. M., Frith, C. D., et al.(2006). Functional specialization within rostral prefrontal cortex (area 10): Ameta-analysis. Journal of Cognitive Neuroscience, 18(6), 932–948.

Green, C. S., & Bavelier, D. (2003). Action video games modify visual selectiveattention. Nature, 423, 534–537.

Green, C. S., & Bavelier, D. (2006a). Effects of action video game playing on thespatial distribution of visual selective attention. Journal of ExperimentalPsychology: Human Perception and Performance, 32(6), 1465–1478.

Green, C. S., & Bavelier, D. (2006b). Enumeration versus multiple object tracking:The case of action video game players. Cognition, 101(1), 217–245.

Green, C. S., & Bavelier, D. (2007). Action video game experience alters the spatialresolution of vision. Psychological Science, 18(1), 88–94.

Green, C. S., Pouget, A., & Bavelier, D. (2010). Improved probabilistic inferenceas a general learning mechanism with action video games. Current Biology, 20,1–7.

Hopfinger, J. B., Buonocore, M. H., & Mangun, G. R. (2000). The neural mechanisms oftop-down attentional control. Nature Neuroscience, 3(3), 284–291.

Hubert-Wallander, B. P., Green, C. S., & Bavelier, D. (2010). Stretching the limits ofvisual attention: The case of action video games. WIREs Cognitive Science, Wiley,1, 1–9.

Hubert-Wallander, B., Green, C. S., Sugarman, M., & Bavelier, D. (2011). Change insearch rate but not in the dynamics of exogenous attention in action video gameplayers. Journal of Vision, doi:10.1167/10.7.1300.

Jenkinson, M. (2003). A fast, automated, n-dimensional phase unwrappingalgorithm. Magnetic Resonance in Medicine, 49(1), 193–197.

Jenkinson, M., Bannister, P., Brady, M., & Smith, S. (2002). Improved optimization forthe robust and accurate linear registration and motion correction of brainimages. NeuroImage, 17(2), 825–841.

Jenkinson, M., & Smith, S. (2001). A global optimisation method for robust affineregistration of brain images. Medical Image Analysis, 5(2), 143–156.

Jenkinson, M., Wilson, J., & Jezzard, P. (2004). A perturbation method for magneticfield calculations of non-conductive objects. Magnetic Resonance in Medicine,52(3), 471–477.

Jha, A., Krompinger, J., & Baime, M. J. (2007). Mindfulness training modifiessubsystems of attention. Cognitive, Affective, & Behavioral Neuroscience, 7,109–119.

Koechlin, E., Basso, G., Pietrini, P., Panzer, S., & Grafman, J. (1999). The role of theanterior prefrontal cortex in human cognition. Nature, 399(13 May), 148–151.

Lavie, N. (1997). Visual feature integration and focused attention: Responsecompetition from multiple distractor features. Perception & Psychophysics,59(4), 543–556.

Lavie, N. (2005). Distracted and confused?: Selective attention under load. Trends inCognitive Sciences, 9(2), 75–82.

Lutz, A., Slagter, H. A., Dunne, J. D., & Davidson, R. J. (2008). Attention regulation andmonitoring in meditation. Trends in Cognitive Sciences, 12(4), 163–169.

Massey, S. C. (2006). Functional anatomy of the mammalian retina. In S. J. Ryan(Ed.). Retina (Vol. 1, pp. 43–82). Elsevier.

Mishra, J., Zinni, M., Bavelier, D., & Hillyard, S. A. (2011). Neural basis of superiorperformance of video-game players in an attention-demanding task. The Journalof Neuroscience, 31(3), 992–998.

Poldrack, R. A., Sabb, F. W., Foerde, K., Tom, S. M., Asarnow, R. F., Bookheimer, S. Y.,et al. (2005). The neural correlates of motor skill automaticity. Journal ofNeuroscience, 25(22), 5356–5364.

D. Bavelier et al. / Vision Research 61 (2012) 132–143 143

Proksch, J., & Bavelier, D. (2002). Changes in the spatial distribution of visualattention after early deafness. Journal of Cognitive Neuroscience, 14, 1–5.

Puttemans, V., Wenderoth, N., & Swinnen, S. P. (2005). Changes in brain activationduring the acquisition of a multifrequency bimanual coordination task: Fromthe cognitive stage to advanced levels of automaticity. Journal of Neuroscience,25(17), 4270–4278.

Raichle, M. E., Fiez, J. A., Videen, T. O., MacLeod, A.-M. K., Pardo, J. V., Fox, P. T., et al.(1994). Practice-related changes in human brain functional anatomy duringnonmotor learning. Cerebral Cortex, 4(January/February), 8–26.

Ramnani, N., & Owen, A. M. (2004). Anterior prefrontal cortex: Insights intofunction from anatomy and neuroimaging. Nature Reviews Neuroscience, 5,184–194.

Rees, G., Frith, C. D., & Lavie, N. (1997). Modulating irrelevant motion perception byvarying attentional load in an unrelated task. Science, 278, 1616–1619.

Romano, J., & Virsu, V. (1979). An estimation and application of the human corticalmagnification factor. Experimental Brain Research, 37, 495–510.

Ruff, C. C., Blankenburg, F., Bjoertomt, O., Bestmann, S., Freeman, E., Haynes, J., et al.(2006). Concurrent TMS-fMRI and psychophysics reveal frontal influences onhuman retinotopic visual cortex. Current Biology, 16(15), 1479–1488.

Schulz, K. P., Bédard, A.-C. V., Czarnecki, R., & Fan, J. (2011). Preparatory activity andconnectivity in dorsal anterior cingulate cortex for cognitive control.NeuroImage, 57(1), 242–250.

Sekuler, R., & Ball, K. (1986). Visual localization: Age and practice. Journal of OpticalSociety of America A, 3(6), 864–867.

Serences, J. T., Yantis, S., Culberson, A., & Awh, E. (2004). Preparatory activity invisual cortex indexes distractor suppression during covert spatial orienting.Journal of Neurophysiology, 92, 3538–3545.

Shadmehr, R., & Holcomb, H. H. (1997). Neural correlates of motor memoryconsolidation. Science, 277, 821–825.

Shapiro, K. (1994). The attentional blink: The brain’s ‘‘eyeblink’’. Current Directionsin Psychological Science, 3(3), 86–89.

Shulman, G. L., Astafiev, S. V., Franke, D., Pope, D. L. W., Snyder, A. Z., McAvoy, M. P.,et al. (2009). Interaction of stimulus-driven reorienting and expectation inventral and dorsal frontoparietal and basal ganglia-cortical networks. TheJournal of Neuroscience, 29(14), 4392–4407.

Shulman, G. L., Pope, D. L. W., Astafiev, S. V., McAvoy, M. P., Snyder, A. Z., & Corbetta,M. (2010). Right hemisphere dominance during spatial selective attention andtarget detection occurs outside the dorsal fronto-parietal network. The Journalof Neuroscience, 30(10), 3640–3651.

Smith, S. M. (2002). Fast robust automated brain extraction. Human Brain Mapping,17(3), 143–155.

Smith, S. M., Jenkinson, M. W., Beckmann, C., Behrens, T., Johansen-Berg, H.,Bannister, P., et al. (2004). Advances in functional and structural MR imageanalysis and implementation as FSL. NeuroImage, 23(S1), 208–219.

Spence, I., Yu, J. J., Feng, J., & Marshman, J. (2009). Women match men whenlearning a spatial skill. Journal of Experimental Psychology: Learning, Memory, andCognition, 35(4), 1097–1103.

Tang, Y. Y., & Posner, M. I. (2009). Attention training and attention state training.Trends in Cognitive Sciences, 13(5), 222–227.

Toepper, M., Gebhardt, H., Beblo, T., Thomas, C., Driessen, M., Bischoff, M., et al.(2010). Functional correlates of distractor suppression during spatial workingmemory encoding. Neuroscience, 165(4), 1244–1253.

Treisman, A. M., & Gelade, G. (1980). A feature-integration theory of attention.Cognitive Psychology, 12, 97–136.

Trick, L. M., Jaspers-Fayer, F., & Sethi, N. (2005). Multiple-object tracking in children:The ‘‘Catch the Spies’’ task. Cognitive Development, 20(3), 373–387.

Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix,N., et al. (2002). Automated anatomical labelling of activations in SPM using amacroscopic anatomical parcellation of the MNI MRI single-subject brain.NeuroImage, 15(1), 273–289.

Vincent, J. L., Kahn, I., Snyder, A. Z., Raichle, M. E., & Buckner, R. L. (2008). Evidencefor a frontoparietal control system revealed by intrinsic functional connectivity.Journal of Neurophysiology, 100(6), 3328–3342.

West, G. L., Stevens, S. A., Pun, C., & Pratt, J. (2008). Visuospatial experiencemodulates attentional capture: Evidence from action video game players.Journal of Vision, 8(16), 1–9.

Woolrich, M. W., Behrens, T., Beckmann, C., Jenkinson, M., & Smith, S. M. (2004).Multi-level linear modelling for FMRI group analysis using Bayesian inference.NeuroImage, 21(4), 1732–1747.

Woolrich, M. W., Jbabdi, S., Patenaude, B., Chappell, M., Makni, S., Behrens, T., et al.(2009). Bayesian analysis of neuroimaging data in FSL. NeuroImage, 45,S173–S186.

Woolrich, M. W., Ripley, B. D., Brady, M., & Smith, S. M. (2001). Temporalautocorrelation in univariate linear modeling of fMRI data. NeuroImage, 14(6),1370–1386.

Worsley, K. J., Evans, A. C., Marrett, S., & Neelin, P. (1992). A three-dimensionalstatistical analysis for CBF activation studies in human brain. Journal of CerebralBlood Flow & Metabolism, 12, 900–918.